EEVblog Electronics Community Forum

Electronics => Metrology => Topic started by: fmaimon on April 14, 2013, 08:21:11 pm

Title: LM399 based 10 V reference
Post by: fmaimon on April 14, 2013, 08:21:11 pm
I'm designing a LM399 based 10V reference. It will be used with Richard's (amspire) general purpose power supply, probably the mk3 one (https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg111557/#msg111557).

Anyway, I want to make it with as high accuracy as I can, so I bought these resistors from ebay, in 75K, 32K and 3k11 values: http://www.ebay.com/itm/10-Vishay-TCC-S102C-S102K-Metal-Foil-0-01-Precision-Resistors-variety-value-/170926783729 (http://www.ebay.com/itm/10-Vishay-TCC-S102C-S102K-Metal-Foil-0-01-Precision-Resistors-variety-value-/170926783729)

They are supposedly used Vishay S102K (http://www.vishaypg.com/docs/63001/63001.pdf) metal foil resistors 0.01% 2 ppm/C. I've measured them with my Agilent 34410A and they are within specs on the resistance, but I didn't measure the tempco.

I also bought some LM399A from ebay and made this:
(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=44356;image)

The PCB layout is this one

(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=44358;image)

What do you think? As I don't have many LM399 and resistors (they are expensive!), I don't want to build one without some feedback.
Title: Re: LM399 based 10 V reference
Post by: free_electron on April 14, 2013, 09:35:44 pm
throw away that op177... ad706 is the way to go.

much lower input noise.
Title: Re: LM399 based 10 V reference
Post by: ivan747 on April 14, 2013, 10:45:17 pm
You will need to pay attention to a lot of details in the power supply in order to achieve the kind of accuracy you will get out of this reference, if you get it right. High spec op-amps and remote sensing will be needed to remove offsets and line regulation over load, respectively. You will also have to take into account the switching noise (mk. 3 is designed to be used with a tracking pre-regulator). If you don't filter the output right, your accuracy and stability will be down into the noise.

Good luck, I'm looking forward to this project  :-+
Title: Re: LM399 based 10 V reference
Post by: codeboy2k on April 14, 2013, 10:50:29 pm
Have you breadboarded this circuit?  It looks like a comparator.   How does it even start??

There is no clear starting current to get the zener into breakdown... perhaps the opamp starts hunting from rail to rail at the start and then it gets going? I am not sure. I don't like it this way, since you're not sure if it will start.

You picked an LM399, so you want stability, so let's give it some.  You should have a constant current source on that board to feed the zener some real, reliable current, about 1.5mA or so. The LM399 can (at its worst) change its output voltage about 2mV for every 1 mA change in input current.   The simplest constant current source is a constant voltage source and a resistor. So use your 15V regulated supply and a 5.6k resistor for about 1.45mA, and feed that into the top of the zener, it's not the best design, but it's ok.  I say it's not the best because the 15V can change, according to the LM317 specs.  i.e. it can change because of load, or temperature, or the 2 divider resistors that set the voltage can heat up and change, so your 15V changes too... When it changes, then your zener current will change (slightly), and the reference will change (slightly). How much it changes you will have to do the math :) If you are concerned about this, then the best love you can give your LM399 is to use another opamp to make a precision current source of 1.5mA and use that to feed the LM399. (there are many examples in almost every datasheet) 

Once you have a constant current, then you can take the reference voltage from the top of the zener and buffer it with a low offset, low drift opamp. You can also use this same opamp to gain up your reference to 10V here, if you wish to.

The LM399 is very stable, but not accurate, so you'll need some way to trim the voltage to your desired reference. That's why I suggested the opamp to buffer and gain it up and at the same time you can trim it to exactly 10V.

The suggestion Free Electron made to use a AD706 is perfect,  low offset, low drift, low bias, no chopper noise, and you get 2 in a single package, so one can be your current source :)  But to make a current source you'll need another shunt reference, it doesn't need to be the same spec as the main reference, because when it changes there's a much, much smaller effect on the main reference.


Title: Re: LM399 based 10 V reference
Post by: Rufus on April 14, 2013, 11:20:50 pm
You should have a constant current source on that board to feed the zener some real, reliable current,

His circuit provides a constant current source as stable as the reference.

I would be surprised if it doesn't start the OPA177 can't drive to ground. It probably sources some current while trying to, if not a bit of pull up on the op-amp output will fix that.

And yes if it is supposed to be a 10v reference it needs +/-5% of trim.
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 15, 2013, 12:11:06 am
throw away that op177... ad706 is the way to go.

much lower input noise.

I've choose the OPA177 as I already have lots of it. But why the AD706 is better than the OPA177, in this case? Is it worth to buy the AD706 as the LM399 is have a typical spec of 7uV RMS noise? In both cases they are at least an order of magnitude better.
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 15, 2013, 12:14:41 am
You will need to pay attention to a lot of details in the power supply in order to achieve the kind of accuracy you will get out of this reference, if you get it right. High spec op-amps and remote sensing will be needed to remove offsets and line regulation over load, respectively. You will also have to take into account the switching noise (mk. 3 is designed to be used with a tracking pre-regulator). If you don't filter the output right, your accuracy and stability will be down into the noise.

I'm aware of the noise of the pre-regulator. I'm just concentrating in one piece at a time, so right now it is voltage reference time...  ;D

Good luck, I'm looking forward to this project  :-+

Thank you. Unfortunately, don't hold your breath too much. Lately I'm not getting much free time, so it will take long.  :(
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 15, 2013, 12:31:57 am
Have you breadboarded this circuit?  It looks like a comparator. How does it even start??

There is no clear starting current to get the zener into breakdown... perhaps the opamp starts hunting from rail to rail at the start and then it gets going? I am not sure. I don't like it this way, since you're not sure if it will start.

You picked an LM399, so you want stability, so let's give it some.  You should have a constant current source on that board to feed the zener some real, reliable current, about 1.5mA or so. The LM399 can (at its worst) change its output voltage about 2mV for every 1 mA change in input current.   The simplest constant current source is a constant voltage source and a resistor. So use your 15V regulated supply and a 5.6k resistor for about 1.45mA, and feed that into the top of the zener, it's not the best design, but it's ok.  I say it's not the best because the 15V can change, according to the LM317 specs.  i.e. it can change because of load, or temperature, or the 2 divider resistors that set the voltage can heat up and change, so your 15V changes too... When it changes, then your zener current will change (slightly), and the reference will change (slightly). How much it changes you will have to do the math :) If you are concerned about this, then the best love you can give your LM399 is to use another opamp to make a precision current source of 1.5mA and use that to feed the LM399. (there are many examples in almost every datasheet) 

Once you have a constant current, then you can take the reference voltage from the top of the zener and buffer it with a low offset, low drift opamp. You can also use this same opamp to gain up your reference to 10V here, if you wish to.

The LM399 is very stable, but not accurate, so you'll need some way to trim the voltage to your desired reference. That's why I suggested the opamp to buffer and gain it up and at the same time you can trim it to exactly 10V.

The suggestion Free Electron made to use a AD706 is perfect,  low offset, low drift, low bias, no chopper noise, and you get 2 in a single package, so one can be your current source :)  But to make a current source you'll need another shunt reference, it doesn't need to be the same spec as the main reference, because when it changes there's a much, much smaller effect on the main reference.

I've not breadboarded it, but I'm pretty confident it will work. It is a constant current circuit. Initially, as the OPA177 output can't go to ground (typ. 2V), we will get an initial current of about 0.6 mA to the zener, through R4. It will start it up and get to about 6.9V. This voltage will be amplified by 1.468 to about 10.13V. Then the current will be (10.13 - 6.9) / 3k11 (R4) = 1.04 mA and will be as stable as the LM399.

I'm really not worried on the accuracy of the reference. 10.13 V will be good enough. The output voltage will be adjusted by the circuit described in this post. (https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg125027/#msg125027)

Now reading the datasheets, why is the AD706 much better in noise than the OPA177? The typical input voltage noise @ 1kHz BW of the 706 is 15 nV/sqr(Hz) and the 177 is about 8 nV/sqr(Hz), both reading the noise graphs.
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 15, 2013, 12:37:01 am
You should have a constant current source on that board to feed the zener some real, reliable current,

His circuit provides a constant current source as stable as the reference.

I would be surprised if it doesn't start the OPA177 can't drive to ground. It probably sources some current while trying to, if not a bit of pull up on the op-amp output will fix that.

That's my plan. First I will etch a board at home. If it works, I will send it to seeedstudio or itead, else, I'll fix it.


And yes if it is supposed to be a 10v reference it needs +/-5% of trim.

As I said in the other post, I don't really care of the accuracy. I will trim it elsewhere in the power supply. I just want it stable. ;D
Title: Re: LM399 based 10 V reference
Post by: BravoV on April 15, 2013, 12:52:29 am
There are 4 purple colored rectangles at the pcb (1 vert & 3 horz), are they cut out holes ?
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 15, 2013, 12:54:49 am
There are 4 purple colored rectangles at the pcb (1 vert & 3 horz), are they cut out holes ?

Yes. I don't want the heat from the regulator getting to the reference circuit.
Title: Re: LM399 based 10 V reference
Post by: ivan747 on April 15, 2013, 01:07:11 am
But to make a current source you'll need another shunt reference, it doesn't need to be the same spec as the main reference, because when it changes there's a much, much smaller effect on the main reference.

A carefully designed current source based on the LM336 voltage reference should work fine. There's a circuit on the datasheet for minimal temperature coefficient. Tolerance doesn't matter, you will have to calibrate the gain of the buffer op-amp to get 10.0000V anyway. Stability is rated at 20ppm per 1000 hours. With temperature coefficient compensated and this stability, you should have a decent current source for the LM399.

Now we have an LM317 (first reference) feeding a LM336 (second reference) with a constant current, only using a series resistor, and finally the LM339, with a proper current source.
Title: Re: LM399 based 10 V reference
Post by: codeboy2k on April 15, 2013, 05:42:32 am
His circuit provides a constant current source as stable as the reference.

I would be surprised if it doesn't start the OPA177 can't drive to ground. It probably sources some current while trying to, if not a bit of pull up on the op-amp output will fix that.
I've not breadboarded it, but I'm pretty confident it will work. It is a constant current circuit. Initially, as the OPA177 output can't go to ground (typ. 2V), we will get an initial current of about 0.6 mA to the zener, through R4. It will start it up and get to about 6.9V. This voltage will be amplified by 1.468 to about 10.13V. Then the current will be (10.13 - 6.9) / 3k11 (R4) = 1.04 mA and will be as stable as the LM399.

Given what you've told me about its startup, then I think I am also convinced it will get enough starting current and will eventually reach regulation.  And I understand that the output buffer drops a voltage across the 3k11 and provides the drive current.  But I see that design decision has problems that are easily remedied.

1) All of the regulating current for the reference comes from the output of the opamp, which is also feeding other circuits in your power supply. You should immediately see this is a problem for maintaining the stability of your reference. This would only be acceptable if you are going to put another voltage follower immediately at J2 before it goes anywhere else on to the rest of your circuits. And if so, why not put that buffer on the reference board itself, so that J2 is the output of a real buffer, not the output of your self-referential voltage reference that also provides it's own current source.

2) R4 provides a path for any voltage or current noise from outside your reference board to get in and throw the reference off course.

3) if J2 is feeding outside, and something outside sinks a large current, say an ADC sample and hold event, then it will steal current that is supposed to be for your constant current source to your reference. I don't know how you can call it a constant current source when you are taking that node off board and allowing other circuit nodes to siphon current away from it. That net feeding your LM399 should be stable and quiescent. Yes, I realize the opamp will provide more drive current when needed, but it cannot recover instantaneously, and do you really want the junction at the top of your current source changing that much due to dynamic loads off-board? I would not.

4) I don't like that your reference has no actually set current.. you really don't know what it will be, just trust that it will stop rising in voltage value as the opamp gives it more and more current until it settles.. and yes, I understand at that point it will be stable. But in my opinion I'd like to actually SET the current to something I specify, and maintain it there.

My design would have a separate op amp for the current source and one more as a buffer. As I said, if your going to the effort of using an LM399, why do something that can mess up your reference?

Actually, I'd guild the lily, and use the second opamp not for merely a current source for the main reference, but as a servo for the LM317 providing your 15V rail.  servo the LM317 to stabilize it over time and temperature, and then you can set your 1.5mA  current reference with a 5k42 low tempco resistor through the LM399. Use the original opamp as a  buffer amplifier where you can dial it up to 10.0000V

(http://i.imgur.com/OEYrtMY.jpg)

Quote
Now reading the datasheets, why is the AD706 much better in noise than the OPA177? The typical input voltage noise @ 1kHz BW of the 706 is 15 nV/sqr(Hz) and the 177 is about 8 nV/sqr(Hz), both reading the noise graphs.

Actually, I'm quite familiar with the AD706 so I know it's a good amp and there's something to be said for going with what you know.  But I just read the datasheets for the OPA177 too, and I agree, it seems quite a lot better in noise, offset and drift specs, and is designed to replace the AD707, which is an even better amp then the AD706.

The only thing the AD706 has over the OPA177 would seem to be its sub-nanoAmp bias currents, which you don't need to worry about. So your choice of OPA177 will work good here.

Sorry for the long post.  I get windy.
Title: Re: LM399 based 10 V reference
Post by: kt315 on April 17, 2013, 12:33:46 am
I am considering building Richard's PSU as well. I do not have a lot of experience with electronics, so excuse my ignorance ...

What are the advantages/disadvantages of using parts like LTC6655B or LT6654A comparing to LM399?

Also, which parts precision would affect accuracy of the reference circuit the most? My current understanding, is that it is critical to get the apm gain right on the reference, hence the divider accuracy would be critical.

Would not it be a problem for precision if the reference is in fact a separate board?

EDIT: In a mean time, off to ebay I go to buy some fancy TFT display ...
Title: Re: LM399 based 10 V reference
Post by: free_electron on April 17, 2013, 04:27:51 am
Zeners are current driven objects. The input cuurent noise is in the femto ampere range for an ad706 and i. The picoamp range fro the op177. Furthermore the ad706 is dc stable. This opamp is built for dc operation. The input bias current is also stable over temperature, that of an op177 is not.

This drift in n put bias current willl shift the operating point of the zener ever so slightly.

Remeber that you are fishing for digits behind the comma here. So the question is: do you want a 10.0000 volt reference , i. Which case the op177 will work. Or do you want a 10.000000 reference that is temperature stable . In which case you will need the ad706...

Title: Re: LM399 based 10 V reference
Post by: codeboy2k on April 17, 2013, 01:11:56 pm
One of the great things about Dave's forum here is that there are great discussions that get one thinking..

I was curious about this discussion, the AD706 vs the OPA177.. So I did some math. I've attached images and Smath Studio output for you viewing pleasure. 

Summary: Even though the initial offset of the AD706 is 100uV and the initial Vos of the OPA177 would seem better at 60uV, the actual output error is dominated by the bias and offset currents through the parallel combination of the (Rfeedback||Rinput), to the point that the total initial error due to offset voltage, offset current and bias current is smaller with the AD706.  Furthermore, a 20C rise in ambient temps will change the output of the OPA177 by ±11.6mV (±6.7mV RMS) ±179uV (±106uV RMS) but will only change ±45uV (±44 uV RMS) when using the AD706.  I did the same with a 5C rise, and the OPA177 will change by ±2.9mV (±1.7mV RMS) ±45uV (±27uV RMS) and the AD706 will change by only ±11uV (±11uV RMS)

Conclusion.. The AD706 is clearly more temperature stable, and its lower input currents even help in reducing total output offset errors.  However, the difference is about 130uV over a 20C change, and a mere 30uV at 5C.  Reducing the feedback and input resistors while maintaining the desired ratio can improve this.  Also, lower input resistors lowers the input noise as well.

And as Free Electron said, the higher bias and offset currents of the OPA177 will ever-so-slightly change the zener reference setpoint, and so that additional zener change will be added to the offset error, and will make the output offset error of the OPA177 even higher than I calculated here.  The AD706 will affect the zener much less.

Also, the OPA177 has better noise specs than the AD706, but I did not do noise calculations. Given the closeness in temperature performance, especially at low temperature changes (ie. when the environment of the reference might be at a stable temperature), and that this can further be lessened by smaller valued resisters, then  perhaps the OPA177 might be a better choice for its noise spec alone, if the error budget allows for the 10s of uV changes due to temperature.  You would be able to achieve 10.0000XX. The AD706 would give one more digit, at 10.00000X

The images attached are my results.  The PDF is the same as the images, but in one PDF file. The ZIP is the source of the Smath Studio input file. Smath Studio is a pretty decent free Mathcad clone and I used it to do all the calculations here and print the images.


[edit: updated the attachments to correct the mistake re: OPA177 Ib and Ios]
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 17, 2013, 04:28:07 pm
Given what you've told me about its startup, then I think I am also convinced it will get enough starting current and will eventually reach regulation.  And I understand that the output buffer drops a voltage across the 3k11 and provides the drive current.  But I see that design decision has problems that are easily remedied.

1) All of the regulating current for the reference comes from the output of the opamp, which is also feeding other circuits in your power supply. You should immediately see this is a problem for maintaining the stability of your reference. This would only be acceptable if you are going to put another voltage follower immediately at J2 before it goes anywhere else on to the rest of your circuits. And if so, why not put that buffer on the reference board itself, so that J2 is the output of a real buffer, not the output of your self-referential voltage reference that also provides it's own current source.

2) R4 provides a path for any voltage or current noise from outside your reference board to get in and throw the reference off course.

3) if J2 is feeding outside, and something outside sinks a large current, say an ADC sample and hold event, then it will steal current that is supposed to be for your constant current source to your reference. I don't know how you can call it a constant current source when you are taking that node off board and allowing other circuit nodes to siphon current away from it. That net feeding your LM399 should be stable and quiescent. Yes, I realize the opamp will provide more drive current when needed, but it cannot recover instantaneously, and do you really want the junction at the top of your current source changing that much due to dynamic loads off-board? I would not.

I really don't see much problem, as long as the output current is within the opamp limits. If I recall correctly, not even the 3458a has an output buffer in it's reference board. But, as I will put an buffer after the board, there is no harm in relocating it.

4) I don't like that your reference has no actually set current.. you really don't know what it will be, just trust that it will stop rising in voltage value as the opamp gives it more and more current until it settles.. and yes, I understand at that point it will be stable. But in my opinion I'd like to actually SET the current to something I specify, and maintain it there.

I don't follow you there. The current is set by R4. Why are you worried about the turn-on transient current if it every circuit will have the same problem? Even the best constant current circuit will do exactly that as it turns on. In the end, it won't matter, as I intend to keep it on as long as I can.


Actually, I'd guild the lily, and use the second opamp not for merely a current source for the main reference, but as a servo for the LM317 providing your 15V rail.  servo the LM317 to stabilize it over time and temperature, and then you can set your 1.5mA  current reference with a 5k42 low tempco resistor through the LM399. Use the original opamp as a  buffer amplifier where you can dial it up to 10.0000V

The circuit, as you drawn is pretty much the same as my circuit, using the LM317 as a high output current buffer. If you put another buffer in there, you will loose that.


One of the great things about Dave's forum here is that there are great discussions that get one thinking..

I agree with that!  :-+

I was curious about this discussion, the AD706 vs the OPA177.. So I did some math. I've attached images and Smath Studio output for you viewing pleasure. 

Where did you get those bias drift numbers for the OPA177? The datasheet, page 2, (http://www.ti.com/lit/ds/sbos008/sbos008.pdf) lists 85 pA/C Ios drift and 60 pA/C Ib drift, a lot worse than AD706 but much better that the ones you've used.
Title: Re: LM399 based 10 V reference
Post by: codeboy2k on April 17, 2013, 05:11:42 pm
I don't follow you there. The current is set by R4. Why are you worried about the turn-on transient current if it every circuit will have the same problem? Even the best constant current circuit will do exactly that as it turns on. In the end, it won't matter, as I intend to keep it on as long as I can.

Yes, I realize R4 sets the current, and it will be as stable as the LM399 itself is.  I'm not worried about the turn-on current per-se, but that you don't know what your current through R4 will be, exactly, in the end. The feedback from the opamp output as it rises to 10.XXX volts will provide a rising current through R4 to the LM399. Once the rising current stops making big enough changes in the zener's output, then the opamp output will also stabilize at 10.XXXX something volts.  I clearly understand this. I think because your application doesn't care about the final voltage, that you can accept it this way. You said you would correct it in the PSU software.

Quote
The circuit, as you drawn is pretty much the same as my circuit, using the LM317 as a high output current buffer. If you put another buffer in there, you will loose that.

I only provided a way to stabilize the +15V power rail from the LM317, using a 2.5V reference instead of the 220 ohm voltage dividers to the ADJ pin.  A stable reference and opamp to servo the ADJ pin.  Thus the +15V power is now stable over time and temperature, and can then be used to provide stable current to the LM399.  At the same time, the rest of your reference circuit would benefit from a more stable power rail.  Its probably not necessary in your application, as I realize you aren't concerned that the final voltage isn't exactly 10.0000 and so you actually have a much more stable current than anything else, since you're using the LM399 itself to provide it's own current source. Just don't let anyone (anything!) take away current from that output, so it's good that you will buffer it.

Quote
Where did you get those bias drift numbers for the OPA177? The datasheet, page 2, (http://www.ti.com/lit/ds/sbos008/sbos008.pdf) lists 85 pA/C Ios drift and 60 pA/C Ib drift, a lot worse than AD706 but much better that the ones you've used.

Oops.. I'm cross-eyed. mistake. my bad. I picked up the numbers I used from the line above the real numbers that you quoted. Yes, the real numbers are 85 pA and 60 pA/C.  Thanks for catching that... I'll fix it and re-upload the pages.
Title: Re: LM399 based 10 V reference
Post by: branadic on April 17, 2013, 06:23:01 pm
Hi,

a few weeks ago I've build the circuit presented in the datasheet of the LM199/399 (Linear Technology) called "10V Buffered Reference" twice.
The references where pre-measured to find the best fitting resistor values. The resistors right now are smd types with 5ppm/K, contrary to the pictures attached.
For stability of them I thermo-glued a LTCC-board-heat-regulator on top of them (http://www.kuhne-electronic.de/en/products/crystal-heater/qh-40-a.html (http://www.kuhne-electronic.de/en/products/crystal-heater/qh-40-a.html)). This solution is much cheaper compared to 1ppm/K resistors and was faster to solve.
BTW: Resistors can be bought over here http://www.rhopointcomponents.com/components/resistors/precision-through-hole.html (http://www.rhopointcomponents.com/components/resistors/precision-through-hole.html)
The LM399 is packaged into a thick walled plastic package on both pcb sides without contact to the slot arms to keep air flow away.
Both boards have know seen a 300h burn-in @ 120°C and output 10,0019V and 10,0034V powered from a lab power supply (stable for more than a week, poorly measured only with a 34410A because of missing a 3458A or null voltmeter) without change in voltage under lab conditions.
You said you want it stable, then take humidity into account too. I've used the LT1001 in CerDIP package sitting in a precision socket.
Title: Re: LM399 based 10 V reference
Post by: ddavidebor on April 17, 2013, 06:44:12 pm
wow, amazing design!
Title: Re: LM399 based 10 V reference
Post by: codeboy2k on April 17, 2013, 06:47:59 pm
Those are nice boards you built.  The heater you linked to is pretty good, it claims to regulate the temperature to with 0.1C

If a reference is made with that heater, and also heating the opamp and resistors, and could keep it all within 0.1C, then it will be really stable, as you've seen.  With a 0.1C stable temperature, the OPA177 from this thread will only change between  500-900 pV at the output.  The AD706 is still better, will change only 200pV.  But both opamps should be stable to 1uV, and give 10.000000, and for sure able to hit 10.00000X (i.e. 10uV, and the 1uV digit would be uncertain)
Title: Re: LM399 based 10 V reference
Post by: branadic on April 17, 2013, 07:33:47 pm
The boards are nothing special, just my expermiental environment with the LM399.
BTW: Current Sources & Voltage References by Linden T. Harrison is very nice "evening-reading".
You can read that the LM199 was designed by Robert Dobkin and introduced in the early 1970s by National Semiconductor, the predecessor of the LM299/399/3999 which had less stringent specifications and covered shorter temperature ranges. He also writes about 90°C chip temperature, that it requires 7,5V to operate the reference and more than 9V for the heater.

Nevertheless, the 10V are impractical in most todays applicatons and a voltage divider (done in the 34410A to produce also 5V out of the 6.95V of the reference) will introduce further tempco and noise. You can overcome the tempco by using a resistor network or expensive low / zero tc resistors, but this is also a bad way.
I agree that the heater is very interesting and as mentioned the LTC6655 but also the LT1236 have perfect characteristics when using them in the LS8 package. Combined with a heater/oven that can give a more practical to use reference voltage for adc or dac applications.
So this is the idea I'll follow next. I've two LTC6655BHLS8-2.5, a few LT1236AILS8-5 and two further heaters.
Voltage references, to be continiued...  :-DD
Title: Re: LM399 based 10 V reference
Post by: codeboy2k on April 17, 2013, 10:13:17 pm
yes, I agree 10V as  a reference inside an instrument is not really necessary anymore.  There are so many good references that it's easy to find one off the shelf that fits and is accurate enough for the task at hand, depending on the error budget allowed.

This volt-nut stuff is (in my opinion) for something on the bench, i.e. a 10V stable benchtop reference that I would use with a K-V divider for research  and prototyping. Or as a stable reference for a precision current source down to the 100 uV or so.  I realize that the OP is using this reference in a precision PSU. 

I don't think I can use a PSU that needed a reference this stable.  If I am building a PSU I would want it settable to 1mV resolution at the best, and probably 10mV is good enough, for my needs. Obviously others have different needs.

My current favorite plug and play reference that can be installed and doesn't need calibration or trimming is the MAX63XX series, with the best chip having 0.01% initial accuracy and 1ppm/C drift. It's expensive, like the LT6655, but I think it's a little better too.

Have you ever used Intersil references? I have not, so I was curious about their FGA technology compared to the buried zeners and bandgaps.

Title: Re: LM399 based 10 V reference
Post by: free_electron on April 18, 2013, 12:15:06 am
throw away the socket... you can get electrical noise due to stress in the socket....
Title: Re: LM399 based 10 V reference
Post by: branadic on April 18, 2013, 06:13:42 am
I was looking forward for some MAX6350MJA, first one's sold on ebay, but tests showed that this were plagiarism. Second time I bought them via a distributor and again got plagiarism, but this time I was sure I could send them back and get back my money.
I don't want to use the plastic package also because of humidity influence and related stress.
Maxim couldn't help me with parts as they don't stock them by now and I would have to buy a dozen oft them (I guess they wrote 100.000 or more in the mail) just for a few one that I need.

throw away the socket... you can get electrical noise due to stress in the socket....
This is a typical "once heard" statement used as an answer for all questions. Stress is less a problem also because of the CerDIP package, air flow is much more critical. But you seem to forget the LT1001 is not the reference but the buffer amp, so I don't care about your "electrical noise due to stress in the socket" argument.

However, there are a few other voltage reference ics in my lab like AD584TH, MAX6126AASA25+, MAX6126AASA50+, AD586LQ, LM334H, REF02DT, ADR444ARZ/ARMZ, enough material for further projects. All that is missing is a good old LTZ1000ACH that I'd like to have as an already used/burned-in board.

BTW: There are a few other threads concering references in eevblog, e.g.:

https://www.eevblog.com/forum/chat/5v-precision-referene/ (https://www.eevblog.com/forum/chat/5v-precision-referene/)
https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/ (https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/)
https://www.eevblog.com/forum/projects/voltage-reference/ (https://www.eevblog.com/forum/projects/voltage-reference/)

And again, there are several links to ready to use boards:

http://www.gellerlabs.com/Voltage%20References.htm (http://www.gellerlabs.com/Voltage%20References.htm)
http://www.voltagestandard.com/ (http://www.voltagestandard.com/)
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 21, 2013, 03:01:05 am
Zeners are current driven objects. The input cuurent noise is in the femto ampere range for an ad706 and i. The picoamp range fro the op177. Furthermore the ad706 is dc stable. This opamp is built for dc operation. The input bias current is also stable over temperature, that of an op177 is not.

This drift in n put bias current willl shift the operating point of the zener ever so slightly.

Could you elaborate a bit more? It is obvious from the datasheets that the AD706 has less drift than the OPA177 (and the OP177, which seems to be exactly the same). Codeboy's calculations shows exactly that with the 177 being about 4 times worse than the 706. What is not obvious is what makes an opamp DC stable? How can I check if an opamp is DC stable and/or built for dc operation by reading the datasheet? AD706's datasheet mentions "High DC Precision" in it's features, but I tend to ignore this "features" and try to find what I want from the specs.
Title: Re: LM399 based 10 V reference
Post by: fmaimon on April 21, 2013, 03:29:23 am
BTW: Resistors can be bought over here http://www.rhopointcomponents.com/components/resistors/precision-through-hole.html (http://www.rhopointcomponents.com/components/resistors/precision-through-hole.html)

Nice find! I've just quoted some resistors. Let's check their prices.
Title: Re: LM399 based 10 V reference
Post by: oilburner on June 17, 2013, 05:51:33 pm
Thank you Branadic, I got this book and it already payed for itself...  I was looking for a MOSFET that could be used for a linear voltage to
current converter and saw the circuit on page 227 that used an ON Semi  MTP3055V.  The original MOSFET I picked was a fast switcher
type that was not as stable with temperature.

Greg
=============
BTW: Current Sources & Voltage References by Linden T. Harrison is very nice "evening-reading".
Title: Re: LM399 based 10 V reference
Post by: branadic on June 17, 2013, 07:13:43 pm
Quote
Thank you Branadic, I got this book and it already payed for itself...
Pretty good stuff isn't it? I still read serveral chapters from time to time.
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 17, 2013, 09:55:18 pm

As I said in the other post, I don't really care of the accuracy. I will trim it elsewhere in the power supply. I just want it stable.
You should consider to stabilize the heater voltage of the LM399 instead of using a (unstabilized?) 18V supply. I have observed 0.5 - 1.5ppm/V output voltage change between 10-16V heater voltage.

Further the LM399 changes output voltage with orientation. Upside down or legs sideways will change output voltage by some ppm's. Current consumption of the heater is a minimum with upside down orientation if you want to save battery power.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: quantumvolt on July 22, 2013, 01:35:22 am
I have a LM399 I will start to play with. In addition to the "staged" version schematic (Linear regulator 15 V, Current source (Resistor or IC),  LM399 Ref., Adjustable Gain Amplifier/Buffer) there are two circuits that uses the reference voltage for current/self bias.

The two circuits discussed in this tread so far seem to be the self bias version and the staged version. What do you think about the "combined voltage/current regulator" version vs. the "self bias" version (compared to the simpler staged version)?

(Edit for pictures)


"Combined"
(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=55525;image)


"Self Bias"
(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=55523;image)


"Staged"
(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=55531;image)
 
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 22, 2013, 09:13:01 pm
The two circuits discussed in this tread so far seem to be the self bias version and the staged version. What do you think about the "combined voltage/current regulator" version vs. the "self bias" version (compared to the simpler staged version)?
Hello,

For the first: you cannot compare circuits which have a higher output voltage than the reference (10V) with a circuit for a lower output voltage (1.018V).

Further you will have to calculate the error budget of the different circuits by yourself. And perhaps publish the results here.

So from my side just some thoughts to the different circuits:

For the combined solution:
+ regulated heater supply (otherwise 0.5 - 1.5 ppm/V drift for input voltage change)
- the load/temperature dependent adjust current influences the zener current
  fortunately the zener itself has a very low impedance of about 0.5 Ohms
  but a 10uA change (e.g. 25-75degrees chip temperature) will result in 5uV (0.7ppm) at the zener.
- the LTC1049 has a relative large noise (2uVpp) against the 7:1 divided noise of the zener (4.2uVpp/7 = 0.6uVpp)
- the LTC1049 can only deliver around 0.5-1 mA at the output. For a buffered reference typically 10 mA are usual.
I would use a LTC2057 instead of the LTC1049.
If you have capacitive loads you will have to do further measures at the output to get it stable.

Self Bias:
- unregulated heater supply (will save battery power but give additional unstability)
+ zener current has lower noise than with other solutions
* the zener current stability depends only on the voltage divider + offset drift and the 5K resistor drift.

Staged:
- unregulated heater supply
- unregulated zener current
- unstable trimming scheme
for this circuit I would use a low noise 14-15V voltage regulator.
up to 20V input voltage a LTC1763 could be used.

With best regards

Andreas




Title: Re: LM399 based 10 V reference
Post by: quantumvolt on July 23, 2013, 12:06:28 am
Thank you.

I understand that noise and drift in the output will be different for 6.95 V divided down / buffered to 1.018 V and 6.95 V amplified up to 10 V. Also the quality of the resistors and ICs will determine the end result. But the resistor arrangement around the op amp on the output as either a voltage divider + voltage follower or a non inverting DC amplifier does not change the operating modus of the circuit.

I was more interested in having opinions on the principles for the 3 solutions. The staged one is "simple" with no "backward interdependence", and the combined solution has "complexity" by making the linear regulator dependent on reference voltage and then letting it feed both heater, reference and op amp.

If no one thinks that any principle is much better then the other, I think I prefer the staged version. Very easy to rig up and set in long time measurement, and then test out refinements by bettering the DC supply / battery, use 18 V regulator for the heater and then step down to 15 V for the reference, use a dedicated current source in stead of a resistor, get better low tempco resistors for the amplifier, try different op amps, put the whole board in a box that is heated to some 40-50 degrees, etc. The simple design makes it easier to see which experimental change produces which effect where ... Imo.

Title: Re: LM399 based 10 V reference
Post by: Andreas on July 23, 2013, 04:27:42 am
But the resistor arrangement around the op amp on the output as either a voltage divider + voltage follower or a non inverting DC amplifier does not change the operating modus of the circuit.

I was more interested in having opinions on the principles for the 3 solutions.
The resistor arrangement has the most influence on the stability of the output voltage. I recommend highest quality precision resistors (either precision wire wound or Vishay metal foil) for this. And yes: they will cost more than the reference itself.

If you look at the schematics of precision DMMs like the HP 34401A (document 34401-90013.pdf) you will always find the self bias circuit to generate the +10V reference voltage for the ADC.
Of course they use a hermetically tight (long term stable) hybrid resistor string for the voltage divider.

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: quantumvolt on July 23, 2013, 06:20:39 am
Thanks again. I guess this is the Agilent 34401A reference schematic:

(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=55645;image)


That makes my choice. I have bought a LTC2400 ADC cheap from fleabay and want to try different chip references at 5 V but also make a 10 V based on the LM399. I hope to make a single range voltmeter with around 0.01% / 1mV accuracy for 10 V measurement.

I will try to use OP177 (have one already) and hand-measured-tempco-selected-matched standard metal film resistors (I have some bands of 1000 pcs. to pick from :phew:). If that does not work, I have some Vishay 0.01% 2.5 ppm/degree that can be used in series/parallel coarse value and then a trim. If necessary I can put a a temperature sensor and a correction table in the microprocessor that will display the measured (corrected) value. All I need is a voltmeter that emulates my 34401A to 4 1/2 digits or better.

If everything else fails, I will buy an AD706 or similar and some decent resistors with the right values.

Is the heater running on 30 V, or have I misunderstood something?

(From the manual: "The instrument precision voltage reference is U403. Resistor R409
provides a stable bias current for the reference zener diode. R408 and
CR404 provide a bias to assure that the reference zener biases to +7 V
during power up.")
Title: Re: LM399 based 10 V reference
Post by: quantumvolt on July 23, 2013, 10:21:56 am
To answer my own question: Yes, the LM399 heater resistor (pin 3 and 4 in U403) in the schematic over does run at 30 V. In this datasheet is another example http://cds.linear.com/docs/en/datasheet/1001fb.pdf (http://cds.linear.com/docs/en/datasheet/1001fb.pdf). With heater voltages from 12-30 V and power (U^2)/R this means that the chip operates on a wide power/temperature range.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 23, 2013, 07:40:58 pm
Hello quantumvolt,

If you have a negative voltage supply, you should use it for the heater. According to my measurements
the power supply rejection ratio of the heater voltage is better the higher the voltage is.
And since heater power is a constant, a higher voltage gives a lower supply current in steady state.

0.01 % longterm stability is a challenge with normal film resistors. They will drift already during soldering up to around  0.1 - 1%. I use capacitive dividers (LTC1043) to divide the 10V input voltage to the input range 0..5V of the LTC2400.
With a good 5V voltage reference and a temperature compensation you can reach below 20ppm/year stability after some pre-ageing.

I think that the OP177 is a better choice for the circuit than the AD706. The important key features are open loop gain, offset voltage drift over temperature and time.
Of course newer chopper OPs like the LTC2057 could get better results.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: quantumvolt on July 23, 2013, 08:32:16 pm
Very good. I have +- 15 V. I did not know that the heater was not "purely resistive".

Due to the odd values of my bands of 1% metal film surplus resistors, the "precision resistors" will be made up from 10-15 resistors in parallel/series where I begin with the large values matching positive and negative tempcos and add smaller and smaller corrections before ending up in a small trim added as "parallel center tap". When I build this "composite resistor" I can check value and tempco after soldering all the resistors together on very small PCBs. The final soldering heat will be on leads added to the PCB's afterwards. I'll do it for fun - if it doesn't work, I'll get a decent divider.

Thanks again. I'll go for the self bias circuit principle with +- 15 V heater / OP177 amplifier, and will be prepared that I may have to get a better ratio divider for amplification to 10 V as a general reference and/or division from 6.95 V to 5 V in order to reference the ADC directly.
Title: Re: LM399 based 10 V reference
Post by: branadic on July 24, 2013, 07:46:00 pm
Quote
I did not know that the heater was not "purely resistive".

Have a look at this:

http://i007.radikal.ru/1202/ad/4d39ad65f42f.jpg (http://i007.radikal.ru/1202/ad/4d39ad65f42f.jpg)

BTW: I was asked for the marking code of the crystal heater. Attached is an image of the board, all parts can be identified by this. But I'm the opinion that it makes no sense to duplicate this board as the circuit is assembled on a ceramic substrate and you won't use FR4.
Title: Re: LM399 based 10 V reference
Post by: babysitter on July 25, 2013, 08:05:10 pm
About 12EUR in germany, this heater. Nice little beast.
Title: Re: LM399 based 10 V reference
Post by: branadic on July 28, 2013, 08:44:42 am
I think a very good design consist of a cascaded LM317 arrangement feeding the heater and a temperature compensated constant current source for the zener (e.g. LM334). Together with a chopper amp such as the named LTC2057 and some pi filter to limit the whiteband noise the design could be completed. I would agree that its worth setting the ouput voltage to 10V nominal and divde them with LTC1043 to 5V or other values.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 28, 2013, 02:14:16 pm
I think a very good design consist of a cascaded LM317 arrangement feeding the heater and a temperature compensated constant current source for the zener (e.g. LM334). Together with a chopper amp such as the named LTC2057 and some pi filter to limit the whiteband noise the design could be completed. I would agree that its worth setting the ouput voltage to 10V nominal and divde them with LTC1043 to 5V or other values.

What do you mean with cascaded arrangement? (schematics).

I did the LM334 solution for my first 2 LM399 References. It is possible but very time consuming to get the tempco of the current source to below 30ppm. I started with the datasheet suggestion of 10:1 resistors (137/1370 Ohms) and ended in a tempco of 305 ppm/K (3600 ppm over a 12 degree span). After some changes to the resistors I ended with 145+1200 Ohms for the one reference and 147+1200 Ohms for the other with a tempco of -30ppm/K.
So its not easy to compensate the +3333ppm/K tempco of the LM334 with a -2mV/K tempco of a diode.

So in the meantime I'm convinced that using good 15ppm/K metal film resistors (RC55Y or RN73) together with the self biased design will reach the same/better result with less effort.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on July 28, 2013, 04:00:19 pm
Quote
What do you mean with cascaded arrangement? (schematics).

"LM317+cascaded"

http://www.acoustica.org.uk/t/3pin_reg_notes4.html (http://www.acoustica.org.uk/t/3pin_reg_notes4.html)

Just one of many possible pages. This will improve PSRR, you maybe remember this thread:

http://www.mikrocontroller.net/topic/295213#3148897 (http://www.mikrocontroller.net/topic/295213#3148897)
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 28, 2013, 06:20:29 pm
Quote
What do you mean with cascaded arrangement? (schematics).

"LM317+cascaded"

http://www.acoustica.org.uk/t/3pin_reg_notes4.html (http://www.acoustica.org.uk/t/3pin_reg_notes4.html)


Ok that makes sense when using mains line powered sensitive equipment.
Of course a good designed ADC or meter will add additional mains line frequency reduction by choosing appropriate integration time.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on July 28, 2013, 08:54:03 pm
Quote
Of course a good designed ADC or meter will add additional mains line frequency reduction by choosing appropriate integration time.

Sure, but you would do at least everything you can as long it's worth but cheap before increasing integration time.
Title: Re: LM399 based 10 V reference
Post by: alm on July 28, 2013, 10:21:47 pm
You're missing the point. By choosing the integration time as an exact multiple of the mains frequency, the ADC can have a much better mains ripple rejection. For 60 Hz, a 16.7ms period will work much better than a 20ms period, for example, even though the integration time is shorter. This is why old meters would have a jumper/front panel setting, or would even have different crystals, for 50 Hz and 60 Hz. Modern meters tend to use a PLL from the mains frequency to produce the sampling clock.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 29, 2013, 04:49:28 am
This is why old meters would have a jumper/front panel setting, or would even have different crystals, for 50 Hz and 60 Hz. Modern meters tend to use a PLL from the mains frequency to produce the sampling clock.

Why not choose a multiple of 100 ms as integration time?
This would give multiple of 5 PLCs on 50 Hz and multiple of 6 PLCs on 60 Hz.

Ok the LTC2400 has to be configured by PIN to 50/60 Hz suppresion (> 110 dB) at Pin 8.
When connecting to VCC the integration time seems to be 80 ms.
(can be measured as spikes on the input of the LTC2400)
The first 80 ms of the 160 ms measurement time are obviously some kind of self calibration mentioned in the data sheet.

With best regards

Andreas.
Title: Re: LM399 based 10 V reference
Post by: alm on July 29, 2013, 05:48:59 pm
Why not choose a multiple of 100 ms as integration time?
This would give multiple of 5 PLCs on 50 Hz and multiple of 6 PLCs on 60 Hz.
That would work. 10 samples per second is a fairly slow sampling rate, though, compared to the 50/60 S/s that are possible with 1 PLC.

Ok the LTC2400 has to be configured by PIN to 50/60 Hz suppresion (> 110 dB) at Pin 8.
When connecting to VCC the integration time seems to be 80 ms.
(can be measured as spikes on the input of the LTC2400)
This pin indeed controls the internal conversion clock to be a multiple of 16.67ms/20ms.
Title: Re: LM399 based 10 V reference
Post by: branadic on August 16, 2013, 05:40:09 pm
My reference is still running, stable at 10.00183 - 10.00184V depending on the labs humidity. I still don't know if this is due to the 34401A or my reference. Attached the actual assembly with the crystal heater on top of the 5ppm/K smd resistors.
But what is hidden to the eye comes visible using an IR-cam.
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 16, 2013, 08:06:47 pm
Good pictures.

By the way: Is the crystal heater not thermally isolated from environment?
Or did you just remove the isolation for the camera?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on August 16, 2013, 08:48:35 pm
Thanks Andreas, we hopefully get the pictures of our small experiment within the next days ;)
The crystal is not thermally isolated with foam directly on its surface, but the complete board is running inside this blue styrofoam box. I just opened it for this pictures.
Title: Re: LM399 based 10 V reference
Post by: babysitter on August 22, 2013, 07:38:37 pm
Today I teached our apprentice how to make PCBs and grabbed his product... and assembled a LM399 Reference ("Portable calibrator" from Datasheet, LM317T preregulator, available OP727R as Amplifier, 0.1% chicken food THT resistors, some C and L to guesstimated-improve EMC and interesting parts hidden in pink foam) during lunch break.

More later...

BR
Hendrik
Title: Re: LM399 based 10 V reference
Post by: branadic on August 23, 2013, 04:30:39 pm
Quote
More later...

What is coming next?
Title: Re: LM399 based 10 V reference
Post by: babysitter on August 23, 2013, 07:08:45 pm
This. Sorry, its in german. Description of my lunch break reference.

http://dg3hda.primeintrag.org/doku.php?id=lm399 (http://dg3hda.primeintrag.org/doku.php?id=lm399)

Greetings
Hendrik
Title: Re: LM399 based 10 V reference
Post by: branadic on August 24, 2013, 01:18:34 pm
I thought there would be something special, cause you worte:

Quote
and interesting parts hidden in pink foam)
Title: Re: LM399 based 10 V reference
Post by: babysitter on August 24, 2013, 03:09:54 pm
Isn't it a nice package ? :)

Well, there is nothing special about it, as the application circuit doesn't call for anything special, right ?
No wire resistors wound by virgins at full moon...

Greetings
Babysitter

Title: Re: LM399 based 10 V reference
Post by: babysitter on August 26, 2013, 11:54:15 am
I received the thermal videos from Branadic, also some background story, and will post it to youtube and will publish a dedicated webpage about his efforts soon.
Title: Re: LM399 based 10 V reference
Post by: branadic on September 02, 2013, 01:20:54 pm
Thanks to Hendrik for setting the videos and the publication online

http://dg3hda.primeintrag.org/doku.php?id=lm399_thermographie (http://dg3hda.primeintrag.org/doku.php?id=lm399_thermographie)
Title: Re: LM399 based 10 V reference
Post by: Andreas on September 22, 2013, 03:03:52 pm
Mhm,

on temperature compensated zeners (1N829A) there exists a "zero tempco current".
Has anyone tested if the LM399 also has some kind of zero tempco current
(near the heater temperature) in the valid range of 0.5 .. 10mA for the zener?
Can the 0.5 - 1 ppm/K be further optimized by this current?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on September 23, 2013, 11:02:11 am
Quote
Has anyone tested if the LM399 also has some kind of zero tempco current
(near the heater temperature) in the valid range of 0.5 .. 10mA for the zener?

I didn't, but an interesting question you ask.
Title: Re: LM399 based 10 V reference
Post by: mwilson on September 24, 2013, 08:56:31 pm
I really dislike buying schmartboards and soldering TSSOP/SOIC packages...

Sorry to not answer your actual question, but it's worth pointing out that SOIC and TSSOP are very different beasts. I'd really recommend you become comfortable hand-soldering SOIC packages. No Schmartboard required... you can buy very inexpensive SOIC adapter boards (for example, I've used many of these: http://www.dipmicro.com/store/PCB-SOIC-SSOP16. (http://www.dipmicro.com/store/PCB-SOIC-SSOP16.) AdaFruit has some as well). There are so many interesting parts that simply aren't available in DIP packages, but if you're lucky are available in SOIC instead of a finer-pitch package. Make sure you use flux and good (which includes being small diameter) solder and you can pretty much just solder SOIC packages one pin at a time.

TSSOP is twice as small (or worse) as SOIC, but with flux and good technique is still pretty doable without needing to buy expensive/bulky Schmartboards. A lot of the adapter boards have SOIC on one side and SSOP on the other side, so you only need to stock one kind of board for a given package pin count and you're covered. It's really not that bad and keeps a whole world of interesting parts open to you.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 04, 2014, 10:15:17 pm
Hello,

I did some further tests regarding the LM399 stability with my LM399 #3 reference after I have removed the socket and soldered the LM399 directly with long leads on the bread board.

Again the results might also be valid for the LTZ1000 reference. So I was not shure wether I should post it here or on the LTZ1000 thread. But since the thermal pictures from LM399 are in this thread it fits better here.
The question from the LTZ1000 thread is still how to achieve the best stability.

I found that on LM399 there is a voltage deviation depending on orientation in room (tilting of LM399). On my LM399 references this is around 3-4 ppm. On a Keithley 2000 I have found even larger deviations of readings depending on flat on table or placed on left or right or top or bottom or front or back side.

The LTZ1000 has a better internal construction. But also some tilting effects can be seen. (see discussion of Dr. Frank and me in September 2010 on the volt-nuts forum).

------

First experiment is LM399 #3 with long leads mounted into a folded box to keep direct air currents away from LM399. (see picture IMG1458w)

Each of the 6 possible orientations (0-5) has been tested with >=15 minutes settling time and 5 minutes measurement time. The 5 minute values are averaged. Measurement setup consists of a 2:1 precision voltage divider and a LTC2400 based ADC. My measurement standard deviation is around 0.25ppm giving up to 2uV  uncertainity with a confidence level of 2 sigma.

The result is on 20140102_LM399_3_boxed.PNG.
Difference between orientation 1 (= pin 1 marker LM399 on bottom) and 3 (= pin 1 marker on top) is 18.7uV or 5.4 ppm as maximum.

The same test with thermal isolation of LM399 #3 by a cotton pad on the solder side and a cotton pad on top of the LM399 fixed with a expanded plastic sleeve (see IMG1452w).

The result is on 20140101_LM399_3_cotton.PNG
The maximum difference is again between orientation 1 and 3 but this time only 4.7uV or 1.4ppm as maximum.

So the thermal isolation seems to minimize the tilting effect on the LM399.

to be continued ...
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 05, 2014, 09:34:15 pm
Hello,

next setup is with the PCB from branadic which was used for the thermal measurements:
For mechanical fixture and air current shielding I put the pcb on one styrofoam sheet and used a second with cutout to fix the pcb within a plastic box. (IMG1465w).


A piece of additional foam (which was used on the thermal measurements) was put on the top of the 2 LM399s. (IMG1464w).

The 2 LM399s are with shortened legs. I call the one with the closed PCB the CH6 and the other with the slotted PCB CH7. The orientations are the same as with my LM399 #3.

0 = upright
1 = pin 2+3 upside (pin 1 marker bottom)
2 = pin 1+2 upside
3 = pin 1+4 upside (pin1 marker top)
4 = pin 3+4 upside
5 = upside down

Unfortunately now I have to measure 2 devices in each orientation. The measurement duration is longer than I can manage within one day when doing my other ageing measurements. Interruptions within measurement are shown as a short dip to orientation zero within the diagram.

The result for shortened legs without slots is shown in 20140104_LM399_CH6_foam.PNG

The maximum difference is about 4.4uV = 1.3ppm which is comparable to the LM399 #3 value with good isolation. Although this is no prove that long and short legs make no difference with good thermal isolation  (since not measured with the same device) it seems to be a indicator for that.
I wish I had equipment with better stability.

The result for shortened legs with slots is shown in 20140104_LM399_CH7_foam.PNG

Here the maximum difference is about 10.1 uV = 2.9 ppm which is significant worse than the CH6 value.
This is what I had expected from the thermo scan pictures since the temperature gradients on the slotted area are much larger than on the non slotted area.

further measurements will follow ...


Title: Re: LM399 based 10 V reference
Post by: Andreas on January 06, 2014, 02:39:05 pm
Hello,

3rd setup of branadic PCB with better thermal shielding similar to LM399 #3:

See picture IMG1467w and IMG1468w.
Cotton pads are on solder side fixed with tape.
And on component side over the LM399s fixed with a foam hose + tape.
All together put into the plastic box from previous setup.

Results:
20140105_LM399_CH6_cotton.PNG
the non slotted board shows 4.1uV maximum difference which is around 1.2 ppm.
So similar to the previous setup.

20140105_LM399_CH7_cotton.PNG
the slotted board shows 3.6 uV maximum difference which is around 1.1 ppm.
Around a factor 3 better than the previous setup.

But stop:
there are shifts in measurement between the two days which I cannot explain by measurement uncertainity.

Mhm, was there enought pre-ageing of the 2 LM399´s?

As branadic stated to me he had only done his thermal tests on the 2 LM399's. A pre-ageing was not necessary for this.
As can be seen in 20140105_LM399_CH_6_7_age.PNG the ageing is about 2.25 uV per day for CH6 and around 3.24uV per day for CH7.

So to say: shi(f)t happens.
I will have to do some correction calculations or repeat the measurements after some run-in phase.
In the mean time I can do some noise comparisons. (e.g. 0.1 - 10 Hz noise).

But anyway:
One result is already clear:
- nothing can replace a good thermal shielding.

There seems to be a second result:
- A slotted board is bad in case there is no good thermal shielding.
- From the first view: if I calculate out the ageing shift it seems to be that with proper thermal shielding the slotted board seems to have a slight advantage.

So the most important question for now is:
- what is the best material with the lowest K-value and at least fair electrically isolation (>> 10 MegOhm) which can be used for a hobbyist to shield a LTZ1000 or a LM399?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 06, 2014, 08:59:52 pm
Hello,

I have done the correction calculation for a linear ageing (linear regression coefficient).

The correction is applied on the averaged measurement values and not for the original 1 minute values.

The spread values with the foam do not really change:
CH6 foam corrected:   4.4 uV = 1.3 ppm (unchanged)
CH7 foam corrected: 10.4 uV = 3.0 ppm (was 10.1 uV)

The spread values for the "cotton" isolation seems even to be better for the slotted PCB:
CH6 cotton corrected:  4.6uV = 1.4 ppm (was 4.1 uV)
CH7 cotton corrected:  2.4uV = 0.7 ppm (was 3.6 uV)

remember that standard deviation for my setup is around 0.25 ppm or 1 uV.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 07, 2014, 07:51:03 am
Thank you, Andreas! It's very interesting result.
I tried to reproduce it with my old triple LM399H 10 V reference connected to the HP34401A DMM. Reference is on-line 24/7 about six months, LM399H without thermal isolation, PCB without slots, HP34401A preheated. Maximum deviation (upright->upside down->upright) is about 0.5 ppm.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 07, 2014, 09:01:35 pm
Hello Mickle,

your experiment orientations are that what I call orientation "0" and "5" where I also have only little differences (up to around 0.5ppm) in my setups.

The pin 1 marker directs to the right side on the photo.
So the critical directions would be the putting them on the left and on the right (connector) sides.
I guess that in this case there would be also about 3-4 ppm difference.

When looking at photos from older HP34401a or Keithley 2000 units the critical directions should be putting them on the left and on the right side. The pin 1 marker shows either to the left or the right side in this case.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 08, 2014, 08:48:00 am
Andreas, you are quite right!
I repeated all the measurements with Datron 1071 in 7.5-digits mode (4 hours preheating). The difference (in critical directions) was 2.1 ppm. Orientation "0" and "5" gives 0.3 ppm.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 12, 2014, 07:09:26 pm
Hello,

as promised now the noise measurements of LM399 in different configurations.

I use a amplifier   x 10000 with a 0.1 Hz .. 10 Hz limitation.
3rd order high pass (+ 10 sec measurement time) and 4th order low pass.
Readout is done on the oscilloscope.
In this case the 5mV / DIV on the scope corresponds to  0.5uV / DIV at the input.
Of cause all measurement are done in a metal cookie box with batteries.

I did 18 measurements in each configuration in upright orientation (see pictures above).
The plastic box + the foam sheets were used to hold the pcb in orientation.
Result is average peak-peak noise and the standard deviation of the 18 measurements.

CH6: cotton (both sides) + foam hose thermal isolation (non slotted board).
average: 2.89 uVpp   stddev: 0.32 uVpp

CH7: cotton (both sides) + foam hose thermal isolation (slotted board).
average: 2.45 uVpp   stddev: 0.27 uVpp

CH6: foam isolation on top (non slotted board).
average: 3.03 uVpp   stddev: 0.29 uVpp

CH7: foam isolation on top  (slotted board).
average: 2.46 uVpp   stddev: 0.21 uVpp

In this case I see no significant change between
good (cotton) and poor (foam) thermal isolation.

CH7 has slightly better values.
But this is most probably due to the individual noise level of the LM399.
(Has nothing to do with the slotted board).

For me it seems that measuring 0.1 .. 10 Hz noise is no good
way to judge the influence of a mounting method
on voltage reference stability.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 12, 2014, 08:05:53 pm
Noise measurements of 6 LM399H in simple one resistor configuration. Used EM Electronics nanovolt-amplifier with LF filtering (0.14-10 Hz).
LM399h No1:  6.6 uV p-p;
LM399h No4:  4.5 uV p-p;
LM399h No6:  5.3 uV p-p;
LM399h No7:  5.6 uV p-p;
LM399h No9:  6.8 uV p-p;
LM399h No10: 6.3 uV p-p.

Noise measurement of triple LM399H with constant current supply and simple filtering (see pictures above):
Pass No1: 0.88 uV p-p;
Pass No2: 0.87 uV p-p.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 12, 2014, 08:55:39 pm

Noise measurement of triple LM399H with constant current supply and simple filtering (see pictures above):
Pass No1: 0.88 uV p-p;
Pass No2: 0.87 uV p-p.

Hello,

That's a very good result: what is the band with of the filter?

Generally I have also large differences on my LM399 references:

The theoretical value on buried zeners with around 1mA is around 0.6ppm (p-p).
So for a 7V reference the theoretical value should be 4.2uVpp.

LM399#1 (LM334 temperature compensated current source) av: 3.85 uVpp s: 0.36 uVpp
LM399#2 (LM334 temperature compensated current source) av: 6.55 uVpp s: 0.35 uVpp
LM399#3 (LT1013 bootstrap current source) av: 3.56 uVpp s: 0.42uVpp

CH6 + CH7 are simple resistor configurations. (But with low noise LT1763 voltage regulator).
But since the dynamic resistance of the LM399 is around 0.5 Ohms
the difference current source / resistor has only very low influence on the 0.1 - 10 Hz noise.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iTist on January 27, 2014, 10:48:31 pm
Hello,

i am following both Voltage Reference threads the LTZ 1000 and the LM399 one.
I am new to the reference tobic only build a LT1027 based one but i want to do the next level step :-)
i think its the LM399. Inspired by the forum i have  a question about the trimming of the 10 Volt output Voltage.
Is it possible to trim the voltage with an DAC?
I have attached an simpified Schematic.


Thanks in advance.

Greetz
Oliver
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 28, 2014, 07:35:38 am
Hello Oliver,

yes this is possible.
At least the FLUKE 7000  voltage standard uses this.
Look for the brochure 10777-eng-01-a.pdf on the Fluke page.

You should use the LM399 output as reference voltage for the DAC.
And of course your feedback resistors should have a very low tempco so that the LM399 TC is not compromized.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 28, 2014, 08:04:29 am
Hello Oliver,
There are many russian voltage standards uses this too.
Title: Re: LM399 based 10 V reference
Post by: iTist on January 28, 2014, 08:15:16 pm
Hi,

thanks for the answers. I made a more detailed schematic (see attachment). Do i need a separate reference output for the DAC (REF which is a bufferd output of the LM3999) or could i use the 10V output. I thought i build a Voltage divider(switched Capacitor with an LTC1043) to feed the DAC.

The 9k and the 20K resistors need to be low TC ones. The 100 ohms and the 1k Resistor aren't so critical is this assumption right?
Sorry for my silly questions but I am just a hobbyist who is interested in Reference Voltages.
Is it better to use one 14 or 16 bit DAC or two (cheaper) 8 bit DACs. For the 8 bit DACs i wouldn't need a µC. I can trim it with simple DIP switches.

Thank you for your comments.

Greetz

Oliver
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 28, 2014, 10:06:35 pm
Do i need a separate reference output for the DAC (REF which is a bufferd output of the LM3999) or could i use the 10V output. I thought i build a Voltage divider(switched Capacitor with an LTC1043) to feed the DAC.

The 9k and the 20K resistors need to be low TC ones. The 100 ohms and the 1k Resistor aren't so critical is this assumption right?

Is it better to use one 14 or 16 bit DAC or two (cheaper) 8 bit DACs.

Of course you can use the 10V output. But what happens to your DAC if anyone tries to feed back some voltage on the output? Or what happens when RF-noise is demodulated at the protection diodes of the DAC input? So it depends on your environment.

If a resistor contributes only 1% to the output voltage the TC can be a factor 100 larger than the one of the output voltage determining resistors. All in all you will want to maintain the (around room temperature usually better as) 1ppm/K spec of the LM399.

Usually the long term stability and the TC are not specified for 8 Bit DACs. If you are only trimming a few ppm´s with the DACs and do the main trimming by other resistors in parallel to the 9K + 20K  the 8 Bit devices may be sufficient.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 02, 2014, 02:01:31 pm
Hi

Here are some pictures of my LM299 reference. I had about 50 of these LM299 in my junk box and figure with not do something with them.

http://www.simonthenerd.com/10Vref.htm (http://www.simonthenerd.com/10Vref.htm)

If I did this project today I would have changed some of the design. I would not have used an LT1013 to amplify the reference voltage to 10V. A chopper op-amp would have been better. I would probably have build the whole thing into an oven. But all these considerations have let me to do a LTZ1000A reference instead.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 02, 2014, 09:56:58 pm
Quote
I had about 50 of these LM299 in my junk box and figure with not do something with them.

I wish I'd had such a junk box too :)

About the oven, there are second hand but cheap double oven stabilized OCXOs such as MV89A out there.

Pictures of the inside: http://www.rbarrios.com/projects/MV89A/ (http://www.rbarrios.com/projects/MV89A/)

One could disassemble the XO circuit and assemble a reference circuit inside. This way the reference is shielded too and if the whole thing is soldered under inert gas condition humidity is also no deal anymore. Could be worth a try.
On the other hand it would be sad about the OCXO itself. You can't have everything :)
Title: Re: LM399 based 10 V reference
Post by: branadic on February 03, 2014, 10:08:30 pm
All the discussion, also the current one in the LTZ thread, leads me to the point:

Use a LMx99 with buffered output, feed the output into a LTC1043 based divide by 1.5V circuit to supply a pwm-dac with 30bit as shown by Andreas (based on the EDN publication) and feedback the output of the dac to the buffer of the reference buffer to realize a trimmable 10V reference output voltage.
This way the gain setting resistors can be fixed values and the complete arrangement is somewhat cheap to realize.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 04, 2014, 06:11:04 am
Hello branadic,

thats why I´m focused on the LM399 at the moment.
I am ageing 16 LM399s to select the best one.
About half of them is already below 1ppm/khr.

First step then is a ADC with LM399 reference.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on February 04, 2014, 08:49:34 pm
This is interesting to know, maybe you want to tell more about what you are planing to do (circuit desciption)?

Why LM399 and not LM199AH/883 with higher temperature specs?

http://www.ebay.de/itm/NS-ST-LM199AH-883-CAN4-Voltage-Reference-/221365837650?pt=Bauteile&hash=item338a6e9f52 (http://www.ebay.de/itm/NS-ST-LM199AH-883-CAN4-Voltage-Reference-/221365837650?pt=Bauteile&hash=item338a6e9f52)
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 04, 2014, 10:08:49 pm
This is interesting to know, maybe you want to tell more about what you are planing to do (circuit desciption)?

Why LM399 and not LM199AH/883 with higher temperature specs?


Up to now:
- LM399 as reference (I simply have them already)
- LTC1043 as divider (*3/4 = 5.2 V reference)
- LTC2400 as ADC input range 10V with 2nd LTC1043 in 2:1 configuration
- processor PIC24FV32KA302 (but im running out of pins so this is subject to change)
- 2*16 digit LC-display    6,5 digits resolution (7,5 digits in slow mode with filtering)
- isolated RS232 interface
- rotary encoder with switch for adjustments
- battery powered up to 48 hrs with 12 NiMH cells
- plastic case with clear lid for display
- optional inner shielding

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on February 05, 2014, 05:41:22 am
Now I'm working on the similar project:
- 3x LM399H as reference. I have a 10-th of them (well aged). 3 have a long-term drift less than 1 ppm/year (~0.3 ppm/sqrt(kHr) )
- ADS1282 in oven
- Input amplifier (like in Sol. 7081) with bootstrap power source
- Compensated metal-foil divider (1:5)
- Plastic case with atmega, LCD e.t.c.

Title: Re: LM399 based 10 V reference
Post by: Andreas on February 05, 2014, 07:03:07 am
Ok,

this will give the high end variant of my design  :-+

Is that a metal plate on the right side within the housing for shielding?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on February 05, 2014, 07:31:50 am
Yes, the metal plate on the right side is guard screen.
Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 05, 2014, 03:40:34 pm
Looks like a nice project Mickle T.

- Simon
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on February 05, 2014, 05:04:51 pm
- LTC1043 as divider (*3/4 = 5.2 V reference)
Is that one or two 1043s? If you got 3/4 with just one IC, I have a puzzle to solve :)  If two, how did you opt to join the *3 and /4 stages? 

I tried to get 10/7 working in LTSpice, but I couldn't get the sim to be stable. I needed 5 ICs and some of them had to switch at the same time; it seemed like a mess to implement in reality.  It had the nice side effect of providing 1, 2, 5, 7, and 10V references, though.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 05, 2014, 07:40:27 pm
Quote
Is that one or two 1043s? If you got 3/4 with just one IC, I have a puzzle to solve :)  If two, how did you opt to join the *3 and /4 stages? 

One 1043, if you stuck have a look on the solution here:

https://www.eevblog.com/forum/projects/building-a-7-decade-voltage-calibrator/msg304819/#msg304819 (https://www.eevblog.com/forum/projects/building-a-7-decade-voltage-calibrator/msg304819/#msg304819)
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 05, 2014, 07:59:42 pm

Is that one or two 1043s? If you got 3/4 with just one IC, I have a puzzle to solve :)  If two, how did you opt to join the *3 and /4 stages? 

I tried to get 10/7 working in LTSpice, but I couldn't get the sim to be stable.

Damned: I should have patended my single LTC1043 *3/4 solution.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on February 08, 2014, 12:09:33 pm
Quote
Damned: I should have patended my single LTC1043 *3/4 solution

Maybe you should have.  :)

I'm nearly finished with the schematics for the adjustable LMx99 reference. You could also say it's some kind of calibrator or transfer standard.

- LMx99 (6.75V ... 6.95V ... 7.3V) is tapped and divided by 3/4 (5.0625V ... 5.2125 ... 5.475V) using LTC1043
- the second tap goes into a LTC2057 booster with gain 1.25 (Z201 5k, 20k)
- the voltage of the divider is feed into the 30bit pwm-dac (modifications given by Andreas)
- the output of the pwm-dac is feed through the 20k resistor into the LTC2057 booster

This gives a digital adjustable reference using standard of the shelf Z201 resistor values.
You could use a DAC such as AD5791 instead, but this is much to expensive.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 08, 2014, 03:10:37 pm
Hello branadic,

Have I understood it right?

you have created a cirquit which is adjustable from around 8.2 to (8.6 + 5.2) V

I thought myself of a similar cirquit.
But with a single resistor divider.
The MAX4053A can be used up to a supply of +/- 8V which is sufficient for a 7V reference. (the LTC1152 with >= +/- 10V)
The output from 0..7V could be fed into a *1.5 amplifier (I thought of 1 or 2 DSMZ resistor dividers) giving 0 .. 10.5V

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on February 08, 2014, 05:03:35 pm
Hi Andreas,

I have drawn a circuit with LMx99 output voltage multiplied by 1.25 and the possibility to adjust that voltage with the pwm-dac to 10.xxxxxxV. Up to know it is just my documentation of an idea and not yet perfect.

branadic
Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 15, 2014, 09:29:52 pm
Hi

In the datasheet for the LM399 (LM199, LM299) there is an application called "Portable Calibrator" (Figure 22 in the datasheet from TI).

http://www.ti.com/lit/ds/symlink/lm399.pdf (http://www.ti.com/lit/ds/symlink/lm399.pdf)

The circuit uses a 200k resistor in series with the reference zener diode. Does anyone know why they used 200k which would set the zener current between 25 uA and 55 uA. In the datasheet they say that the operating current is between 500 uA to 10 mA. Why do they then set the current in their application for a much lower value?

- Simon
Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 15, 2014, 09:33:59 pm
Ah...well it may be the combined current for the reference and the heater...probably.

- Simon
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on February 15, 2014, 09:51:37 pm
The circuit uses a 200k resistor in series with the reference zener diode. Does anyone know why they used 200k which would set the zener current between 25 uA and 55 uA.
That 200k isn't setting the zener current; the opamp output is driving the zener through a 5k resistor.
Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 15, 2014, 09:56:24 pm
Ahh, thanks for the input. :)
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 15, 2014, 10:41:25 pm
Hello,

the 200K is only for start up in case the op-amp has a negative input offset voltage.
I also use a similar resistor in my LTZ1000A circuit.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 16, 2014, 07:26:49 am
Hi

Okay...so you did not use the "standard" application note schematic in the LT1000(A) datasheet. That sounds interesting since most people stick to that circuit. :) Cool.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 16, 2014, 10:14:09 am
Hello,

you will find this pull-up resistor in the HP3458A reference board circuit.
Also the Datron reference circuit uses a pull up.
The LM399 too.
So probably in the datasheet of the LTZ they simply forgot to add one.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: SeanB on February 16, 2014, 10:18:18 am
Or in the LT test unit the opamp had a positive offset and was thus self starting. Get one that is not so and you will have it sitting stably at zero a lot of the time until noise turns it on.
Title: Re: LM399 based 10 V reference
Post by: SimonSatCom on February 16, 2014, 12:11:27 pm
Andreas are you referring to this board?

http://www.maxmcarter.com/vref/ (http://www.maxmcarter.com/vref/)

And btw...does anyone have any pictures or schematics of the high precision version of the HP3458A reference board?

- Simon
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 16, 2014, 06:32:25 pm
Yes, thats it.

I do not believe that the high precision version differs from a standard version.
Except that the pre-ageing of the reference at the facility has been either done for a longer time
or by selecting the better boards.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 09, 2014, 08:54:36 pm
Hello,

now after about 1500 hours the LM399 CH6 + CH7 seem to stabilize.
So I plan to repeat the tilt measurements.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on March 09, 2014, 09:23:51 pm
What a bummer, Dave would say. LM399 CH7 is drifting incredibly and hasn't stopped yet even if it seems to settle to smaller drift rates.
Is it just a variance or is there a correlation to the lead lengths? Your guess is as good as mine, but would be interesting to investigate :)
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 09, 2014, 09:40:15 pm
Just to clarify:

lead length is the same (short leads) for CH6 + CH7.

CH7 uses a slotted board area.
CH6 uses non slotted board.

But I do not think that this has directly something to do with ageing drift.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: babysitter on March 10, 2014, 08:14:25 am
Let me take the role of the "Bild" Newspaper:

SLOTS IN PCB INCREASE AGING DRIFT
Effect discovered by amateur scientist contradicts prior scientific work considering slots to be advantageous. We all gonna die!


Title: Re: LM399 based 10 V reference
Post by: branadic on March 14, 2014, 07:51:50 am
Quote
Let me take the role of the "Bild" Newspaper:

SLOTS IN PCB INCREASE AGING DRIFT
Effect discovered by amateur scientist contradicts prior scientific work considering slots to be advantageous. We all gonna die!

 >:D Don't, there are people out there that could believe in that, see below. One exemplar is not of statistical relevance, thats for sure.

Quote
So, it seems that the slots are counter productive in this use case.  Maybe, it would be better to let your LM399(s) be on a separate PCB from the rest of the circuitry.  That would make the whole LM399 assembly easier to insulate.

No, there is no significant number of examples that proof such an argument. It was just a loud thought of mine, but this larger drift of one exemplar could be nearly everthing and is not constrained by an effect of slots in the board. Maybe I should delete my thought in my previous post? Damn.  :palm:
Title: Re: LM399 based 10 V reference
Post by: macfly on April 06, 2014, 10:29:22 am
Has anyone an information about the temperature stability of the thermal regulator inside the LM399 ?
This is  a possible question to Bob Dobkin ...

Regards,

macfly
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 06, 2014, 07:34:53 pm
Hello macfly,

not so good as in the LTZ1000.

you can see this in my tilting experiment.
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg360779/#msg360779 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg360779/#msg360779)

Further the stability is dependant on heater supply voltage.

You can get good results with stabilized heater voltage and good thermal management.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: macfly on April 06, 2014, 08:20:24 pm
Hello macfly,

not so good as in the LTZ1000.

you can see this in my tilting experiment.
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg360779/#msg360779 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg360779/#msg360779)

Further the stability is dependant on heater supply voltage.

You can get good results with stabilized heater voltage and good thermal management.

With best regards

Andreas

Hello Andreas,

very interesting, but not really an answer of my question. But I asked with some thought's unspoken.
Yesterday, I tested my 5 pieces of LM399H in a manner, I have not read anywhere:

1. Supply the zener with ~1mA  - but do NOT supply the heater.
2. Measure the output voltage
3. Put the heater on, wait for thermal balance.
4. Then re-adjust the zener current, to get rid of the zener's dynamic resistance.
5. Measure ouput voltage again.

With these two measurements you may approximately calculate the temperature coeffcient of the device.
If we assume ~90°C working temperature with a stability of +/-0.1 degrees Celsius, the best of my LM399
has a temperature coefficient of only ~0.06ppm/°C.
Now you understand my first question about the stability of the regulation?
I was very astonished about these result.   

Regards,

macfly
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 06, 2014, 09:34:32 pm

Now you understand my first question about the stability of the regulation?


Hello macfly

Not really. (how are the exact measurement conditions and what do you mean when speaking of temperature coefficient).

The unheated reference should be in the around 50ppm/K range. (Similar to the LTZ1000).
The heated reference near room temperature is much better.

On my LM399#2 i measured 8uV maximum voltage difference (including hysteresis + noise) in a range of 8-46 degrees celsius environment temperature (measured outside the card box). Note: the voltage (right y-axis) is measured as difference in mV between 2 LM399 references. One at room temperature the other within a "thermal chamber".

So the temperature "coefficient" according to "box method" would be 0.008mV / 6860 mV / (38 K) = +/- 0.015ppm/K.
The temperature "slope" is around -0.07ppm/K as best fit straight line.

But most of the measured error is probably due to thermocouples at the solder joints.
Since I get similar results with a LTZ1000 based reference.

With best regards

Andreas





Title: Re: LM399 based 10 V reference
Post by: macfly on April 07, 2014, 05:48:23 pm

Now you understand my first question about the stability of the regulation?


Hello macfly

Not really. (how are the exact measurement conditions and what do you mean when speaking of temperature coefficient).

The unheated reference should be in the around 50ppm/K range. (Similar to the LTZ1000).
The heated reference near room temperature is much better.



Hello Andreas,

attached you will find my measurement results. What me surprised is, that the tempco (calculated out of the Vz @ 20°C and Vz @ 90°C)
is more then ten times lower than the 50 ppm/K. With my guess of a regulator stability of 0,1K, one of the LM399 has nearly the same tempco
as a LTZ1000 !

Regards,

macfly

Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 07, 2014, 07:33:09 pm
I have a very different results of the LM399 TempCo measurements.

Title: Re: LM399 based 10 V reference
Post by: Andreas on April 07, 2014, 08:07:32 pm
Hello macfly,

the results are too good to believe.
That brings me to the idea to select the references according to raw tempco.

I have looked up my measurements: unfortunately I did not document the values carefully.
But the difference of 2 of my references is at least 4-5mV unheated / heated.

@macfly + Mickle:
Did you use National Semiconductor parts or LT parts?
On my side all are National Semiconductor parts up to now.

@macfly:
Did you kelvin sense of the zener or is the heater current affecting the zener voltage?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Vgkid on April 07, 2014, 09:33:03 pm
It seems that it will be an even better reason for me to follow this thread, my 6.5 digit meter had a 200uv offset at 10 v.
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 08, 2014, 04:48:51 am
Parts from National Semiconductor with the same date code, kelvin connection.
Title: Re: LM399 based 10 V reference
Post by: Vgkid on April 08, 2014, 06:04:44 am
What are the differences in the H vs. AH specced parts?
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 08, 2014, 06:40:39 am

That brings me to the idea to select the references according to raw tempco.


Hello,

when thinking twice the better idea would be to test the tempco without heater around 90 degrees.
There is a large possibility that the tempco is not linear but either a parabolic or s-shaped curve (due to compensation of the zener with a Vbe).
So it can be that at 25 degrees and 90 degrees the voltage is accidently the same but in worst case with a large drift around 90 degrees.

It seems that it will be an even better reason for me to follow this thread, my 6.5 digit meter had a 200uv offset at 10 v.

200 uV = 20 ppm is below the 35 ppm/year as in the specification of a good 6.5 digit meter. (Which one is it?)
If it includes the first year drift I would not bother about.
Later the drift should go down to 1-2 ppm / year for a good instrument.

What are the differences in the H vs. AH specced parts?
The A-type is selected for a smaller (overall) tempco and smaller initial tolerance (National Semiconductor only). See datasheet.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Vgkid on April 08, 2014, 03:53:55 pm
I should have said that there was a 200uv offset between the +\- readings on the 10 v scale, I had to use a 9v battery to get a stable enough reading.
My meter is a Schlumberger Solartron 7065, mine is about 33 years old.
Title: Re: LM399 based 10 V reference
Post by: macfly on April 08, 2014, 04:48:14 pm
Hello macfly,

the results are too good to believe.
That brings me to the idea to select the references according to raw tempco.

I have looked up my measurements: unfortunately I did not document the values carefully.
But the difference of 2 of my references is at least 4-5mV unheated / heated.

@macfly + Mickle:
Did you use National Semiconductor parts or LT parts?
On my side all are National Semiconductor parts up to now.

@macfly:
Did you kelvin sense of the zener or is the heater current affecting the zener voltage?

With best regards

Andreas

Hello Andreas,

1. ... That brings me to the idea to select the references according to raw tempco -> No - this was my idea  8)
2. ... the results are too good to believe -> yes, they are too good - and  that was the reason for my post.
3. ... Did you use National Semiconductor parts or LT parts? -> I used NS parts.
4. ... Did you kelvin sense of the zener or is the heater current affecting the zener voltage? -> For sure.

At weekend I will repeat the measurement.

@Mickle
Thanks for your measurement results - they are just like that, what I expected.

Regards,
macfly
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 08, 2014, 09:16:13 pm
Hello,

I have now modified my LM399 ageing board with 2 jumpers to switch off the heaters.
There are 16 LM399 from National Semiconductors with 2 different datecodes (12 + 4)
Room temperature was 23 deg C so with around 2 K self heating of the 7mW zener I guess 25 deg for the cold value.

The measured values cold + hot are as follows:

With best regards

Andreas



Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on April 09, 2014, 03:19:55 am
Interesting that the lower TC modules tend to be < 6.9V.  Makes me wonder if the others would have a lower temp Co at a lower current.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 09, 2014, 04:32:21 am
Interesting that the lower TC modules tend to be < 6.9V.  Makes me wonder if the others would have a lower temp Co at a lower current.

Hello,

very good eyes !!!

so the "ideal" voltage on my references seems to be around 6875 mV.

In the ageing cirquit I use simple 6K8 Metal film resistors (#4-11) or 10K + 22K in parallel (#12-19)
along with a 15V low drop regulator.

with best regards
Title: Re: LM399 based 10 V reference
Post by: macfly on April 09, 2014, 07:21:15 am
Hello Andreas,

thanks for yur measurement results. This nearly confirms my measurements.
And as Galaxyrise observed, my LM399 with the lowest TC has the lowest Vz.

Next step to do is changing the zener current and looking what happens ....

Regards,

macfly
 
Title: Re: LM399 based 10 V reference
Post by: branadic on April 09, 2014, 06:38:55 pm
Somewhat off topic but interesting for all of us. I'm currently looking for precision resistors and here is what I found up to now:

Distributor   Resistor type   Value   Toleranz   TC   Stability   Price
HBE-Shop   Y14535K00000V9L   5K   0,005%   ± 0,05 ppm/°C   to ± 0.005 % at 70 °C, 2000 h at rated power   34,94 €
HBE-Shop   Y145320K0000V9L   20k   0,005%   ± 0,05 ppm/°C   to ± 0.005 % at 70 °C, 2000 h at rated power   37,81 €
HBE-Shop   Y00625K00000T0L   5K   0,010%   ± 1 ppm/°C   ± 0.005 % at 70 °C, 2000 h at rated power   18,44 €
HBE-Shop   Y006220K0000T9L   20k   0,010%   ± 1 ppm/°C   ± 0.005 % at 70 °C, 2000 h at rated power   not in stock
HBE-Shop   Y07855K00000T9L   5K   0,010%   ± 2 ppm/°C    to ± 0.005 % at 70 °C, 2000 h at rated power   20,53 €
HBE-Shop   Y078520K0000T9L   20k   0,010%   ± 2 ppm/°C    to ± 0.005 % at 70 °C, 2000 h at rated power   21,14 €
                  
Burster   1142   5k   0,010%   ± 2 ppm/°C   0.01% over years   23,13 €
Burster   1142   20k   0,010%   ± 2 ppm/°C   0.01% over years   23,13 €
                  
Rohpoint   8G16D 5K   5k   0,100%   ± 3 ppm/°C   ±35ppm/10,000hours ±50ppm/26,000 hours   4,52 €
Rohpoint   8G16D 20K   20k   0,100%   ± 3 ppm/°C   ±35ppm/10,000hours ±50ppm/26,000 hours   5,30 €
Rohpoint   8G16A 5K   5k   0,010%   ± 3 ppm/°C   ±35ppm/10,000hours ±50ppm/26,000 hours   6,80 €
Rohpoint   8G16A 20K   20k   0,010%   ± 3 ppm/°C   ±35ppm/10,000hours ±50ppm/26,000 hours   7,95 €


The hermetically Vishay types seem to be unobtainium and have long waiting time. Most of the Vishay parts seem to be not in normal stock, a simple of the shelf (these are my words because its standard datasheet value and not custom made) DSMZ divider needs at least 3 months to arrive. Are you guys at Vishay kidding me? For a proof of concept this is much to long and inadmissible. The price is for sure fair:

1-24 pc.   25-49 pc.   50-99 pc.
21,05 €   15,30 €   12,40 €

Someone with other and better resisitor sources out there?
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 09, 2014, 07:05:05 pm
Hello Branadic,

The Vishay parts are usually with around 12 weeks delivery time. (from Israel).
They are manufactured on demand only. (with except of a handfull of stock parts).
If you ask Vishay Germany for parts in sample quantities you are usually redirected to Powertron / Teltow.
They have also own precision resistors.
Of course you can have quicker turnaround times but at a extra cost.

At your prices I guess that HBE-Shop (a FARNELL reseller) has prices including VAT. (and the others are adding VAT).

SMD-Parts (VSMP) are available at DigiKey. Trimming to the desired resistance is included.
http://www.digikey.de/product-search/de/resistors/precision-trimmed-resistors/66806?k=vishay%20vsmp (http://www.digikey.de/product-search/de/resistors/precision-trimmed-resistors/66806?k=vishay%20vsmp)

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on April 09, 2014, 09:02:53 pm
What is ignored are the longterm / load life stabilities:

Vishay VSMP: to ± 0.03 % at 70 °C, 2000 h at rated power
Vishay Y1453: to ± 0.005 % at 70 °C, 2000 h at rated power
Vishay Y0062: ± 0.005 % at 70 °C, 2000 h at rated power
Vishay Y0785: to ± 0.005 % at 70 °C, 2000 h at rated power
Rohpoint 8G16: ±35ppm/10,000hours ±50ppm/26,000 hours
Burster 1142: 0.01% over years (whatever that finally means)

That shows that VSMP is a comparable bad choice and 8G16 seem to be best, beside the "higher" TC.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 10, 2014, 09:15:44 pm
Hello branadic,

in my opinion the shelf life stability is the more interesting spec.
(rarely found in the data sheet).
Usually you would not use a precision resistor near the rated power.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on April 11, 2014, 12:15:30 am
...just sticking my head in to get updates on the thread....
Title: Re: LM399 based 10 V reference
Post by: branadic on April 13, 2014, 06:32:49 am
Quote
in my opinion the shelf life stability is the more interesting spec.
(rarely found in the data sheet).
Usually you would not use a precision resistor near the rated power.

I agree in both points, but unless such values are printed in datasheet you need to compare the one that are. And the load life stability is a good parameter to compare them and better than nothing, right?
I would assume that a resistor with ±35ppm/10,000hours ±50ppm/26,000 hours is more stable compared to the one with ± 0.005 % at 70 °C, 2000 h at rated power.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 22, 2014, 08:55:26 am

I have done the correction calculation for a linear ageing (linear regression coefficient).

The correction is applied on the averaged measurement values and not for the original 1 minute values.

The spread values with the foam do not really change:
CH6 foam corrected:   4.4 uV = 1.3 ppm (unchanged)
CH7 foam corrected: 10.4 uV = 3.0 ppm (was 10.1 uV)

The spread values for the "cotton" isolation seems even to be better for the slotted PCB:
CH6 cotton corrected:  4.6uV = 1.4 ppm (was 4.1 uV)
CH7 cotton corrected:  2.4uV = 0.7 ppm (was 3.6 uV)

remember that standard deviation for my setup is around 0.25 ppm or 1 uV.


I now have repeated the tilt measurements on the well isolated (cotton) LM399s:

Again the slotted PCB (CH7) seems to be better.
CH6 cotton (aged):  4.1uV = 1.2 ppm   (zener voltage = 2 * 3428 = 6856 mV)
CH7 cotton (aged):  1.9uV = 0.5 ppm   (zener voltage = 2 * 3452 = 6904 mV)

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on April 22, 2014, 02:23:56 pm
What are the LM399's driving currents?
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 22, 2014, 03:17:10 pm
Hello,

simple 7K5 pull up resistors (metal film) on battery supplied, stabilized supply of 14.08V for CH6 + CH7
so around 1mA.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: turbo! on April 22, 2014, 05:52:07 pm
This voltage reference is used in a precision balance I tore down just the other day. The resistor that attaches to it is a 15.00k Ohm, 0.5%, 50ppm/K

Pay close attention to the ground plane "tongue" dips under the resistors. The ordinary 10 ohm resistor right next to it feeds the heating element for the can.

Perhaps the  10 ohm resistor is meant to be a heating element as well to warm up the trace beneath it so the entire cluster remains at the same temperature? The best way to avoid EMF voltage generation aside from using exotic matching is to have the affected components all be at the same temperature.

The six banded resistor then connects to ground through the shunt resistors on stand-offs which are 0.1%. 

The permissible voltage variation span is fairly wide. 6.8 to 7.1v. For this scale, it doesn't really matter what the exact voltage is, because the offset adjustment is made with a calibration. What matters is that it STAYS where it is, whatever that value may be.
(http://i.imgur.com/DkssRWb.jpg)
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on April 22, 2014, 06:06:21 pm
@turbo

R43 and R44?
Title: Re: LM399 based 10 V reference
Post by: turbo! on April 22, 2014, 06:18:39 pm
@turbo

R43 and R44?

43 feeds the heating element. I don't see why its needed, but I think the more important function is to act as a heater itself to heat the copper trace. 
44 is 15kohm 0.5% 50ppm, then it connects to the other shunt. The big diode looking thing on stand offs is a Vishay 0.1% resistor. So, I think they're connected as a divider. The precision scale is functioning, so I'm not going to do too much autopsying. This is a 1980s design by the way.

Interestingly, you'll see the stand-offs for the big shunt uses intentionally thinned out pattern to connect to ground. I think this is to minimize thermal transfer from the trace.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 22, 2014, 06:58:31 pm
but I think the more important function is to act as a heater itself to heat the copper trace. 

Interestingly, you'll see the stand-offs for the big shunt uses intentionally thinned out pattern to connect to ground. I think this is to minimize thermal transfer from the trace.

Hello,

perhaps only to limit the inrush current at start up (see datasheet).
In steady state the 20mA * 10 Ohms = 0.2V  -> 4mW will not produce significant heat. (perhaps 2-4 degrees C)

Interestingly the 10 ohms will not significantly dampen down inrush current.
But the output voltage of LM399 has a relative large dependancy on heater voltage/current.
So perhaps the 10 Ohms might be a compensation for this.

Do you know how large the heater supply voltage is? is it +15/0V or +15/-15V?

The thinned out pattern is mainly due to wave soldering.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on April 22, 2014, 07:07:00 pm

But the output voltage of LM399 has a relative large dependancy on heater voltage/current.



I've heard this before. Is there a mechanism that causes this?
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 22, 2014, 07:18:02 pm
I've heard this before.

Perhaps in one of my threads.
After finding out that the LM399 is sensitive to tilting, I decided to measure all effects that might have an influence.
The PSRR of the heater voltage was one of the measurements not mentioned in the datasheet.

For the reason I can only guess: perhaps imperfections in the temperature governor to increase stability?
Since there is no capacitor on the chip it can only be a p-governor.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 22, 2014, 07:52:37 pm
Some words about tilting.
One of my simple statistical voltage references with the Dewar flask isolation has no orientation effects at all, i.e. 0.0 ppm in any direction. LM399 heater current < 5 mA.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 22, 2014, 08:32:59 pm
Hello Mickle,

you are always good for a surprise.  :-+

Your idea with the dewar is thermal isolation in perfection.
And 5 mA is really a very low value. Is it for 15 or 30V heater supply?

Is it possible to get more detail photos from the isolation construction?

What is the metal block on the photo? is it containing the voltage divider resistors? thermally stabilized?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: casinada on April 22, 2014, 09:27:46 pm
Mickle,
Are the banana connectors stolen from old HP equipment? :)
Title: Re: LM399 based 10 V reference
Post by: turbo! on April 23, 2014, 12:41:08 am
(http://i.imgur.com/DkssRWb.jpg)

For those who asked about the circuit above used in an electronic balance.
It takes about 4-5 seconds to settle down to 6.863v.

It runs off of 24v DC from a linear power supply. The 10 ohm resistor feeds the heater and the 15k ohm resistor carries the zener current.

The heater current starts around 12mA and stabilizes to 10mA, so the total resistance is around 2.4k ohm. It's dissipating close to 1/4W which is more than enough to become noticeably warm.
The 15k ohm resistor had about 17v, and with 7v regulation, it adds up right. This comes out to only a tad over one mA of reverse current through zener. The surface temperature runs around 45°C after about half an hour with the cover on.
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on April 23, 2014, 03:44:24 am
I've heard this before.

Perhaps in one of my threads.



Probably, I've read it several places. Also I think DiligentMinds said Bob Dobkins (the designer) said it too during his interview.
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 23, 2014, 06:05:38 am
Hello Andreas,

Stabilized 15V heater supply.
Aluminum block is filled with paraffin and contain a PTFE board with 14 TC-selected and aged hermetic foil resistors for stat. divider. Dividers tempco is ~ 0.04 ppm/C, so there is no thermal stabilization.
Isolation construction is very simple: only Dewar and huge foamed plug in it  :)

Mickle T.
Title: Re: LM399 based 10 V reference
Post by: acbern on April 23, 2014, 07:21:56 am
can you comment on the paraffin. is this for moisture? your resistors are already hermetic. or any other (mechanical/thermal) reason?
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 23, 2014, 08:05:16 am
Paraffin is only for temperature equalization. S5-61 foil resistors have a significant TC difference.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 23, 2014, 08:31:59 pm
Hello Mickle,

on your cirquit diagram on volt nuts I have seen that you also used 10 Ohms resistors
for your 3*LM399 cirquit in series with the heater (on a 14.5V supply).
Now we have the weight scale also with the 10 Ohms (but with a 24V supply).

Is there a special reason for the 10 Ohms resistor in the heater line of the LM399?

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 24, 2014, 05:02:55 am
Hello Andreas,

I think this is an accidental coincidence. There is no special reason for the 10 Ohms resistor.

Mickle T.
Title: Re: LM399 based 10 V reference
Post by: macfly on April 24, 2014, 10:44:03 am
Paraffin is only for temperature equalization. S5-61 foil resistors have a significant TC difference.

Hello Mickle,

very smart, your design!
What are these 'S5-61' resistors?
Where is the manufacturer and what specs they have (TC, long-time stability)?

Regards,

Macfly
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on April 24, 2014, 02:08:17 pm
Hello Macfly,

S5-61 is one of the old russian hermetic foil resistors, equivalent to the Vishay. I don't have full specs.

Mickle T.
Title: Re: LM399 based 10 V reference
Post by: bingo600 on April 30, 2014, 08:00:26 pm
Hi guyzz

I have bought two of Branadic's LM399 boards. 
Schematic is the same as in the LM399-DS : Buffered Reference With Single Supply - but with a LT1001 in Cerdip.
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg219527/#msg219527 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg219527/#msg219527)

And have also gotten the "econistors" from Rhopoint (Outchhh $$)
I couldn't get the 7.5K for the heater , but have some 6.49K 5-ppm resistors.

Can i use one of those for the heater ?

I was thinking of using a 7815 for the psu , a led to get the 7815 in regulation , a diode for reverse polarity protection , and feeding it 18v. Maybe from my HP-E3610A.
Is a LM317 a better choice (I have the 7815 in the drawer).
 

I'm an analog noob , so go easy on me

TIA
Bingo
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 30, 2014, 08:56:43 pm

I couldn't get the 7.5K for the heater , but have some 6.49K 5-ppm resistors.

Can i use one of those for the heater ?


Hello,

the heater of the LM399 does not need a 7.5K resistor.
I think you speak of the zener pull up resistor.
(the blue resistor on branadics photo).

In this case a 6.5K will do its job.
You will have slightly more than 1mA at the zener.
-> ageing drift increases slightly.

For the 15V supply you should chose one with low tempco and low ageing drift
to minimize changes on output voltage.

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 01, 2014, 06:11:05 am
7815/7915 works very well. I didn't find any measurable impact on the reference output.
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 01, 2014, 06:45:09 am
VRE310J have a large TC (even typical 0.6 ppm/c is too large). Low-frequency noise - no comments, see picture  :)
Resistrors - selected russian metal-foil S5-61, 0.1 ppm/C.
Title: Re: LM399 based 10 V reference
Post by: Vgkid on May 01, 2014, 07:09:09 am
That is an impressive increase in stability.
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 01, 2014, 07:55:03 am
Side note: 0.1 ppm/C is only after selection. All resistors in my reference have a 5 ppm/C class (max tempco). Typical is 10 times better. By selection is possible to achieve 0.1 ppm/C (20-50 C range).
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 01, 2014, 03:20:00 pm

In this case a 6.5K will do its job.
You will have slightly more than 1mA at the zener.
-> ageing drift increases slightly.
Andreas

Thanx Andreas & Mickle

I'll use 6.49k for the Zener pullup , a 7815 and some Red Wima's

/Bingo
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 10, 2014, 07:34:56 pm
Hello,

the 10 Ohms resistor in series to the heater of the LM399 in some designs did not get out of my mind.
I decided to make some measurements on PSRR (power supply rejection ratio)
with different resistors in series to the heater.

I choosed  0 R, 10 R, 22 R and 47R.

From my LM399 ageing box I decided to choose LM399#13 (see April 9.th) it is the one which has the largest TC between heater off and heater on. Additionally I will do the measurement with LM399#14 which is the one with the lowest TC from heater off/on measurement.

Measurement conditions:
Zener in parralel with 100nF is supplied from battery with LT1763 stabilisation to 14.08V via a 6K8 1% metal film resistor.
Kelvin sensing of zener voltage with 2:1 precision divider and measured with 24bit ADC (LTC2400).
Heater is supplied from programmable power supply and fed via the chosen resistor to the heater.
47uF + 100nF additional in parallel to the heater (avoid oscillations due to series resistor).
Start was with 12V increments 0.25V to 30V. Then back to 9V and finally increasing to 12V again.

X-axis is heater voltage
left Y-axis is 2:1 divided zener voltage in mV (red dotted line)
right Y-axis is local PSRR gradient in ppm/V (green dots)
(But due to measurement noise the PSRR gradient is nearly not usable)

Here the results for LM399#13:

Amazingly this device cannot read the datasheet  :(
The heater should work downto 9V if used without series resistor.
But below 9.1V there is a large change in zener output voltage.
Of course with additional series resistor the minimum working voltage has to be increased.
I decided to "zoom" and cut off the "odd" measurement values.

Evaluation between 12 and 30V

  0R: 34uV   (17uV * 2)
10R: 37uV   ignoring the "event" due to bad contact
22R: 38uV
47R: 37uV

So there is no enhancement in PSRR for the heater on the zener output with a series resistor.
Best seems to be when using no series resistor.
Since PSRR decreases on higher voltages the heater voltage should be choosen as high as possible.

with best regards

Andreas

Edit: continued on:
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg443181/#msg443181 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg443181/#msg443181)
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 10, 2014, 08:08:38 pm
I did the first LM299 ref.
Thanx for letting me buy the boards Branadic  :-+


See schematic in post below , this one has an error

Had to put a 200ohm trimpot in series with the 9k feedback , now i can hit 10.00000v  - using approx 9k140 ohm

I have added a RC , and a feedback cap to Branadic's version , my EE Guru said it was a good idea.

/Bingo
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 10, 2014, 09:40:15 pm
my EE Guru said it was a good idea.

The standard cirquit would tap R3 on the other side of R5.
So voltage loss due to R5 is canceled out.

Edit: see http://www.ti.com/lit/ug/tidu032a/tidu032a.pdf (http://www.ti.com/lit/ug/tidu032a/tidu032a.pdf) figure 17

Usually R5 can be choosen around 100R or even somewhat below
without loosing stability with large capacitive loads. (C2 necessary).

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 10, 2014, 10:05:47 pm
my EE Guru said it was a good idea.

The standard cirquit would tap R3 on the other side of R5.
So voltage loss due to R5 is canceled out.

Usually R5 can be choosen around 100R or even somewhat below
without loosing stability with large capacitive loads. (C2 necessary).

With best regards

Andreas

You're absolutely correct Andreas.

My EE has just corrected me with exact same info.
I have updated the schematic.

Edit: See newer schematic below

Thanx to all you electronics (volt) Gurus.

/Bingo
Title: Re: LM399 based 10 V reference
Post by: Hideki on May 10, 2014, 10:15:25 pm
If you remove all the small green junctions where a single wire simply connects to a component, it would make me very happy :)

Absolutely keep them for three-way junctions. If you are drawing this using "Wire" in Eagle... don't! Use "Net" instead, and the junctions will for the most part automatically show up when they are needed.
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 10, 2014, 11:09:19 pm
If you remove all the small green junctions where a single wire simply connects to a component, it would make me very happy :)

Absolutely keep them for three-way junctions. If you are drawing this using "Wire" in Eagle... don't! Use "Net" instead, and the junctions will for the most part automatically show up when they are needed.

Thanx for the NET tip , will use that for new schematics

My EE suggested to rearrange the trimpot , should create better tempco i think.

(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=93182)

/Bingo
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 11, 2014, 05:37:17 am
My version of the quick-and-easy LM399 voltage reference  :)
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 11, 2014, 07:34:22 am
Hello Mickle,

just one question.
From where can I get those tiny dewar cylinders.
Do you have a source with reasonable pricing?

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: macfly on May 11, 2014, 08:29:15 am
Side note: 0.1 ppm/C is only after selection. All resistors in my reference have a 5 ppm/C class (max tempco). Typical is 10 times better. By selection is possible to achieve 0.1 ppm/C (20-50 C range).

Hello Mickle,

would you let the cat out of the bag, where we can buy S5-61 foil resistors?

Regards,

Macfly
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 11, 2014, 08:51:57 am

Hello Mickle,

would you let the cat out of the bag, where we can buy S5-61 foil resistors?


Hello,

You could also order some VHP100/VHP103 or from the HZ series resistors from vishay and select for TC.
Most probably the S5-61 resistors are no longer manufactured. Otherwise you could google after them.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 11, 2014, 08:55:59 am
My version of the quick-and-easy LM399 voltage reference  :)

Mickle ... Easy  :phew:
Is that transistor arrangement for controlling the Zener from the 10v , meaning super stable ?
Why the transistors , to avoid loading of the Vref Out ?

Edit: Beautifull finish  :-+
        Is there an "outside screw" that connects to the trimpot on the 10ppm adjust ?

/Bingo

Ps: I just modifyed my LM299 to implement the changes that are in the last schematic.

Here is a graph from the night (before the R3/C5 mod , and the trimpot mod)
Blue (K2015) is the LM299 , strange that the LM299 (Board and/or trimpot) seems to have a negative tempco.

Well see what the next day's graphs reveals

/Bingo
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 11, 2014, 09:09:43 am

Mickle ... Easy
Is that transistor arrangement for controlling the Zener from the 10v , meaning super stable ?
Why the transistors , to avoid loading of the Vref Out ?


Hello,

the transistors are just a standard cirquit to give a short circuit proof output with at least 10-15mA current.
In case you want to attach a thermal rms converter for precise AC calibrations.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 11, 2014, 09:49:04 am
From where can I get those tiny dewar cylinders.
Do you have a source with reasonable pricing?
Hello Andreas,

These tiny dewars was taken from "Giatsint" TCXO (~10$), but I don't know is it currently available or not.

Mickle T.
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 11, 2014, 10:17:35 am
Hello Mickle,

thanks for the information.

When I look for dewar I only get to pages with LN2.
And the lowest cost dewar is around one HP34401A.
$10 sounds reasonable.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 11, 2014, 11:14:19 am
Blue (K2015) is the LM299 , strange that the LM299 (Board and/or trimpot) seems to have a negative tempco.

Interesting would be also the raw value of the voltage.

Which type of trimmer do you use?
AccuTrim series of vishay?

Otherwise you should use the "standard precision" trimming scheme shown e.g. in the AD587 datasheet (figure 3+4).
extra 10-100k trimmer from the output to ground and the tap via a "precision resistor" (RT in datasheet) to the center tap of the 9+20K divider.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 11, 2014, 11:56:25 am
Andreas

Else i think you should look at Thermo Cups

http://www.aliexpress.com/item/Chinese-tea-cup-280ml-261-280-elegant-tea-art-double-layer-glass-water-cup/1375175051.html?s=p (http://www.aliexpress.com/item/Chinese-tea-cup-280ml-261-280-elegant-tea-art-double-layer-glass-water-cup/1375175051.html?s=p)

The above one seems to big (280 ml) , but ....

http://www.aliexpress.com/item/free-shipping-by-CPAM-10-pcs-lot-Bombs-Away-Shot-Glasses-Missile-Shot-Glss-44ml-Double/671805147.html (http://www.aliexpress.com/item/free-shipping-by-CPAM-10-pcs-lot-Bombs-Away-Shot-Glasses-Missile-Shot-Glss-44ml-Double/671805147.html)

http://www.aliexpress.com/item/Free-Shipping-Set-of-6-hand-blown-double-walled-glass-Espresso-Coffee-mug-80ml-teacup-Thermo/1760091959.html (http://www.aliexpress.com/item/Free-Shipping-Set-of-6-hand-blown-double-walled-glass-Espresso-Coffee-mug-80ml-teacup-Thermo/1760091959.html)

http://www.aliexpress.com/item/4pcs-1-7-fl-oz-50ml-Heat-Resisting-Glass-Double-Wall-Cocktail-Wine-Tea-Cup-Mug/1545114611.html (http://www.aliexpress.com/item/4pcs-1-7-fl-oz-50ml-Heat-Resisting-Glass-Double-Wall-Cocktail-Wine-Tea-Cup-Mug/1545114611.html)


Not perfect , but maybe usable

/Bingo
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 11, 2014, 11:58:50 am
Blue (K2015) is the LM299 , strange that the LM299 (Board and/or trimpot) seems to have a negative tempco.

Interesting would be also the raw value of the voltage.


Which type of trimmer do you use?
AccuTrim series of vishay?

Otherwise you should use the "standard precision" trimming scheme shown e.g. in the AD587 datasheet (figure 3+4).
extra 10-100k trimmer from the output to ground and the tap via a "precision resistor" (RT in datasheet) to the center tap of the 9+20K divider.

with best regards

Andreas

Quote
raw value
Is that the LM299 Vout ?

The trimmer has CERMET written on it (but from ebay .. so ??) , i have 100 & 50 ohm also but only 0805 1% smd's in "General values" (csd-electronics.de).
I might measure the current value on the 200R , and try to replace it with a resistor & a 50ohm trimmer.

I don't have any of the stable trimmers.

Will have a look on the 587 DS (thanx)


/Bingo
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 11, 2014, 01:19:37 pm
Not perfect , but maybe usable

I have seen similar tea cups already but:
The silver reflector on the vacuum side is missing.

But perhaps its worth a try.
The dewars from Mickle seem to be no longer available.
Would be interesting if a thermal coffee mug
(stainless steel with vacuum) gives similar results.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on May 11, 2014, 03:54:12 pm

Would be interesting if a thermal coffee mug
(stainless steel with vacuum) gives similar results.



I've been using 1 and 2 L coffee carafes as LN2 dewars now for a while. They silvered glass ones work great. All the others (stainless steel) not so great. Look for the word "Airpot" on ebay.
Title: Re: LM399 based 10 V reference
Post by: macfly on May 11, 2014, 08:19:36 pm

Hello Mickle,

would you let the cat out of the bag, where we can buy S5-61 foil resistors?


Hello,

You could also order some VHP100/VHP103 or from the HZ series resistors from vishay and select for TC.
Most probably the S5-61 resistors are no longer manufactured. Otherwise you could google after them.

with best regards

Andreas

Andreas,

be sure, I studied the complete VISHAY precision resistor program, before I ask such a question.
Buying 10 or more VISHAY resistors of VHPxxx is nearly impossible for a hobbyist like me.
The russian resistors are military surplus, I think. Perhaps you can buy twenty S5-61 resistors for one VHPxxx?
They are also very interesting for applications with lower requirements.

Best regards,

macfly
Title: Re: LM399 based 10 V reference
Post by: bingo600 on May 12, 2014, 06:11:12 am
A bit of graphs (K2015 (blue) is the LM299) , red is the SVR-T

The LM299 sems quite stable , but it's strange that the board seems to have a negative tempco.
PCB is on the table aka. "open air".

I'll prob do as Andreas suggests , measure LM299 out and the 10v out.

/Bingo
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on May 12, 2014, 03:23:03 pm
I've had a raw LM299 hanging on magnet wire from a HP power supply for about a month now -- just swinging in the air over my bench. I've been really surprised how stable its voltage is for an unheated style of Vref. If I'm remembering correctly, the LM299 has a negative tempco.
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 12, 2014, 07:31:30 pm
If I'm remembering correctly, the LM299 has a negative tempco.

Hello,

from my LM399 most have a positive tempco see:

https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg422367/#msg422367 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg422367/#msg422367)

@bingo:
I do not think that the deviation comes from the LM299.
Except if you have a defective device.

In my ageing box I had also large deviations from one LM399.
The cause was a bad contact of the heater supply.

By the way: Do you use good thermal isolation for the LM299.
On both top and bottom side of the PCB?

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Andreas on May 12, 2014, 09:27:59 pm
Hello,

this is 2nd part of:
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg441913/#msg441913 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg441913/#msg441913)

now I have measured LM399 #14.
#14 is the one of my ageing box with the lowest voltage difference between unheated + heated.
Test setup is the same as previous post.
14V via 6K8 fix for the zener, 9-30V variable for the heater with variable series resistor.

There are some differences to mention:

PSRR (power supply rejection ratio) on #14 (60uV) is nearly 2:1 compared to #13 (34uV)
-> there seems to be no correlation between PSRR and TC (temperature coefficient)

#14 works downto 9V heater voltage as specified in the datasheet. (#13 had some problems below 9.1V)

According to tempco (slightly negative on #14) the output voltage rises if heater voltage is too low with series resistor. (see unzoomed picture of 10R series resistor).

Again on the diagrams:
X-axis is heater voltage
left Y-axis is 2:1 divided zener voltage in mV (red dotted line)
right Y-axis is local PSRR gradient in ppm/V (green dots)

Results for 12-30V heater voltage change of LM399 #14

  0R: = 60uV
10R: = 64uV
22R: = 62uV
47R: = 67uV

So again any resistor on the heater seems to worse the PSRR slightly.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on May 13, 2014, 10:01:28 pm
Well, its official.... I hate how the LM399's datasheet shows the example circuits in reverse to how the die is wired.    :palm:
Title: Re: LM399 based 10 V reference
Post by: branadic on May 14, 2014, 08:17:17 am
Quote
When I look for dewar I only get to pages with LN2.
And the lowest cost dewar is around one HP34401A.
$10 sounds reasonable.

Hello Andreas,

I found dewar available from different suppliers:

https://www.thgeyer.de/store?PGM=search&GOTO=V1&OPEN=23233030303030333136&TICKETID=0012984908&randomseed=140514101031#opened (https://www.thgeyer.de/store?PGM=search&GOTO=V1&OPEN=23233030303030333136&TICKETID=0012984908&randomseed=140514101031#opened) (see 9.032 011, 100ml, also available without the blue jacket, so the price could be somewhat lower)

http://www.carlroth.com/catalogue/catalogue.do?lang=de-de&act=showBookmark&favOid=000000040000907600020023 (http://www.carlroth.com/catalogue/catalogue.do?lang=de-de&act=showBookmark&favOid=000000040000907600020023) (see LE85.1)

It's not that cheap compared to the disassembled TCXOs, but anyway available.

Regards, branadic
Title: Re: LM399 based 10 V reference
Post by: babysitter on May 14, 2014, 11:51:49 am
http://www.laborhaus.de/98/112/AD183/MTAxMQ==/KGW%20Isotherm%201011%20Laborhaus%20Scheller.html (http://www.laborhaus.de/98/112/AD183/MTAxMQ==/KGW%20Isotherm%201011%20Laborhaus%20Scheller.html)

The replacement glass part might save a few € against the full part and also the life of a poor nice oscillator !
Title: Re: LM399 based 10 V reference
Post by: branadic on May 14, 2014, 06:35:23 pm
Quote
The replacement glass part might save a few € against the full part and also the life of a poor nice oscillator !

Thats what I meant with:

Quote
(see 9.032 011, 100ml, also available without the blue jacket, so the price could be somewhat lower)

It could be cheaper to directly ask http://www.kgw-isotherm.de/ (http://www.kgw-isotherm.de/) for a quote of a replacement glas as they are the original manufactor:

http://www.kgw-isotherm.de/produkte/dewar/00.html (http://www.kgw-isotherm.de/produkte/dewar/00.html)
Title: Re: LM399 based 10 V reference
Post by: branadic on May 15, 2014, 06:00:26 pm
Okay, it's exact the same price at KGW-Isotherm direct 35,51€. The plug itself is about 23,04€ pretty expensive.
Title: Re: LM399 based 10 V reference
Post by: babysitter on May 15, 2014, 06:25:19 pm
One could top it off with a Petri dish? At work I am using some customized thermostats out of - you guessed right - tuner boxes embedded in styrodur-like foam. In the 36 to 42 °C range I get the innards flopping around in a ~0.07° window.

Title: Re: LM399 based 10 V reference
Post by: doktor pyta on May 16, 2014, 08:31:48 am
Check this out:
https://www.eevblog.com/forum/projects/vishay-bulk-foil-drift-after-soldering/ (https://www.eevblog.com/forum/projects/vishay-bulk-foil-drift-after-soldering/)
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 16, 2014, 08:16:09 pm

I found dewar available from different suppliers:



The replacement glass part might save a few € against the full part and also the life of a poor nice oscillator !


Hello,

thanks for sharing.
This may be one of the next projects:
Either 4 LM399s or 1 LTZ1000 within a dewar.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on May 23, 2014, 02:02:15 am
I've really enjoyed these volt-nut threads and though I'm a newb, I've started down my own path of designing an LM399 reference.

The idea I'm exploring is to avoid any sort of trim on the output op-amp's voltage divider (just two fixed value resistors to set the gain).  Rather, I'll achieve trim by adjusting the current driven through the LM399 zener.

Not only that, but I'll control the current source for the LM399 zener via a DAC, so that you can calibrate it via software.  This means I can immerse the entire circuit in an oil bath and not have to crack it open while calibrating to tweak a pot, etc.

So, the first step down this path is coming up with a current source.

My first attempts were a traditional op-amp with current sense resistor on the low side.  However, I realized having the current sense resistor on the low side would introduce current error one way or another.

So I tried wrapping my brain around making a high-side current source, and this is what I came up with:

(http://i.imgur.com/ijmXitI.png)

(Here, the DAC is represented by a voltage divider, and its reference represented by a simple zener).

Using this setup, a drift in the LM317 of +/- 0.1V results in a current source error of only 2.3 uA.  Not bad!

I'm initially choosing an LM317 target of 16.25V (using R values of 100 and 1200), as that allows operation from three 9V batteries discharged down to about 6.5V each.  (the batteries would be needed while e.g. putting the reference in a faraday cage -- I have terrible noise pickup at my house).

Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on May 23, 2014, 02:19:51 am
Hrmm, oops, 16.25 volts is too low for full range output of the DAC in this example.  That's partly because the LM399 is represented by an 8.2V zener, rather than a 6.95V zener.  So it looks like the minimum voltage supply I need is about 17.5V.

(http://i.imgur.com/tQTAfYq.png)
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on May 23, 2014, 04:09:50 am
Uh oh, 10 minutes at the breadboard and I'm officially addicted to volt-nutting.  Even the most trivial LM399 circuit stops all the digits on my HP 3478A dead still!   The reading hasn't budget an LSB in 10 minutes now!

I fear I'll be justifying upgrading to a used 34401A all too soon...

Fun!
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 23, 2014, 09:50:59 pm
Rather, I'll achieve trim by adjusting the current driven through the LM399 zener.
Good Idea: so you will have more fun at adjusting: Trim a bit, wait 15 minutes for thermally equalizing. Trim again. w. w. w.

Using this setup, a drift in the LM317 of +/- 0.1V results in a current source error of only 2.3 uA.  Not bad!
Too much: with 1 Ohms for the zener this is 2.3uV output voltage change.
Requirement of datasheet is 1mA +/- 0.1% for long term stability.

So I tried wrapping my brain around making a high-side current source, and this is what I came up with:
The LT1001 is no rail/rail OP. The cirquit violates at minimum the input common mode range below 2mA.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on May 23, 2014, 10:22:08 pm
Thanks Andreas

Requirement of datasheet is 1mA +/- 0.1% for long term stability.

Ah, I missed that part of the spec.  Do we know if they list this part of the spec simply because that is the only condition under which they tested long term stability, or are they saying that exactly 1mA is the sweet spot for long term stability because they've tested many different currents and found that to be best?
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 24, 2014, 04:54:03 am
https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/msg404500/?topicseen#msg404500 (https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/msg404500/?topicseen#msg404500)
https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/msg427366/?topicseen#msg427366 (https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/msg427366/?topicseen#msg427366)

It seems to be the later

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: echen1024 on May 24, 2014, 05:08:17 am
I am using a double walled glass teacup I purchased in China. Seems to work quite well.
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on May 25, 2014, 02:03:35 am
I came across an interesting current source in the LM4050 datasheet:

(http://i.imgur.com/FrPYXcC.png)

I thought this arrangement might be interesting for use with driving an LM399:

(http://i.imgur.com/VCRZ6MP.png)

This struck me as being interesting, because this greatly reduces the parts count, and also because the only pins which need to touch a voltage regulator are the op amp supply pins.

The 1000:1 ratio of the LT1029's 5V to R1's 5000 ohms translates into a 1000:1 ratio of LT1029 voltage drift to LM399 zener current, and also allows us to drive exactly 1mA through the LM399.

An R2 of 6.95k would allow us to also drive 1mA through the LT1029 (all of the stability specs in the LT1029 datasheet specify 1mA).  Welwyn makes a 0.1% 15ppm/C 6.98k resistor for $1.92.  http://www.digikey.com/product-detail/en/RC55Y-6K98BI/985-1056-1-ND/2401921 (http://www.digikey.com/product-detail/en/RC55Y-6K98BI/985-1056-1-ND/2401921)

If we replace the zeners with voltage sources we can start to do some analysis.  It appears a drift of 1mV in the LT1029 would result in 200nA drift in our current source:

(http://i.imgur.com/NyjtVAD.png)

BOM costs thus far:
LM399A: $10
5k Vishay Foil: $20
LT1029A: $5
R2 (6.98k): $2
LT1001 (x2): $13

Total: $50.


This circuit seems to work, but I do admit I'm having a hard time convincing myself that it wouldn't suffer from start-up problems (what would prevent the op amp from just staying at 0V?).
Title: Re: LM399 based 10 V reference
Post by: fmaimon on May 25, 2014, 02:09:06 am
Read the first post of this thread. It's a better way to create a constant current as stable as the LM399 it is powering.
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on May 25, 2014, 02:48:13 am
Read the first post of this thread. It's a better way to create a constant current as stable as the LM399 it is powering.

ah, taking another look at your schematic, I think I finally understand it :)
Title: Re: LM399 based 10 V reference
Post by: CaptnYellowShirt on May 25, 2014, 04:15:09 am
Does anyone have experience with using current regulating "diodes" for this purpose?

http://www.mouser.com/Semiconductors/Discrete-Semiconductors/Diodes-Rectifiers/Current-Regulator-Diodes/ (http://www.mouser.com/Semiconductors/Discrete-Semiconductors/Diodes-Rectifiers/Current-Regulator-Diodes/)
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 25, 2014, 05:23:04 am
Read the first post of this thread. It's a better way to create a constant current as stable as the LM399 it is powering.

That is still the best circuit for creating a stable reference.
If you look at the HP34401 service guide you will find exactly this ciruit. (Additionally with some startup pull up resistor).

If you do not want to use the 10V output, but create only a stable current source for the LM399 you could even use somewhat cheaper resistors (RC55Y with 15ppm/K) instead of S102 or Z201.

Does anyone have experience with using current regulating "diodes" for this purpose?
These are simple J-FETs where Gate + Source is connected together. -> you will never satisfy the requirements of a precision circuit.

If you have very much time you could try to use a LM334 current source and use the ciruit with TC compensation of the datasheet. With individual adjustment of the resistors you can get a final TC for the current source of below 50 ppm/K. But it will take several temperature cycling to get you there.

With the above cirquit of fmaimon you will get instant stable output depending on your resistors.
But where is more fun?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 25, 2014, 03:11:32 am
After not having much luck with LTZ1000, I thought maybe parallel LM399 was more my style.  Having something I could leave powered off most of the time appealed to me, too.

I ordered 10 LM399A from digikey, hoping I would get several that behaved like Andreas's star performers.  Turns out the Linear parts aren't quite like the National Semiconductor units, my lowest reference is 7.03V and highest is 7.1  (this seriously foiled my resistor plan, too.)

I then built a crude aging oven (an electrical box with nightlights under it, housed in a box made from a ceiling tile), wired my 10 references up in serial with a constant current driver, attached a keithley 2000 scan card, and fired it up.  There's also a RTD inside hooked up to another meter (a GDM 8261), and I wrote a simple program to talk to both meters and take readings.

 I can now confirm
There is a large possibility that the tempco is not linear but either a parabolic or s-shaped curve (due to compensation of the zener with a Vbe).
(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=99065;image)
The x axis is temperature of the RTD, y axis is ppm change from a row I choose based on what the operating temperature would be in normal use.  The actual die temperature will be lagging the RTD temperature by a variable amount, accounting for some of the 'S'-ness of the curve (the experiment was only intended to compare TC of the different units and not compute an accurate TC curve.)  Imagine my surprise when one of them turned downwards at around 80C! Too bad for me I only have one like that.  Just like with the other reports in this thread, my lowest voltage unit was the best.  Absolute voltage isn't represented on this chart, but lower voltages corresponded to lower tempco.

My range of unheated tempco appears to match Andreas exactly.  Were those units also LM399A?

Has anyone tested if the LM399 also has some kind of zero tempco current
(near the heater temperature) in the valid range of 0.5 .. 10mA for the zener?
I have tested 0.6, 1.0, 1.1, and 1.5mA.  The resulting graphs were effectively identical.
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on June 25, 2014, 04:15:35 am
Hi Galaxyrise,

I've got some data on the LMx99/A series from Bob Pease, as soon as I can dig it out of my archives I will try to post it, may be difficult in the form it is.
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 25, 2014, 06:52:32 pm
My range of unheated tempco appears to match Andreas exactly.  Were those units also LM399A?

I have tested 0.6, 1.0, 1.1, and 1.5mA.  The resulting graphs were effectively identical.

Hello,

my LM399H are all non "A" versions from National Semiconductor.
Good to know that the A-Versions currently stocked on DigiKey have the upper limit.
So I´ll probably better try the non A version.
The typical ageing spec of LT is somewhat lower than the NS spec was.

With LT1027 parts I have also made the experience that the C-grade devices (LT1027CCN8-5) in average had lower temperature gradient around room temperature than the 5 B-grade devices (LT1027BCN8-5) that I ordered later.
On non heated devices I can imagine that they tweak the (average) tempco at the temperature extremes giving more temperature gradient around room temperature (S-shaped curve). But with heated devices ....?

1N829A Zeners have their minimum T.C. around 7.5 mA +/- 3 mA

p.s.: do you have a photo from your measurement?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 25, 2014, 08:10:25 pm
1N829A Zeners have their minimum T.C. around 7.5 mA +/- 3 mA

p.s.: do you have a photo from your measurement?

I'll let it cool off tonight and then try 7.5mA tomorrow to see if it's any different.  I think that would be awfully high from a long-term stability standpoint, though. 

Since I'm letting it cool down anyway, I can easily photo anything you'd like, but I'm not sure what you're asking for...  Could you explain what you'd like to see?

I've got some data on the LMx99/A series from Bob Pease, as soon as I can dig it out of my archives I will try to post it, may be difficult in the form it is.
Neat! I look forward to it.  I did get the impression he played around with this part a lot.
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 25, 2014, 09:08:21 pm
Hello,

I just can´t imagine how your "ageing oven" from night lamps and tiles looks like.
sounds interesting.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 26, 2014, 05:06:04 am
I just can´t imagine how your "ageing oven" from night lamps and tiles looks like.
sounds interesting.
My pleasure to do a bit of show and tell :)

Things not clear from the pictures: The inside is roughly 6" square; the tile was originally 24" square.  The pieces of tile are tacked together with finishing nails.  The lid is another section of ceiling tile inside a cardboard box cut and taped to be an exact fit. I made the lid pretty tall on a whim, but that extra layer of cardboard looks to do a fair amount of good!

The bulbs are 15W each, and they're plugged in to a dimmer (the white box with the slider.)  As you can see from where the slider is, the bulbs aren't quite run at full power (the oven gets to around 140C at full power.) Heating 100C above ambient has a time constant around 30 min.

The metal platform is a steel plate from an old keyboard with three 2" bolts for legs.  I burned a paper towel and the smeared the ashes all over the bottom of that plate, blacking it.  (That was a huge help.) On top of the plate is a piece of perforated aluminum (http://www.homedepot.com/p/MD-Building-Products-3-ft-x-3-ft-Aluminum-Venetian-Bronze-Lincane-Sheet-57015/202525469) serving as my oven grill.  The holes in it were also handy for tying things down.  Then sitting on the plate but surrounding the grill is the 4x4" electrical box.  The wires going into the oven are a combination of white wire-wrapping wire and red enameled wire, both 30 gauge.
Title: Re: LM399 based 10 V reference
Post by: Vgkid on June 26, 2014, 01:31:47 pm
Certainly an interesting build, how stable can you get it?
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 26, 2014, 02:00:07 pm
Certainly an interesting build, how stable can you get it?
It's not regulated at all; it varies over a 5C range as ambient conditions change.  Good for aging parts, not so ideal for accurate TC characterizing.

I may update it some day with a peltier module+fan instead of the lamps and make controlled environment out of it.  (But peltier modules typically don't like going to 125C, hehe)
Title: Re: LM399 based 10 V reference
Post by: jlmoon on June 26, 2014, 02:10:05 pm
That's an awesome looking ez-bake oven..  ;D
Title: Re: LM399 based 10 V reference
Post by: Vgkid on June 26, 2014, 05:40:27 pm
Certainly an interesting build, how stable can you get it?
It's not regulated at all; it varies over a 5C range as ambient conditions change.  Good for aging parts, not so ideal for accurate TC characterizing.

I may update it some day with a peltier module+fan instead of the lamps and make controlled environment out of it.  (But peltier modules typically don't like going to 125C, hehe)
good enough, i might try it in the future.
Title: Re: LM399 based 10 V reference
Post by: jlmoon on June 26, 2014, 06:11:54 pm
I was wondering how I was going to age some parts I have.. without spinning the electric meter off the wall.  Great idea Galaxy
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 26, 2014, 08:36:04 pm

My pleasure to do a bit of show and tell :)


This is how I like it: simple + effective.  :-+

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 29, 2014, 12:45:27 am
Thank you all for the kind words!  And yes, it is like a low-power, old-style Easy-Bake Oven, hehe.  That collection of LM399s even vaguely resembles a cookie...

I had some technical difficulties when I first reconfigured the experiment, and then I opted to measure more currents--thus the delay.  I must say I'm very impressed with the robustness of these parts: given the errors that I've made, I'm surprised they all still function properly. 

Attached are voltage (as ppm difference from 90C) vs temperature at 0.6mA, 7.5mA, and 9.5mA.  Also is voltage vs temperature for #8 at different currents.  I corrected the graphs for temperature lag as best I could, which gets rid of that erroneous S bend at the beginning.  You can see the difference in the shape of the 0.6mA graph, which is the same data as the graph in my earlier post.

As Andreas expected, there is a visible difference at higher currents; a very non-linear shift.  It looks like most of my units can not be "zeroed" out at 90C within the 0.5 - 10mA range.  Following my curve-fit guess for how the parabola moves with current, #8 would "zero" at around 8.5mA.

Also not surprising, at 9.5mA there is a substantial amount of self-heating; raising the RTD by 12C or so when the oven lamps are off.

Edit: Removed charts that showed a decrease in the temperature of peak voltage with increase in current.  I had only accounted for half of the self heating.
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 29, 2014, 09:11:50 am
Hello,

amazing result.

Thanks for sharing

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on June 29, 2014, 05:16:34 pm
Hello everyone,

Galaxyrise, nice set of curves, good repeatability within your limits.  I have found one note that Bob Pease sent me so far, you may have a hard time reading his hen scratch, this was part of a letter he sent me back in 2003, he did not mention just what meters were involved in the test nor the exact date of the test.  The 'graph's (I'll use that term loosely) indicated a repeatable drift in the measurements of the LM399 (he did not mention what the suffix was, an H or AH).  The test was over some months long but Bob didn't mention just how long, usually a long term drift measurement was about 10-12 months.

The maximum TCV of a LM399H is 2 PPM/°C, if you are seeing anything higher than this, there is a problem with the test setup.  The LMx99 is very linear, regardless of operating current, the internal temperature is regulated at ~90°C.  The operating ambient temperature range for the LM399/A is 0°C to +70°C.  The internal temperature of the LMx99/A is self regulating within the specified ambient temperature range.  Your measurements should indicate a linear TCV with change in ambient temperature.  All those curves seem to indicate problems with the test setup, even with repeatability of the curves, it indicates the error sources are relatively stable themselves.  Yes, the LMx99/A run at 90°C so you are going to see heat radiating off of them, the zener generates much less self-heating as is the design intention.  The heater runs at about 300mW at 30V, the zener would run at ~70mW at 10mA, at 1mA only ~7mW.  The heater compensates for the zener's heat as well, mostly leaving just the TCV of the zener itself.  The drift spec indicates a zener current of 1mA +/- 10%.  I have used the LMx99/As and I have never seen anything beyond a consistent TCV with them.  With varying ambient temperature, I have measured nothing more than the drift of the zener, usually under 1 PPM/°C for a LM399/A version.  Thermal EMFs can be really difficult to control and they tend to be nonlinear in nature.

I have attached Bob's note to me plus a National data sheet from 1999 which gives some interesting specs.  I also attached the LM199 data sheet from Linear Tech from 1990.  Linear made some improvement in the TCV over the older National part, but we're talking a few tenths of a PPM/°C here.

I remember someone mentioning the old 1N827/A voltage reference zener which dates back to the 1950s from Motorola, this special zener could exhibit a TCV of 5 PPM/°C but this wasn't the only Motorola zener that could do that; the 1N4569/A, 1N4574/A, 1N4579/A, 1N4584/A, 1N4779/A, 1N4784/A, 1N939/A/B, 1N4769/A, 1N4774/A, 1N2624/A/B, 1N2169/A, 1N2170/A, 1N2171A and the 1N945/A/B.  These zeners all had 5 PPM/°C TCV and voltages between 6.2V and 11.7 V nominally.  The zener currents varied between 0.5mA and 10mA for best performance and the temperature extremes could be as wide as -55°C and +150°C, pretty impressive I'd say.
Title: Re: LM399 based 10 V reference
Post by: jpb on June 29, 2014, 05:32:01 pm
The maximum TCV of a LM399H is 2 PPM/°C, if you are seeing anything higher than this, there is a problem with the test setup.  The LMx99 is very linear, regardless of operating current, the internal temperature is regulated at ~90°C.  The operating ambient temperature range for the LM399/A is 0°C to +70°C.  The internal temperature of the LMx99/A is self regulating within the specified ambient temperature range.  Your measurements should indicate a linear TCV with change in ambient temperature.  All those curves seem to indicate problems with the test setup, even with repeatability of the curves, it indicates the error sources are relatively stable themselves.  Yes, the LMx99/A run at 90°C so you are going to see heat radiating off of them, the zener generates much less self-heating as is the design intention.  The heater runs at about 300mW at 30V, the zener would run at ~70mW at 10mA, at 1mA only ~7mW.  The heater compensates for the zener's heat as well, mostly leaving just the TCV of the zener itself.  The drift spec indicates a zener current of 1mA +/- 10%.  I have used the LMx99/As and I have never seen anything beyond a consistent TCV with them.  With varying ambient temperature, I have measured nothing more than the drift of the zener, usually under 1 PPM/°C for a LM399/A version.  Thermal EMFs can be really difficult to control and they tend to be nonlinear in nature.

I was under the impression that for this experiment the heaters weren't wired in - but looking back through the posts I can't see where I got that impression from now (i.e. the zeners just connected as zeners).

Perhaps Galxyrise can clarify this.
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on June 29, 2014, 05:47:19 pm
Hello jpb,

The point of using an LMx99/A is to utilize the internal heater to stabilize the reference zener's operating temperature, hence the exceptionally low TCV.  Attempting to measure the LMx99/A's zener TCV without the heater energized is an invalid test of the part and only erroneous readings will result if that is the case.

In the graphs, I noticed that the readings were referenced to 90°C, it would appear to infer that the LM399s were not being operated with the heater energized and that while the LM399's storage temperature does extend to +150°C....unenergized.....then the conditions and data inferred by the graphs are invalid.  If the LM399 had been operated with the heater energized and the ambient was as high as 100°C, then the chips were being operated improperly under those conditions, that does not appear to be the case though.

It would be appreciated if Galaxyrise could clarify his test conditions for us.
Title: Re: LM399 based 10 V reference
Post by: jpb on June 29, 2014, 06:31:12 pm
Hello jpb,

The point of using an LMx99/A is to utilize the internal heater to stabilize the reference zener's operating temperature, hence the exceptionally low TCV.  Attempting to measure the LMx99/A's zener TCV without the heater energized is an invalid test of the part and only erroneous readings will result if that is the case.

Again, as I'm not Galaxyrise I can't claim any deep knowledge, but my understanding was that by finding which devices had flattest curves around 90C good devices could be selected without having to monitor the heated devices for a very long period - but this is just a guess on my part. I think the point was to find devices (and current levels) which would be optimal when the heaters were used in the final circuit.
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 29, 2014, 06:42:33 pm
Attempting to measure the LMx99/A's zener TCV without the heater energized is an invalid test of the part and only erroneous readings will result if that is the case.

The heater is not energized for the above; jpb is correct.  And as such, those charts are not the TCV you'd get from a "fully operational" LM399.  Sorry if I confused anyone with that!  But I wouldn't call the readings "erroneous", either. I believe they represent the TCV of just the reference half.

I would expect that the TCV of a fully powered LM399 would essentially be a measure of the change in internal temperature as external temperature changes. It's hard to guess if the A units are 1ppm/C because the heater regulates better, the reference TC is better, or both, but 140ppm corresponds to a 10C shift on my "worst" unit. It would be very hard to notice any non-linearity in my graphs over that 10C range, I think. You'd need a reference with a zero TC current and a very good experiment to see that gentle curve--the curve would only be 7ppm high over the 70C operating range!

So I think my results are consistent with your experiences and expectations from the datasheet.

The story is that I was rigging these LM399 to age at 125C (as suggested by Dobkin himself), and I wanted to be able to pick the most stable references of my 10 over time when aging was done.  I logged the warmup period to help correct for temperature variations when evaluating time stability.  Then I saw that downward turn on one reference, and the rest is in the posts here :)

It wouldn't surprise me at all if there remain artifacts from my test setup as I'm no expert in measuring this kind of thing and the setup wasn't even intended to measure it in the first place. But everything has been behaving consistently and repeatably, so at this point I'm inclined to think the shape and movement of the curves are reasonably representative of truth. 
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 29, 2014, 06:49:49 pm
I have attached Bob's note to me

Thanks! An enjoyable read. I also struggle with how to judge long-term stability of LM399 when my measuring gear is also LM399-based, hehe. 

Quote
plus a National data sheet from 1999 which gives some interesting specs.  I also attached the LM199 data sheet from Linear Tech from 1990.  Linear made some improvement in the TCV over the older National part, but we're talking a few tenths of a PPM/°C here.
   The NS sheet does have interesting extras, like that 1000hr aging chart.  And a TO-92 part, teehee! 
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on June 29, 2014, 07:54:22 pm
Galaxyrise,

You are welcome, I think I have some more data from Bob, it may be buried in old e-mails from him or I just can't find the hard copy of it.  Bob usually liked hard copies instead of e-mailing.  His handwriting though was a stinker to decipher at times, particularly schematics which was helter-skelter drawn.  By the way, Jim Williams of Linear Tech was also schematic drawing challenged.

Sorry, if I had come across this thread earlier, I might have been able to save you some work.

Your assumption about the 'zener' diode in the LM399 is understandable but it is still an invalid assumption.  The buried zener circuit was designed to operate at the chosen temperature of 90°C and was optimized for that temperature only, deviating from that temperature even a few degrees will not 'show' the TCV of the zener accurately.  Its characterization depends on its operation at the specified 90°C only internally.  Outside of that temperature range, the buried zener will have a composite TCV curve made up of the entire voltage reference circuit not just the buried zener.  It cannot be characterized in the same manner that you would apply to a regular zener diode unfortunately.

If you have several LM399s available, you could parallel them using the appropriate summing circuit to get an improved composite characteristic, it doesn't just work for reducing noise.
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 29, 2014, 10:10:10 pm
Your assumption about the 'zener' diode in the LM399 is understandable but it is still an invalid assumption.  ... Outside of that temperature range, the buried zener will have a composite TCV curve made up of the entire voltage reference circuit not just the buried zener.
Characterizing the entire reference circuit is exactly what I thought I was doing (as it's what matters in the end), but I hadn't really thought about that in detail. I was just reading through National Semi's App Note 161 (http://www.ti.com/lit/an/snoa589c/snoa589c.pdf), and I can really see what you mean.  The current changes, especially, basically don't affect the zener at all!  (Though it does affect the circuit as a whole.) I suspect I will run a similar experiment when aging is done, but with the heaters on (and a narrower temperature range, obviously.)  It will be interesting to see if it correlates to the unstabilized data at all, or if the charts I produced are just a curiosity. 

Quote
If you have several LM399s available, you could parallel them using the appropriate summing circuit to get an improved composite characteristic, it doesn't just work for reducing noise.
That's the plan!  I've seen where Pease talked about running them in groups of 4, and I was figuring 4-6 based on how they looked in the latter half of aging.
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on June 30, 2014, 07:10:29 am
I have found one note that Bob Pease sent me so far, you may have a hard time reading his hen scratch, this was part of a letter he sent me back in 2003, he did not mention just what meters were involved in the test nor the exact date of the test.  The 'graph's (I'll use that term loosely) indicated a repeatable drift in the measurements of the LM399 (he did not mention what the suffix was, an H or AH).  The test was over some months long but Bob didn't mention just how long, usually a long term drift measurement was about 10-12 months.

Wonderful!

I have tried my best to read Bob's hen scratch and have typed it up in a google doc:

https://docs.google.com/document/d/1fpuR_R9YvK-08QuEIy1pIZLdUrmNANFGbx4z22E5e68/edit?usp=sharing

That document is publicly editable, so let's have our community collectively come up the best interpretation of his hen scratch :)
Title: Re: LM399 based 10 V reference
Post by: branadic on June 30, 2014, 09:56:00 am
Quote

Note to Ed Pettis

If you look at the LM399’s you might see a PATTERN for the first couple weeks, and the last week:

(figure)

- Now if they all look like this - is that plausible? (No)

- Is it likely that the measurement system’s Reference V drifted (figure) ? Yes.
We keep alternative references and if we see that everything is drifting like (figure) then we know these parts are really quite good.  If we had to, we could crank in some corrections and show that these parts are actually more stable than the measurement system!  But as we are only talking about ~ + and - 7 PPM.

Title: Re: LM399 based 10 V reference
Post by: Andreas on June 30, 2014, 07:12:57 pm

Characterizing the entire reference circuit is exactly what I thought I was doing (as it's what matters in the end), but I hadn't really thought about that in detail.

 or if the charts I produced are just a curiosity. 

Hello,

I believe that you can improve the T.C. behaviour around the heater setpoint significantly with the "zero T.C." zener current.

Of cause higher currents may create a worse long term stability.
On the other side you should get better short therm (less noise with good thermal shielding) and better temperature stability. (lower overall T.C.). Noise should decrease by a factor of 2 if using a factor 4 higher current similar to a array of 4 LM399 with 1 mA each.

A good test would be probably the "tilting experiment" in comparison of 1 mA and "zero T.C." current.
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg360779/#msg360779 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg360779/#msg360779)

This test uses the imperfection of the LM399 that the heater is not symmetrical to the zener and the temperature sensing element.

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on July 02, 2014, 06:00:44 am
I think the shift of the top of the TCV curve in the last graph is due to self-heating of the LM399.  If you turn on the current and take a voltage reading then turn off the current, and allow some thermal settling time between measurements-- I think these shifts will go away.
I started the 7.5mA and 9.5mA tests by letting the references heat the oven to get a sense for the self-heating, but now that you suggest it, I'm pretty sure I didn't compensate for it nearly enough.  Your measuring method sounds like a much better way to eliminate the error. It would be harder to do over the entire range, but so long as the parabola is still true, I would only need to spot check a few points to get a rough sense for the shift.  I will endeavor to do that this weekend. 
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on July 02, 2014, 04:52:57 pm
Unfortunately, zener diodes, buried or otherwise, are just plain noisy, very careful processing at the wafer level can go a long way to minimizing wideband noise and even paring down shot noise a bit but they are noisy, just like the gas regulator tubes back in the good ole days.  These references also have additional circuitry in parallel with the zeners which add their own bit of noise to the output too.  While filtering can decrease overall wideband noise, it has its own set of compromises too.  In the case of a voltage reference that is running for relatively long periods of time, a longer time constant in the filter is not so much of an issue.

If you are using the reference as a short term on source, then the time constant becomes much more important.  The filters would require PWW (not necessarily tight tolerance) resistors and polypropylene capacitors but even these capacitors will impart small voltage errors if the resistances are of high values.  Active filtering with appropriate op amps could possibly provide a good compromise.  Multiple summed voltage references (again with PWW resistors) will generally provide the least amount of total noise.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 02, 2014, 08:08:04 pm

Since the internal circuitry of the LM399(A) shunts extra current around the Zener, there is no way to decrease the awful noise spec by increasing current-- so you may as well run it at the 1.0mA current to get the best long term stability.


That´s a good point, I did not recognize this.
I made 2 measurements with LM399#19 out of my ageing PCB.

One with a 6k8 pull up to 14.1V stabilized from a battery. (around 1 mA)
the other with 6K8 + 1K in parallel (around 8 mA)
All packaged into a cookie tin can and amplified *10000 with a 0.1 ... 10 Hz 4th order bandpass.
So the scaling on the oscilloscope is 1sec/div on time axis and 0.5uV/div in y-axis.

So there is no visible difference between 1 mA and 8 mA.

On the other side I do not understand from where do you get the figure of 10uVpp of noise.
Measured with which bandwidth?
With correct setup of 0.1 ... 10Hz bandwidth I usually measure values from 2.5uVpp to 4uVpp as average with std deviation of 0.2 .. 0.4. Only one LM399 is around 7uVpp.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 03, 2014, 06:57:59 am
Hello,

if I look at the current LM329 datasheet it looks more like 3-4uVpp (Figure 15)
http://www.ti.com/lit/ds/symlink/lm329.pdf (http://www.ti.com/lit/ds/symlink/lm329.pdf)

How could you live without noise measurements up to now?
How will you detect the "stinkers" of your references?

I am not aware of a TI appnote for measuring noise of references.
(Although perhaps there is one for low noise voltage regulators using a (lower noise) voltage reference to compensate DC)
Usually only cirquits of OP-Amps are shown for this purpose at TI.
But this is more easy: no DC-Offset and the DUT is part of of the amplifier.

I was inspired of LT AN124
http://cds.linear.com/docs/en/application-note/an124f.pdf (http://cds.linear.com/docs/en/application-note/an124f.pdf)

But since I only wanted to measure down to 1uVpp I used standard precision OP-Amp cirquits like LT1037 and LT1012 for the 2 amplification and filter stages.
The result is a noise floor below 200nVpp.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on July 04, 2014, 04:08:04 am
Yes, that tantalum capacitor was a real special, something on the order of $1,200 each but at the time, it was the best solution to the circuit requirements.  I rather doubt very many of those were built.

Every now and then, on eBay, some Chinese manufacturer will put up a large capacitance polypropylene capacitor for a good price.  Not too long ago, I remember seeing a 400uF poly for $14 and change.  They don't show up all the time but they have specs better than the tantalum capacitor Jim was stuck with.  They are not too small but they are not the elephant in the room any more either.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 04, 2014, 05:15:44 am
You can also use standard (85 deg) electrolytic capacitors (35V)
if you select for lowest (< 20nA) leakage current @10V after 48 hours forming time.

But you will have to bias the capacitor all the time even when not using the cirquit.
Otherwise you will have excess noise for the first 24-48 hours when measuring.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Mickle T. on July 04, 2014, 02:39:23 pm
Quote
I used 3 each tantalum capacitors in parallel, they were 68uF/16V, they couple the input from the device under test to a 25 ohm 0.1% 25 ppm resistor to the inverting input.
Cutoff frequency is 31.2 Hz, but not 0.1 Hz.
Title: Re: LM399 based 10 V reference
Post by: branadic on July 04, 2014, 05:23:22 pm
I've dublicated the

 1/f noise amplifier by Andreas. (http://www.mikrocontroller.net/topic/207061#2060389)

Attached is a picture off my board. For its input cap (1.000µF||2.200µF) I've measured electrolytic caps with same value  (1.000µF and 2.200µF) f different vendors and temperature type to find the one with lowest leackage current. Best cap I found were 85°C types by Yageo. The combination of 1.000µF/25 || 2.200µF/35V for the low noise amp showed a leackage of <5nA after 24h on a 9V block. 105°C types were decades worse.
The amp is somewhat hand sensitve through its aluminium case and needs to be placed with the DUT in a shielded cookie box.

I've also build the 10Hz - 100kHz amp that is shown in AN83 (http://www.mikrocontroller.net/topic/250656#2666994) in different versions (60dB for oscilloscope applicaton, 80dB version combined with a STM32 board and its 12bit ADC).

I've used Oscon Caps (330µF, 6.3V) as input cap, but found that SMD tantal caps (330µF, 10V) do work as well. For higher voltages I've got 3 pretty expensive FFB54D0117KJC (http://www.digikey.com/product-search/en?lang=en&site=us&KeyWords=FFB54D0117KJC&x=10&y=6) 110µF, 75Vdc polyester caps that can be paralleled.

Both amps are pretty neat.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 04, 2014, 05:35:54 pm
Quote
I used 3 each tantalum capacitors in parallel, they were 68uF/16V, they couple the input from the device under test to a 25 ohm 0.1% 25 ppm resistor to the inverting input.
Cutoff frequency is 31.2 Hz, but not 0.1 Hz.
Hello,

That also explains the "low noise" of the Fluke 515A which is specced with up to 0.01% rms of range.

I would not connect such a low impedant capacitive source to a sensitive unbuffered LTZ1000 reference.
The last time I did this I got a hysteresis shift on the output voltage.
With a LM399 (1 Ohms differential impedance) you will get 4% less amplification since the 25 Ohms and 1 Ohms are in series.

With best regards

Andreas


With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 04, 2014, 05:46:09 pm
Attached is a picture off my board.

Hello branadic,

I would remove the socket of the first input stage to further reduce the noise.

And its always a good idea to do some thermal isolation (e.g. a piece of cotton or cloth)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on July 04, 2014, 05:53:16 pm
Quote
I would remove the socket of the first input stage to further reduce the noise.

Thanks, could be an option, I will keep that in mind and try that maybe some day.

Quote
And its always a good idea to do some thermal isolation (e.g. a piece of cotton or cloth)

The board is assembled in a standard aluminium case together with the batteries and some cotton, but this is not shown here.
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on July 05, 2014, 09:15:55 am
I think the shift of the top of the TCV curve in the last graph is due to self-heating of the LM399.  If you turn on the current and take a voltage reading then turn off the current, and allow some thermal settling time between measurements-- I think these shifts will go away.
Confirmed.  I took spot measurements to locate the peak of the parabola, and it appears to be right around where it was at 1mA.  I'm going to remove the errant graphs lest they confuse someone who doesn't read the whole thread
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on July 06, 2014, 06:10:06 am
Simulation: before and after
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on July 06, 2014, 02:09:10 pm
Sorry, it's my mistake  :-[
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 07, 2014, 06:43:58 pm
Hello,

although the discussion gets now rather off topic:
The LT1028 has very good voltage noise specs but rather bad current noise.
Above 200-300 Ohms input resistance the current noise dominates on a LT1028.
So just flipping the resistors will not help in this case for a good flicker noise amplifier.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on July 08, 2014, 01:05:57 am
These two notes from Linear Tech should help with op amp selection, one is older, the other an updated version.
Title: Re: LM399 based 10 V reference
Post by: TonyGreene on July 08, 2014, 09:21:47 am

although the discussion gets now rather off topic:

With best regards

Andreas


EEVblog Electronics Community Forum
*
A Free & Open Forum For Electronics Enthusiasts & Professionals



I have removed my postings from this thread.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 13, 2014, 05:43:06 pm
although the discussion gets now rather off topic:

I have removed my postings from this thread.
It´s your decision.

I mentioned this to suggest to you to spin off a thread with a appropriate title.
In my opinion nobody will seek for a flicker noise amplifier within a LM399 thread.

And also I think that a flicker noise amplifier is more complex than it looks on the first glance.
So it would be worth a own thread.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Andreas on July 13, 2014, 05:43:33 pm
Hello,

I have attached a updated ageing chart of the CH6 + CH7 LM399 devices.
If you remember this was the "slot or not" PCB of branadic.
The drift is now scaled in ppm.

CH6 with short legs and no slots
Ch7 with short legs and slots

Ageing drift now seems to stabilize more and more after 180 days on both devices.
CH6 seems to have more low frequency noise (standard deviation of drift).
CH7 seems to have larger ageing drift.

See also:
PCB
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg361683/#msg361683 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg361683/#msg361683)
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg402498/#msg402498 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg402498/#msg402498)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: MK on September 04, 2014, 08:40:06 pm
I intend to build 2 or 3 lm399 10 V references, what are the suggestions for the number of ppm span for a trimpot in the chain. I expect to pre-age them then select of pad the volt setting resistors and to have some low value resistors of various values in the chain with 1-2 ohms with a 100 ohm pot in parallel for setting to final value periodically. If i make the tim range too small I will have to cut or solder too often, if I make the trim range too large hysteresis in setting the value will limit the stability? what are the thoughts about a best span to choose?
Title: Re: LM399 based 10 V reference
Post by: Andreas on September 04, 2014, 09:17:09 pm
Hello,

after proper pre-aging of the LM399 the further drift is in the range of about 1-2ppm/year for the LM399.
So the trimming range is mainly dependant on the drift of your voltage divider from 7 to 10 V.
I would use a trimming scheme similar to the "Standard Cell Replacement" cirquit on page 5 of the datasheet.

http://www.linear.com/docs/3317 (http://www.linear.com/docs/3317)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Vgkid on September 05, 2014, 01:34:30 am
If you need more stability in the trim range, check out the trimming methods discussed in an42.
Title: Re: LM399 based 10 V reference
Post by: MK on September 06, 2014, 08:11:29 am
I decided to use the fluke 732A timming technique in the end.
Title: Re: LM399 based 10 V reference
Post by: TiN on November 06, 2014, 04:21:34 am
Revive the dead....

(http://dev.xdevs.com/projects/kx/repository/revisions/b48a270f148bbeea16e2e96c06f2f958fae90a5c/entry/img/vref_s00/vref_s00.png)

Dual LM399 + 7V>10V/20V ratio using LTC2057HV.
BOM is to be much lower cost than my LTZ version :)

Resistors - PTF56 25ppm/C
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 06, 2014, 07:04:27 pm
Schematic is hidden here:

https://www.eevblog.com/forum/projects/project-kx-diy-calibrator-reference-sourcemeter/msg544604/#msg544604 (https://www.eevblog.com/forum/projects/project-kx-diy-calibrator-reference-sourcemeter/msg544604/#msg544604)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: MK on November 07, 2014, 09:28:52 am
After my conversation with Bob Dobkin, my take-away was that there is nothing wrong with building a voltage reference around some paralleled LM399A's [or even just one of them really].  The more you parallel, the lower the DC-10Hz noise will be.  Because they run so hot, hysteresis is not much of an issue, and after a long burn-in period they are only going to drift 1-2ppm/year [if left on 24/7], which is really not that bad.  Additionally, there is an opportunity to build a voltage reference that has no batteries [or at least the batteries don't need to hold the reference circuit up for more than a few hours]-- Bob said that these devices don't drift if they are off.  So, you could build a voltage reference that you leave off most of the time, and then only turn it on an hour or so before you are going to use it [or calibrate it].  In this use case, the voltage reference could hold it's calibration within 1ppm of the SI volt for many years, which is important for hobbyists, because calibrations can be rather expensive [especially if they are using a JJA].

My burn-in recommendation for the LM399(A) is fairly radical-- it would involve building a very well insulated oven that was controlled at 150C [+/- 5C].  The LM399(A)'s are placed in this oven for ONE YEAR [and now you can see the need for excellent insulation!]  This will provide "artificial aging" that approximates about 67 years of natural aging, and could result in references that drift less than 1ppm/year, even if you left them on all the time-- and much less if you only turn them on when you need them.  To "soften" the die-attach [which can cause sudden jumps of 1ppm or so-- like a "pop"], the references are placed in a live circuit [no need for accurate resistors in this circuit], and then that is placed in a freezer, and the power is cycled on/off [one minute on 1 minute off] for 90-days.  This way, if there are any bubbles or micro-cracks in the die bonding material, these will be "worked out" over this period.  For obvious reasons, it is more economical to use this burn-in procedure on a hundred or more references at a time-- and because it takes so long, one has to have great patience.

Bob Dobkin said that you should parallel at least 6 of the LM399(A)'s, but I think 4 of them would be sufficient, and this is still less cost than an LTZ1000(A) based reference.  You get a sqrt(N) reduction of noise, so this would economically reduce the noise by a factor of 2.  The next level would be 9 devices for a noise reduction factor of 3, and then 16 of them for a noise reduction factor of 4.  9 devices would exceed the cost of an LTZ1000(A) reference, so 4 LM399(A)'s is about the economic limit for this technique.

Bob also said that the current limiting resistor for the LM399(A)'s Zener should be tied to the stable 10V output.  This causes a start-up problem, but in the two threads I think this issue has been solved in different ways-- and all of them should work.

In the LTZ1000 thread, you can see one of my early designs of a PWM circuit for the 7V-to-10V boost circuit.  Since then, I have refined the technique and it is very much simplified [using only one 32-bit PWM stage, and one filter].  No critical resistor ratios or absolute resistor values are needed, and thus there will be no resistor-related drift [at least in the boost circuit].  I have some more work to do on the digital side of things, and of course I need to build and test a statistically large enough population of these references-- but once I have done that successfully I can post my findings.  Since the boost circuit could apply to either an LM399(A) based or an LTZ1000(A) based reference, I will probably start a thread just for that and link to the two voltage reference threads.
A good answer to the starting issue is to use a fet with a significant gate threshold, thus the circuit is already starting to receive zener current even when the output is pegged to low. this also reduces the current demand and the heat issue in the opamp that is required to have low drift and shifts in output.

Title: Re: LM399 based 10 V reference
Post by: Andreas on November 24, 2014, 10:20:26 pm
Hello,

Just another updated ageing chart of the CH6 + CH7 LM399 devices.
If you remember this was the "slot or not" PCB of branadic.
The drift is now scaled in ppm.

CH6 with short legs and no slots
Ch7 with short legs and slots

Ageing drift is now after nearly 11 months on both devices.
still the same:
CH6 seems to have more low frequency noise (standard deviation of drift).
CH7 seems to have larger ageing drift.

See also:

https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg361683/#msg361683 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg361683/#msg361683)
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg402498/#msg402498 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg402498/#msg402498)
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg478496/#msg478496 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg478496/#msg478496)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: metacollin on December 11, 2014, 05:34:53 am
It looks like you've managed to collect data for nearly (or in reality?) every day of a year.  Fantastic stuff.  I just wanted to say, I for one definitely appreciate the effort you've made But the results are turning out to be pretty interesting so far!   Great work and dedication!  It's helped me make a decision for my own design, using quantitive real-world measurements, and that always gives an engineer the warm fuzzies ;). 

For us data nerds, I was wondering if you might be willing to share the raw data now or at some later stage (in any format of your choosing really)?  There is always some interesting analyses that could be run when this much quality data. Only if it's not a hassle or problem though, if it is, well, you've already donated tons of effort, so don't feel obligated.  And I'm asking entirely so I can nerd out with it using matplotlib for fun, nothing worth you going through trouble.  But you have my gratitude if you do chose to release it.  Anyway, thank you!
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 11, 2014, 12:41:46 pm
Hello,

Just another updated ageing chart of the CH6 + CH7 LM399 devices.
If you remember this was the "slot or not" PCB of branadic.
The drift is now scaled in ppm.


...

With best regards

Andreas


Hello Andreas,

I went through the whole thread again, but could not find any description, how your monitoring system works, and on which volt reference it is based.
Therefore, I could not draw any conclusion, which drift you really display here, i.e. really the drift of the DUT, or the drift of the monitoring system.

Would you mind explaining your practical setup, and how you are able to determine the absolute drift of the LM399s?

Thanks

Frank
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 11, 2014, 10:54:35 pm
I went through the whole thread again, but could not find any description, how your monitoring system works, and on which volt reference it is based.
Therefore, I could not draw any conclusion, which drift you really display here, i.e. really the drift of the DUT, or the drift of the monitoring system.

Would you mind explaining your practical setup, and how you are able to determine the absolute drift of the LM399s?

Thanks

Frank

Hello Frank, Ken,

the LM399 CH6 + CH7 charts are simple measurements of raw data (without offset compensation) out of my best 24 bit ADC (ADC#13).
Offset of this device is about 6uV which drifts around 1uV over the year. (negligible seasonal change).

ADC#13 is a well aged LTC2400 based device with a AD586LQ voltage reference and a temperature sensor.
T.C. is compensated by a 3rd order correction curve.

So this measurement is relative to ADC#13.

ADC#13 stability can be found by the following charts of my "daily" automated measurements.
These attached charts are done with offset compensation.
Offset is measured once at beginning of the measurement and subtracted from the following readings.

If I set day 0 at the beginning of LM399 CH6 + CH7 measurements you can see that during same time
ADC13 does not drift more than about +/- 1 ppm against two LTZ1000A references (blue and green)
which is also negligible against the 10ppm of the LM399 CH7 reference.

For the LTZ1000A references I try to get reliable calibrations.
But the only thing that I can say from  comparisons to other instruments like Keithley 2000 with calibration history
or Fluke 5520A is that the drift of the LTZ devices is up to 2 ppm/year against those devices.
But I cannot tell wether the calibrator drifts or the LTZ1000A.
I will still need some years to make a final decision.

With best regards

Andreas




Title: Re: LM399 based 10 V reference
Post by: TiN on December 12, 2014, 12:03:58 am
Andreas,

I am also interested how is switching done between different channels?
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 12, 2014, 08:54:45 am

Hello Frank, Ken,

...
ADC#13 is a well aged LTC2400 based device with a AD586LQ voltage reference and a temperature sensor.
T.C. is compensated by a 3rd order correction curve.

..
ADC#13 stability can be found by the following charts of my "daily" automated measurements.
These attached charts are done with offset compensation.
Offset is measured once at beginning of the measurement and subtracted from the following readings.

If I set day 0 at the beginning of LM399 CH6 + CH7 measurements you can see that during same time
ADC13 does not drift more than about +/- 1 ppm against two LTZ1000A references (blue and green)
which is also negligible against the 10ppm of the LM399 CH7 reference.

For the LTZ1000A references I try to get reliable calibrations.
But the only thing that I can say from  comparisons to other instruments like Keithley 2000 with calibration history
or Fluke 5520A is that the drift of the LTZ devices is up to 2 ppm/year against those devices.
But I cannot tell wether the calibrator drifts or the LTZ1000A.
I will still need some years to make a final decision.

With best regards

Andreas

Hello Andreas,

although the AD586 is quite mediocre concerning TC and ageing, your setup with TC compensation and comparison (frequent calibration?) against your LTZ1000As gives good stability / uncertainty for your ADC#13, I think.

The Keithley 2000 probably is based on a LM399, and the Fluke 5520 may be based on a single SZA263 or LTFLU, see T.C. and 1 year uncertainty specifications, latter are on the order of 11 .. 30ppm/year.
Therefore it is obvious, that your DIY LTZ1000A very probably are much more stable than the Keithley and Fluke instruments.
The LTZs at <= 50°C, really are on the order of <1ppm/yr, as your 1yr. comparison between these both LTZ#1, LTZ#2, and your stable LM399s indicate.

To definitely decide about the stability of your LTZs, also within 1 year or less, you are simply lacking more references of the same stability grade. A number of >=4 in total would be better.

A circular comparison, on a regular basis, with similar references of other volt-nuts would also do the job.

What is needed, is a traveling LTZ standard, and a group of volt-nuts willing to do the comparison.

What do you think about that?

Frank
Title: Re: LM399 based 10 V reference
Post by: babysitter on December 12, 2014, 09:27:04 am
Offering my travelling LTZ for this kind of project !
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 12, 2014, 08:28:11 pm

Hello Andreas,

although the AD586 is quite mediocre concerning TC and ageing, your setup with TC compensation and comparison (frequent calibration?) against your LTZ1000As gives good stability / uncertainty for your ADC#13, I think.
.....

What do you think about that?

Frank


I am also interested how is switching done between different channels?

Hello Frank, Babysitter,

together we have 5+1 references that would give a nice calibration party in Franks lab.
Wouldnt it?

How about end of next week? Thursday/Friday?

By the way: for ADC13 I am doing onlydaily measurement of offset and subtract it usually from the measurement results.
Tempco correction coefficients have been calibrated only once after having built the device.
The only thing that I have done is recalibrating once the reference voltage at 25 deg after the initial drift phase.

@TiN:

I am using a simple battery supplied relay multiplexer with 7 ground referenced inputs and two independent outputs.
Every output can be either connected to ground or to one of the 7 inputs.
So its possible to measure input voltage or input voltage differences between 2 references.

The relays are bistable signal relays (TQ2 with bifurcated contacts) so there is no heating during normal operation.
Control is done by a isolated RS232 connection+a microprocessor.
The 9V block lasts for about one week of operation without recharging.

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: wiss on December 13, 2014, 11:10:22 am

What is needed, is a traveling LTZ standard, and a group of volt-nuts willing to do the comparison.

What do you think about that?

Frank

Where in BRD are you guys located? I have some vague memory of Frank in Frankfurt am Main?

I just built a few LM399-based ref-boards where the 10V seems to be with in 10 ppm, but the Zener-voltage should be as stable as the LM399 can be (using high-Z voltmeter).

Next Saturday I will go by car from Puttgarden to Zwickau..
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 13, 2014, 02:38:58 pm
Hello everybody!

That happening then would be called Traveling Standards & Poor Engineers, I think. ;D ;D

Well, I'm available next Friday evening and also on Saturday, so everybody is welcome.
Maybe we find a common target date, here.

Or send me a PM, when you'd like to show up.

Frank
Title: Re: LM399 based 10 V reference
Post by: TiN on December 13, 2014, 02:56:03 pm
Thanks for details. I was thinking about making DAQ system to logging as well.
So far decided to stop with KI2002 + 2001-TCSCAN card (which i happen to snag long ago).

As for VREF party, I wish to participate as well. I think it's worth to create separate thread and make it bigger? :)
I am about to order PCB revision with fixes for my LTZ reference, and dual LM399 version with both direct and 10V/20V outputs.
I think i will have all hardware ready to go at new year. I could ship 2-3 boards each somewhere, and you guys have party and measure it, and return refs back after.
As commitment , I will send free bare boards as well.

On my side, here in Taipei, I can only measure stuff with calibrated K2001 (cal Feb/14 by local Tek official service) and K2400 (same time cal) and 2002 (cal in 2007).
Going to calibrate my gear again in few month, as already have set of VPG HZ resistors for Keithley calibration specs, but I would like to "import" proper voltage standard into my home first. I think 3 LTZ boards in hermetic box shipped over would serve this job well.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 04, 2015, 08:24:39 pm
Day 365 of 2 LM399 ageing (see above)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 04, 2015, 09:02:30 pm
Well done, well recorded  :-+
Title: Re: LM399 based 10 V reference
Post by: DLWarr on January 05, 2015, 01:56:13 pm
I am just starting to read this "LM399 based 10V Reference" Subject,,,, am on pg 7,, But am wondering what spice model anyone might be using for the zener.. Although I use Proteus,,, I don't see anything in LTSpice on the LM399.... You all have done such a good job describing the device's in's and outs... Never seen so much activity for a 4 terminal device... Any response on this will be most appreciated..
Title: Re: LM399 based 10 V reference
Post by: rob77 on January 05, 2015, 02:03:51 pm
I am just starting to read this "LM399 based 10V Reference" Subject,,,, am on pg 7,, But am wondering what spice model anyone might be using for the zener.. Although I use Proteus,,, I don't see anything in LTSpice on the LM399.... You all have done such a good job describing the device's in's and outs... Never seen so much activity for a 4 terminal device... Any response on this will be most appreciated..

it's just a resistive heating element a diode and a zener inside - you can simulate it using regular resistor, diode and zener... the reason why people using LM399 is the extraordinary stability of that buried zener (and you can't simulate that anyway...)
Title: Re: LM399 based 10 V reference
Post by: MK on January 05, 2015, 06:48:24 pm
The simplified schematic for the circuit is provided in the National documentation, but for modelling the general behavior of the supporting circuitry then a simple existing zener in your package of choice is most likely close enough. Remember that the "Zener" in the 399 is an active circuit and that the noise does not go down for increasing current, so approx 1mA is enough drive, that is useful if you want it battery powered to reduce ground loop issues.
Title: Re: LM399 based 10 V reference
Post by: DLWarr on January 06, 2015, 03:16:21 pm
Wonderful then,,,,, Any thoughts on using a the "Portable Calibrator" circuit WITH   " Walt Jung's, Analog Devices, Build An Ultra-Low-Noise Voltage Reference... Electronic Design 6/24/93"....  Of course substituting both op-amps with the LTC2057...  Like everyone in the world, I'm looking for low drift,, low noise...   all in a 3 terminal package...aaahhhhhhhh. Any thoughts will be appreciated, thanks...
Title: Re: LM399 based 10 V reference
Post by: rf-design on January 06, 2015, 04:36:54 pm
I did not understand how the flicker or 1/f noise of the reference is measured. I could understand that a chopper amp could reduce the impact of the needed voltage amplification. But the amp could only amplify a voltage difference of two potentials or one potential to a ground reference point. But the reference is the DUT itself have an offset of 7V against the chopper amplifier. So there should be second reference which significant lower or well characterized flicker noise to counter the DUT reference voltage. The difference should also be smaller than the requested gain of the measurement.

So what kind of offset source is used?
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 06, 2015, 11:15:34 pm
Wonderful then,,,,, Any thoughts on using a the "Portable Calibrator" circuit WITH   " Walt Jung's, Analog Devices, Build An Ultra-Low-Noise Voltage Reference... Electronic Design 6/24/93"....  Of course substituting both op-amps with the LTC2057...  Like everyone in the world, I'm looking for low drift,, low noise...   all in a 3 terminal package...aaahhhhhhhh. Any thoughts will be appreciated, thanks...

The EDN cirquit has a edge frequency of 1.6 Hz. (time constant 0.1 s).
So the filtering is only for wideband noise (> 10 Hz).

Since your integrator filters that noise already a extra filter in this frequency range is rather useless.
For precision measurements your integration times are usually minimum 2 seconds.

I did not understand how the flicker or 1/f noise of the reference is measured. I could understand that a chopper amp could reduce the impact of the needed voltage amplification. But the amp could only amplify a voltage difference of two potentials or one potential to a ground reference point. But the reference is the DUT itself have an offset of 7V against the chopper amplifier. So there should be second reference which significant lower or well characterized flicker noise to counter the DUT reference voltage. The difference should also be smaller than the requested gain of the measurement.

So what kind of offset source is used?

I use a large (3200 uF) electrolytic capacitor selected for low leakage current.
Noise floor is <0.2 uV together with a 1K input impedance and a LT1037 OP-Amp
in the first amplifier stage.
See also AN124 of Linear Technology.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: babysitter on January 07, 2015, 06:13:43 am
Congratulations for 1 year of logging, Andreas!
Title: Re: LM399 based 10 V reference
Post by: rf-design on January 07, 2015, 08:30:11 am
Quote
I did not understand how the flicker or 1/f noise of the reference is measured. I could understand that a chopper amp could reduce the impact of the needed voltage amplification. But the amp could only amplify a voltage difference of two potentials or one potential to a ground reference point. But the reference is the DUT itself have an offset of 7V against the chopper amplifier. So there should be second reference which significant lower or well characterized flicker noise to counter the DUT reference voltage. The difference should also be smaller than the requested gain of the measurement.

So what kind of offset source is used?

I use a large (3200 uF) electrolytic capacitor selected for low leakage current.
Noise floor is <0.2 uV together with a 1K input impedance and a LT1037 OP-Amp
in the first amplifier stage.
See also AN124 of Linear Technology.

With best regards

Andreas

Andreas, thanks for the AN. So the measurement is limted to 0.1Hz. I thought to oberserve the total drift spectrum. So only the integration time of the voltage logging sets the sample rate and all further effects could be observed.

I have further doubts that the dielectric absorption set the lower limit above the reference. The 24h settling could mean that the interface traps very similar to the source of high 1/f noise in MOS gates set again a limit on the 1/f noise measurement floor which is otherwise rejected by the hybrid amplifier (LT1012, Q1+Q2, LT1097). The reported floor was achieved with a $400 wet slug tantal.

To me an alternative is to measure the low 1/f noise amplified voltage difference between pairs of the same type of reference. With 10 references you have 45 pairs of drift measurements which are not limited by 0.1Hz. From the pairs you can calculate the indidividual noise spectra.

BR
Reiner
Title: Re: LM399 based 10 V reference
Post by: wiss on January 07, 2015, 08:52:57 am
My home-lab spent some time over Christmas measuring delta-voltages, every 16 minutes 4 measurements are taken between 3 references:

https://www.eevblog.com/forum/projects/yet-another-%28lm399%29-volt-reference/15/ (https://www.eevblog.com/forum/projects/yet-another-%28lm399%29-volt-reference/15/)
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 07, 2015, 08:29:00 pm

I have further doubts that the dielectric absorption set the lower limit above the reference.

From the pairs you can calculate the indidividual noise spectra.


Dielectric absorption (DA) is only a problem for measuring leakage current.
On AC-Signals a 0.1% error of the noise voltage due to DA should be no problem.

Leakage current itself contributes to the noise floor.
Of course it might be a good idea to let the input capacitor always under power to minimize leakage.

I doubt that one can calculate individual noise from pairs.
I'd rather average >9 references to get a lower noise voltage (sqrt(number)) and measure the D.U.T. against the average.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: rf-design on January 08, 2015, 01:33:53 pm
I doubt that one can calculate individual noise from pairs.

Indeed you can :) Click :

www.intersil.com/data/an/an177.pdf (http://www.intersil.com/data/an/an177.pdf)

The calculation is not difficult!

The measurement is a pairwise sample DUT noise density over frequency. Because the assumption is that the samples are statistical independend the noise power at sa specific frequency is simply the addition of the noise power of the DUT pair.

If you have n=4 samples you can make n*(n-1)/2=6 pairwise measurements. I will call the reference sample spectral noise power at a specific frequency simply a,b,c,d.

What you measure with the pairs is the sum of the noise powers

a+b
a+c
a+d
b+c
b+d
c+d

You get the noise power of the first sample reference by the following expression:

a=(1/3)*(((a+b)+(a+c)+(a+d))-(1/2)*((b+c)+(b+d)+(c+d)))

a=(1/3)*((3*a+b+c+d)-(1/2)*(2*b+2*c+2*d))

a=(1/3)*((3*a+b+c+d)-(b+c+d))

a=(1/3)*(3*a)

a=a

For the general case n the expression is:

a=1/(n-1)*(sumwith(a)-1/(n-2)*sumnot(a))



The application note from LT does not give a calculation example. It is only stated that you have to use a much better reference as a DUT partner.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 08, 2015, 07:36:57 pm
What you measure with the pairs is the sum of the noise powers

Hello,

what do you want to measure?
wideband noise > 10 Hz or
1/f noise 0.1 .. 10 Hz or
drift (below 0.1 Hz).

For wideband noise you are right: the power (effective value) adds.
The voltage adds only geometrically. (square root of the squared sum).

For 1/f noise you usually can only measure the peak-peak voltage value.
And if you really have done such measurements you now that
from measurement to measurement you have a lot of variation which
makes it practically not possible to measure differences.

Except when one of the 2 references has much lower noise.
-> my suggestion to average at least 9 references as
    "low noise reference" to have less than 10% error for the DUT.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: rf-design on January 09, 2015, 10:54:53 am
What you measure with the pairs is the sum of the noise powers

Hello,

what do you want to measure?
wideband noise > 10 Hz or
1/f noise 0.1 .. 10 Hz or
drift (below 0.1 Hz).

For wideband noise you are right: the power (effective value) adds.
The voltage adds only geometrically. (square root of the squared sum).

For 1/f noise you usually can only measure the peak-peak voltage value.
And if you really have done such measurements you now that
from measurement to measurement you have a lot of variation which
makes it practically not possible to measure differences.

Except when one of the 2 references has much lower noise.
-> my suggestion to average at least 9 references as
    "low noise reference" to have less than 10% error for the DUT.

With best regards

Andreas

Andreas,

my thought about the reference noise problem was to avoid in general a requirement to have better reference to characterize or measure other references. I know that similar principles are also used for phase-noise measurements at levels where the best economical sources are less than 10dB away. As the amplifiers could be build with 1/f noise and drift much less than the expected reference noise and drift a differential measurement seems to be an easy solution.

Because for 10 references you have 45 pairs you could not measure all pairs at the same time the result will be dependend on the statistical nature of the drift and noise process. So it is assumed that the noise spectrum, which is hole picture of short drift and 1/f noise, is stationary. So otherwise the calculation of the last post is not valid.

The measurement sample could be every 100ms over 2^14=16384 samples. So a pair measurement need less than 30min. I know that it is difficult to interpret a noise spectrum extending down to 306uHz. Otherwise if the noise spectrum follow a 1/f rule the noise power integrated over a frequency range with a fixed factor of the upper limit to the lower limit will give an equal noise power. So there is no further gain in noise reduction to measure longer. If the spectrum follow (1/f)^k with k<1 than gain is small. Finally the 1/f or short time stabilty set for most systems a lower limit. And I know that it is difficult to differentiate that from an againg which typical goes only in one direction.

BR
Reiner
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 13, 2015, 07:27:56 pm
Hi,

Can i also play?  ;)

This is a portable reference i am designing, is uses a LM399.
I've done my best to keep the noise low and the current to the heater as low as possible.
Therefore, I use three 9V batteries, it gives me a lower heater current and longer live of the batteries.
Only 2 euros each, for alkaline here in the Netherlands.

The corner of the filter is about 1 a 1,5 Hz, this is low enough to filter out most of the noise.
If you are calibrating a digital multimeter, choose the longest integration time, 1 to 2 seconds.
There are also elements for the protection of the circuit on the output circuit.

I now own a calibrated Agilent 3458a, later i wil show a short term plot of the portable reference.
(http://www.bramcam.nl/NA/NA-Porta-Cal/NA-Porta-Cal-01.png)
Shoot at it!

Kind regarts,
Blackdog
Title: Re: LM399 based 10 V reference
Post by: MK on January 13, 2015, 08:31:44 pm
Hi Blackdog, with a rai to rail output of a cmos amp and with a FET for Q1, it should self-start without the 180K R1 resistor and the output should  be slightly more stable too.

regards.
Title: Re: LM399 based 10 V reference
Post by: wiss on January 13, 2015, 09:01:34 pm
You should definitely remove R1 (me without an hard empirical proof).
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 13, 2015, 09:16:03 pm
Hi,

I know that R1 should be removed, but i need to do some test.
Today, i received the last components, and from other references i had build using the same schematic setup
R1 was not necessary, we will see wats come out of it...
Thanks for the remarks.

A little test print, with the LM399, LM368 10V reference and a 2K Rhopoint 8E16 who is putting the current in de reference
drift within 2PPM with long cable's to a Tektronix DMM4050 Multimeter, not bad :-)

Kind regarts,
Blackdog
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 18, 2015, 09:21:04 am
LM399H with the external oven :D
- Noise (0.1-10 Hz) ~ 2.6 uV p-p;
- TempCo ~ 0.25 ppm/C;
- T(oven) = 52 C;
- Long-term stability is unknown.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 18, 2015, 03:36:43 pm
LM399H with the external oven :D
- TempCo ~ 0.25 ppm/C;

a) Is the tempco including transformation to 10V?
b) what is the tempco of the zener alone (unheated). (Did you select for tempco?)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 18, 2015, 03:40:45 pm
1. There is no transformation to 10 V.
2. Zener have a broken heater and have a huge TC ~ 70 ppm/C.
Title: Re: LM399 based 10 V reference
Post by: ManateeMafia on January 18, 2015, 03:55:25 pm
Mickle,

This is the first time I have seen that voltmeter in your posts. Is this another one of your neat projects?
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 18, 2015, 04:05:50 pm
Yes it is. This is a simple homemade 8.5-digits DVM/voltage comparator.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 18, 2015, 04:17:24 pm
Hello Mickle,

I think it´s time for a teardown.
Isnt it?

Which ADC
which Reference
which processor.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: wiss on January 18, 2015, 04:45:12 pm
Yup!!!  :-/O
Title: Re: LM399 based 10 V reference
Post by: Vgkid on January 18, 2015, 04:49:01 pm
I agree with the above statements.
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 18, 2015, 05:01:00 pm
Quote
Which ADC
which Reference
which processor.
ADC - Mark-space type with convergence accelerating signal.
Reference - external 7-14 V.
Inguard CPU - Atmega8.
Outguard CPU - AMD Geode 300 MHz.
Some photos is here: http://bbs.38hot.net/forum.php?mod=viewthread&tid=102351&extra=page%3D1&page=3 (http://bbs.38hot.net/forum.php?mod=viewthread&tid=102351&extra=page%3D1&page=3)
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 18, 2015, 07:59:11 pm
Hello Mickle,

nice project.
What do you mean with "mark-space" type?
I guess it is some kind of first order sigma delta modulator?

On the pictures the reading changes only by 1 count (0.1uV).
Which would be better than on a HP3458a.
But how much do the values really change?
Do you have a standard deviation figure for the readings?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: wiss on January 18, 2015, 08:08:25 pm
Mickle, could you plot the DVM measuring it`s own ref for half an hour or so? (and post here)
Youtube video would be ok if no computer interface  >:D
Title: Re: LM399 based 10 V reference
Post by: babysitter on January 18, 2015, 08:47:24 pm
My mother today emailed me that she considers the LM399 thermography movie by Andreas and branadic "art" :)
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 18, 2015, 09:09:18 pm
Hello Andreas,
"Mark-space ADC" is a term proposed by Solartron (Pat. US3942172A, US4340883 e.t.c.). I have used the structure of the ADC, the average between Solartron DMMs and Valhalla 2720GS with some improvements. Noise is much more than 0.1 uV  :)
This project is suspended now. I will continue to deal with them, since I can afford to buy an LTZ and make the internal reference.

Mickle.
Title: Re: LM399 based 10 V reference
Post by: wiss on January 18, 2015, 09:15:02 pm
That looks very much like my 7075! Total noise about +- 2 uV at 14 V range, no averaging.
Title: Re: LM399 based 10 V reference
Post by: MK on January 19, 2015, 07:16:17 am
Well done Mickle, that  is quite an acchievment. I went to the 38hot link you gave, unfortunately no images shown for me in the UK? I am impressed that you can read and write chinese as well.
Title: Re: LM399 based 10 V reference
Post by: quarks on January 19, 2015, 07:54:45 am
MickleT, great work as always. Please keep sharing.
Many thanks
quarks
Title: Re: LM399 based 10 V reference
Post by: macfly on January 19, 2015, 05:07:31 pm
Hi Mickle,

I agree with the other knowledge-hungry guys: please show us your great work in detail.   :clap:

Best regards,

macfly
Title: Re: LM399 based 10 V reference
Post by: branadic on January 20, 2015, 08:48:23 am
My mother today emailed me that she considers the LM399 thermography movie by Andreas and branadic "art" :)

Why art?
Title: Re: LM399 based 10 V reference
Post by: janaf on February 10, 2015, 04:15:17 pm
FYI there are LM399 (NS) on the bay at $4.60. I bought some a few months ago and they where OK, even if I have not done any deeeeep measurements on them. Seller Polida2008. I have no affiliation whatsoever.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 24, 2015, 06:52:21 pm
Galaxyrise posted a link to AN-161 (National Semiconductor / TI), which I found very interesting and didn't know before. We had a discussion about lead length in this thread. Here is what the AppNote says:

"...Thermocouple effects can also use errors. The kovar leads from the LM199 package from a thermocouple with copper printed circuit board traces. Since the package of the 199 is heated, there is a heat flow along the leads of the LM199 package. If the leads terminate into unequal sizes of copper on the p.c. board greater heat will be absorbed by the larger copper trace and a temperature difference will develop. A temperature difference of 1°C between the two leads of the reference will generate about 30 ?A. Therefore, the copper traces to the zener should be equal in size. This will generally keep the errors due to thermocouple effects under about 15 ?V.
The LM199 should be mounted flush on the p.c. board with a minimum of space between the thermal shield and the boards. This minimizes air flow across the kovar leads on the board surface, which also can cause thermocouple voltages. Air currents across the leads usually appear as ultra-low frequency noise of about 10 ?V to 20 ?V amplitude..."

This confirms what I already stated here:

http://dg3hda.primeintrag.org/doku.php?id=lm399_thermographie (http://dg3hda.primeintrag.org/doku.php?id=lm399_thermographie)

Keep lead lengths short.
Title: Re: LM399 based 10 V reference
Post by: Vgkid on February 24, 2015, 08:06:49 pm
What if you did the opposite, keeping the leads at the longest, but using a styrofoam sheet(a thick one, going the whole length) as a spacer.
Title: Re: LM399 based 10 V reference
Post by: senso on February 24, 2015, 08:16:47 pm
FYI there are LM399 (NS) on the bay at $4.60. I bought some a few months ago and they where OK, even if I have not done any deeeeep measurements on them. Seller Polida2008. I have no affiliation whatsoever.

Also bought 4 LM399 from that sellers, still in the sealed bag he put them in, I need to make some small pcb's for them :palm:
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 24, 2015, 08:45:23 pm
A temperature difference of 1°C between the two leads of the reference will generate about 30 ?A.

Keep lead lengths short.


Ok, I have finally evaluated the measurements = .csv tables of Branadic from the LM399 thermograpic measurement.

Long leads without slot:   around 5 degres stray within pad  + 1,7 degrees from pad to pad
Long leads with slots:      around 5-6 degres stray within pad + 2,3 degrees from pad to pad
Short leads without slot: around 9-11 degres stray within pad + 3,1 degrees from pad to pad
short leads with slots:     around 12-15 degres stray within pad + 6,6 degrees from pad to pad

So from the measurement the conclusion is that with long leads the stray (=thermoelectric voltage) within pad is lower than with short legs. And the thermoelectric voltage between different pads is lower without slot.
If there should be mechanical issues the slots should be at least around 15 mm from the reference to give the pads the possibility to equalize the temperature.


On the thermography with short leads I have seen up to 5 deg C more difference within pad and 1,4 deg C more difference between the soldering pads
So what is better: having up to 5 K * 40uV/K = 200uV voltage difference within the soldering joint (or 52uV between 2 pads) or have longer leads which can be thermally isolated by other means? So I´m still not shure wether long (good thermally isolated) legs or short legs are really better.

For me only one thing is clear: thermal isolation of the legs and solder joints as good as can.
And no slots (thermal mass on the PCB) to equalize the temperature of the solder joints.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on March 01, 2015, 10:41:49 pm
Sadly, I blew up my best LM399.  (I'm still not quite sure why that experiment destroyed things, either.)

Of the references that survived that experiment, I took 5 with fairly similar voltages and wired them in parallel, averaging their voltages with 5ppm/C 1k resistors.  To supply the reference current, I used a LT1236 and some 5ppm/C 3k resistors.  I think all the support circuitry drifts are attenuated by at least 1000:1, but I used 5ppm stuff anyway ;)

Inspired by Mickel T's super-insulated reference and the Bob Dobkin comment to keep the heater current  down, I also put them in parallel thermally as shown in the attached pictures.  (I've replaced that proto board with a proper PCB, but didn't take another picture before taping it all back together.) With a 30V heater supply, the combination of 5 LM399 draw about 15.5mA heater current at around 15C ambient; 3.1mA per LM399, which is a decent reduction in heater current, I think.

I monitored the setup overnight along with ambient temperature changes to get a rough feel for the relationship between heater current and ambient temperature: about 150uA per C.  I then put the foam cube inside a cooler and monitored the results.  At its peak, the heater current was 20.25mA and the voltage shift looked to be around -26uV; a fine result, I think, and suggests I'm getting better than the nominal 0.5ppm/C tempco.

I want to do the 7V to 10V conversion with LTC1043s, but that involves much faster rise times than I've designed for yet.
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 02, 2015, 07:17:25 am
Hello,

interesting results.
Which output voltage do the references have?
(I have always to think of the "sweet spot voltage" between unheated and heated device.)

It is not unusual that the LM399 near room temperature
is much better than over the whole temperature range.

My LM399#2 (National Semiconductor with 6860 mV output)
has around 9uV shift over 10 deg C to 45 deg C.
A part of it (the "hysteresis") are obviously thermocouple induced voltage shifts.

To your sudden die of the reference:
It is winter. Perhaps you have zapped it by ESD.

This is one reason why I place a 100nF capacitor (SMD) directly on the
pads over the zener output and over the heater input.
Besides ESD it also rejects some mains line noise for the case
that I use a linear power supply instead of batteries.

With best regards

Andreas



Title: Re: LM399 based 10 V reference
Post by: paulie on March 02, 2015, 10:29:04 am
FYI there are LM399 (NS) on the bay at $4.60. I bought some a few months ago and they where OK, even if I have not done any deeeeep measurements on them. Seller Polida2008.

I was worried because those were not available a week or two back but you were right about him relisting after the chinese holiday. Since these were available for slightly less than the cost of my meter (self imposed budget for that project) I ordered one. Two friends also got in to take advantage of his combined shipping so cost was even less. Your lead was greatly appreciated. Everybody else wanted $10-$15. Some even charged 2x that or even more than the cost of an LTZ1000. Typical Ebay tricks.
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on March 02, 2015, 05:42:51 pm
To your sudden die of the reference:
It is winter. Perhaps you have zapped it by ESD.
For the charts I produced awhile back in this thread, I had the 10 LM399 in series driven by a 1mA current source, so they'd all have the same current during measurements.  I had my DMM switching between them via switching card.  I ran like this for days.

I tried to improve the resolution of this measurement by using another 7V reference on a different supply.  So the switching card was now moving the ground point of that extra reference and the DMM was measuring the difference between each LM399 and the 7V reference.  After some minutes of this, the first LM399 in the chain would fail short.  By the time I caught it, three were dead. 

There was definitely LM399-killing potential in the circuit, but I'm pretty fuzzy on how it was getting carried over from the last to the first LM399.  I wonder if the capacitors you suggest would have prevented the damage.

(I have always to think of the "sweet spot voltage" between unheated and heated device.)
The resulting average is 7.0689.  None of the units I got were under 7V. I didn't record normal operating voltage on them individually, but I believe they ranged from 7.05V to 7.08V. 
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 02, 2015, 09:06:43 pm

I tried to improve the resolution of this measurement by using another 7V reference on a different supply.


What kind of supply?
- Battery?
- linear supply
- switch mode supply

Especially the switch mode supplies partly have rather large capacities (around several nF) between mains line and output. These could have forward biased the internal diode (do not forward bias above 0.1V) within the LM399.
And also linear supplies may have some 10 to 100 pF (parasytic) capacities.
10 Zeners in series (70V) would also violate the max 40V spec of the internal diode if the heater is supplied from a common source.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on March 02, 2015, 10:59:51 pm
Quote
10 Zeners in series (70V) would also violate the max 40V spec of the internal diode if the heater is supplied from a common source.
They weren't. I did try to take that 40V spec into account, at least, and the failures were closely correlated to the DMM scanning and not the static setup (which had been running for days before I did the scan.)

What kind of supply?
Both the LM399s and the 7V were on linear supplies. 

Quote
And also linear supplies may have some 10 to 100 pF (parasytic) capacities.
Some capacitance-related answer certainly seems likely.  Parasitic to earth ground, you mean?  Hmm... Perhaps I'll try to recreate it with resistors instead of LM399s and see if I can track it down.  Seems likely to be educational at the least!  I don't want to clutter up this thread with it, though; I'll put something in Beginners if I do play with it some more.  Thanks for your input!

Title: Re: LM399 based 10 V reference
Post by: TiN on April 03, 2015, 07:13:37 am
Time to play with LM399 (actually LM199 from one of 2001 boards).
To do: ratio units for conversion 7V -> 10V, 7V -> 20V.

(https://lh3.googleusercontent.com/qB9VUH-Avnczu3_S312N1JlYOMUsvRSsfsr65oulqmY=w368-h228-p-no)

Meanwhile during debugging: some values (http://xdevs.com/datalog/) pretty far from 10.00000VDC  :box:
Title: Re: LM399 based 10 V reference
Post by: Gyro on May 25, 2015, 07:22:19 pm
I Just wanted to share a photo of the LM399 (die marked 199A) die that I found on the web. I couldn't see it posted in the thread anywhere.

The die looks about as asymmetrical as it's possible to get with the heater down the right hand edge. Not sure but I assume the buried zener is the circular deep well in the center??? There's another one bottom right in the heater section (the chip schematic shows two in the heater).

Anyway, hopefully of interest....
Title: Re: LM399 based 10 V reference
Post by: rikkitikkitavi on May 28, 2015, 07:13:17 pm
Gents, and experts on LM399 ...

I have a couple of these desoldered from decommisioned equipment .

Unfortunately I only had eyes on the OP27, OPA627s, MAX420s etc on the boards, not the LM399s and therefore didnt focus on these little fellows.

A couple of them however fell of the boards easily when I used my hot air gun (1500 W) to remove the opamps. These later chips have all been tested satisfactory with regards to operation but I have no idea about actual temperature exposure but that they came of the board very easily (thin covar leads are bad thermal conductors compared to an 8 pin DIP leads) indicates comparatevly low thermal stress.

I would estimate that the  equipment in place would have been in operation for several years, all chips are datecoded early-mid 90ies.

In your opionion, if these LM399 are in working order at power applied, do you think they are "finished ageing to reach maturity" or did I start something bad by desoldering? Any opionion or guestimate is welcome, if not for learning something new...


And yes, I have already slapped myself...

regards,

Rickard
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 28, 2015, 08:25:15 pm
Hello,

my experience with references:
every stress to the package (including soldering) may (= most probably, or sometimes may not) start a new ageing cycle.

The chip and the mounting plate of the package (and the epoxy die attach) have different thermal expansion.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: plesa on July 02, 2015, 06:54:37 pm
Last week I finished one of TiN LM399 10V ref using Vishay UXB0204 thin film resistors and LM399 (non A) reference.
Current output voltage is slightly off, I'm waiting for Vishay 1280G pots.
I surprised how stable it is during burnin period (<1ppm). Zener current is set to 4.5 mA ( in datasheet it is set about 1mA )
Title: Re: LM399 based 10 V reference
Post by: TiN on July 03, 2015, 04:58:34 am
Wow, cool. You did more progress on this module, that I had  ;D
Would be happy to see it in good use with some measurements.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 03, 2015, 06:41:56 pm
Zener current is set to 4.5 mA ( in datasheet it is set about 1mA )

There is no advantage in setting the zener current above 1 mA.
(see statement of Ken).
Maximum stability of LM399 will be at 1 mA.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: plesa on July 03, 2015, 08:23:52 pm
Wow, cool. You did more progress on this module, that I had  ;D
Would be happy to see it in good use with some measurements.
Thanks, I do not think so, I also needs to spend more time on LTZ1000. For measurement I needs to finish the climate chamber and some MUX. I would like to assemble additional three PCB with LM399AHs.All with Vishay UXB resistors and same enclosure.

Zener current is set to 4.5 mA ( in datasheet it is set about 1mA )

There is no advantage in setting the zener current above 1 mA.
(see statement of Ken).
Maximum stability of LM399 will be at 1 mA.

With best regards

Andreas

It was my dilema. I read the response from Bod Dobkin about 1mA Zener current provided by Ken, but based on the noise comparison in LTZ1000 datasheet I decided to increase it at least for first try. It can be interesting to make noise measurement dependency on zener current like it is in LTZ1000 datasheet. So compare something like 10uA-> 20mA can confirm this. I know that I find reason to buy decent decade box sooner or later and this is the case ;-)
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on July 03, 2015, 10:44:07 pm
@Plesa #334,

You should be using either Bourns 3250 or 3290 wire wound pots (these are military qualified) if you want long term stability, film type pots are not nearly as stable over the long term (2-3 years compared to decades) and also require a minimum amount of current to flow through the wiper to keep it in 'good shape'.  The WW pots do not really have to have any current flowing in the wiper but 1uA-3uA is suggested.  The Vishay 1280G pots may have a bit better resolution but they are not nearly as stable long term and also cost a lot more than the WW pots.

The main drawback of the LM399/A references is noise compared to the LTZ references, about 4-5 times higher but at the same time they can be just as stable over time as the LTZs.  If your application does not need the lower noise level of the LTZ then the LM399/A is certainly the lower cost way to go.  As Bob Dobkin said, the higher current level does reduce noise but increases drift, there is no way around that tradeoff, if a little higher drift is okay then you are set with the higher current but noise is a very important spec and sets the minimum accuracy floor of the reference, you cannot filter the noise out from a zener so if you need low noise you must start with a low noise source.
Title: Re: LM399 based 10 V reference
Post by: plesa on July 05, 2015, 08:31:58 am
@Plesa #334,

You should be using either Bourns 3250 or 3290 wire wound pots (these are military qualified) if you want long term stability, film type pots are not nearly as stable over the long term (2-3 years compared to decades) and also require a minimum amount of current to flow through the wiper to keep it in 'good shape'.  The WW pots do not really have to have any current flowing in the wiper but 1uA-3uA is suggested.  The Vishay 1280G pots may have a bit better resolution but they are not nearly as stable long term and also cost a lot more than the WW pots.

The main drawback of the LM399/A references is noise compared to the LTZ references, about 4-5 times higher but at the same time they can be just as stable over time as the LTZs.  If your application does not need the lower noise level of the LTZ then the LM399/A is certainly the lower cost way to go.  As Bob Dobkin said, the higher current level does reduce noise but increases drift, there is no way around that tradeoff, if a little higher drift is okay then you are set with the higher current but noise is a very important spec and sets the minimum accuracy floor of the reference, you cannot filter the noise out from a zener so if you need low noise you must start with a low noise source.

Hi Edwin,

thanks!! I just ordered tube of 10k  3250 and we will see how they will work;-)
I'm going to build also the LTZ1000 reference to compare them.
So I will go down with Zener current to 1mA.

Title: Re: LM399 based 10 V reference
Post by: Andreas on July 05, 2015, 08:26:09 pm
I surprised how stable it is during burnin period (<1ppm).

Hello,

I am not surprised. At least after some run in time the LM399 is very stable.
See the LM399 CH6 + CH7 of branadics thermal test.
Picture below is now after 1.5 years.
Blue curve is my LTZ1000A#2 shown as comparison.

see also old recordings:
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg580443/#msg580443 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg580443/#msg580443)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: plesa on July 06, 2015, 08:22:14 pm
Thanks Andreas.

I put reference metal enclosure into two polystyrene insulation box and I will continue in measurement.
If someone will be interested in the atmospheric pressure dependency, it can be also added later on.
Hermetic chamber is almost ready.
Title: Re: LM399 based 10 V reference
Post by: F64098 on September 27, 2015, 07:41:30 am
Hello everybody,

what do you think about this offer:

http://www.ebay.de/itm/LM199AH-883-Manu-NS-ST-Package-CAN4-Voltage-Reference-/191046638531 (http://www.ebay.de/itm/LM199AH-883-Manu-NS-ST-Package-CAN4-Voltage-Reference-/191046638531) ?

Nice price, but when i look at the bottom and see those small rings of plastic at the outlets of the pins, it looks like a counterfeit, because i can't find similar looking items when searching the whole internet.
Your opinions?

So it may be better to buy this, slightly more expensive but pre-aged ones:

http://www.ebay.com/itm/1x-LM199AH-883-Precision-Reference-LM199-/111015694020 (http://www.ebay.com/itm/1x-LM199AH-883-Precision-Reference-LM199-/111015694020) ?

Best regards

Frank
Title: Re: LM399 based 10 V reference
Post by: plesa on September 27, 2015, 08:15:24 am
Sorry, but it is nonsense to buy references on ebay if you can buy them from manufacturer.
You do not know history and reason why it is on ebay and references on pictures does not look to be in best shape.
Buy several (10x) LM399AH from Linear for 9 USD/pc and operate them at 125°C for one month and select pair (2-6) of lowest tempco references in your circuit.
Or Linear is offering free samples for both versions of LM399, so you can have 4 pieces for free.

I have to investigate one fake LTZ1000 in different thread.
Title: Re: LM399 based 10 V reference
Post by: Andreas on September 27, 2015, 10:18:25 am

what do you think about this offer:

Your opinions?

So it may be better to buy this, slightly more expensive but pre-aged ones:

Price is much too high. (Compared to new LT parts from DigiKey).

Datecode 1209 on a NS-Reference?
I bought my last references from NS in 2008.
That was about the time when they were discontinued.
So in 2012 genuine NS parts are very strange.

If you solder the parts a new ageing cycle may begin.
So it is useless to buy pre aged parts which were already soldered many times.

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on September 27, 2015, 10:05:46 pm
Regarding the myth of surplus LTZ1000/A parts, according to Linear Tech, there are no surplus parts or overstocks, every one of them are accounted for, including the ones sold through their online shop.  LT keeps close track of the LTZs that they sell to the OEMs, they trade data back and forth, the OEMS do not, I repeat, do not have surplus or overstock to sell to anybody, they buy just what they need and nothing more.

Unfortunately, the condition of used parts removed from equipment is available before purchase, they may be in good shape but odds are they are not, de-soldering can cause changes or damage to the chip, therefore, if you are trying to achieve spec in an Vref, buy the LTZ1000/A from a reputable source and forget about the sellers on eBay or elsewhere, they are selling you a pig-in-a-poke.
Title: Re: LM399 based 10 V reference
Post by: ez24 on September 28, 2015, 03:17:59 am
fmaimon

This post is very long and old.  It is hard for me to understand what is going on.  Can you give a summary of what has happened over the years or start a new post on what you have learned.  (for new comers)

thanks
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on October 09, 2015, 09:03:20 am
DC-Cal Lite v1.0 - 10V/10mA/1mA calibrator, based on the AD5791B DAC.
0.1-10 Hz noise: 0V - 0.4 uV p-p, 10V - 4...5 uV p-p.
0-10 V INL <1 ppm (measured with Solatron 7081 DMMs).
Output resistance 0.003 Ohm (DC voltage mode).
TempCo ~ 0.1 ppm/C.
Title: Re: LM399 based 10 V reference
Post by: lukier on October 09, 2015, 09:57:16 am
Wow, that is a very nice piece of kit and seriously expensive DAC. Wouldn't it be cheaper to do PWM-DAC, like Fluke does in their calibrators?

What is the stuff in the front end in the 8th picture, in the white/yellow box. Some kind of electrometer sensitive circuit with teflon insulated wires? It's not some weird battery with balancing outputs, is it?
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on October 09, 2015, 10:17:22 am
White/yellow box is a 7 decades PWW voltage divider / resistor network from 1 Ohm to 9 MOhm.
For example: 0...1 uV rande on the N1A nanovoltmeter.
Title: Re: LM399 based 10 V reference
Post by: bingo600 on October 09, 2015, 12:04:42 pm
Mickle

Why do you have two trafos on the DC ref ?

One for analog, and one for digital ?

/Bingo

Title: Re: LM399 based 10 V reference
Post by: Mickle T. on October 09, 2015, 12:07:19 pm
Yes, it is. One for analog, and one for digital.
Title: Re: LM399 based 10 V reference
Post by: Vgkid on October 09, 2015, 02:27:00 pm
Very cool, it is interesting to see what Mickle shows.
Title: Re: LM399 based 10 V reference
Post by: plesa on October 09, 2015, 04:54:50 pm
Wauu, interesting piece of equipment. MickleT are you going to share schematic?
Title: Re: LM399 based 10 V reference
Post by: Andreas on October 09, 2015, 07:18:27 pm
Wauu, interesting piece of equipment. MickleT are you going to share schematic?

Not only the schematic. Also the PCB layout is highly interesting.
Mickle: do you have a RS274 Gerber file for that.
(and of course the software.)
Which processor are you using?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on October 09, 2015, 07:49:09 pm
Schematic is a simplest. ATMega8, some opamps, resistors and that's all. PCB files is not yet ready, because of some critical mistakes in it. May be some later.
Title: Re: LM399 based 10 V reference
Post by: Andreas on October 09, 2015, 09:16:51 pm
Thanks for sharing.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: quarks on October 11, 2015, 01:30:54 pm
DC-Cal Lite v1.0 - 10V/10mA/1mA calibrator, based on the AD5791B DAC.
0.1-10 Hz noise: 0V - 0.4 uV p-p, 10V - 4...5 uV p-p.
0-10 V INL <1 ppm (measured with Solatron 7081 DMMs).
Output resistance 0.003 Ohm (DC voltage mode).
TempCo ~ 0.1 ppm/C.

Very nice, thanks a lot for sharing
Title: Re: LM399 based 10 V reference
Post by: bingo600 on October 17, 2015, 03:07:20 pm
Guyzz

Would any of the "Volt Gurus" care to have a look at my 731B graphs , and explain why it's moving the voltage "up", even w. same temperature.

https://www.eevblog.com/forum/testgear/is-my-'new'-fluke-731b-stabilizing-or-going-crazy/msg779636/#msg779636 (https://www.eevblog.com/forum/testgear/is-my-'new'-fluke-731b-stabilizing-or-going-crazy/msg779636/#msg779636)

TIA
/Bingo
Title: Re: LM399 based 10 V reference
Post by: branadic on November 09, 2015, 12:30:02 pm
I had a phonecall with Prema today. They currently have our Prema 5017 for cal in their labs. After 2 years without calibration the gear is still within the 90 days spec. Pretty good.  :-+
I asked if they used a "special" treatment before they solder the reference into the boards. I was told that each reference is running in a batch with several other LMx99s for 1 year in a box in normal condition, taking data each month to calculate the drift for each of it. After 1 year the references were selected for drift for the different DVMs, the bad ones for low accuracy devices, the good ones for high accuracy devices.
They also did similar for the input voltage devider/resistor network and selected for T.C., low T.C. for high accuracy devices, worse ones for low accuracy devices.
I found that kind of interesting to hear and thought to share this info with you. Sad that they don't build DVMs anymore.
Title: Re: LM399 based 10 V reference
Post by: TiN on November 09, 2015, 03:30:04 pm
Thanks for info, good to know!  :-+
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 09, 2015, 08:30:35 pm
taking data each month to calculate the drift for each of it.

Mhmm,

monthly measurements leave many open points.

attached 2 LM399 references monitored now about 650 days.

The red LM399 (CH6) has lower drift over time.
But obviously has a problem with "popcorn noise".
Jumping around 1-2 ppm within few days.

The green LM399 (CH7) had a large drift in the begin.
But now has calmed down to less than 1 ppm.
So I would trust more the green one.

The blue LTZ1000 is only for reference
(to show the instability of the measurement system)

So how exactly did they select the voltage references:
Including the measurement of the noise?.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on November 09, 2015, 09:34:09 pm
Quote
So how exactly did they select the voltage references:
Including the measurement of the noise?.

The references were measured against a Fluke calibrator and drift against the calibrator were calculated. That's all the info I got.
Title: Re: LM399 based 10 V reference
Post by: Theboel on November 25, 2015, 12:57:48 am
Hello All,
As far as I understand for LTZ1000 lowering working temperature can increase the stability is this also work for LM399 ? if this lowering temperature also work with LM399 could somebody share their experience and how to do it (current limiter maybe ?)
Thank You
Title: Re: LM399 based 10 V reference
Post by: Macbeth on November 25, 2015, 01:27:32 am
I believe all the thermal regulation is built in to the LM399, unlike the much simpler LTZ which requires complicated external support circuitry that can be tweaked?
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 25, 2015, 06:56:18 am
(current limiter maybe ?)

Hello,

this is no good idea and will lead to instabilities.
(how do you get the inner temperature of the LM399?)

The only practical way is to leave the heater unconnected and put the LM399 in a own oven.
You could also use the LM329 (= LM399 without heater) which is cheaper but unfortunately no longer sold in metal can case.

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Gyro on November 25, 2015, 10:48:22 am
Quote
if this lowering temperature also work with LM399 could somebody share their experience and how to do it (current limiter maybe ?)

The LM399 datasheet actually has (had) data on the permissible value range of a resistor in series with the heater to limit its peak warm-up load on the power supply. Unfortunately this only increases the warm-up time, the internally regulated die temperature remains the same.

Edit: I suppose you could feed the heater circuit via a series pass element controlled by a temperature sensor attached to the case, but I suspect that this would be extremely unstable due to the small thermal mass and coupling to the sensor. You'd be a lot better not using the heater and building an external oven instead.
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on November 25, 2015, 11:50:00 am
(current limiter maybe ?)

Hello,

this is no good idea and will lead to instabilities.
(how do you get the inner temperature of the LM399?)

The only practical way is to leave the heater unconnected and put the LM399 in a own oven.
You could also use the LM329 (= LM399 without heater) which is cheaper but unfortunately no longer sold in metal can case.

With best regards

Andreas

Hi Andreas,

I've looked at the LM329 information and just as a side note - the Linear datasheet has exactly the same picture of LF noise there as in the LM399 datasheet (from Linear)  :palm: , and the TI datasheet has a different (from the old National datasheet) picture with much lower noise level. I suspect that the noise level on an unheated LM329 should be considerably lower than on the LM399 at 90C.

Cheers

Alex

Title: Re: LM399 based 10 V reference
Post by: branadic on November 25, 2015, 02:58:25 pm
Don't trust pictures of a single exemplar, better trust the maximum limits.
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 25, 2015, 10:19:23 pm
I suspect that the noise level on an unheated LM329 should be considerably lower than on the LM399 at 90C.

I suspect that they have only "beautified" the diagram.

25 deg -> ~ 300K
85 deg -> ~ 360K
why is sqrt(300K) = 17.3 significant lower than sqrt(360K) = 19
when aproximating the noise as Johnson noise?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on November 25, 2015, 11:02:47 pm
I suspect that the noise level on an unheated LM329 should be considerably lower than on the LM399 at 90C.

I suspect that they have only "beautified" the diagram.

25 deg -> ~ 300K
85 deg -> ~ 360K
why is sqrt(300K) = 17.3 significant lower than sqrt(360K) = 19
when aproximating the noise as Johnson noise?

with best regards

Andreas

Johnson noise calculations won't work here. Just have a look at the noise density graph available in the LM399 datasheet for 25C and 90C - the increase is about 1.5 times (and most likely more in the low 1/f region).

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on November 26, 2015, 11:51:26 am

You could also use the LM329 (= LM399 without heater) which is cheaper but unfortunately no longer sold in metal can case.

With best regards

Andreas

Hi Andreas,

I've just bought some NOS top grade LM129A from eBay and there are more available (http://www.ebay.com/itm/381061192142?).

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 27, 2015, 08:53:21 pm

Johnson noise calculations won't work here. Just have a look at the noise density graph available in the LM399 datasheet for 25C and 90C - the increase is about 1.5 times (and most likely more in the low 1/f region).

Cheers

Alex

Hello Alex,

attached 4 pictures of noise measurement of my LM399#20
2 with heater on
2 with heater off.
all other: same conditions. Measured in cake box with additional thermal shielding.
X-axis 1 sec/div, Y-axis 0.5uV/div

So you should be able to tell which are with heater active and which with heater disabled.

$15 for a half LM399 is a bit too pricey for me.

With best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on November 27, 2015, 09:44:30 pm

So you should be able to tell which are with heater active and which with heater disabled.

$15 for a half LM399 is a bit too pricey for me.

With best regards

Andreas

1) No, I can not tell the difference, sorry! Which means the datasheet is not to be trusted  ;D .

2) It is $15 for 2 and I've agreed a better price for 10 lots (20 devices in total).

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 27, 2015, 09:59:36 pm

1) No, I can not tell the difference, sorry! Which means the datasheet is not to be trusted  ;D .


Hello,

the stray from measurement to measurement is relative large on LM399.
Most is due to thermocouple effects on the (KOVAR) legs.

And also the stray from device to device is large.
I have devices with 2-3 uVpp (Like LM399#20)
and others with around 9 uVpp (LM399#2).

Interestingly the LM399 out of a HP34401A also measured 2-3uVpp.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Gyro on November 28, 2015, 07:17:28 pm
I guess even if the noise level difference is difficult to spot, it's likely that there will be an improvement in long term drift at lower temperature. Personally I find that more of a worry to me due to infrequent opportunities to check calibration.

Chris
Title: Re: LM399 based 10 V reference
Post by: TiN on December 25, 2015, 04:06:58 am
I could not help it, curiosity finally won. Test of LM399 in liquid nitrogen.

(http://xdevs.com/doc/xDevs.com/KX/VREF/lm399_77k.jpg)

Only zener is connected via onboard 3x10.7K resistors. Input power +12VDC from linear Agilent PSU.
Sampling 34970A. Room zener voltage was 6.915VDC, cold voltage - 7.0319VDC. Measured thermocouple Type T temperature -194.72°C.
Title: Re: LM399 based 10 V reference
Post by: SeanB on December 25, 2015, 06:36:19 am
Now put it back into a drift check, and see if you have reset the drift curve from the thermal shock.
Title: Re: LM399 based 10 V reference
Post by: TiN on December 25, 2015, 10:49:28 am
I want to get more data, so will back it to ambient on monday. Taking readings every minute. So far it was about 3uV peak-peak but with many "popcorn"-style jumps few tens of uV back and forth. Tried more current also caused bigger jumps. I have CSV for today, will plot data later today.

Also LN2 temp is not that stable, it vary about 0.1K over minutes. Dewar is not pressured, so likely LN2 getting liquid air and oxygen causing temperature to vary.
Title: Re: LM399 based 10 V reference
Post by: Theboel on December 25, 2015, 12:21:34 pm
@ TiN
wow You are faster and cooler I been thinking to put LM399 in the freezer with -3 C and turn off the heater to see what happened, just curious because my freezer always on 24/7 
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 25, 2015, 02:51:59 pm

Also LN2 temp is not that stable, it vary about 0.1K over minutes. Dewar is not pressured, so likely LN2 getting liquid air and oxygen causing temperature to vary.

Liquid N2 normally is quite stable, indeed.

In fact, pure liq-N2 (~77.2K) condenses air, and therefore, the oxygen part of this mixture will slowly raise the temperature, towards the 90K of liquid air. But that takes hours to reach maybe 80K, depending on the volume of the dewar, and the temperature reading will not flicker on minutes basis, instead goes  steadily upwards. Anyhow, you should cover the opening of the dewar with cork, let a small hole anyhow, for the evaporating gas.
Also, changes of atmospheric pressure may slightly change the temperature of the liquid N2, but that also does not happen within minutes.

Your sensor is not of best quality, and very sensitive. A PT100 will deliver much more stable readings.

Frank
Title: Re: LM399 based 10 V reference
Post by: TiN on December 25, 2015, 03:38:54 pm
Here is the data, over 7 hr span. This is rather experiment to see if it's worth to spend any time on this direction, or forget it. As there are ways to get these temperatures for 24/7 without pouring LN2, if there are visible benefits of doing so. Will see back monday how it behave, after which I will leave same module to log two more days with ambient temperature, so we can compare both datasets.

Dewar is Air Liquide TR26 with ~6 liters of LN2, not pressurized.

(http://xdevs.com/doc/xDevs.com/KX/VREF/vref_ln2_lm399.png)

Those jumps are ~2ppm each. Perhaps due nitrogen cavitation and bubbles?

CSV with all the wobbles (http://xdevs.com/doc/xDevs.com/KX/VREF/ln2_lm399_nplc100.csv).
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 25, 2015, 04:09:55 pm
You should not leave this dewar open! That's very important!
Otherwise, Air and humidity will enter the vessel.
If you put a cork or similar on top, not tight, of course, the evaporating gas causes a slight overpressure, which will prevent any gases entering  from outside.

26l of liquid nitrogen.. That's a lot of stuff. The temperature should be rock stable.
Not pressurized simply means that it is always under atmospheric pressure.

The fluid boils a little bit, but no 'cavity' will be inside..

Also test a better thermometer, that will give directly stable results.
Title: Re: LM399 based 10 V reference
Post by: TiN on December 25, 2015, 04:15:32 pm
There is hardfoam cap on it, so it's not fully open. I have pressure valve device for it, but seal O-ring is broken so I did not use it. Perhaps I can repeat test next weekends, but will need bring few of my Keithley's to lab. There is nothing better than 34970 atm :)

Also there is actually about 5-6L of nitrogen in it only as for now.
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 25, 2015, 05:10:17 pm
Thermo elements usually have very low thermoelectric voltages at lower temperature.
This type T has -15uV/K only.
The fluctuations you see, are equivalent to 3uV only.
That easily explains these pseudo temp measurement changes.

I used a special Kondo effect thermo couple (*), with much higher sensitivity down to 4.2K.
It showed such dips also, afair.
So I replaced it with a high grade PT100 down to about  25K, and a special diode thermometer down to 1.4K, which both gave much more stable measurements.

It is not necessary to pressurize liq. N2, otherwise you want to create lower temperatures by pumping on the cryogenic liquid.

Frank

(*) NiCr/FeAu, I think
Title: Re: LM399 based 10 V reference
Post by: TiN on December 25, 2015, 05:20:23 pm
I'm more concerned about voltage zener reading, not the temperature itself.
Have Honeywell HEL-705 1k RTDs so can use those.
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on December 25, 2015, 05:30:12 pm
I'm more concerned about voltage zener reading, not the temperature itself.
Have Honeywell HEL-705 1k RTDs so can use those.

So the noise on the LM399 output should have nothing to do with any changes of the N2 temperature.
Maybe the silicon does not like temperature shocks.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 25, 2015, 05:54:39 pm
I don't think it's the temperature that cause the 10 µV steps. The steps are rather uniform an look quite fast. So it's more an chip internal thing similar to popcorn noise.

I don't think a low temperature stabilized reference is really practical, as this also causes quite some trouble with humidity and even condensation. One nice thing with a heated reference is that relatve humudity is low (e.g. less than about 50% for 10 degrees over ambient). So I would definitely prefer something like 40-50 C so well above ambient, but not as high as the oroginal LM399.

Also the  way to low noise is using a rather large reference or several in parallel, thus quite some heat source.
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on December 25, 2015, 06:17:46 pm
So I would definitely prefer something like 40-50 C so well above ambient, but not as high as the oroginal LM399.

I've bought some LM129A zeners to try for that very reason. One interesting side of using the LM129/329 is that (unlike the LM399) it is graded by tempco, with the "A" grade is under 10ppm/C from -55C to +125C and the "D" grade is under 100ppm/C from 0C to +70C. In the LM399 all that range is randomly spread.

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on December 25, 2015, 10:03:51 pm
Hmm, interesting, thank you. I was also thinking that higher noise of the 399 is due to the low Zener current. In the JFET based reference it is not a problem though. I already have results close to the buried Zener in the noise and tempco, if the long term stability is good, with a simple oven it would be an interesting alternative approach.

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: branadic on December 27, 2015, 05:09:42 pm
Somewhat off-topic, but I found LM329 spice model, which means a model of LM399 without the heater:

http://www.diyaudio.com/forums/power-supplies/236257-lm329-ltspice-model-needed.html (http://www.diyaudio.com/forums/power-supplies/236257-lm329-ltspice-model-needed.html)

Attached you find the Subcircuit (.sub) and the Symbol (.asy) for LTSpice.

There is another Spice Model simulating additional parameters: http://www.ece.utah.edu/~ece2280/reg_zener.lib (http://www.ece.utah.edu/~ece2280/reg_zener.lib)
Title: Re: LM399 based 10 V reference
Post by: Macbeth on December 27, 2015, 06:33:29 pm
Thanks, that reminds me that I have the other half of my rubbish attempt at drawing up those schematics in LTSpice.

Find attached my heater schematic. I'm sure it is laughable. Perhaps someone who knows what they are doing can combine the models and make a proper netlist?

Title: Re: LM399 based 10 V reference
Post by: Andreas on January 06, 2016, 11:02:22 pm
Hello,

again a ageing chart of 2 LM399 references.
now after 2 years (more than 730 days) 24/7 operation.  8)

And probably the last time. Since I may need the multiplexer
channels for two brand new LTZ1000A references soon.

see also:
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg796829/#msg796829 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg796829/#msg796829)
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg705039/#msg705039 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg705039/#msg705039)
https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg566721/#msg566721 (https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/msg566721/#msg566721)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 07, 2016, 08:26:37 am
Thanks for sharing Andreas. For my understanding, the blue curve measuring the LTZ#2 indicates the drift and noise of the ADC's reference? Which reference is used for this reference?
Thus, the remaining error of the ADC's reference is still within the curves for the LM399s? I'm asking because there seems to be some common mode behavior between the blue and green curves but also a much smaller fraction between green and red curves.


Maybe you can explain the diagram in more detail? Thanks.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 07, 2016, 11:57:29 am
Hello branadic,

LTZ#2 was intended to show the drift of the ADC reference.
LTZ#2 itself has most probably a drift of around -2 ppm / year.
But since the calibration standards that I use to calibrate LTZ#2
also drift it will need some years until I know the whole truth.

By the way ADC13 drifts around the same amount as LTZ#1.
LTZ#1 calibration values show a lower drift than LTZ#2.
(so I should have used this as reference, but who knows in advance).

The common mode effects are partly due to environment
humidity + temperature. Temperature is compensated within ADC.
But nothing is perfect. And over the time also the contacts of the
measurement setup get loose and have to be moved/tightened again.

Attached the ageing curves of ADC13 compared to my
"stable" standards over the same time window.

The conclusions from my side are:

you will need minimum 200 days (half a year) run in time
 until you can start qualifying the LM399.

In this case I would prefer LM399 CH7 which shows the higher stability.
(less standard deviation)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on January 07, 2016, 02:13:08 pm
Hello branadic,

LTZ#2 was intended to show the drift of the ADC reference.
LTZ#2 itself has most probably a drift of around -2 ppm / year.
But since the calibration standards that I use to calibrate LTZ#2
also drift it will need some years until I know the whole truth.

...
With best regards

Andreas

Will we three be able to check that coming Saturday?

Frank
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 07, 2016, 04:07:21 pm
I hope at least getting 2 new 8.5 digit measurement points on the diagrams  8)
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on January 07, 2016, 06:56:02 pm
I hope at least getting 2 new 8.5 digit measurement points on the diagrams  8)

Andreas, certainly. But this was a two-fold question:

..Will we three be able to check that coming Saturday?

Frank

Anyhow, I'll carry my new 34465A also..

Title: Re: LM399 based 10 V reference
Post by: branadic on January 07, 2016, 07:00:51 pm
Hi Frank,

I'm with you on saturday, together with Keithley 2002 (last Cal in 11/2015) and my Prema 5017 SC (not Cal'ed yet). What else I come up with, well, let's see.
Title: Re: LM399 based 10 V reference
Post by: Dr. Frank on January 07, 2016, 07:22:51 pm
Hi Frank,

I'm with you on saturday, together with Keithley 2002 (last Cal in 11/2015) and my Prema 5017 SC (not Cal'ed yet). What else I come up with, well, let's see.

GREAT!   :-+ :clap: :-+ :clap: :-+ :clap:

Then let's have a contest.
Title: Re: LM399 based 10 V reference
Post by: babysitter on January 08, 2016, 08:46:43 am
Klingt als hättet Ihr was vor! :)

Title: Re: LM399 based 10 V reference
Post by: bingo600 on January 08, 2016, 09:06:09 pm
Klingt als hättet Ihr was vor! :)

Agreed ....

What are you guyzz up to  :popcorn:

/Bingo
Title: Re: LM399 based 10 V reference
Post by: branadic on January 11, 2016, 11:33:04 am
Nothing fancy, only a small comparison between meters and voltage references :)
Title: Re: LM399 based 10 V reference
Post by: bingo600 on January 11, 2016, 07:48:36 pm
Nice gear.

What's the "Red 7-Seg" one ?
A Prema ?

/Bingo
Title: Re: LM399 based 10 V reference
Post by: branadic on January 11, 2016, 07:52:56 pm
Yes, a Prema 5000 and beneath a Prema 5017 SC.
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 14, 2016, 08:12:06 pm
Hello everybody,

after reading the whole thread and some other threads and articles, i finished my first revision of a LM399H based reference. Or is it AD587LQ based?
At the end, i don't really know.
Few weeks ago i soldered a 100% copy of an SVR-T on a small chinese breadboard and since that time, i got 10,000.00V, drifting between 9,999.99 and 10,000.01V, depending of the room temperature.
I used simple 25ppm resistors and a 100 ppm trimmer with 200 ohms for the trimming of the AD 587.
Quick and dirty T.C. compensation is done with a 100k thermistor in series with a 220k fixed resistor (50 or 100ppm, don't know).

So i decided to use the combination of the LM399 and the AD587 as described in Joe Gellers patent, to add the stabilities of both parts.
I tried my best to avoid the known constructive errors, but i'm not sure, if i should use a ground plane on both layers for a better heat distribution
or if i should omit it for a better temperature isolation. 
Please have a look at the pictures of both layers and tell me your opinions.

Thanks in advance

Frank
Title: Re: LM399 based 10 V reference
Post by: branadic on January 14, 2016, 10:35:50 pm
So you want to use SMD resistors? Do you plan to use some thermal stabilisation for them?
As you can read in this thread I've finally used a pickaback board with the gain setting SMD resistors (5ppm/K) and a crystal heater on top of them, to avoid temperature changes. Seems to work quite well.

However, I would spend at least some buffer amp for the LMx99.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 14, 2016, 11:05:00 pm
Hello,

I would thermally decouple the 85 deg LM399 from the room temperature NTC and AD587.

You have anyway at minimum 2-3 deg C temperature difference between NTC and the chip of the AD587.
The LM399 will worse that if thermally coupled.

The AD587 has a bad PSRR. (10ppm/V) So I would not put the 15V stabilisation too far away from the cirquit.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 15, 2016, 07:30:49 pm
So you want to use SMD resistors?

Yes, and i have to use 0805, because my favorite vendor Reichelt sells the complete E96 row only in 0805.
Next step would be 5ppm resistors. Where do you buy them in germany?

Quote
As you can read in this thread I've finally used a pickaback board with the gain setting SMD resistors (5ppm/K) and a crystal heater on top of them, to avoid temperature changes. Seems to work quite well.

Maybe the next edition has a temperature stabilizer on board, with the heating resistors placed on the "right" positions.
But it's not easy to get a high precision temperature stabilizer and then find out the right positions for the heaters and the sensor.

Quote
However, I would spend at least some buffer amp for the LMx99.

In this configuration, the AD587 is the "buffer" for the LM399H.

Best regards

Frank
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 15, 2016, 07:43:52 pm
I would thermally decouple the 85 deg LM399 from the room temperature NTC and AD587.

In my opinion, there is enough decoupling to avoid hot spots and enough coupling to reach a kind of thermal balance for the whole unit. I will place the unit in a TEKO tin-box and fill it with molten paraffine. This little box will be isolated with minimum 25-30mm foam and placed in another box, maybe heated with a simple PTC-heater or a heating-foil to ~40°C

Quote
You have anyway at minimum 2-3 deg C temperature difference between NTC and the chip of the AD587.

The parrafine should avoid such temperature differences.

Quote
The AD587 has a bad PSRR. (10ppm/V) So I would not put the 15V stabilisation too far away from the cirquit.

I already finished a LM317 tracking preregulator with 2 LM329 in series for voltage stabilization as a small module.
It will be placed in the outer box.

With best regards

Frank

Title: Re: LM399 based 10 V reference
Post by: lars on January 16, 2016, 07:41:55 am
For the bad PSRR at DC of the AD587 I would say Analog Devices is very conservative. For my SVR´s with AD587LQ I have I have measured 0.1-0.3ppm per Volt around 15V in. I just did a quick check for 6pcs of AD587, 2pcs REF102 and 2pcs LT1021 and all were 0-3ppm for a quick change of 15 to 20 to 15Volt in. The REF102 and LT1021 was not better than the AD587 even if the spec says so.

Lars
Title: Re: LM399 based 10 V reference
Post by: lars on January 16, 2016, 07:49:11 am
Hello Frank

Will be interesting to see how the result for the LM399+AD587 combination will work. I have thought of testing this but with the SVR-T I think some of the benefits vanished. I think Joe did the combination to get away with the worse temperature coefficient of the AD587LQ compared to the LM399.

Lars
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 17, 2016, 02:43:26 pm
Will be interesting to see how the result for the LM399+AD587 combination will work.
I'm also a little bit curios...

Quote
I have thought of testing this but with the SVR-T I think some of the benefits vanished.
Indeed! My breadboarded and not perfectly compensated unit runs stable since 7 weeks, varying only 15-20µV over the changing of room temperature.

Quote
I think Joe did the combination to get away with the worse temperature coefficient of the AD587LQ compared to the LM399.
Yes, but i think, most of the TC is produced by the resistors and the amp and not by the unheated zener.
Maybe there will be only a small effect to the total TC.

Frank
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 17, 2016, 05:09:12 pm
Hello,

I am missing the cirquit diagram for your PCB.

One problem might be to set the current stable (1mA within 0.1%) through the LM399 zener.
The AD587 has a 4K source resistance at the NR pin.
The voltage at this pin is most probably around 7V-7.5V with some tolerance.
(Just guessing from the +3/-1% trimming range).

For the resistor network:
from the datasheet the T.C. is trimmed in some way.
So either the zener current is adjusted to "zero T.C. current"
or the resistor network itself is done non-linear to compensate for the T.C. of the zener.
Would be interesting how different AD587 behave for the 2 halves of cirquit.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on January 17, 2016, 05:13:01 pm
Indeed! My breadboarded and not perfectly compensated unit runs stable since 7 weeks, varying only 15-20µV over the changing of room temperature.
That could easily be variations in the DMM if you haven't compensated for it
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 17, 2016, 06:36:48 pm
I am missing the cirquit diagram for your PCB.

Mine is too big for posting.
It's like http://patentimages.storage.googleapis.com/US7382179B2/US07382179-20080603-D00000.png (http://patentimages.storage.googleapis.com/US7382179B2/US07382179-20080603-D00000.png)
with
R3 = 3k
R4 = 820R
R7 = 3k01
R8 = 200R
R9 = 7k87

I added a 1µF for noise reduction on #101 and the TC compensation resistors from the SVR-T.

Quote
One problem might be to set the current stable (1mA within 0.1%) through the LM399 zener.
All resistors are 0,1% with 25ppm.
 
Quote
The AD587 has a 4K source resistance at the NR pin.
The internal source will be nearly total suppressed by the LM399.

Quote
Would be interesting how different AD587 behave for the 2 halves of cirquit.
I bought several LM399 and AD587, but they are still on their way to my work bench.
As a fresh infected volt-nut my stock is already virgin.

With best regards

Frank
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 17, 2016, 08:38:00 pm
Looks like a really clever circuit to use the amplifier and stable resistors in the AD587 for the LM399 reference. The trimming part may need adjustment for individual units. I would have used less current here, just to keep selfheatung lower - but thats a minor detail.
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 18, 2016, 01:11:41 pm
I would have used less current here, just to keep selfheatung lower - but thats a minor detail.

I have taken the values from the SVR-T which are proven as good for the perfect stability of these units.
But you are right, there are some trim circuits out in the web with much higher values and e.g. a 1 megohm resistor between the trimmer and the trim input.
I will get enough PCBs to test all variations.

Frank
Title: Re: LM399 based 10 V reference
Post by: Macbeth on January 18, 2016, 09:29:16 pm
You should also consider paying Joe Geller a small royalty fee for each of these circuits you build-- (it *is* patented after all)...  I don't know how much he would want, but if you get in touch with him you can negotiate a number...

Surely royalties would apply to each one he actually sells commercially rather than builds for own use?

Also, what of Lars contribution (if I read his earlier post right)?

I also liked that Geller Labs was so open with design, schematic and BOM´s (including my NTC compensation with both first and second order compensation that Joe could have kept for himself)
Title: Re: LM399 based 10 V reference
Post by: lukier on January 19, 2016, 11:54:09 pm
At first, I thought like you did-- if not for commercial use, then no fees, right?  Well, I was wrong.  Evidently, it's OK to experiment [research] with patented ideas, but as soon as you put them into actual use, then you are benefiting from the patented design, and the owner of the patent is entitled to compensation. 

Interesting. I didn't know that. I also always assumed the commercial connection.

So, if I employ a patented idea for my own use (e.g. hobby, some gadget for the home etc) how I am benefiting? It often might not have a financial aspect at all, so how should I compensate the owner?
Title: Re: LM399 based 10 V reference
Post by: Kean on January 20, 2016, 07:01:43 am
Patent law is never that simple, especially internationally - there are exemptions in many countries for private/non-commercial use.
See http://www.wipo.int/meetings/en/doc_details.jsp?doc_id=256317 (http://www.wipo.int/meetings/en/doc_details.jsp?doc_id=256317)

In this case the patent is probably in the USA, and the USA don't appear to have this exemption, but many people interested in use of the described methods will be outside the US and so it gets awfully murky.  I guess they could sue you if you ever entered the US.

In any case, I fully support the concept of contacting the creator of any unique idea (patented or otherwise not openly licensed) and offering a small donation as a private user.  In most cases they'll be very happy of the appreciation and make a license exemption of some sort, and if not then you might not want to further associate yourself with such people or their "creations".
Title: Re: LM399 based 10 V reference
Post by: quarks on January 20, 2016, 09:01:30 am
About Patent law, it looks like a very interesting situation, because Joe Geller is the patent owner and he also seems to be a registered patent attorney
see last paragraph in link
http://www.gellerlabs.com/about.html (http://www.gellerlabs.com/about.html)
Title: Re: LM399 based 10 V reference
Post by: Theboel on January 20, 2016, 12:46:14 pm
I just email Mr Joe ask about his patent and here what he reply,

Hi Anton,

   There were no international patent rights, just the U.S. Patent.  Also, I allowed the U.S. Patent to expire, which means that particular patent is now in the public domain.

Title: Re: LM399 based 10 V reference
Post by: Macbeth on January 20, 2016, 11:02:13 pm
That's great, because Joe Geller is obviously a gentleman. However, the slimey sharks that patent such bullshit as a flashing cursor, shopping cart, hyperlink and all happen to legally exist in Texas relying on what one would consider an extreme version of US patent law. Really - asking them for "permission" to breath is enough for a lawsuit or at least a demand for 1c for each breath for your entire life.
Title: Re: LM399 based 10 V reference
Post by: babysitter on January 21, 2016, 04:35:05 pm
Hi there,

just in case you want to run some reference at lower temperatures - at https://www.mikrocontroller.net/topic/386300#new (https://www.mikrocontroller.net/topic/386300#new) in a german microcontroller forum a guy sells crystalless OCXO "carcasses" (just the PCBs, temp. programmable by external resistor), believed to be 50-100 deg for 2, 50€ per quartet.

Title: Re: LM399 based 10 V reference
Post by: barnacle2k on January 23, 2016, 08:39:12 am
Completely unrelated to previous posts:
i want to know if i killed my LM399:

What happened forgot to connect the heater negative terminal.
In the DS it says:
Quote
The substrate is electrically connected to the negative terminal
of the temperature stabilizer. The voltage that can be applied to either
terminal of the reference is 40V more positive or 0.1V more negative than
the substrate.

so if the heater current was flowing through that (parasitic) diode, the substrate would be elevated by the voltage drop.
Resulting in a negative potential on the negative terminal of the reference of more then the 0.1V allowed in regard to substrate.


Output is half decently stable but is extremely susceptible to input voltage variations and noisy. 
Before i chase other possible causes i'd like to know if i destroyed the LM399 with that mishap.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 23, 2016, 11:32:24 am
Hello,

how much current does the cirquit take?
(if the heater is off then the cirquit is relative instable).

What is extreme noisy (uVpp)? Which Frequency range did you measure?

Power supply rejection should be below 2 ppm/V above 15 V heater voltage.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: GK on January 23, 2016, 12:31:24 pm
I need a precision trimmed, bipolar reference of +/-7.00V for a precision powersupply I am working on. The reference needs to have low output impedance and low noise.
Here is what I came up with. Filtering is incorporated into the trimmed amplifiers such that the effects of capacitor leakage current is mitigated. The 10uF non-polar will be polypropylene film. No currents are injected into ground, eliminating errors due to IR voltage drops. The -7.00V amplifier provides the bias current for the LM399 zener. The 1N4148 and the 10k to -12V ensure start-up.
     
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=196029)
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 23, 2016, 01:10:38 pm
Hello,

just a question:
how do you ensure that the -12V starts up equally/before the +12V.
(otherwise you could have the oops case above).


I would most probably use 2 schottky diodes to keep Pin 2  near ground level.
And a additional at the output of the negative voltage regulator.

For the 6K8 resistor I would also use at minimum a good metal film (RC55Y with 15ppm/K).

With best regards

Andreas
 
Title: Re: LM399 based 10 V reference
Post by: GK on January 23, 2016, 01:34:58 pm
It doesn't matter how the +/-12V supply rails come up (no damage will result) and a run-of-the-mill 50ppm TC for the 6k8 is fine. The DC resistance of the LM399 is 1 ohm or better IIRC.

7V/6800 = 1.03mA.

(1.03/10e6) * 50 (ppm) = 51.5nA per deg. C.

51.5nA * 1 ohm = 51.5nV per deg. C.

That's more than 2 orders of magnitude less error than the ~7uV deg. C (1ppm) specified drift of LM399 itself.
 


 
Title: Re: LM399 based 10 V reference
Post by: barnacle2k on January 23, 2016, 10:15:27 pm
how much current does the cirquit take?
(if the heater is off then the cirquit is relative instable).

What is extreme noisy (uVpp)? Which Frequency range did you measure?

Power supply rejection should be below 2 ppm/V above 15 V heater voltage.

Thank you for your reply Andreas, i was looking for those members that actually blew up a few LM399. [TiN?, ...]
Circuit takes ~20mA. (including OP)
I isolated the LM399 from rest of the circuit now and tested the PSSR, stabilizer connected to Vin, Rv=2.21k.
10.5Vin - Vz=6.990935V
10.0Vin - Vz=6.990745V
9.50Vin - Vz=6.990546V
9.00Vin - Vz=6.990344V

55ppm voltage difference in a 1Vin change (10.5V to 9.5V).  :--
I think the magic smoke is out of this vintage 1971 lm199.
Title: Re: LM399 based 10 V reference
Post by: TiN on January 24, 2016, 08:28:09 am
No, I blew only one LTZ1000 2 years, which is dissected and photographed pages pages away :)
I think one LM399, which had fun in boiling LN2 ain't much useful, but would not call that blown  :-DD

Also since we talking dead stuff, here are graphs from my cryo torture:

(https://xdevs.com/doc/xDevs.com/LM399_LN2/air_heater_1.png) (https://xdevs.com/doc/xDevs.com/LM399_LN2/air_heater.png)

(https://xdevs.com/doc/xDevs.com/LM399_LN2/air_no_heater_1.png) (https://xdevs.com/doc/xDevs.com/LM399_LN2/air_no_heater.png)

(https://xdevs.com/doc/xDevs.com/LM399_LN2/ln2_1.png) (https://xdevs.com/doc/xDevs.com/LM399_LN2/ln2.png)

CSV-file, heated LM399 (https://xdevs.com/doc/xDevs.com/LM399_LN2/heated_lm399_nplc100_2d.csv)
CSV-file, unheated LM399 (https://xdevs.com/doc/xDevs.com/LM399_LN2/air_no_heater_lm399_nplc100_2d.csv)
CSV-file, LN2 submerged LM399 (https://xdevs.com/doc/xDevs.com/LM399_LN2/ln2_lm399_nplc100_2d.csv)

Even though some jumps are present (dirty lab environment with lots of gear around randomly running), overall conclusion I already got made for myself.
Datalogger: HPAK 34970A with 34901A card. Temperature reading -  Type T AWG36 thermocouple wire soldered directly to pin 1 of LM399. Current resistor 7.5KOhm (@ambient temperature), supply +15VDC from HPAK linear bench PSU.

Note different scale on ppm's (red on right). Time scale in minutes, sample taken every minute.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 24, 2016, 01:46:04 pm
55ppm voltage difference in a 1Vin change (10.5V to 9.5V).  :--
I think the magic smoke is out of this vintage 1971 lm199.

Hello,

I also fear this.
I have looked up PSRR measurement on my LM399#3.
It is around 0.7ppm/V in the range 8.5 to 15 V.

But the difference is that I use a constant current source
(1V over a 1K resistor with a LT1013 + BF245A and the LM399 as reference).

On the other side 20mA is at the lower limit of a LM399.
My LM399 take usually 25-30mA at room temperature at 9.35V with some additional isolation.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 24, 2016, 05:14:07 pm
PSRR comes from 2 or 3 contributions:
1) The temperature regulator part
2) the OP buffering the outout if used
3) the current source to drive the zener part of the LM399 - often this a resisitor from the scaled up voltage.

With modern voltage regulators, the PSRR is not that important any more. Even a 7812 can keep the supply resonable stable - with only a few mV in change 55 ppm/V are not a probem at all.
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 24, 2016, 05:47:01 pm
On the other side 20mA is at the lower limit of a LM399.
My LM399 take usually 25-30mA at room temperature at 9.35V with some additional isolation.
I'm actually testing some LM399H. The older ones (NOS) from 1989 take about 28mA @15V at room temperature, the newer ones (used) from 1998 are running at 35mA.
Strangely, the older ones are increasing their Zener voltage for ~100ppm, when the heater is powered, the newer ones are decreasing for ~85ppm.
Both types are from NS.

With best regards

Frank
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 24, 2016, 07:35:34 pm
Hello Frank,

You should compare heater power. 30mA @ 9.35V are around 300mW.
At 15V you should expect around 20 mA at room temperature.

is there a correlation of zener voltage to tempco on your references?
see here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg422367/#msg422367 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg422367/#msg422367)

the "sweet spot" seems to be at 6875 mV.
Above I have positive T.C.
Below negative T.C.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: F64098 on January 26, 2016, 11:28:41 am
is there a correlation of zener voltage to tempco on your references?
Yes, like your's  and i wonder about the much higher deviation from cold (18°C + 2K self heating) to hot of mine.

Quote
the "sweet spot" seems to be at 6875 mV.
Would be nice, to get pre-selected parts from some ebay sellers  ::)

With best regards

Frank
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 30, 2016, 08:36:20 am

Would be nice, to get pre-selected parts from some ebay sellers  ::)


Hello,

keep on dreaming.

In my dream world the manufacturer sends all his production of LM399s to his primary precision instrument manufacturer.
This one selects out the best parts and stamps his personal number (something like 1826-1249) on it.
The rest of production is sent to the other instrument manufacturer.
He selects out the best of the rest and marks them. (with something like SL40057).
(that is my personal explanation that I already found three instruments of this manufacturer which have more than 5 ppm difference when tilting from the left to the right side whereas the other manufacturers instruments have below 5 ppm tilting effect from back to front side (the same orientation regarding the LM399 on this instrument)).

The rest is then sent back to the IC manufacturer and is stamped with LM399.
So how can you tell me that someone will get those LM399 at the sweet spot.

The most probably reason that I have some of them is that I all bought them when NS went out of production. So the manufacturers all selected among the LT parts at this time.

I have put several measurements into one single graph.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: F64098 on January 31, 2016, 11:28:56 pm
In my dream world the manufacturer sends all his production of LM399s to his primary precision instrument manufacturer.
.....
The rest is then sent back to the IC manufacturer and is stamped with LM399.

That sounds too bad, to be false...

Quote
I have put several measurements into one single graph.

Very nice!
I got another batch from a german seller and their behaviour is much better, than the others.
But the seller increased the price shortly after i bought them, so it's not economic to buy another batch.

Best regards

Frank

Title: Re: LM399 based 10 V reference
Post by: Andreas on February 02, 2016, 07:15:07 am
In my dream world the manufacturer sends all his production of LM399s to his primary precision instrument manufacturer.
.....
The rest is then sent back to the IC manufacturer and is stamped with LM399.

That sounds too bad, to be false...


Updated diagram.

By the way: when I leave out my references out of the diagram it gets obvious.

That what I would expect from datasheet 6.95V +/-4% for the NS parts is that
68% of the measured devices (gaussian distribution) lies within 6950 +/- 90mV.
(4% = 3 sigma).
If I count in the 2nd diagram there is a large gap...

So either the process is not well controlled (sounds not plausible to me)
or someone did some cherry picking...

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: F64098 on February 02, 2016, 12:52:48 pm
So either the process is not well controlled (sounds not plausible to me)
or someone did some cherry picking...

It seems, we have to look out for parts with the magic datecode...  8)

Frank
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 03, 2016, 10:25:03 pm

It seems, we have to look out for parts with the magic datecode...  8)


Hello,

after reading this thread all sellers will re-stamp their parts accordingly. (and raise the price).  >:D

by the way: today I measured 2 LM399 from LT which I bought from DigiKey recently (DateCode 1351).

results attached and overview picture updated.

The only hope that I have is that perhaps the LM399A are the way to go.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: fluxamp on February 24, 2016, 09:07:01 pm
Some more data points for your plot Andreas: I got some LT LM399AH from DigiKey with date code 1502 and did some quick measurements. Room temperature was at 19C, measured with Agilent 34410A at 10nPLC

LM399AH #1:
unheated: 7.077,21V
heated:      7.086,82V
ppm/K:       19.2

LM399AH #2:
unheated: 7.074,50V
heated:      7.085,28V
ppm/K:       21.5

LM399AH #3:
unheated: 7.057,67V
heated:      7.065,89V
ppm/K:       16.4

LM399AH #4:
unheated: 7.060,37V
heated:      7.068,70V
ppm/K:       16.7
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 24, 2016, 09:32:31 pm
Hello,

thank you for the data.
But my hopes that the LM399AH will be better are destroyed.
The output voltage and then the T.C. are both too high.
Attached: updated chart with LM399AH

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on February 24, 2016, 09:53:36 pm
Thanks for the update Andreas. Seems like I need to measure my mixture of LMMx99 too (also some National Semiconductor LMx99), to give it a further statistics.
Maybe you want to share the complete data table?
Title: Re: LM399 based 10 V reference
Post by: TiN on February 25, 2016, 08:06:37 pm
Better in Excel format :)
You guys get me interested too.

(https://doc.xdevs.com/doc/xDevs.com/KX/REFb/lm399s_1.jpg) (https://doc.xdevs.com/doc/xDevs.com/KX/REFb/lm399s.jpg)

Some tests without heater:

(https://doc.xdevs.com/doc/xDevs.com/KX/REFb/lm399s.png) (https://doc.xdevs.com/doc/xDevs.com/KX/REFb/lm399s.xlsx)

Perhaps better to bodge up test jig with SMU to ramp zener current from 100uA to 10mA in 100uA steps and with Heater on/off switch?


Title: Re: LM399 based 10 V reference
Post by: fluxamp on February 28, 2016, 04:43:20 pm
Thought I'd share my build of the simple LM399 based transfer. It uses the self-biasing circuit posted in the first few posts.
I used a double-walled soup container/flask from Ikea as a housing and SMA connectors for unbuffered 7V zener voltage and 10V main output.

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=204698;image)

The pink stuff is packaging material from Farnell that I used to isolate the board from the metal walls. At the top of the board sits an LM317 and underneath the white foam is the LM399. The gain setting resistors are all located below.

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=204700;image)

The blue coated resistors are Vishay PTF56 (5ppm/K) that are mounted on little standoffs. That way I didn't have to bend the legs. Unfortunately, I had a bit of brain-freeze when ordering them, so I needed to parallel them up to get the right values  :palm: The back of the board only holds the LTC2057 and some more foam.

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=204702;image)

Currently the unit sits in a lab corner powered 24/7 to observe the thermal behavior. Its intended purpose is momentary measurements so it's not going to be powered for any extended time. But I thought it'd be interesting to see whether the LM317 generates enough heat to upset the LM399 or influence the gain setting resistors. After half a week I opened the lid and the inside was a few K warmer than the outside of the case but not by a lot. For the next build I'll try some different things.
Title: Re: LM399 based 10 V reference
Post by: TiN on February 28, 2016, 04:52:51 pm
Clean build! Is that wire triax from ebay? I was just making pair of cables using very similar looking low-noise triaxial.
Title: Re: LM399 based 10 V reference
Post by: fluxamp on February 28, 2016, 05:00:01 pm
Thanks! The white wire is "only" coax. My boss gave it to me so I have no clue where it comes from, sorry.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 28, 2016, 10:25:57 pm
Perhaps better to bodge up test jig with SMU to ramp zener current from 100uA to 10mA in 100uA steps and with Heater on/off switch?

Hello,

what do you want to do with that?
below 0.5mA the behaviour is undefined (according to datasheet).
above that you should measure the differential resistance of the zener (around 0.5 Ohms).
I bet that the 1826-1249 has near 6.9 V (+/- 0.04) output at 1 mA with heater on.

Thought I'd share my build of the simple LM399 based transfer.

Interesting housing.
First I thought that also the lid is some metal.
But obviously it is isolating plastic.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on February 28, 2016, 11:19:10 pm
Perhaps better to bodge up test jig with SMU to ramp zener current from 100uA to 10mA in 100uA steps and with Heater on/off switch?
what do you want to do with that?
below 0.5mA the behaviour is undefined (according to datasheet).
above that you should measure the differential resistance of the zener (around 0.5 Ohms).

Just a reminder that the reference section current is not the zener diode current.  Current through the buried zener is always around 250uA in normal operating conditions.  This is called out nicely in National Semi's App Note 161 (http://www.ti.com/lit/an/snoa589c/snoa589c.pdf (http://www.ti.com/lit/an/snoa589c/snoa589c.pdf))
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 07, 2016, 10:09:41 am
Another simple DMM tester (now with the stupid mistake on the label  ;D ).
Voltage output 0...10 V with 1 ppm resolution and 1 ppm INL. TempCo -0.2 ppm/K. Output resistance 0.002 Oh. LF noise (0.1-10 Hz): 10 V - 2.4...2.8 uV p-p, 0 V - 0.2...0.25 uV.
Current output 0...10 mA with 1 ppm resolution. TempCo -3 ppm/K
3wire-resistance: hermetic foil 1kOhm +3 ppm/K, 10 kOhm +/-1 ppm/K, hermetic PWW 100 kOhm -2,6 ppm/K, 1 MOhm <10 ppm/K.

Title: Re: LM399 based 10 V reference
Post by: necessaryevil on May 29, 2016, 09:08:20 pm
I'm impressed Mickle. But what mistake do you mean?
Title: Re: LM399 based 10 V reference
Post by: ebclr on May 29, 2016, 09:20:23 pm
https://www.taobao.com/view_image.php?spm=a312a.7728556.2015080703.1.7UZkf9&pic=HEYVDllCEghRV1dcWwocCUQOFx0XTEJfW1lXDBZFAlpcMAl9eXE7MRRUDzQuE1BhaERiZwNbQlpeXV0OSFM=&title=TE0zOTkgUENC&version=2&c=MWIzOTMwMTNhMzgzMDI0OWY4MzY3ZGNiNjE5OWM4NDA%3D&itemId=14154303091&shopId=58559233&sellerRate=25555&fv=9 (https://www.taobao.com/view_image.php?spm=a312a.7728556.2015080703.1.7UZkf9&pic=HEYVDllCEghRV1dcWwocCUQOFx0XTEJfW1lXDBZFAlpcMAl9eXE7MRRUDzQuE1BhaERiZwNbQlpeXV0OSFM=&title=TE0zOTkgUENC&version=2&c=MWIzOTMwMTNhMzgzMDI0OWY4MzY3ZGNiNjE5OWM4NDA%3D&itemId=14154303091&shopId=58559233&sellerRate=25555&fv=9)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on May 29, 2016, 09:28:38 pm
The mistake is just labeling: "DC Voltage 0-10.0... mA"
Title: Re: LM399 based 10 V reference
Post by: TheAmmoniacal on May 30, 2016, 06:24:06 am
https://www.taobao.com/view_image.php?spm=a312a.7728556.2015080703.1.7UZkf9&pic=HEYVDllCEghRV1dcWwocCUQOFx0XTEJfW1lXDBZFAlpcMAl9eXE7MRRUDzQuE1BhaERiZwNbQlpeXV0OSFM=&title=TE0zOTkgUENC&version=2&c=MWIzOTMwMTNhMzgzMDI0OWY4MzY3ZGNiNjE5OWM4NDA%3D&itemId=14154303091&shopId=58559233&sellerRate=25555&fv=9 (https://www.taobao.com/view_image.php?spm=a312a.7728556.2015080703.1.7UZkf9&pic=HEYVDllCEghRV1dcWwocCUQOFx0XTEJfW1lXDBZFAlpcMAl9eXE7MRRUDzQuE1BhaERiZwNbQlpeXV0OSFM=&title=TE0zOTkgUENC&version=2&c=MWIzOTMwMTNhMzgzMDI0OWY4MzY3ZGNiNjE5OWM4NDA%3D&itemId=14154303091&shopId=58559233&sellerRate=25555&fv=9)

What is this PCB?
Title: Re: LM399 based 10 V reference
Post by: Macbeth on May 30, 2016, 07:07:29 am
...bad link...

What is this PCB?

Link doesn't work for me. A better one https://world.taobao.com/item/14154303091.htm (https://world.taobao.com/item/14154303091.htm)

Looks like a way of using up lots of reclaimed S102 10K000's. Perhaps there is a glut of them on the taobao market?
Title: Re: LM399 based 10 V reference
Post by: Dave on May 30, 2016, 11:14:00 pm
Very clean build, Mickle. :-+
What did you use to machine the openings in the case and how did you put the text/symbols on it?
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 31, 2016, 05:52:08 am
LCD opening was made by CNC milling machine. All others - by hand drill.
Title: Re: LM399 based 10 V reference
Post by: Dave on May 31, 2016, 05:04:28 pm
What about the writing and symbols on the front panel?
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 31, 2016, 05:15:00 pm
Laser printer film and 3M adhesive  :)
Title: Re: LM399 based 10 V reference
Post by: Sbampato12 on May 31, 2016, 06:13:36 pm
LCD opening was made by CNC milling machine. All others - by hand drill.

Wich are these banan receptacles? (Manufactures, code, or where to buy?)
They look very good!
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on May 31, 2016, 06:57:27 pm
Sorry, don't know code/mfr of these connectors :(
Title: Re: LM399 based 10 V reference
Post by: ebclr on May 31, 2016, 07:15:35 pm
Ve se este serve

https://world.taobao.com/item/522839915395.htm?spm=a312a.7700714.0.0.4vuSFA#detail
Title: Re: LM399 based 10 V reference
Post by: kridri on June 08, 2016, 11:45:34 am
For my coming voltage reference I was doing my research on ovens and I found a lot of references to two papers of Jim Williams. But when google it or look in the library (University library with connections to other universities) I can't find them. Does some one have them or knowing where I can find them?

It is specific about these papers:
- Temperature controlling to microdegree (MIT 1971)
- Portable wide range chopper stabilized temperature controller (MIT 1974)
Title: Re: LM399 based 10 V reference
Post by: blackdog on June 09, 2016, 10:01:38 am
Hi,  :)

Kridri,
Is this wat your looking for?
www.bramcam.nl/Diversen/EDN-Split-a-temperature-degree-to-10uC.pdf (http://www.bramcam.nl/Diversen/EDN-Split-a-temperature-degree-to-10uC.pdf)

Kind regarts,
Bram
Title: Re: LM399 based 10 V reference
Post by: kridri on June 09, 2016, 12:36:21 pm
Hi,  :)

Kridri,
Is this wat your looking for?
www.bramcam.nl/Diversen/EDN-Split-a-temperature-degree-to-10uC.pdf (http://www.bramcam.nl/Diversen/EDN-Split-a-temperature-degree-to-10uC.pdf)

Kind regarts,
Bram

Hi Bram
Apparently Flinstone has deleted his message.  He also posted the same article like you. I already read it and wanted to read the thesis/paper for which that article is written. But i can't find it online.

Greetings

Kris
Title: Re: LM399 based 10 V reference
Post by: kridri on June 09, 2016, 03:04:47 pm
@ kridri : I have looked for these 2 articles too, I was never able to find them. Especially nothing on Worldcat, it means that unfortunatelly there's no copies of these papers publicy available.

In google scholar, if you type the title and include citations in the search results, you can discover that most of citing papers were written by Jim Williams himself. The others were written in the 70's so there no chance that you can get a copy by contacting the authors.

https://scholar.google.fr/scholar?cites=9898204615121267227&as_sdt=2005&sciodt=0,5&hl=fr

I thought about contacting the MIT library, but as there's nothing in their catalog about this paper, I doubt they could do anything. Maybe the best chance is to contact the Linear Technology staff? I'm quite sure Jim would have kept a copy of this paper in his library.

BTW, if you're looking for papers about sub-milidegree temperature control, plenty were published in the 70's in the scientific journals. BTW-bis, another companion-paper for the MIT oven : "McDerrnott, James. 'Test System at MIT Controls Temperature of Microdegrees." Electronic Design (January 6,1972)" (Not available online).

I also searched in google scholar and via my university library but no hits.  About contacting MIT library, unfortunately you can only contact them as a student or alumi. Otherwise you must visit them. I will send LT a mail with the question and post the result here!
Thanks for thinking with me!

PS: Do you have some good references about sub-milidegree temperature control?
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on July 12, 2016, 07:55:59 pm
Another DIY poor man's multifunction calibrator :) Based on the AD5791B.
DCV 0.1, 1, 10, 100 V. DCI 10 mA, 1 A. Ohms 1k, 10k, 100k, 1M.
LF Noise (0,1-10 Hz): 2.2...2.8 uV p-p (10V/10V), 0.4 uV p-p (0V/10V).
Title: Re: LM399 based 10 V reference
Post by: David Hess on July 12, 2016, 09:41:50 pm
Laser printer film and 3M adhesive  :)

I like to do the same thing.  Print reversed on an overhead transparency sheet, spray with 3M 77 adhesive, flip over, and apply.  Then the Mylar sheet protects the printing.
Title: Re: LM399 based 10 V reference
Post by: Kean on July 12, 2016, 10:43:35 pm
Laser printer film and 3M adhesive  :)

I like to do the same thing.  Print reversed on an overhead transparency sheet, spray with 3M 77 adhesive, flip over, and apply.  Then the Mylar sheet protects the printing.

Yep, and a Silhouette Curio paper cutter can be used to cleanly cut the outline and holes based on alignment marks printed on the sheet, all imported via DXF if necessary.
I've also been using the Curio to cut custom gaskets for seals around connectors, CNC'd parts, etc.
Title: Re: LM399 based 10 V reference
Post by: rob77 on August 19, 2016, 10:15:33 pm
today I've got 10 small cuties in my mail ;) not the original LM399 but close enough (hopefully) :D
does anyone have experience building a reference with "eastern block" MAB399 clones of LM399 ? they have slightly higher voltage but otherwise should be quite similar. so i assume i can safely use the circuits for lm399 , the only difference would be slightly different resistor values because they have 7,25V nominal zener voltage instead of 6,95V.

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=249184)
Title: Re: LM399 based 10 V reference
Post by: TiN on August 29, 2016, 11:18:13 pm
Hmm, what these could be?

(https://doc.xdevs.com/doc/xDevs.com/KX/REFS/kxsb_top_1.jpg) (https://doc.xdevs.com/doc/xDevs.com/KX/REFS/kxsb_top.jpg)

Schematic is same as on original LM399 boards I had, but layout was slightly revised. Both LMx99's are to be assembled on top side now.

(https://doc.xdevs.com/doc/xDevs.com/KX/REFS/kxsb_b_1.jpg) (https://doc.xdevs.com/doc/xDevs.com/KX/REFS/kxsb_b.jpg)

As often, not without minor mistake, such as missing silkscreen for LMx399's  :-//
Title: Re: LM399 based 10 V reference
Post by: Macbeth on August 29, 2016, 11:34:51 pm
TiN, I never understood how the Bourns trimmer was connected. It just seems to be hanging in the breeze?

Also this circuit should work with a single LM399? Did you find the use of two of them really worth it? (Though Linear does supply 2x for free as samples of course!  :-DD )
Title: Re: LM399 based 10 V reference
Post by: TiN on August 29, 2016, 11:43:39 pm
Yes, it's not connected to anything. I just think that leaving empty PCB space is a crime  ::).
Idea (bad one?) behind that is that during debug/tweaking you may want to put a pot there and wire to any fixed resistor in circuit, so you can tweak it to some required value for test. Then once you get the value, pot removed and proper fixed resistor installed instead.

And yes again, circuit will work just fine with single ref. Zener outputs combined together via R3 and R23 to REFIN node of scaling opamp.
There is also possibility to measure both zener outputs via sense outputs (Pin 1 and Pin 4 on J2 connector for top zener, and two tiny exposed vias in bottom right corner for bottom zener).
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 30, 2016, 08:09:35 am

Schematic is same as on original LM399 boards I had, but layout was slightly revised. Both LMx99's are to be assembled on top side now.


Hello,

that reminds me of a "event" that I had recently with one of my LM399´s:
I recognized a large drift on my LM399#2 of about 1.5 ppm within 3 days.
I then recognized that I had put the LM399#2 accidently up side down after a measurement.

When putting back to normal orientation, it needed 2-3 weeks to get back to the
previous reading.
So it might be better not to put the LM399 on the bottom side.
(I dont know if it would stabilize in this orientation).

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: TiN on August 30, 2016, 09:30:23 pm
To my shame, I never really tested LM399-based ones. Just built few boards to test overall function, but no ppm-nitpicking was done.
Time to correct this.
Title: Re: LM399 based 10 V reference
Post by: Nuno_pt on August 30, 2016, 09:54:36 pm
Well, looks like I may move into this adventure to check my K2000.

Since I can´t find any cheap lab around here, official or amateur.

I wish I live near TiN. :-)
Title: Re: LM399 based 10 V reference
Post by: Theboel on August 31, 2016, 02:16:55 am
Hello All,

I have 16 pcs LM399 that I connect to 30v supply for heater and cheap 30K ohm 20ppm resistor for zener after 6 month I found 4 of them are very close the difference about a few uV (according to my 3478A) and then after another 3 month I found the difference between them are 2uV.
My Initial plan are put them in condition like that for aging purpose before I can buy a decent resistor for 10v standard but then I realize I maybe made a mistake, I need to de-soldering and solder again and I read somewhere in this thread that procedure will cycling the aging process again ?  :-//
How long I need to aging this LM399 with lets say edwin resistor before I can calibrated it against "real volt standard"

Thank You
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 31, 2016, 01:17:25 pm

how long I need to aging this LM399 with lets say edwin resistor before I can calibrated it against "real volt standard"

Hello,

this is very individual on the LM399 sample.
Some devices need only 1-2 kHr to stabilize, others need up to 6-12 months (and some stabilize never).
You should make yourself a drift limit (per month or per kHr) and then decide when it is time to calibrate.
A good device will stabilize to around 1-2 ppm/year.

See also diagram here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg836912/#msg836912 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg836912/#msg836912)

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Theboel on September 04, 2016, 11:29:54 am
Hi Andreas and Diligentminds,

Thank you for your time to respond.
I believe time and patient is our enemy for voltage reference.
and for resistor type I prefer use Edwin resistor because the price and availability is best in this parts of the world except for my LTZ1000 I already wait for 3 month for my vishay but I think I still need to wait it looooonger.
btw I just to made sure for one thing. are some one have experience about aging for LM399 or LTZ1000 with "whatever resistor You have" and then when You can get the "proper" resistor are they need to re aging again like a brand new parts ?

Thank You
Title: Re: LM399 based 10 V reference
Post by: Macbeth on September 04, 2016, 04:38:14 pm
I'm trying to follow TiNs LM399 circuit and scratching my head. There are fairly specific resistor networks used for the op-amp gain but they don't add up?  :-//

e.g.

RG1 is made of 4 x 3.75K (R8-R11) in parallel which are in series with 3 paralleled 466 ohm (R12-R14) = 1093 ohm in total

RF1 is made of 2 x 16.5k (R18,19) in series which = 33k ohm for the 20V gain, or is in parallel with 2 x 11.655k (R20,21) which = 13.66k for the 10V gain.

Calculating gain 1+RF1/RG1 gives 31.19 or 13.5 for the 20V and 10V options respectively which are totally out of bounds.  :wtf:

(https://xdevs.com/doc/xDevs.com/KX/VREF/vref_s00_1.png)

I'm sure I must be thick and am missing something obvious, but I've looked through the thread and can't find anything :-//

TiN or anyone please explain and put me out of my misery  :-DD IIRC, Plesa made some of these boards?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 04, 2016, 09:21:18 pm
The way how the current to the LM399 reference is provided looks suspicious. AFAIK the resistors that provide the current (R15 and R16) should go directly to the LM399 and not to the point common to both. As shown the circuit should not work well  - its essentially only one of the LM399 active. With the extra positive feedback, the divider also can die different.

As for the output, i would have preferred an extra isolation resistor, so the output can stand a capacitive load.
Also having filtering at the (+) input of the OP might be a good idea.
Title: Re: LM399 based 10 V reference
Post by: Macbeth on September 04, 2016, 10:26:38 pm
I figured I would let that stuff pass for now. I just don't understand how the RF/RG resistors which seem to be very specifically tuned can possibly work. I mean I get stupid gains of >200V or >90V which are just not possible.  :scared:

The zener current feed R15-R17 is 7485 ohms, and each LM399 has their own 542 resistor to balance things out, so 8027 ohms. With a 20V output that would give 1.6mA to the zeners. 10V only 0.4mA. There are also the optional 200k bias resistors R2 and R22 which could donate 85uA?

I'm not sure of the -24V supply either, but that seems to be an option R5.

But I'm pretty much not an analog guy and am probably making a fool of myself   :popcorn:
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 05, 2016, 07:04:45 pm
With all the resistors populated the amplification does not look right, in addition to using negative and positive feedback together. Sometimes the many parallel resistors can be just optional for variations or maybe different sizes.

Normally it also would not make sense to have so many resistors in parallel and than use different values.
If at all it might be useful to use the same type of resistor several times to get similar properties or reduce the BOM. So for the 7 to 10 V step something like using two resistor in parallel again one of the same to get the amplification of 1.5 as a first approximation and than only add one in parallel to the two to adjust to the right value.

The funny way to provide the current also seems to be in the board, just in the schematics.
Title: Re: LM399 based 10 V reference
Post by: TiN on September 06, 2016, 01:22:18 pm
Well, you guys got me out into the light. I assembled board for test, as I really don't remember any thought when I draw this thing 2 years ago.

1. There is large impact from variable input voltage. Tens to hundreds ppm on output per 1V.
2. I got now why current supply to zener must be two separate resistors, fixed in hw now.
3. Resistors values are incorrect, they need to be matched specific to used LM399's. Currently for test I used 10K + network with total R=4k0855 to get 10V

Btw, even with these mistakes/errors seem to get result TCV about 0.9ppm/K:

(https://xdevs.com/doc/xDevs.com/KX/REFS/399lmx.png)

My datalog python snake was running ACAL DCV on 3458? for each change of temperature >0.3K, to remove DMM's TCV (which is ~0.3ppm/K on my box).

Board itself with bodges:

(https://xdevs.com/doc/xDevs.com/KX/REFS/lmface_1.jpg) (https://xdevs.com/doc/xDevs.com/KX/REFS/lmface.jpg) (https://xdevs.com/doc/xDevs.com/KX/REFS/lmbot_1.jpg) (https://xdevs.com/doc/xDevs.com/KX/REFS/lmbot.jpg) (https://xdevs.com/doc/xDevs.com/KX/REFS/lmres_1.jpg) (https://xdevs.com/doc/xDevs.com/KX/REFS/lmres.jpg)

Let increase current for each zener to ~1.5mA (2K resistors) and test again :)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 06, 2016, 04:18:42 pm
With the LM399 it does not really help to increase the reference current. The reference is not just a zener, but a circuit that sends a fixed current trough the actual zener and the rest only flows through parallel transistors.
So it is mainly a thermal effect from an increased current.

If the two references are close in voltage, the circuit with just one resistor for the current might still work somewhat, with slightly uneven current for both. But the it should be more sensitive to the resistors (that set the current). In the correct circuit the resistors to set the current and those to average between the 2 are not critical.

I don't think there is a big need for an extra low noise voltage regulation in this circuit. By far the dominant noise source are the two LM399 followed by the LTC2057. At least the higher frequency (e.g. > 200 Hz) noise of the reference could be filtered with a capacitor from the (+) input of the OP to ground.
Title: Re: LM399 based 10 V reference
Post by: TiN on September 06, 2016, 04:45:55 pm
With 2K current feed resistors Vz for top one is 7.07705V, bottom 7.06236V, average output after pair of 499 resistors is 7.069731V.

LDO is to keep input VCC to module stable for now at +15V.
Title: Re: LM399 based 10 V reference
Post by: Mickle T. on January 10, 2017, 09:35:55 am
Just another DIY poor man's selfcal DVM & voltage/resistance calibrator :) ADS1256+AD5971B AD5791B
1 ppm INL, 0.25 ppm p-p noise (DCV output), 3 ppm INL (DCV input).
Title: Re: LM399 based 10 V reference
Post by: HighVoltage on January 10, 2017, 10:15:52 am
Just another DIY poor man's selfcal DVM & voltage/resistance calibrator :) ADS1256+AD5971B
1 ppm INL, 0.25 ppm p-p noise (DCV output), 3 ppm INL (DCV input).

That looks much better than a poor man's DIY.
Very nicely done.
Title: Re: LM399 based 10 V reference
Post by: Vgkid on January 10, 2017, 02:47:50 pm
That is really well done.
Title: Re: LM399 based 10 V reference
Post by: quarks on January 10, 2017, 05:41:54 pm
Just another DIY poor man's selfcal DVM & voltage/resistance calibrator :) ADS1256+AD5971B
1 ppm INL, 0.25 ppm p-p noise (DCV output), 3 ppm INL (DCV input).

Looks very good, thanks for sharing
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 10, 2017, 09:10:25 pm
but where did you hide the LM399.
In one of the black boxes near the statistical divider?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Harfner on January 10, 2017, 09:50:39 pm
but where did you hide the LM399.
In one of the black boxes near the statistical divider?

With best regards

Andreas

The LM399 is most probably hidden in the white cube above the black boxes. The third picture does show the main board from below and there you can see a white cover as well. That is a protection against air currents. 
Title: Re: LM399 based 10 V reference
Post by: chickenHeadKnob on January 11, 2017, 12:43:08 am
Quote: "Just another DIY poor man's selfcal DVM & voltage/resistance calibrator :) ADS1256+AD5971B"

Part number for DAC is AD5791B, and is found here:http://www.analog.com/en/products/digital-to-analog-converters/da-converters/ad5791.html (http://www.analog.com/en/products/digital-to-analog-converters/da-converters/ad5791.html)

For those of us that like to play along at home.  8)
Title: Re: LM399 based 10 V reference
Post by: julian1 on January 11, 2017, 05:57:44 am
Quote: "Just another DIY poor man's selfcal DVM & voltage/resistance calibrator :) ADS1256+AD5971B"

Part number for DAC is AD5791B, and is found here:http://www.analog.com/en/products/digital-to-analog-converters/da-converters/ad5791.html (http://www.analog.com/en/products/digital-to-analog-converters/da-converters/ad5791.html)

For those of us that like to play along at home.  8)

The inexpensive one on mouser is AUD $84 and the priciest variant is $323. Do you know what the grade differences are?
Title: Re: LM399 based 10 V reference
Post by: Echo88 on January 11, 2017, 04:00:41 pm
The pricy AD5791SRU-EP operates from -55°C to +125°C, while the standard A/B-grade-AD5791 work from -40°C to +125°C.

Page 17: http://www.analog.com/media/en/technical-documentation/data-sheets/AD5791-EP.pdf (http://www.analog.com/media/en/technical-documentation/data-sheets/AD5791-EP.pdf)
Page 28: http://www.analog.com/media/en/technical-documentation/data-sheets/AD5791.pdf (http://www.analog.com/media/en/technical-documentation/data-sheets/AD5791.pdf)
Title: Re: LM399 based 10 V reference
Post by: branadic on January 20, 2017, 05:58:06 pm
Quote
Just another DIY poor man's selfcal DVM & voltage/resistance calibrator :)

Pretty nice  Mickle T. :-+

I could today measure my a few years running National Semiconductor LM399H using our new Keysight 3458A with semiautomatic readings (Keysight Interactive IO). With 100 NPLC and within 486s I could read:

min:  10.00173439V
max: 10.00173691V
mean: 10.0017357892213V
std: 517.228098173513nV
peak-peak: 2.52µV

Hopefully in near future fully automatic reading is possible. Also didn't know that there is an extra digit using GPIB interface.
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on January 20, 2017, 10:27:16 pm
You do realize that anything past the 6th decimal place in those voltage readings are essentially garbage, no accuracy to speak of?  You realize those calculated amounts you posted are also mostly garbage numbers, assuming that the figures used to calculate them are good to the 6th decimal place, that puts the limit of the calculations well short of the 13 decimal places you display and the std. calculated value isn't really any good past the 2nd or 3rd decimal so why even bother putting them there?  Really, sloppy work as Bob Pease would say.
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 21, 2017, 12:12:17 am
You do realize that anything past the 6th decimal place in those voltage readings are essentially garbage, no accuracy to speak of?  You realize those calculated amounts you posted are also mostly garbage numbers, assuming that the figures used to calculate them are good to the 6th decimal place, that puts the limit of the calculations well short of the 13 decimal places you display and the std. calculated value isn't really any good past the 2nd or 3rd decimal so why even bother putting them there?  Really, sloppy work as Bob Pease would say.

Yes, I can just hear Bob Pease about now winding up for another lecture on this subject :-DD  I was on the receiving end of a couple of those and learned my lesson.

Even if this device were nulled to a JJ-Array -  is only going to be around +-0.3ppm uncertainty on a good day, or +-3uV on a 10V scale.  It is a good practice to try to keep the digits of precision reasonable and within the limits of measurement resolution, no more; even for intermediate calculations.  Otherwise you'll tend to generate un-beleivable results.

Be careful of that extra magical digit on GPIB on HP / Agilent / Keysights - that is -generally- false noise data and not noted for adding -true- accuracy or resolution to your measure.  You don't get something for free there.  Even Keysight will tell you that.

The 399's can certainly be pretty good performers even though they can be fairly noisy.  Nice devices for 6 digit meters and devices that need power cycling - otherwise their relative high noise can be a problem.

Title: Re: LM399 based 10 V reference
Post by: The Soulman on January 21, 2017, 12:42:32 am
The 399's can certainly be pretty good performers even though they can be fairly noisy.  Nice devices for 6 digit meters and devices that need power cycling - otherwise their relative high noise can be a problem.

Ok that seems to be general consensus, noob question: why not L-C filter the heck out of them at the expense of load regulation?
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 21, 2017, 05:06:24 am
The 399's can certainly be pretty good performers even though they can be fairly noisy.  Nice devices for 6 digit meters and devices that need power cycling - otherwise their relative high noise can be a problem.

Ok that seems to be general consensus, noob question: why not L-C filter the heck out of them at the expense of load regulation?

The problem is that it is difficult and nontrivial to filter with a low enough corner frequency to remove low frequency noise without introducing more low frequency noise in the form of drift do to the high impedance required.

But I am not sure what specification MisterDiodes is looking at.  Burried zener references like the LM399 are lower noise than bandgap references but all references are pretty noisy.
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on January 21, 2017, 05:28:40 am
I am quite sure MisterDiodes was referring to the LTZ compared to the LM399, the LM399 being somewhat more noisier and yes, while you can accomplish some filtering of a reference's output, it must be done very carefully or it can end up degrading the output's performance.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 21, 2017, 09:08:22 am
One can filter the higher frequency noise components like > 100 Hz and this sometimes makes sense (some ADCs convert some of that noise to the relevant low frequency range). However there is not much a classical low pass filter can do about the low frequency (e.g. < 1 Hz) range and this noise is often the dominant part due to a strong 1/f component. The obvious way would be a second LM399 or similar to average, or a reference with a higher current (more current to a LM399 does not help).

There is some sense in using 1 or 2 more digits for intermediate results, but not that many. You gain rather little (minute amount of rounding errors), but it is also essentially no extra effort for a computer. For human readings, those extra readings are more distracting than good.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 21, 2017, 09:42:13 am
Quote
You do realize that anything past the 6th decimal place in those voltage readings are essentially garbage, no accuracy to speak of?  You realize those calculated amounts you posted are also mostly garbage numbers, assuming that the figures used to calculate them are good to the 6th decimal place, that puts the limit of the calculations well short of the 13 decimal places you display and the std. calculated value isn't really any good past the 2nd or 3rd decimal so why even bother putting them there? Really, sloppy work as Bob Pease would say.

Noone said anything about the accuracy of the readings right? They are within the given 4ppm of the gear. But resolution carries information too.

I'm really happy that you are working accurate and that you are in the position to criticize. I'm sure Bob Pease would be very proud of you.  :-+
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 21, 2017, 10:15:23 am
Hmm,

here in the volt nut section we don´t throw away any gear nor any digits  >:D

with best regards

Andreas

P.S: at least the long term stability seems to be very high. Allan deviation goes down after 1 minute.
We should have a longer measurement (perhaps with the connectors covered by a piece of cloth).
Title: Re: LM399 based 10 V reference
Post by: Gyro on January 21, 2017, 07:13:00 pm
I am quite sure MisterDiodes was referring to the LTZ compared to the LM399, the LM399 being somewhat more noisier and yes, while you can accomplish some filtering of a reference's output, it must be done very carefully or it can end up degrading the output's performance.

Something I've wondered about for a while, is it possible that too low an input impedance filter (ie. low R in an RC filter) could degrade a reference Zener - I guess I'm thinking LTZ rather than LM399 where the Zener is buffered. Given that the breakdown is at least partially avalanche, would having a capacitor more or less directly across the Zener, feed extra energy into the junction as it breaks down noisily, possibly shifting its characteristics over time?

The nearest analogy I can think of is a GM tube, where parallel capacitance across the tube needs to be minimised to avoid shortening its life. A simple Neon bulb oscillator (series R, parallel C) would be another.

I don't think I've ever seen anything written on this in relation to Zeners. Sorry if it's a little OT.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 21, 2017, 08:38:38 pm
The reference side of the LM399 / LM329 is not just a zener-diode, but a buffered circuit, that internally set the zener current and provides a low impedance output.
This is why there is a limited effect of having a large cap directly in parallel to the LM399 and not much sense in running it with much more than 1 mA. The internal circuit also makes the LM399 less sensitive to changing current than a normal reference zener.

Capacitance directly parallel to a zener is usually not a problem, but also of limited effect due to the low impedance.

Even with the LM399 filtering is difficult, though it can be done to a limited degree. The 0.1 Hz -10 Hz range often used in DS is just much more convenient to measure. But usually it is more the 1 mHz-100mHz range that is really causing trouble.
Title: Re: LM399 based 10 V reference
Post by: Gyro on January 21, 2017, 09:05:01 pm
Sure, as you say, the 399 is buffered and the Zener current controlled, and it would also be silly to have a very low R in an RC filter.

The question I was asking really relates to the LTZ or discrete Zeners (and therefore getting seriously OT). I was just curious about the effect that low impedance parallel capacitance might have on the physics of Zener / avalanche breakdown (at the microstructure level) and whether it could cause permanent shifts in the breakdown voltage. Probably of academic relevance only -  I just wondered if operating a Zener in the lowest possible capacitance environment could subtly improve its long term drift.
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 22, 2017, 11:09:30 am
I am quite sure MisterDiodes was referring to the LTZ compared to the LM399, the LM399 being somewhat more noisier and yes, while you can accomplish some filtering of a reference's output, it must be done very carefully or it can end up degrading the output's performance.

That must be it; I was thinking of in comparison to bandgap references.  The LTZ1000 has an advantage in that the zener operating current can be controlled while other zener references like the LM399 are fixed.  It is easy enough to distinguish bandgap and zener references by their output noise which is handy since manufacturers do not always distinguish them by type in their selection guides; for the later, noise is at least 5 to 10 times higher.

Other than the LTZ1000, do any integrated zener references allow adjustment of the zener current?  Even the 723 regulator fixes its zener current internally.

The question I was asking really relates to the LTZ or discrete Zeners (and therefore getting seriously OT). I was just curious about the effect that low impedance parallel capacitance might have on the physics of Zener / avalanche breakdown (at the microstructure level) and whether it could cause permanent shifts in the breakdown voltage. Probably of academic relevance only -  I just wondered if operating a Zener in the lowest possible capacitance environment could subtly improve its long term drift.

I have never heard of any such effect even though shunt capacitance is sometimes added to reduce susceptibility to EMI.  The junction is irrevocably altered or damaged depending on perspective when initially operated in breakdown so I doubt the lower AC impedance from a shunt capacitor will have any additional effect and the junction's series resistance places a lower limit on the impedance seen by the junction anyway.
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on January 22, 2017, 06:23:38 pm
The easiest and most effective way of reducing LM399 noise is to parallel them, up to about five, but then you're getting close to the cost of an LTZ which will have a lot lower noise than even five paralleled LM399s.  So if you really need that low of noise, just get an LTZ, it's easier.  A LM399 is good for about 6-7 digits, with a little extra higher frequency filtering to help keep that last digit from bobbling around too much.  The high frequency Gaussian noise can be filtered some and averaging the output of the Gaussian noise portion will be good enough to clean up the LSD but it will have no effect on the 1/f noise, that cannot be averaged out unless you employ a very long averaging period (no 100 NLPC is no where near long enough for 486 seconds).

The LM399 will likely have achieved its very low drift term after several years of operation, that isn't the problem, it is measuring the DC value hidden underneath all of that noise.  I would have confidence in Mickle T's readings to probably a few PPM; 10.001735 ±.00005 at most, everything after that is essentially junk buried in noise and no amount of math is going to improve that, the actual reading accuracy is the limit and that cannot be changed by any manipulation of numbers.  The uncertainty of the measurement also puts concrete limits on any measurement or calculation, the ±4 PPM accuracy of the 3458A is not the only limit at work here.  There certainly would be less question of the measurement if it was an LTZ1000/A because of the much lower noise present.
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 22, 2017, 06:29:05 pm
I am quite sure MisterDiodes was referring to the LTZ compared to the LM399, the LM399 being somewhat more noisier and yes, while you can accomplish some filtering of a reference's output, it must be done very carefully or it can end up degrading the output's performance.
Sorry - didn't make that clear.  LM399 Noise spec is 20uV, 0.5ppm TC, drift is typically lower to mid ppm first year, although can improve with age - they are also tolerant of frequent power cycles.  LTZ1000 noise spec is 1.2uV, 0.05ppm TC, drift is typically low ppm per year; usually gets better with time.

You get what you pay for.

If you're building a Vref for ADC or DAC in the 20 bit +  range,usually LTZ is better choice.  '399's can be good performers but noise is an issue - you -could run 5 or more '399's in parallel to reduce noise (more than that is diminishing returns), but at that point might as well run an LTZ to lower parts count & cost.  Trying lowpass filter a precision Vref can be a fool's errand - The best way to combat noise is do everything possible to NOT generate it in the first place.

Title: Re: LM399 based 10 V reference
Post by: Andreas on January 22, 2017, 06:48:51 pm
Sorry - didn't make that clear.  LM399 Noise spec is 20uV, 0.5ppm TC, drift is typically lower to mid ppm first year, although can improve with age - they are also tolerant of frequent power cycles.  LTZ1000 noise spec is 1.2uV, 0.05ppm TC, drift is typically low ppm per year; usually gets better with time.

You get what you pay for.


Sorry,

but you are comparing apples with pies.

the 20uV(eff) is broadband noise of the LM399. (10Hz - 10kHz).
Whereas the 1.2uV(pp) is 1/f noise (0.1-10Hz).

My LM399 typically are around 4 uVpp (0.1-10Hz).
(some better and some also worse).

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 22, 2017, 07:00:42 pm
Hmm,

here in the volt nut section we don´t throw away any gear nor any digits  >:D

with best regards

Andreas


Hmmm indeed - There would be a lot of Professional Engineers / Metrologists / VoltNUts that would debate the issue of fantasy digits with you - but for for good engineering practice it is best to always keep track of your measurement uncertainty, measurement confidence, and keep your results reasonable to your test setup without adding "phantom accuracy" along the way. 

Which by the way is something not up for debate - as a professional engineer, if I were to submit measurement results to an ISO 900x quality control audit team, the very first thing they check is the equipment used to make a measure and basic measurement accuracy, the last calibration date, the next calibration due date, the calibration history, any extrapolated calibration estimates based on past calibration cycles, and so on. 

Then the very next thing that would be examined is any subsequent calculations based on those measurements, and at any point was -any- apparent accuracy increased by the improper use of extra digits - for instance using 13 digits of precision to calculate a measurement made with a 6ppm accurate DMM.  That would be cause for immediate loss of certification right there.

Everything gets boiled down to double-check that a certain testing procedure, if we are delivering a device that operates to some PPM accuracy - really does operate to that level, and is traceable back to NIST at every step.  The whole idea is to cross-check at every step that accuracy is not lost or gained.  High Resolution is not data if accuracy isn't there to match.

That is the only Head's Up there, and absolutely no offense intended to you or Branandic, and I apologize if it was taken that way.  It wasn't meant to.  Keep Building and Have Volt Nut Fun!  But along the way it is helpful if you can be a student of good metrology lab practice, and learn what passes for data - and not - in the real world.

The biggest thing to keep in mind is your initial measurement uncertainty, as that limits the rest of your calculations.  As I wrote in my Price of Chasing PPM thread - the best way to measure a precision Vref is via 732a/b, mechanical null meter and Kelvin Varley divider.  Not only can you do that measure under battery power only, you get your 10Hz bandwidth limiting done for you.  It's much harder to get a good .1 to 10Hz bandwidth measure from a 3458a.  It can be done, but the noise the DMM adds is an issue - besides the fact a 3458a isn't really considered the best voltage transfer device for absolute measures.

And OF COURSE the best way to INCREASE measurement accuracy, DECREASE uncertainty and INCREASE measurement confidence is to acquire more equipment  :-DD  So I will 100% agree with you there.  More accurate equipment always helps!

Now the mark of a REALLY good Volt Nut is if their life savings is depleted buying lots of 732a/b AND they get buddies to do the same and store everything at VoltNuts house...
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 22, 2017, 07:09:08 pm
Sorry - didn't make that clear.  LM399 Noise spec is 20uV, 0.5ppm TC, drift is typically lower to mid ppm first year, although can improve with age - they are also tolerant of frequent power cycles.  LTZ1000 noise spec is 1.2uV, 0.05ppm TC, drift is typically low ppm per year; usually gets better with time.

You get what you pay for.


Sorry,

but you are comparing apples with pies.

the 20uV(eff) is broadband noise of the LM399. (10Hz - 10kHz).
Whereas the 1.2uV(pp) is 1/f noise (0.1-10Hz).

My LM399 typically are around 4 uVpp (0.1-10Hz).
(some better and some also worse).

with best regards

Andreas

The comparison still applies:  at 10Hz and below (what you normally need), LTZ is still much lower noise, and at 4uV '399 noise you'd still need 4~ 5ea '399s in parallel to get the noise performance of an LTZ (TC is still 10x worse on '399) - and at that point it's lower BOM count and cost to just build an LTZ in the first place IF THAT is appropriate to the application.  AND you get better TC also.

There are of course times when a '399 will work - and of course use it! But for lowest noise - which is what we're after most of the time - the LTZ's will get you there faster and cheaper a lot of the time, as a practical matter.

No matter what, you still get what you pay for...
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 22, 2017, 08:05:19 pm
To get the noise down from 4 µV to 1.2 µV , that is a little more than a factor of 3, one needs something like 10 or 12 of the LM399. Reduction of noise goes with the square root.

Things can get even more in favor for the LTZ1000, as the LTZ1000 at higher current has a low noise corner and thus the 0.1 -10 Hz value is not yet fully 1/f noise, but still has quite some white noise contribution. So for the often more relevant (but not so easy to measure) 0.01-1 Hz range the LM399 should be at 4 µV again while the LTZ1000 should be below 1 µV, more like 0.5 µV.

If can still find a few (like 4) of the LM129 (essentially LM399 without the heater part), there would be the option to use a few of them together (series or parallel) and add temperature stabilization (to a lower temperature than the LM399) by hand - this way temperature stabilization could also apply to the scaling resistors. Still this needs a lot of testing, compared to the more or less ready made LM399, that needs only little support circuit.
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 22, 2017, 08:19:48 pm
To get the noise down from 4 µV to 1.2 µV , that is a little more than a factor of 3, one needs something like 10 or 12 of the LM399. Reduction of noise goes with the square root.


Correct.  I was going with the benefit of the doubt, comparing a group of lower noise '399's - which can be as low as a few uV on the 10Hz spectrum if you're lucky.  Really though it is diminishing returns after you get about 4 or 5 '399's in parallel, and averager amp issues, etc.  Plus even with parallel '399's the TC isn't anywhere near as good as LTZ.

Title: Re: LM399 based 10 V reference
Post by: Andreas on January 22, 2017, 08:21:52 pm
Hello,

Ok now we are near truth for the noise.
LM399 is typically around factor 2-3 more noisy than a LTZ1000.
1.2uVpp typ against 4uVpp typ 1/f noise
and
4uV eff typ against 7uV eff typ for the broadband noise
(4uV taken from the 40nV/sqrt(Hz) from the diagram with bandwidth of 10kHz).

For the T.C. we have still to work on. (have you really measured your T.C.s?)
From my measurements at least near the typical "lab temperature range" my LM399
have around 7-10uV change over a > 30 deg C range. (0.05 ppm/K)

A undadjusted LTZ1000A (without R9) showed 35uV change over a 22 deg C range (0.2 ppm/K)
so not all LTZ1000 have the typical 0.05ppm/K value.

So the price/performance ratio of the LM399 is not too bad.

With best regards

Andreas



Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 22, 2017, 10:23:26 pm
..If you search and sort you can find '399's that are -only- 2X to 3X as noisy as LTZ, and maybe run those in parallel, then you could have somewhat low noise like an LTZ gives you.  No matter what - If you're sorting thru '399's to find the lower noise units you still have to pay for the majority of the '399's that are much noisier than that - so in a production situation I wouldn't say that '399's are really that profitable compared to an LTZ if noise is a factor.  TC still spec'd at 10X worse of course on '399, and in that there is no real comparison.  We normally get well within all datasheet specs on our LTZ's.

For what I do most of the time, <very> low noise and TC are top priorities in a Vref anyway. Long term drift rate is somewhat important but on a production situation these will all come with a required calibration interval customer maintains, whatever that is - 12 month, 6 month, whatever.

For production BOM pricing, you need to go with just datasheet specs (not cherry picking VRefs) - and accounting will you still tell you: You get what you pay for.  Period.  Generally LTZ gets you what you need for lowest noise for a better profit margin at the end of the day.

If a '399 works for your application, then that is great.  But if you need the whole package of (much) lower noise, better TC and generally low drift rate - the LTZ is a much better choice in general.

The '399 can certainly settle down to a low drift rate - and several 3456a meters we've got on line haven't needed a cal adjustment for decades (even with pots!!), so that is a testament that for some <higher noise tolerant> applications where they can be just the ticket. 



Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on January 22, 2017, 11:47:57 pm
After checking two versions of the LM399 data sheets (an old National Semi and newer Linear Tech), both had very similar data, in the National a scope photo showing nearly two divisions (5uV/division) of noise (0.1Hz - 1.0 Hz, 2 minutes/div.) and the LT showing a probable strip chart recording showing virtually identical noise spanning 25 sec./div.  Both displaying the typical spec of 7uV of low frequency noise.  As we all know, some units will have a bit lower noise and others will have more noise, the LM399 version has a maximum spec of 50uV noise, the LM199 has a maximum of 20uV.  The bulk of the production will be around the 7uV noise spec, certainly you can sort out for lower noise but the fact is that the manufacturer has already gone through and pulled the best performing chips for the higher priced versions, the LM199 and military types so if you happen to find a LM399 with noise significantly lower than 7uV, count yourself as lucky, there probably aren't too many of them around.
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 23, 2017, 03:49:59 am
OH that's right - I had forgotten the older NS datasheet has slightly different specs.  It still shows you if you have the opportunity to pick up a working strip chart recorder, they can be very useful at 10Hz!  Null meters will tend to have a strip recorder output driver, and that's handy.  Plus they make fascinating museum pieces for friends and family to ogle. 

The 399's we had tested before were certainly more around the 5uV ~ 10uV + noise range (some quieter), but it's been a while since we used any.  To get down to 2 or 4uV noise you'd certainly have to sort thru a bunch.  They have their application but aren't going to replace LTZ's

Our customers these days need the faster high bit rate, higher bit count ADC's, and you really need to have as low a Vref noise as possible - and so we normally stick to the LTZ's for that, and using PWW to keep noise to a minimum too (Edwin's resistors work fine!).
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 23, 2017, 04:35:39 am
OH that's right - I had forgotten the older NS datasheet has slightly different specs.  It still shows you if you have the opportunity to pick up a working strip chart recorder, they can be very useful at 10Hz!  Null meters will tend to have a strip recorder output driver, and that's handy.  Plus they make fascinating museum pieces for friends and family to ogle.

My old vector CRT DSOs are enough of a museum piece for show and tell.

What is the best way to make those nice datasheet quality nV/sqrt(Hz) versus frequency measurements extending to low frequencies which show 1/f noise without a dedicated analyser?  I usually make do with RMS and peak-to-peak spot noise measurements.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 23, 2017, 04:37:00 pm
To get the noise density data of a reference / voltage sources one usually uses a low noise AC coupled amplifier and than digitize the data and let a computer (e.g. PC) do the FFT (or similar). The tricky part can be AC coupling with very low frequency and still low noise. Something like 0.1 Hz is relatively easy, 0.01 Hz and lower start to get tricky.  Sometimes one can look at the difference of two more or less equal references - elimination AC coupling. For amplifiers it gets easier, as the DUT can do the first step of amplification itself and no possibly critical AC coupling is needed.

Usually the ADC/digitizer does not have to be especially good quality if the amplitude is well adjusted.

For the very low noise region (e.g. batteries) there is a correlation method, using two channels and lot more math.
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on January 23, 2017, 05:54:00 pm
To David Hess:

I quite agree with MisterDiodes method of measuring low frequency noise (or DC voltage for that matter):

"The biggest thing to keep in mind is your initial measurement uncertainty, as that limits the rest of your calculations.  As I wrote in my Price of Chasing PPM thread - the best way to measure a precision Vref is via 732a/b, mechanical null meter and Kelvin Varley divider.  Not only can you do that measurement under battery power only, you get your 10Hz bandwidth limiting done for you.  It's much harder to get a good .1 to 10Hz bandwidth measure from a 3458a.  It can be done, but the noise the DMM adds is an issue - besides the fact a 3458a isn't really considered the best voltage transfer device for absolute measures."

If you don't have a 732A/B available, you could possibly substitute a stable low noise battery as the reference source and null it against the DUT, since you are interested in the noise for this particular measurement, the precise DC voltage is unimportant and just becomes a drift component.  Most null meters have an output available and this already has limited bandwidth, fairly simple setup, you get the noise voltage out of the null meter and a good null meter will have very low AC noise, usually something on the order of 0.2uV - 0.3 uV.

Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on January 23, 2017, 08:33:18 pm
That is a very good point, Edwin - If the most important thing you're after is low noise and not absolute value on your Vref, you just need a -quiet and stable- voltage source to null against.

Handy stuff to know, if you haven't acquired a Null meter in your collection yet:

The most beautiful thing about a Null meter is the input impedance goes essentially to infinity at null (not exactly there but well into GOhms range), and for an economical price you get very good accuracy on the 1uV scale or lower (2% or better accuracy on a 1uV scale or lower ain't shabby!) - and all on battery power as well.  Self calibrating for the most part, even with the 7-decade Kelvin-Varley divider at one side.  No switching noise to worry about, no CPU digital clock noise, no weird ADC artifacts mucking up the results.  People worry about stiction on the meter needle but really that is very, very hard to see if you've ever used one, as long as you keep your measures to reasonable accuracy limits.  Another technique is to flip your meter connections over to spot the true null - at that magical point you will see no real needle deflection as you swap your + and - connections over, and that helps you see thermal problems as well.

Once you hit the sweet spot on your KVD dials, you'll see the meter needle drift slightly and equally around zero point, and you'll get a very accurate idea of your 10Hz and below noise levels.  You'll probably get started on the 10uV or 20uV or higher scales until you get things dialed in on the KVD and everything warmed up.  Later on with an LTZ you can usually switch down to 1uV meter scale and watch the noise gently from there, if you have a very quiet voltage source to work with.  That's when you know things are really working right!

Watch your shield / guard and grounds!

Keep your device under test in a well-shielded box, but as you get into lower PPM, you find that a shielded, low noise testing room situation will be a big help also.  As much as possible switch off all digital power supplies nearby, and look for any wall-warts too!  Watch out for mains-powered LED and Flour. lighting!  Those buggers can really create problems.  Incandescent or even battery-power incandescent is a good way to light, as long as you keep the heat farther away.  LED's running on pure DC can work also.

Keep a lookout for power supplies plugged in on the -other- side of the wall also!

Handy Tip: Setting the Null meter up so you can see the needle through an indoors testing room window is helpful, if that can work for you.  OR take a peek thru an cracked-open door.  That way you can non-thermally and non-drafty check on things while you get other stuff done during the day without approaching the test bench.

Another trick we've used:  Use the Null Meter 1V strip chart output to drive a quiet analog optic fiber cable driver connected to a fiber cable, and now you can use a quiet analog fiber receiver and your favorite DMM (maybe in the next room) to monitor and log measurements, if you must.  Not quite as accurate and adds a bit of TC but let's you coarse-check on the noise test without going anywhere near the shielded testing area, and no possible ground loops.  Much, much quieter than any wired / shielded copper or RF connection.

You might have to re-zero the meter now and then during the day, but if your lab temp is constant that isn't a huge issue.  If you see your meter needle drifting off-center don't assume your Vref is drifiing; that usually just means you have to touch up the meter zero point a little - which has no effect on noise measurment.  If you have to adjust the KVD, that means either your Vref or voltage source has drifted, so you'll have to sort that out - but for a noise test that probably isn't long enough time scale to worry about.

These are old & slow techniques, but they always work without firmware upgrades or digital noise - and you get stable, accurate and repeatable results down to low PPM with some practice.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 24, 2017, 07:46:19 pm
Hello,

my opinion:
if noise is your key spec you have to measure every single device.
At least if there are no 100% production tested guaranteed maximum values given (only typical).

Also the noise diagrams seem to be mostly not measured:
how do you explain that a LM329 and a LM399 of LT have exactly the same peaks and dips over time on their low frequency noise diagram.

And how can a LT1021 have exactly the same noise pattern as a LT1236?
Ok they have a similar chip (except the trimming) but how do they get absolutely synchronous noise?
 :-//  :bullshit:  :popcorn:

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Cerebus on January 24, 2017, 08:18:14 pm
Once you hit the sweet spot on your KVD dials, you'll see the meter needle drift slightly and equally around zero point, and you'll get a very accurate idea of your 10Hz and below noise levels.

I've gone back to using analogue meters on the bench for anything where spotting low frequency (say 0.1 Hz to 5 Hz) effects is necessary. Following a needle drifting about is so much easier than mentally re-creating the trend from a digital display.
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 24, 2017, 08:52:47 pm
if noise is your key spec you have to measure every single device.
At least if there are no 100% production tested guaranteed maximum values given (only typical).

Noise takes a lot of time to test so if you want a device with a maximum noise specification, then you have to either pay extra for it or grade it yourself.  For some devices they do a fast medium frequency spot noise measurement and assume that it correlates with low frequency noise.

Quote
Also the noise diagrams seem to be mostly not measured:
how do you explain that a LM329 and a LM399 of LT have exactly the same peaks and dips over time on their low frequency noise diagram.

And how can a LT1021 have exactly the same noise pattern as a LT1236?
Ok they have a similar chip (except the trimming) but how do they get absolutely synchronous noise?

The graphs only represent typical values and for devices which share the same design; the differences are not significant.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 24, 2017, 09:08:19 pm
The graphs only represent typical values and for devices which share the same design, the differences are not significant.

Has the LM329 (a unheated device) really the same noise as a heated LM399?
Is the noise in the diagram typical for the heated or the unheated device?

 :-//

By the way: does anyone have a datasheet for the LTFLU-1 with noise diagram?

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 24, 2017, 09:41:26 pm
I have a copy of a NS LM399 datasheet from 1999. It shows noise densitiy curves for 25 C and heated chip. The noise for the cold version only about half the value. The LF noise curve shown is for the heated chip (and different from the one shown below) - looks 8 µV_pp for the 0.01 -1 Hz range.

They also show an internal circuit - identical the the LM329 for the reference part.
So i would expect a lower noise for the relatively cold LM329/LM129 compared to a hot LM399.

The noise density curve for some reason slightly lower for the  LTs version of the LM329, but not much.
Title: Re: LM399 based 10 V reference
Post by: dr.diesel on January 24, 2017, 10:07:28 pm
By the way: does anyone have a datasheet for the LTFLU-1 with noise diagram?

I think the last LTFLU datasheet was last seen in Jimmy Hoffa's back pocket.
Title: Re: LM399 based 10 V reference
Post by: Theboel on January 25, 2017, 01:59:31 am
I have a copy of a NS LM399 datasheet from 1999. It shows noise densitiy curves for 25 C and heated chip. The noise for the cold version only about half the value. The LF noise curve shown is for the heated chip (and different from the one shown below) - looks 8 µV_pp for the 0.01 -1 Hz range.

They also show an internal circuit - identical the the LM329 for the reference part.
So i would expect a lower noise for the relatively cold LM329/LM129 compared to a hot LM399.

The noise density curve for some reason slightly lower for the  LTs version of the LM329, but not much.

I struggle to understand why the cold zener chip has less noise compare to hot one ?
and second question how to limit the temperature of LM399 ? what about current limiter for heater 
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 25, 2017, 01:34:35 pm
I struggle to understand why the cold zener chip has less noise compare to hot one?

I wondered about this a little bit when looking at the LM339 datasheet which has different noise curves at two different temperatures but did not consider it seriously until you mentioned it thereby ruining my digestion.  The noise from zener and avalanche behavior is primarily shot noise which has no dependence on temperature.  Does avalanche current multiplication have a temperature coefficient?

My guess is that this has to do with the non-monotonic behavior of avalanche breakdown noise at low current (https://archive.org/details/NoiseBehaviourOfZenerDiodes) which I suspect is do to irregularities which allow different areas of the junction to breakdown at different voltages.  Avalanche rated rectifiers and transient voltage suppressors are specially processed to prevent this to reduce spot heating.  This would also explain why some temperature compensated zener diodes are significantly noisier than others at low currents.

If the reference was not already low noise, a change in avalanche noise with temperature would never be noticed and the LTZ1000 should not display this behavior when operated at a higher current.

Quote
and second question how to limit the temperature of LM399? What about current limiter for heater?

Based on the schematic, it looks to me like the temperature regulation circuit can be set to a lower value by using a fixed voltage below 9 volts.
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on January 26, 2017, 02:49:51 am
I just checked my 1980 National semiconductors handbook regarding the noise figure.
LM129 noise powerdensity ( nV/sqrt(Hz)) is reflected in both LM399 and LM3999 spec when not operating with heater .

One of the app notes on LM129 suggested that the powersupply rejection might be improved by lowering operating current ( that is increasing bias resistor).
In addition it is possible to further improve the performance by incorporating the device into a bridge like configuration.
Namely form a bridge out of two biasing resistors supplying current to thedevice and a resistor respectively.The resistor is sized to reflect the dynamic impedance of the device and output is taken from the bridge ( device and the resistor).

One of those .. ahaaa  moments.
Title: Re: LM399 based 10 V reference
Post by: TiN on January 26, 2017, 04:15:53 am
Quote
By the way: does anyone have a datasheet for the LTFLU-1 with noise diagram?
There is a datasheet? I surely missed it.  :-X
Title: Re: LM399 based 10 V reference
Post by: HighVoltage on January 26, 2017, 10:44:00 am
Quote
By the way: does anyone have a datasheet for the LTFLU-1 with noise diagram?
There is a datasheet? I surely missed it.  :-X
I have tried to get one and gave up.
I don't think it is publicly available.
Title: Re: LM399 based 10 V reference
Post by: zhtoor on February 15, 2017, 05:04:20 am
Hello everybody,

especially edwin, andeas, mrdiodes and mickle t. !

am new to this subject and i want to setup a small metrology lab. as a startup excercise i have the following in mind:-
(please bear with me, i am a newbie!)

1. acquire a bunch of LM329's (commercial ones from TI or LT)
2. use some kind of delamination procedure (methods required here!) to strip them off the epoxy/plastic encapsulation.
3. characterize / select these de-laminated parts for an averaging type voltage reference in a custom heating assembly.
4. ie; trying to do an LTZ1000 on the cheap, using selected LM329's.

since all of you are from advanced countries with lots of hands-on experience, please do guide me.

regards.

-zhtoor
Title: Re: LM399 based 10 V reference
Post by: chickenHeadKnob on February 15, 2017, 10:46:41 am

1. acquire a bunch of LM329's (commercial ones from TI or LT)
2. use some kind of delamination procedure (methods required here!) to strip them off the epoxy/plastic encapsulation.
3. characterize / select these de-laminated parts for an averaging type voltage reference in a custom heating assembly.
4. ie; trying to do an LTZ1000 on the cheap, using selected LM329's.

No need to delaminate, it would be a counter productive exercise. Epoxy packaging is fine providing your oven is reasonably sealed. It should keep the parts and circuit board at a steady humidity.  For the best effect the oven should be continuously powered (hot). Thermal cycling is bad for drift and when the oven cools it will have the propensity to draw in outside air and moisture. LM329's are quite noisy compared to LTZ1000
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 15, 2017, 04:44:03 pm
If you are after a low cost, low noise 10 V source, the Chinese 2DW23x (from the right source) might be an option: With an adjusted current one can get a low TC and thus the oven around it would not be very critical. They are supposed to be low noise and with a voltage of about 5.3 V for just the zener and thus 10.6 V for two in series, the divider towards 10 V would be less critical than going from 7 to 10 V with a resistive divider / amplifier.

Stripping hte LM329 is likely not such a good idea - more like getting old / used LM129, with some risk in sorting through fakes.
Title: Re: LM399 based 10 V reference
Post by: Conrad Hoffman on February 15, 2017, 05:52:35 pm
Just curious- you can still buy all manner of 1Nxxxx series zero TC zeners. I've made some really decent references with those, an OP-07 and a few resistors, the standard self-biased circuit. Is a 2DW23x really any better than 1N821 or 1N45xx or 1N935-938, or just cheaper?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 15, 2017, 07:24:34 pm
At least the specs for the classical zener refs like 1N821 show quite a high noise level. However there is usually only one number in the 1 µV/Sqtz(Hz) and without noting a frequency. If this is for low frequency (like 0.1 -1 Hz) this noise figure would be not that bed. If it is the higher frequency (e.g. kHz) range, this noise level would be poor. 

The 2DW23x are lower voltage zeners (e.g. 5.3 V), and these tend to have less noise. However the lower voltage zeners also usually have a higher differential resistance and this makes a reference circuit more sensitive to resistor drift. So I would expect an trade off between low noise and long term stability. At least in some cases the noise for the 2DW23x is very low, even lower than something like an LTZ1000 or similar. However the TC is not very low and even with an adjusted current to get zero linear TC, there is a considerable second order term - so an oven is more or less needed to get it really stable.

The case of starting with something like 10.5 V form 2 zeners and than divide down to 10 V could more than compensate for a higher sensitivity to the current setting resistor, because drift of the divider would be less important than in the 7 -> 10 V amplifier case that is often used with the LM399.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 15, 2017, 08:08:45 pm

2. use some kind of delamination procedure (methods required here!) to strip them off the epoxy/plastic encapsulation.
3. characterize / select these de-laminated parts for an averaging type voltage reference in a custom heating assembly.


Hello,

I also do not understand why you want to delaminate the epoxy.
I would either use the metal can devices or put the epoxy types into a hermetically sealed housing.
Perhaps with some baking of the epoxy before sealing (with glass).

To reach the LTZ 1/f noise spec you would have to average ~16 LM329 devices (selected for low noise of ~4-5 uVpp)
There was a teardown with pictures of a commercial LMx29 reference here in the board somewhere.
I would start there. perhaps you get the handbook with schematics.

good luck

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Andreas on February 15, 2017, 08:20:33 pm
Just curious- you can still buy all manner of 1Nxxxx series zero TC zeners. I've made some really decent references with those,
Hello,

did you measure 1/f noise of the zener?
I got some 1N829A of different manufactures.
NOS of Motorola and ST (cheap) and some new APD (5 EUR / piece).

I measured different noise from device to device (typical 1.7-3uVpp for Mot+APD)
and very different between ST (up to 22uVpp) and the others.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: zhtoor on February 15, 2017, 08:37:49 pm

2. use some kind of delamination procedure (methods required here!) to strip them off the epoxy/plastic encapsulation.
3. characterize / select these de-laminated parts for an averaging type voltage reference in a custom heating assembly.


Hello,

I also do not understand why you want to delaminate the epoxy.
I would either use the metal can devices or put the epoxy types into a hermetically sealed housing.
Perhaps with some baking of the epoxy before sealing (with glass).

To reach the LTZ 1/f noise spec you would have to average ~16 LM329 devices (selected for low noise of ~4-5 uVpp)
There was a teardown with pictures of a commercial LMx29 reference here in the board somewhere.
I would start there. perhaps you get the handbook with schematics.

good luck

with best regards

Andreas

Hello.

first of all, thanks for the response.

second.

i wanted to delaminate the epoxy to establish a better temperature distribution to the references being averaged, and also to give me a bigger
latitude to confine a relatively large number of these in a small space (16 to 64pcs, dont know yet).

regards and thanks.
Title: Re: LM399 based 10 V reference
Post by: zhtoor on February 15, 2017, 08:45:15 pm
Just curious- you can still buy all manner of 1Nxxxx series zero TC zeners. I've made some really decent references with those, an OP-07 and a few resistors, the standard self-biased circuit. Is a 2DW23x really any better than 1N821 or 1N45xx or 1N935-938, or just cheaper?

Dear Sir,

thanks for the response.

1N829 is available at ~USD 12 a piece., quite expensive.

http://www.digikey.com/products/en/discrete-semiconductor-products/diodes-zener-single/287?k=1N829&k=&pkeyword=1N829&pv69=80&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25 (http://www.digikey.com/products/en/discrete-semiconductor-products/diodes-zener-single/287?k=1N829&k=&pkeyword=1N829&pv69=80&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25)

which is probably a selected 5.2V zener in series with a regular forward biased diode, both selected to form a concave curve of voltage variation vs. temperature,
and the valley coinciding with around 25 degC.

if the same thing could be done by characterizing a bunch of 5.2V zeners and regular -2,2mv/degC diodes, could be interesting.

Regards and thanks.
Title: Re: LM399 based 10 V reference
Post by: Conrad Hoffman on February 15, 2017, 09:14:33 pm
Just curious- you can still buy all manner of 1Nxxxx series zero TC zeners. I've made some really decent references with those,
Hello,

did you measure 1/f noise of the zener?
I got some 1N829A of different manufactures.
NOS of Motorola and ST (cheap) and some new APD (5 EUR / piece).

I measured different noise from device to device (typical 1.7-3uVpp for Mot+APD)
and very different between ST (up to 22uVpp) and the others.

with best regards

Andreas

I haven't, but a friend who built commercial equipment did, and he said something interesting. It seems if you select the quietest parts from a batch, those will also have the least long term drift.

The zero tc diodes seem to have gotten quite expensive, so maybe selection isn't a good strategy these days. It does make me wonder if one could pot together pairs of conventional (read cheap) zeners that were selected for noise- roll your own!
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 15, 2017, 09:20:05 pm

I haven't, but a friend who built commercial equipment did, and he said something interesting. It seems if you select the quietest parts from a batch, those will also have the least long term drift.

Hello,

thats also something that I read somewhere.
But I still do not know if the 1/f noise or the popcorn noise is meant.
I guess its the popcorn noise since it is related with impurities / imperfections of the silicon.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: zhtoor on February 15, 2017, 09:21:50 pm
Just curious- you can still buy all manner of 1Nxxxx series zero TC zeners. I've made some really decent references with those,
Hello,

did you measure 1/f noise of the zener?
I got some 1N829A of different manufactures.
NOS of Motorola and ST (cheap) and some new APD (5 EUR / piece).

I measured different noise from device to device (typical 1.7-3uVpp for Mot+APD)
and very different between ST (up to 22uVpp) and the others.

with best regards

Andreas

I haven't, but a friend who built commercial equipment did, and he said something interesting. It seems if you select the quietest parts from a batch, those will also have the least long term drift.

The zero tc diodes seem to have gotten quite expensive, so maybe selection isn't a good strategy these days. It does make me wonder if one could pot together pairs of conventional (read cheap) zeners that were selected for noise- roll your own!

that is *EXACTLY* what i had in mind sir !

please tell me how !

regards.
Title: Re: LM399 based 10 V reference
Post by: zhtoor on February 15, 2017, 09:30:19 pm

I haven't, but a friend who built commercial equipment did, and he said something interesting. It seems if you select the quietest parts from a batch, those will also have the least long term drift.

Hello,

thats also something that I read somewhere.
But I still do not know if the 1/f noise or the popcorn noise is meant.
I guess its the popcorn noise since it is related with impurities / imperfections of the silicon.

With best regards

Andreas

sort of "harvesting" through these:-

http://www.digikey.com/product-detail/en/fairchild-on-semiconductor/1N5232BTR/1N5232BFSCT-ND/458917 (http://www.digikey.com/product-detail/en/fairchild-on-semiconductor/1N5232BTR/1N5232BFSCT-ND/458917)

5.6V zener $24/1000 pcs.

maybe you could characterize these for noise performance and tempco curves vs. temperature.

regards.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 16, 2017, 04:17:45 pm
Form comparison of the Chinese DW232 from different sources, there seems to be different levels of noise. If you get the wrong type of zener, you might end up with 1000 relatively noise zeners and not a single low noise one. If you are combining diodes it might also be an option to combine two zeners and one forward diode to get something like a 11,8 V (2x5.6+0.6) reference. One could also do it with a transistor instead of a diode (much like the LTZ1000 internal circuit) and this way reduce amplifier noise and drift.

As long as the parts operate at a constant power level, it should not be so bad having them on different chips. Just make sure the temperature is steady - so a good thermal setup is essential to avoid LF noise from temperature fluctuations.
Title: Re: LM399 based 10 V reference
Post by: julian1 on February 17, 2017, 07:25:16 am


i was looking at NZX6V2 which is available @ digikey, and according to its datasheet,
the tempco is +1.5 to 2.5mv/C at 1 to 5ma zener current. Now if this zener family is selected/characterized
for noise performance and paired with a regular silicon diode or a lownoise transistor connected in diode config,
having a tempco of ca. -2.2mv/C, there shall be a "sweet spot" between 1ma and 5ma where these tempco's
shall cancel out completely, and now if properly aged components are used, we *may" have one lownoise (?)
source of steady reference.


Just to to note - this is exactly what the 2DW23x component is.  Like this - but ignore the resistor.

(http://blog.julian1.io/public/images/dw232/2dw232-schematic.jpg)

Two diodes in series in a back to back configuration. So at a certain current - somewhere between 5-30mA depending on device - the tempco of the two diodes are in balance with one another.

The nice thing about the 2DW23x versus implementing the same thing using two discrete diodes, is that the diodes are on the same silicon substrate - which minimises small temperature gradients and differences.

At that point it is also still beneficial to ovenize or stabilize the device temperature to create a better reference - since the TC0 balance point is subject to second-order effects and is kind of "peaky".
Title: Re: LM399 based 10 V reference
Post by: enut11 on February 21, 2017, 03:56:45 am
Hi All
I have just taken delivery of four LM399 ICs and want to use one on a 10v ref PCB.
Is there a way of testing the chips to select the best one before soldering to the PCB?
enut11

Edit: I think I have answered my own question. There is no simple way to simulate the conditions of a ref circuit without testing each one in turn in the same ref circuit. So I think I need to socket the zener.
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on February 21, 2017, 07:07:55 am
Enut11,

I think this comes to the question of "pre ageing " components... it does not necessarily work.
Without soldering.. there is no way of knowing just how the component will behave post final assembly.
My two cent-i volts of noise
Title: Re: LM399 based 10 V reference
Post by: David Hess on February 21, 2017, 11:27:05 am
You could test them for temperature coefficient , noise level, and for popcorn noise.  Testing for popcorn noise can take a while.  Long term drift is more difficult because it is affected by mounting.
Title: Re: LM399 based 10 V reference
Post by: mimmus78 on February 21, 2017, 12:00:54 pm
Any interesting project with a blown heater lm399? I connected backwards ...

Inviato dal mio Nexus 6P utilizzando Tapatalk

Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on February 21, 2017, 12:09:41 pm
Any interesting project with a blown heater lm399? I connected backwards ...

Inviato dal mio Nexus 6P utilizzando Tapatalk

If the voltage reference part is still OK, you have now essentially LM329 in a metal can. Measure the tempco at room temperature, you might got lucky and have now a nice stable low current voltage reference, possibly with a somewhat lower noise than a heated LM399. If the tempco is high (10-20ppm/C) but reasonably linear around 25C, you may attempt to compensate for it with some diodes, as I did for my LT1021 mini-reference (https://www.eevblog.com/forum/metrology/lt1021cmh-based-mini-reference/).

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 21, 2017, 08:56:57 pm
You could test them for temperature coefficient , noise level, and for popcorn noise.  Testing for popcorn noise can take a while.  Long term drift is more difficult because it is affected by mounting.

I second this.

But in every case I would do longterm testing.
The best would be soldering in final configuration.
(see also attached picture with much larger stray of the daily measurements against the soldered solution).

Then monitor the devices at least daily and record drift for at least some kHr.
 + calculate standard deviation over the last 2 kHrs to see which reference has lowest day to day drifts.
similar to here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829)
with best regards

Andreas

Edit: corrected link
Title: Re: LM399 based 10 V reference
Post by: mimmus78 on February 21, 2017, 10:54:43 pm
Thanks guys,

I will do some of this experiment when I will free from my ltz project that is sucking me up :-).

I wonder how Andreas manage to track down his references day by days for years.
Title: Re: LM399 based 10 V reference
Post by: enut11 on February 23, 2017, 12:43:19 am
Hi all
I am sorting through some LM399 chips and measuring Vz. Variation so far is 6.86v to 7.15v.

Does this mean anything apart from needing different trimming resistors to get the required 10v buffered output?
enut11
Title: Re: LM399 based 10 V reference
Post by: mimmus78 on February 23, 2017, 01:16:09 am
So this is noise of the lm399 with blown heater as read by 3458a at 100NPLC ... 3 times better than the other lm399 with heater.
Title: Re: LM399 based 10 V reference
Post by: mimmus78 on February 23, 2017, 01:17:18 am
Hi all
I am sorting through some LM399 chips and measuring Vz. Variation so far is 6.86v to 7.15v.

Does this mean anything apart from needing different trimming resistors to get the required 10v buffered output?
enut11

The ones with lower voltages seems to be better ... just read thorough the thread :-)
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on February 23, 2017, 03:35:59 am
Mimmus,

If you think that lm399 with a blown heater is a better choice for You, You should get hold of an LM329 ( zener half of LM399 no heater ) and run a test with it.
As a comparison

Element14 offer LM399 rated at 2% initial accuracy, and a temp coefficient of 0.3ppm/C at AUD 18.60

LM329 is offered as

Texas instruments product  5% initial accuracy  and 50ppm/C temp cpefficient AUD 1.76
and
Linear Technology product unspecified initial accuracy  and 15ppm/C temp coefficient AUD 5.00

Even the most rudimentary of ovens should be able to bring in the 50ppm /C  specification into check.

Title: Re: LM399 based 10 V reference
Post by: Theboel on February 23, 2017, 04:17:07 am
I believe every LM399 has their own characteristic when You measure it with 3458A. I have 16 pcs LM399 that already run 24/7 for 1 year even the voltage is "different" the closed one are 3,4mV different, noise is also different. data sheet number is the worst condition and I do not found any exceed the datasheet. so in my opinion if You like to know how good or bad the LM399 with or without heater the best way is compare the healthy one with heater on and then turn the heater off.
surely I like to know how good the LM399 if the heater is turn off when You measure with 3458A.   
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on February 23, 2017, 09:15:50 am
Mimmus,

If you think that lm399 with a blown heater is a better choice for You, You should get hold of an LM329 ( zener half of LM399 no heater ) and run a test with it.
As a comparison

Element14 offer LM399 rated at 2% initial accuracy, and a temp coefficient of 0.3ppm/C at AUD 18.60

LM329 is offered as

Texas instruments product  5% initial accuracy  and 50ppm/C temp cpefficient AUD 1.76
and
Linear Technology product unspecified initial accuracy  and 15ppm/C temp coefficient AUD 5.00

Even the most rudimentary of ovens should be able to bring in the 50ppm /C  specification into check.

The main difference is that current LM329 offers are all in a plastic case, so the long term stability is not guaranteed, especially if the humidity varies. LM399 is always supplied in a metal hermetic case and provides for a much better stability, even if used not heated. I have some LM129AH in a metal case and selected for a low tempco, one day I'll try to build a reference on 3-4 of these. My mini-reference based on LT1021CMH chip (https://www.eevblog.com/forum/metrology/lt1021cmh-based-mini-reference/) in a metal case so far is quite stable (over last three weeks the drift is less than 2ppm and that is a conservative estimate) even with a couple of thermal cycles from 15C to 40C and some hours travelling in my backpack outdoors (<5C) .

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: mimmus78 on February 23, 2017, 02:35:01 pm
I believe every LM399 has their own characteristic when You measure it with 3458A. I have 16 pcs LM399 that already run 24/7 for 1 year even the voltage is "different" the closed one are 3,4mV different, noise is also different. data sheet number is the worst condition and I do not found any exceed the datasheet. so in my opinion if You like to know how good or bad the LM399 with or without heater the best way is compare the healthy one with heater on and then turn the heater off.
surely I like to know how good the LM399 if the heater is turn off when You measure with 3458A.

Well I can only compare it with a good one LM399 (got from a broken K2000). This heated
LM399 has not drifted at all in 2 months so I'm not going to touch it.

Anyway this blown LM399 seems enough stable ... I think I will find the zero tempco point and
than use it in some oven project I have in mind. Bob Pease does not recommend to run LM399
without the heater as it leads to instability ... I guess this cannot be applied when the heater is
blown.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 23, 2017, 03:36:24 pm
Thanks to Andreas here is a 24h + x measurement taken with Keysight 3458A at 100NPLC of my may years running LM399.
There is some related drift, but I can't say if this is due to a change of temperature in the controlled lab, if it is drift by the reference itself or if it is due to the drastical change in weather from yesterday morning (calm)  til this morning (stormy with 80km/h) and thus a change in ambient pressure. The ambient monitoring system is currently in use in another lab :(
However, after removing the drift component by a parabolic function the mean is 10.0017577V with a standard deviation of 1µV. My reference obviosly shows popcorn noise in the order of ~3µV.
I plotted Allan Variance before and after removing the drift component... need help to interprete the charts. I attached the original data, so maybe someone wants to analyse them and give some feedback?

EDIT:
Removed reference and added a short at the input of the 3458A, to see how noise compares in this situation. After a small thermal settling time of the short the noise is only a few hundred nV (still 100NPLC @10V). So the popcorn is proven to be from my LM399. Seems it's time to activate my LTZ and take some 24h hour measurements.
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on February 23, 2017, 10:14:52 pm
I think I will find the zero tempco point and
than use it in some oven project I have in mind. Bob Pease does not recommend to run LM399
without the heater as it leads to instability ...
zero TC temperature, you mean?  From my measurements in the past, that is likely to be quite warm, and most of the ones I measured didn't have a zero slope in the temperature range I put them through.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on February 23, 2017, 10:23:25 pm
... or if it is due to the drastical change in weather from yesterday morning (calm)  til this morning (stormy with 80km/h) and thus a change in ambient pressure.

The air pressure here at my place in London dropped from approx 1018 mbar on Wednesday midday to 995 mbar by midday today - that's a 2.25% change, not insignificant.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 24, 2017, 05:35:38 pm
Quote
The air pressure here at my place in London dropped from approx 1018 mbar on Wednesday midday to 995 mbar by midday today - that's a 2.25% change, not insignificant.

I agree with that. So even though I haven't monitored the ambient myself there is still the possibility to compare to online weather data. So here are the data available on the web for my location in the time I measured.
Title: Re: LM399 based 10 V reference
Post by: enut11 on February 25, 2017, 02:25:15 am
In Reply #533 I mentioned that I purchased four LM399 chips and that I wanted to run a simple test to see which was the best to run with in a ref circuit.

Also, IconicPCB mentioned that "I think this comes to the question of "pre ageing " components... it does not necessarily work.
Without soldering.. there is no way of knowing just how the component will behave post final assembly."

Nevertheless I did not want to just pick one at random so I ran a 12Hr bedding-in test on each one with the following results:

LM399a: Was quite stable for most of the time but showed a drop-off towards the end. Practical variation was ~30 uV.
LM399b: Was more stable for longer?. Practical variation was ~20 uV
LM399c: Showed more instability over the 12 hours. Practical variation was >40 uV
LM399d: A bit better than 'c' but still a large variation over the 12 hrs. Practical variation ~40uV

Which one would you choose to be part of a buffered LM399 10v voltage reference?
enut11

PS
1) The socketed circuit board that I used for testing these devices has been in use for some time and has logged up many hours testing other devices prior to this test
2) I did not do any temperature or humidity logging as I am not set up to do so. On average, temperature range was 25-28C and RH 60-70%
3) I used a blister pack around the test PCB to minimise the effects of air drafts
4) All measurements were done on my HP3456A using the Legacy HP XL Logger developed by @IanJ and modified by @Bud (available on the Forum)
Title: Re: LM399 based 10 V reference
Post by: Edwin G. Pettis on February 25, 2017, 04:14:53 am
You cannot predict any future performance of these chips based on such a short test run.  According to Bob Pease, they usually ran these chips (for guaranteed drift rates) for 1,000 hours.  Bob gave me a set of four LM399 chips with guaranteed drift rates with charted drifts over the 1,000 hour period.  While some inference could be assumed after 250-500 hours with reasonable accuracy for the future drift rate, Bob said they never used anything less than 1,000 hours of operation to guarantee long term drift.  Generally, after 250-500 hours you could usually see any chips that were not likely to perform for a given drift rate in the future, these most likely would still meet maximum drift rate later though.  Another factor which also took time was noise, with time the noise could either go up, go down or stay the same, again many hours are needed to tell which way a chip was headed.  There is no reason not to use sockets for this initial drifting/noise test, the act of soldering the chip into a PCB could affect the zener's crystal lattice, care is required as is with the LTZ in this case.  Soldering/de-soldering a chip is not really recommended as it will likely affect stability.
Title: Re: LM399 based 10 V reference
Post by: enut11 on February 25, 2017, 04:41:01 am
Hi Edwin
At least a 1000 hours is only 42 days, not really that long. Looks like I need a few more mule boards for testing.

I am currently mapping the Vz to see if it shows any difference.

Thanks for the feedback. It will be interesting to compare my initial findings with those after 1000 hours.
enut11
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 25, 2017, 08:43:32 am
To see the stability of the reference, one should look at the reference voltage only, not the voltage scaled to 10 V. The resistors used for scaling from 7 to 10 V are a second thing, they can drift just as much as the LM399 - this is especially true for the very beginning, when there will be humidity released form the board. A short 12 h test could only identify rare very bad units with maybe excessive popcorn noise.
If at all the slightly higher noise level for the later 2 refs might be relevant, but for the low frequency noise there are also effects like thermal EMF effects and thus temperature fluctuations or air pressure variations that can contribute. So even for the relevant noise the tests might be to short and not well enough thermally shielded.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 25, 2017, 03:35:12 pm
Hello,

of course a 12 hrs test is too short.

On day 65 I would not have thought that LM399#CH7 is more stable than LM399#CH6
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg402498/#msg402498 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg402498/#msg402498)

But when I compare today standard deviation of the measurements over 1 kHr then CH7 is about 0.25 ppm and CH6 is > 0.5 ppm
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829)

And the conditions are undefined.
- too many contacts (seebeck effect) due to "kleps" (I would prefer direct soldering on the PCB)
- pins of LM399 not shielded against air draft
- no decoupling capacitor (10-100nF) against EMI across the LM399 zener nor across the heater
- no measures to fix the tilting angles of the LM399 (yes it is sensitive against tilting).

By the way: do you know the T.C. of the HP3456A and the according environment temperature which was during measurement?

And perhaps you could increase resolution by measuring the difference of your references in 100mV range.
So you should be able to tell which reference drifts most against the average value of drift.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: enut11 on February 25, 2017, 10:49:31 pm
To see the stability of the reference, one should look at the reference voltage only, not the voltage scaled to 10 V. The resistors used for scaling from 7 to 10 V are a second thing, they can drift just as much as the LM399 - this is especially true for the very beginning, when there will be humidity released form the board. A short 12 h test could only identify rare very bad units with maybe excessive popcorn noise.
If at all the slightly higher noise level for the later 2 refs might be relevant, but for the low frequency noise there are also effects like thermal EMF effects and thus temperature fluctuations or air pressure variations that can contribute. So even for the relevant noise the tests might be to short and not well enough thermally shielded.

All noted, thank you very much. Need to improve my quality control :D

I have started to log Vz inside an insulated cardboard box. This should take care of the effects of air currents on the output.
enut11

Edit: The Vz seems quite noisy. Will have to look at noise reduction as per @andreas' suggestion
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 26, 2017, 09:49:54 am
It is already an free air space of more than something like 5 mm over the board that there can be convective and unstable air currents can appear, that can cause low frequency temperature fluctuations and thus via thermal EMF (e.g. at the OPs) lead to low frequency additional noise. So for really sensitive circuits one should have a cover relatively close on top.

It should not be so bad with the LM399, as the critical part is all inside the chip. Still at least the area right around the LM399 should be more closely covered as there is quite a temperature gradient and the pins are able to generate thermal EMF.

For the outside the cardboard box and the plastic container are only a partial fix. A much better solution would be a metal can, as they reduce temperature gradients and also reduce EMI.

The capacitors are the reference are more against EMI, they don't provide significant filtering as the output impedance of the LM399 is already quite low. looking at the output impedance curve of the LM399/LM329, I would even be careful with low ESR capacitance in parallel to the zener - something like 1-10 Ohms in series to the cap should be beneficial. Getting some filtering for the 10 V reference is a different thing - usually a capacitor at the OP and a resistor between the reference and the OP. This is filtering noise above something like 1 kHz.
Title: Re: LM399 based 10 V reference
Post by: branadic on March 02, 2017, 10:21:43 pm
Measurement is still running and it seems like there is really some pressure sensitivity as can be seen in the pictures below. However this is not the whole story.

I now have a sensor system with TM112, SHT25 and MS5611 that needs some code to run it synchronous with the dmm readings. This way I can directly plot the dependence from temperature, humiditiy and ambient pressure and compensate the readings for that to see the real drift.
Title: Re: LM399 based 10 V reference
Post by: branadic on April 12, 2017, 07:22:55 am
I measured my LM399 with a Keysight3458A together with some additional ambient sensors (SHT25 for temperature and humidity, TMP112 for measuring temperature at selected points and MS5611 for temperature and pressure). Thanks to Andreas I'm able to get all measured values with an aquisistion software in one file.

In memoriam my reference is inside a styrofoam box, the reference itself is within an additional plastic case, that is filled with cotton. I used 5ppm/K SMD resistors for the 7V --> 10V translation, that have a crystal heater on top, to keep them all at same temperature (see also https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg219527/#msg219527 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg219527/#msg219527)).

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=57776;image)

First of all I ran all ambient sensors close together, to see if the single temperature values are tracking each other. This is true, but they show different absolute values. However, this is fine for relative measurements though.

After this measurement I put the TMP112, which at least has lower resolution, inside the styofoam box to catch their inner temperature.
Beside the popcorn noise of around 5µV I tend to see an influence of humidity in my measurements, which fits best with the change in reference voltage. So this comes either from my reference or from the 3458A.

Next step is to turn off the crystal heater and measure the change in behavior.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 12, 2017, 09:52:16 am
Hello,

perhaps you should try to correlate humidity with a PT1-filtered humidity value.
typical time constants for DIP8-Epoxy-packages are 3-7 days.
(to find the best fit value I correlated several filtered values with different time constants).

Your divider resistors/DMM may vary.

With best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on April 12, 2017, 07:01:23 pm
Quote
perhaps you should try to correlate humidity with a PT1-filtered humidity value.

Why and how?
If I compare time domain data (humidity vs. reference voltage) you can see that there is no time lag but a visible similarity. The problem in the reference voltage over humidity plot is, that the noise and the popcorn noise are hiding the correlation somewhat.
I once replaced LT1001 in plastic DIP packages with CERDIP package, even though the photo still shows the plastic package. I still have no explanation why the reference should have a humidity influence, except the FR4 circuit board and the components mounted on it.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 12, 2017, 07:14:28 pm
How about the FR4 epoxy influencing the 7->10V voltage divider resistors?
What coating do the divider resistors have? Some (rare) are glass passivated others with epoxy.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on April 14, 2017, 10:15:48 am
I keep it like occam's/ockham's razor: "...Among competing hypotheses, the one with the fewest assumptions should be selected..."
So before I do any further hypotheses I will do additional measurements in the sense of an exclusion procedure.

That's why I turned of the crystal heater (thermal controller) that is heating my resistors and now measure how the reference behaves while still monitoring ambient parameters. There are several active controllers within the circuit: the power supply, the reference itself and the crystal heater for the resistors, each effecting the complete result of output voltage of the reference. The latter is now turned off, so in a few days I'm somewhat more wise about its influence.
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on April 14, 2017, 04:27:39 pm
Keep an eye on those SMD divider resistors - they sometimes can pick up FR4 board stresses that show up on the divider ratio output.  For precision, on FR4 we'll usually use TH resistors or if we have to use SMT they will go on something like a better stability Rogers pcb material, etc.

Just a suggestion:  You might want to monitor the '399 output directly, before the booster circuit - just to see if you have a resistor stability issue.

Andreas is correct - humidity effects are seen over many days or weeks / months.  The water absorption rate into plastics is definitely there, but fairly slow.  You normally don't see big humidity effects instantly on a circuit - unless you're using a humidity sensor, etc.
Title: Re: LM399 based 10 V reference
Post by: branadic on April 15, 2017, 08:53:06 am
Quote
You might want to monitor the '399 output directly, before the booster circuit - just to see if you have a resistor stability issue.

You can be shure that this is one of the very next steps.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on April 15, 2017, 02:37:05 pm
Keep an eye on those SMD divider resistors - they sometimes can pick up FR4 board stresses that show up on the divider ratio output.  For precision, on FR4 we'll usually use TH resistors or if we have to use SMT they will go on something like a better stability Rogers pcb material, etc.

The problem with discrete through hole parts is that it's difficult to get the thermal bonding that comes for 'free' with SMD parts. Also discrete parts aren't going to come from the manufacturer with a tempco tracking guarantee in the way that some SMD parts do, at least not without handing over many shekels to someone like Vishay or Edwin.

With that in mind, do you have any take on using DIP resistor networks such as Vishay's TDP series that do have tempco tracking guarantees. Is moving to a DIP package as good at avoiding board stress issues as moving to fully discrete parts, or is it a halfway house?
Title: Re: LM399 based 10 V reference
Post by: branadic on June 28, 2017, 08:33:04 pm
Today I present the results of the last 1111 hours of my LM399 with 10V booster with heater on that is on top of my 5ppm/K SMD resistors for the booster. Measured with our 3458A (100NPLC).

Mean: 10.0017711V
Std: 3.4µV

The total difference from mean is about +/-1.5ppm incl. all spikes, not bad at all.

Using the environmental sensors SHT25 and MS5611 I calculated temperature coefficient in the order of 0.8µV/K and humidity coefficient in the order of 0.5µV/%rH or expressed as dew point temperature coefficient 0.875µV/K using magnus formular. There is also some small amount of pressure coefficient in the order of -0.25µV/hPa.
I will try compensating the effects the next days.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 29, 2017, 04:26:31 am

wow nearly 7 weeks (continously?) measurement.  8)

Hmm,

how much of the coefficients belong to the LM399 and which part is for the 3458A?
And how much do they add or compensate?

You need more (different) instruments/refererences to distinguish.   >:D

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on June 29, 2017, 07:42:14 am
Quote
wow nearly 7 weeks (continously?) measurement.  8)

Yes, almost continuously. There was only a small break of a few hours in between, where I couldn't measure the reference, but that's it.

Quote
how much of the coefficients belong to the LM399 and which part is for the 3458A?
And how much do they add or compensate?

I can't tell you. One sad thing is, that there is now multiplexer inside the 3458A and damn, I once sold my Prema 2080 when I got my Prema 5017 SC. That gear would have been perfect to compare multiple references on the 3458A.

Here are the real calculated coefficients instead of reading them out of a diagram using a linear fit:

Temperature sensor inside my styrofoam box with the reference:
TMP112 = 782.2192909802102e-09     10.00175078714701

Ambient conditions:
SHT25_Temp = 702.0809476528713e-09     10.00175529680888
SHT25_Hum = 252.2596185775486e-09     10.00175926263134
SHT25_Dew = 874.2006798035875e-09     10.00176195005415e+01
MS5611_Temp = 649.5007437887549e-09     10.00175629266645
MS5611_p = -252.3944060096974e-09     10.00201462316427

Next step, compensate for temperature, humidity and pressure.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on June 29, 2017, 10:43:45 am
I performed the compensation for temperature, afterwards for humidity and at least for ambient pressure. Attached are the single steps as a series of diagrams.

The compensation is resulting in:

Mean: 10.001790V
Std: 2.58µV
Title: Re: LM399 based 10 V reference
Post by: branadic on June 29, 2017, 01:54:23 pm
Last but not least the Allan deviation and the FFT of the compensated voltage.
Interpreting the charts by my self I would say, that only random walk and aging is visable.

Assuming a linear behaviour for aging/drift during the 1111h I can calculate a value of <2µV/1000h  :-+
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on June 29, 2017, 02:43:35 pm
Very cool!

In order to calculate those coefficients, do you need an environment where changes can be isolated?  I.e. You can change temperature while holding humidity and pressure constant?   In any air conditioned scenario, humidity and temperature are going to be difficult to isolate.
Title: Re: LM399 based 10 V reference
Post by: branadic on June 29, 2017, 04:47:33 pm
Quote
In order to calculate those coefficients, do you need an environment where changes can be isolated?

No you don't. This is similar to compensating a sensor for ambient influences. You would measure its characteristic curve first. Afterwards you would keep the sensor steady and would change ambient conditions without changing the main variable the sensor is sensitive for. For example, measuring ambient influences (temperature, humidity, pressure) on a inclination sensor would be done keeping the sensor in its zero position.
This is the same with a voltage reference, keeping the zener in zero position without changing current or voltage across it equals the sensor in zero position. Now you can change ambient condition. All you need for compensation is a reference sensor that is only sensitive to the variable that you want to compensate.

I first compensated for the most significant variable, temperature. After this I compensated the temperature compensated values for humidity (the second significant variable) and at least I compensated the temperature and humidity compensated values for ambient pressure.

What I can't compensate for is drift, as I have no idea by what amount my reference and the 3458A drifted. Maybe the drift on one of them is compensating the drift of the other. The only way to get away with that are multiple references that you compare with each using one 3458A or having multiple 3458As measuring a single reference.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on June 29, 2017, 05:00:07 pm
To calculate a compensation for different influences one needs sufficient changes in the parameters. Usually one needs to have a balance of enough changes in temperature, humidity etc. to get a measurable effect and small enough changes to assume the effects are linear.

Compensating for one effect at a time (e.g. start with temperature as the possibly largest) is only a first approximation. It might be ok for a data-set that is very long. A more accurate calculation would use regression with all parameters in a single model. Some software could also give you an idea on how accurate the estimated parameters can be. However some of those error estimates also don't work - because they assume a wrong model.

Especially the effect of humidity can shown quite some delay and thus might not be well described by a simple linear (and instant) approximation. So the effect of humidity can be tricky.
Title: Re: LM399 based 10 V reference
Post by: branadic on June 29, 2017, 08:42:53 pm
Correct but as you can see the temperature changes by 6K, humidity by 45%rH and pressure by 28hPa. Even though the experiement is running in a temperature controlled lab it's enough change to make a raw estimation of what is goning on.
I agree, a better way would have been a climate chamber, but currently all of our chambers are in use. This way you would perform two measurements, one with varied temperature and constant humidity, a second one with constant temperature and varied humidity. I don't know of any commercial climate chamber that allows for controlling pressure on the DUT. We have a calibrated pressure controller, but it's to much work to set up a system to control ambient pressure for a DUT inside a climate chamber.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 29, 2017, 09:16:52 pm
I performed the compensation for temperature, afterwards for humidity and at least for ambient pressure. Attached are the single steps as a series of diagrams.

The compensation is resulting in:

Mean: 10.001790V
Std: 2.58µV

Mhm
What does "Mean" mean here.
Is it interpolated to 0 deg C and 0% rH and 0% pressure or what?
from the diagrams the value should be more near 10.00177xV.

Thats why I normalize all my interpolations to 25 deg C. (for components).
For instruments it should be more 23 deg C.

It gets not clear to me:  you did the correlation one after the other
or all coefficients together
or incremental (with correction of the previous coefficients)?

On the first view the dew point temperature seems to be better than rH for humidity.
But since the time constants for humidity are in the 3-7 day range a correlation with a pt1 filtered dew point temperature would be also interesting.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on June 30, 2017, 07:02:55 am
Quote
Is it interpolated to 0 deg C and 0% rH and 0% pressure or what?

Exactly, even though it's only an academic value I can correct the values to "any condition" I want.
However, for me my measurements are showing, that each precision voltage reference should contain sensors for temperature, humidity and pressure! I have evaluated different sensors and will use SHT25 and MS5611 for my next designs. Also I got an own feeling of what a case and the electrical connections through the case should look like. I think that is a very important point, if you want to do precision stuff you must run all the way through the dessert and do all the mistakes on your own to develop a gut instinct.

Quote
It gets not clear to me:  you did the correlation one after the other or all coefficients together or incremental (with correction of the previous coefficients)?

1. I first compensated for the most significant variable, temperature.
2. After this I compensated the temperature compensated values for humidity (the second significant variable) and
3. at least I compensated the temperature and humidity compensated values for ambient pressure.

With your words: the incremental way.

Quote
On the first view the dew point temperature seems to be better than rH for humidity.

Dew point is a calculated value including temperature and humidity using Magnus approximate formula, I wouldn't use that to compensate for humidity.

Quote
But since the time constants for humidity are in the 3-7 day range a correlation with a pt1 filtered dew point temperature would be also interesting.

Still I don't understand what you mean by "pt1 filtered". Maybe you can give me an example?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on June 30, 2017, 07:19:04 am
.....
Quote
But since the time constants for humidity are in the 3-7 day range a correlation with a pt1 filtered dew point temperature would be also interesting.

Still I don't understand what you mean by "pt1 filtered". Maybe you can give me an example?
The effect of humidity is usually not instant. It is not air humidity but more like the humidity contend in the board and parts that influences the circuit. This is delayed against the air humidity. A pt1 filter (= simple first order low pass filter) could be used as a first estimate. However there is still the time constant (e.g. 5 days) that is an additional parameter. Using a filtered humidity is a little like using an intentional very slow sensor.
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 30, 2017, 06:15:48 pm
Still I don't understand what you mean by "pt1 filtered". Maybe you can give me an example?

Hello,

Example: you want a time constant for the low pass filter of 3 days = 259200 seconds.]
Your measurement samples are 100 NPLC so each after 4 seconds (with zero adjustment active).

You calculate the slope K of the PT1 filter between the measurement points.
e.g.  K = 4 s/ 259200 s
You calculate the filtered values from the measurement values M[sample]
Init: PT1[0] = M[0]
PT1[1] = PT1[0] + (M[1] - PT1[0]) * K
...
PT1[n+1] = PT1[n] + (M[n+1] - PT1[n]) * K
for all samples.

you repeat the same for 7 days time constant so that you have 2 pt1 filtered rows with different time constants.
then you calculate the correlation coefficients (together).

From the ratio of the 2 coefficients you can check wether you are nearer to 3 or to 7 days.
Repeat the same for 2 new estimations of time constants until you get no improvement in standard deviation for the corrected values.

With best regards

Andreas




Title: Re: LM399 based 10 V reference
Post by: branadic on August 06, 2017, 05:14:01 pm
I stopped measuring my LM399 for the moment. Here are the latest results. It seems that dew point temperature represents best the change in output voltage of my LM399.
I think I will now perform some hardware modification on the reference, such as a die-cast case or similar to minimize humidity influence. Also thermal isolation will be improved to reduce ambient temperature influence. Both are resulting in dew point temperature without any lag as it seems.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 06, 2017, 06:43:41 pm
Interesting result.

For me it seems that there is a 1 ppm drift over time additionally to the dew point sensitivity.
It also would be interesting if the result is due to the scaling resistors or the LM399 itself.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on August 06, 2017, 07:38:41 pm
Quote
For me it seems that there is a 1 ppm drift over time additionally to the dew point sensitivity.

Yes, you are right Andreas, there is drift in the order of 1ppm.

Quote
It also would be interesting if the result is due to the scaling resistors or the LM399 itself.

Not a measurement I will perform yet.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: 2N3055 on August 06, 2017, 09:49:05 pm
Hello everybody!

I was wondering, dew point influence without time delay, could that also be thermal influence?

Thermal conductivity of 0% saturated air as opposed to 100% saturated air (at dew point) is 28% more (0.032 W/m*K at 0% vs 0.025 W/m*K at 100% humidity) at 90°C (temp of LM399). At room temp, there is practically no difference, but as temperature rises, there is difference between humid and non humid air. This can create thermal conductivity gradients between hotter and colder parts of circuit, and emphasize temperature gradients... Cool parts are being cooled better than hot parts when it's humid...

Usually not a problem but at ppm levels maybe it is...

Just an idea...

Regards,

Sinisa

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on August 07, 2017, 05:29:29 pm
For the thermal conductivity it is the absolute humidity or dew point that is important. So 90 % RH at 90°C is far from what one could get from room air. Under normal conditions the partial pressure of water is more like in the 10-20 mbar range, thus something like 1% abs water contend this one would expect a water contribution in the 0.5% range. So it would be a limited effect on the thermal properties.
Title: Re: LM399 based 10 V reference
Post by: 2N3055 on August 07, 2017, 06:55:36 pm
For the thermal conductivity it is the absolute humidity or dew point that is important. So 90 % RH at 90°C is far from what one could get from room air. Under normal conditions the partial pressure of water is more like in the 10-20 mbar range, thus something like 1% abs water contend this one would expect a water contribution in the 0.5% range. So it would be a limited effect on the thermal properties.

I should have known you guys already thought about it...  :palm:

Regards,

Sinisa
Title: Re: LM399 based 10 V reference
Post by: Spikee on August 25, 2017, 04:12:31 am
Would anybody here be interested in buying pre-aged LM399AH's ? As in with a minimum of 1k on time in ambient environment or using the accelerating aging method ?
Title: Re: LM399 based 10 V reference
Post by: pelule on August 25, 2017, 12:23:28 pm
Interested, send a PM
/PeLuLe
Title: Re: LM399 based 10 V reference
Post by: branadic on September 13, 2017, 06:17:12 pm
Almost done with modifications on my LM399. Main modification was that the reference was put into a Hammon aluminium die cast, filled with cotton batting for thermal insulation  and some selfmade true copper feedthrough. Therefor some installation copper wire (1.5mm²) was glued (Roth Thermokitt) (https://www.carlroth.com/de/de/Chemikalien/A-Z-Chemikalien/T/Thermokitt-Roth-1100-%C2%B0C/Thermokitt-Roth-1100-%C2%B0C/p/0000000700032ab000040023_de) into a brass part with external thread, so that I can screw the selfmade feedthrough into the Hammond die cast. Further I can now measure reference voltage as well as amplified reference voltage. Hoplefully I can start new measurements within the next days.

Will add some pictures soon.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on September 29, 2017, 04:53:54 pm
Now here is a picture of my selfmade copper-feedthrough. As already mentioned the mainbody is made of brass which was coated with chemical tin. The copper wire is glued with Thermokit inside the main body.
My LM399 reference is sitting inside its Hammond aluminium die cast with the copper-feedthrough installed to supply power for the reference and crystal heater on top of the gain setting resistors and to measure the reference voltage as well as the amplified reference voltage. The reference is now again running and monitored by 3458A. Reference voltage is very stable. It currently seems like ambient influence is now fully cancelled out, but with the additional thermal isolation inside the Hammond case I can observe a drift presumably of the gain setting resistors 6.95V --> 10V. Maybe I have to remove some of the wadding to allow some heat dissipation? Some further pictures will follow soon.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Gyro on September 29, 2017, 05:47:27 pm
You could save yourself a fair amount of effort on the feedthroughs. Hollow bore glass feedthroughs are readily available on ebay for running copper wire through.

https://www.ebay.co.uk/sch/i.html?_odkw=glass+feedthrough&_osacat=12576&_from=R40&_trksid=p2045573.m570.l1313.TR0.TRC0.H0.Xglass+hermetic+feedthrough.TRS0&_nkw=glass+hermetic+feedthrough&_sacat=12576 (https://www.ebay.co.uk/sch/i.html?_odkw=glass+feedthrough&_osacat=12576&_from=R40&_trksid=p2045573.m570.l1313.TR0.TRC0.H0.Xglass+hermetic+feedthrough.TRS0&_nkw=glass+hermetic+feedthrough&_sacat=12576)

Of course, if using a hammond encloure you'd need to solder then into drilled out a brass screws, or glue them. Some of them are very small.
Title: Re: LM399 based 10 V reference
Post by: branadic on September 29, 2017, 06:12:33 pm
I tried something similar with ceramic feedthroughs before without success, because the inner metallization broke while soldering. Maybe because of the high temperature gradients and the involved stress. So this solution is fine for me and with only small effort as I needed the brass body with the outer thread anyway. The rest is straight forward, just a thick copper wire and the Thermokit. As I'm an engineer for devoloping new sensors based on assembling technology this solution perfectly fits my profession :)
This is not a hermetically solution, but close enough to increase time constant for humidity to a maximum amount.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: ManateeMafia on September 29, 2017, 10:45:14 pm
I bought some from that seller. They look ok but I haven't tried soldering them yet to see if they can handle the heat.
Title: Re: LM399 based 10 V reference
Post by: d-smes on October 04, 2017, 10:26:26 am
You can also buy off-the-shelf feedthrough EMI filters with silver-plated copper lead.  NOT hermetic, but epoxy sealed.  For example: http://www.mouser.com/Tusonix-CTS/Passive-Components/EMI-Filters-EMI-Suppression/EMI-Feedthrough-Filters/_/N-bw7oz?P=1z0zkur (http://www.mouser.com/Tusonix-CTS/Passive-Components/EMI-Filters-EMI-Suppression/EMI-Feedthrough-Filters/_/N-bw7oz?P=1z0zkur)
Title: Re: LM399 based 10 V reference
Post by: branadic on October 10, 2017, 04:43:20 pm
As mentioned my reference is running in it's new case with my selfmade feedthrough. I first monitored the reference output to see the behaviour to ambient changes. Nothing visible.
Afterwards I monitored the amplified output (~267h) and was observing a signficant drift.
I than switched back to the reference output and monitored its ouput till today (~95h). As you can see, the reference voltage is pretty stable and no influence of ambient is visible.
However, the amplified 10V output is drifting. Thus, the constantly heated resistors (crystal heater on top of 5ppm/K SMD resistors) seem to drift.
I will go on measuring the reference output another few days. I will then try to remove some of the batting inside the case to see, if this gives any improvement in amplified output. Maybe I have to allow for some heat dissipation to decrease the drift of the resistors. If this still doesn't give any improvement the only way I see is to quit the SMD resistors + crystal heater approach and go back to classical leaded resistors.

Attached are some picture of the measurement setup and the measurement results.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on October 10, 2017, 05:04:22 pm
SMD resistors are sensitive to board stress and thus indirectly may react to humidity. The main effect of a partially sealed case is slowing down changes in humidity. So it can take a long time to get a stable value.
Title: Re: LM399 based 10 V reference
Post by: branadic on October 22, 2017, 01:57:30 pm
SMD resistors are sensitive to board stress and thus indirectly may react to humidity. The main effect of a partially sealed case is slowing down changes in humidity. So it can take a long time to get a stable value.

I don't think that this is the case with my setup. I'm currently running the reference with crystal heater turned of. After some recovery time (a few days) with opposite drift direction the 10V output seem to have stabalized around 10.001802 with noise in the order of <±5µV at first sight. Updated diagrams following soon.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on October 23, 2017, 10:46:52 am
As announced attached is a diagramm how the readings have recovered with crystal heater (on top of the ±5ppm SMD resistors) turned of.
Not sure if I will leave it as it is or if I will change the resistors to wired ones. I found that RS Components do have resistors from Alpha Serie MCY for a decent price and ordered some 10k and 20k that I now have on my desk. I had in mind to parallel the 10k with some ±25ppm/K resistors  to trim the output to 10V. Maybe I will wait some additional time before I decide this step.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on November 01, 2017, 12:01:33 pm
Did it, replaced the gain setting resistors by Alpha MCY ±2.5ppm/K (10k || selected 82k ±25ppm : 20k). Instead of the crystal heater I placed some 33k NTC to the connections. So the story goes on.
Will use this 10V-reference for the AD5791 board that is coming within the next days, so that I can make some automated linearity measurements on my Prema 5017 together with the 3458A, but that's another story of another thread.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on November 08, 2017, 09:26:35 pm
As reference is measured with 3458A I found a source for some uncommon Vishay foil resistor values with 5ppm/K on ebay (http://www.ebay.de/itm/0-1-5ppm-0-5W-Very-High-Precision-Vishay-SFERNICE-Foil-resistor-values-100K/142274715625?hash=item21203bdfe9:m:mNw3RbcbyxSLc6Chz8bA45w). So I ordered some 82k2 to replace the 82k, 25ppm/K metal film resistor, that is paralleled to 10k and that is gaining the reference voltage with another 20k to 10V. I hope to get a little closer to 10V and get it a little more stable. Actual diagrams below.
This modification is the last one I will perform on this LM399 before I connect it to AD5791 that is now on my desk.  :-+

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 09, 2017, 09:04:15 pm
Hello Branadic,

what are those large voltage dips on your measurement?
What happens if you have a 100nF capacitor across the zener?
Or is it the output scaling amplifier?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on November 09, 2017, 09:42:46 pm
Andreas,

beside amplified popcorn noise I again do have a temperature influence on the 6,95V --> 10V amplifier. This is why I will replace the 82k 25ppm/K resistor by a 82k2 5ppm/K one, that is in parallel to a 10k 2.5ppm/K in the amplifier. Hope to get the resistor by the middle of next week.

The original 5ppm/K smd resistors where fine, after they where preaged incircuit (with crystal heater on) and after crystal heater turned of. However, I wanted to see what difference I can achieve by through hole components, so I replaced them by AE 2.5ppm/K resistors (10k and 20k).

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on November 16, 2017, 04:16:15 pm
Vishay SFERNICE Foil resistor 82k2 received today, so I stopped the current measurement (updated graphs in former comment) and replaced the 82k metal film resistor. Here are the first measurement points captured. I will update the graphs from time to time, so we get an idea of what has changed.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 16, 2017, 07:32:55 pm
beside amplified popcorn noise I again do have a temperature influence

Hello branadic,

I did not mean the 0.5ppm popcorn noise but the larger (short) spikes of around 4 ppm especially at the beginning of the measurement.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on November 16, 2017, 08:12:22 pm
Quote
I did not mean the 0.5ppm popcorn noise but the larger (short) spikes of around 4 ppm especially at the beginning of the measurement.

Got you, but I don't know where they are coming from. Maybe some external emf coming via power line into the lab power supply and thus into reference voltage? There is a lot of test equipment, test benches but also a photo table top with background lights inside the lab, that is used irregular and that I can't control. Mains line is also messy by the maschines running inside our building. So I wouldn't pay to much attention on that spikes as the reference is not running on batteries.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on November 24, 2017, 07:04:16 pm
Updated the graphs above. After the weekend I will stop measurements on this LM399 for a while and will turn over to the 4xLM399 of Philipp and see how stable this unit is. I'm quite happy with the performance by now. Even though the temperature changes haven't been very big it's obvious that the most temperature sensitive part of the reference is the gain amplifier.

Compare reference output from post 619 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1320837/#msg1320837)
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=376251;image)
and amplified output from post 627 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1350823/#msg1350823)
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=373143;image)

I think I will perform some investigation on this monitoring the difference between reference output and amplified output soon, to make this sensitivity somewhat more visible.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on November 24, 2017, 08:24:46 pm
The amplified reference does not look that bad. The sharp peaks going down should be popcorn noise from the reference itself. It somewhat depends on how stable the temperature was how good the performance of the 7 to 10 V amplifier is. Unless the temperature was really stable the performance looks good.
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 24, 2017, 09:20:15 pm
Hello branadic,

was the measurement on different time / different measurement instruments?
It looks a bit confusing to me for a direct comparison. (85K vs. 170K samples).

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: branadic on November 24, 2017, 09:43:48 pm
Hello Andreas,

reference output was from post 619 and amplified output from post 627, two different measurements as I do only have one 3458A and can't measure both outputs at the same time. Thus I will perform a measurement with the difference between reference output and ampflied output.

By the way: 85.000 samples = 94,44 hours as measurement rate is 0,25 Hz (100NPLC with Auto Zero on)

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on November 29, 2017, 09:26:34 pm
With the latest data given, some dependancy can be plotted:

1. voltage vs. temperature
2. voltage vs. humidity
3. voltage vs. dew point temperature (includes temperature and humidity)
4. voltage vs. ambient pressure

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on December 03, 2017, 08:26:38 pm
I added the dependancy plots for the raw reference output https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1320837/#msg1320837 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1320837/#msg1320837)

While the raw reference voltage is almost insensitive to changes in temperature, humidity and pressure, the amplified 10V output indicates a correlation. I'm currently not sure if the resistors haven't yet stabilized in their new environment or if it's the LT1001 which is a CDIP type. But as they are all within a Hammond aluminium die cast and covered with batting I can't find a simple and thus logical explanation. Any ideas on this?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 03, 2017, 08:54:26 pm
The voltage versus temperature slope is around 3 µV/K. This would be ratehr high for an LT1001 after amplification by a factor of about 1.5.  So it is either is rather poor OP or it's more like the resistors. The resistor dependent part is about 3 V or the 10 V.  So it would be something like a 1ppm/K TC matching for the resistors.
Title: Re: LM399 based 10 V reference - First 650 hours of data
Post by: Insatman on December 12, 2017, 09:45:03 am
I built this 10volt LM399 based reference on perfboard inside a cast aluminum enclosure.  The LM399 is insulated inside a cube of Styrofoam about 1" square.  The +18v supply is from a decent quality wallwart unit (from digikey).  Attached is voltage data taken approx every 24hours on an HP3457a meter.   The average value for "overnight" is logged.  The typical standard deviation is less than 7.0E-6 during an overnight run.   Also attached is as-built schematic.   All parts are thru-hole.   The resistors marked LTC are low temp co.  Typically 15-25pm.  The pot used for adjusting the voltage was the best I had rated at 100ppm.  The measurement cable used is shielded twisted pair with gold pomona plugs on both ends. 

Question here is this.  Is the upward slope of this LM399 typical for >500 hours of operation?  I doubt the meter is drifting but it is possible I suppose. 
Title: Re: LM399 based 10 V reference
Post by: TiN on December 12, 2017, 12:38:09 pm
Maybe it's your R7 trimmer resistor affecting the stability? I'd use smaller value resistor just for finetune so effect from trimpot would be minimal, with further replacement to  proper fixed  resistance network.
Title: Re: LM399 based 10 V reference
Post by: David Hess on December 12, 2017, 02:24:40 pm
While the raw reference voltage is almost insensitive to changes in temperature, humidity and pressure, the amplified 10V output indicates a correlation. I'm currently not sure if the resistors haven't yet stabilized in their new environment or if it's the LT1001 which is a CDIP type. But as they are all within a Hammond aluminium die cast and covered with batting I can't find a simple and thus logical explanation. Any ideas on this?

Thermocouple effects?

Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.
Title: Re: LM399 based 10 V reference
Post by: branadic on December 12, 2017, 03:10:16 pm
Quote
Thermocouple effects?

Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.

Well, the LT1001 is in a precision socket, but everything inside the Hammond die cast is filled with batting. I don't use the offset trim pins, they are left open.

I've started measuring the difference between reference voltage and amplified voltage, thus the amp itself. Here are current diagrams. The FFT indicates 1/f-noise as well as the popcorn (burst) noise of the reference. Temperature influence on the gain stage is also clearly visible.  :--

-branadic-
Title: Re: LM399 based 10 V reference
Post by: David Hess on December 12, 2017, 06:40:12 pm
Quote
Thermocouple effects?

Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.

Well, the LT1001 is in a precision socket, but everything inside the Hammond die cast is filled with batting. I don't use the offset trim pins, they are left open.

The increase in offset voltage drift actually applies to any offset so trimming to zero offset reduces the offset voltage drift to zero as well but given the already low trimmed offset voltage of the LT1001, I doubt this is the problem.

In the past I have built a test circuit to grade amplifiers for offset voltage and offset voltage drift but that should not be necessary here with the LT1001.

Quote
I've started measuring the difference between reference voltage and amplified voltage, thus the amp itself. Here are current diagrams. The FFT indicates 1/f-noise as well as the popcorn (burst) noise of the reference. Temperature influence on the gain stage is also clearly visible.

I do not think the FFT is showing anything useful about the amplifier.  Maybe operate the circuit with the reference output disconnected and the amplifier input grounded to get just the performance of the amplifier?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 12, 2017, 08:12:32 pm
One would expect the noise to be mainly from the LM399 reference. Compared to OPs the noise of the LM399 is really high (100 nV/Sqrt(Hz) and around 10 Hz 1/f noise cross over) Only for the higher frequencies (e.g. > 100 Hz) where filtering is practical the reference noise can be reduces.

On the other hand drift and TC would be most likely to a large part due the two main resistors that set the gain. Here it depends on the resistor types. The TC can still be quite a bit better than the nominal TC of the resistors, as the second oder TC component can often compensate and only about 1/3 of the 10 V is from the amplified part. Drift of the OP would be in the 1 µV/K order of magnitude and thus 0.1 ppm /K range for the output.

The FFT and other data only show the very low frequency noise part. This 1/f and maybe thermal variations.
Title: Re: LM399 based 10 V reference - First 650 hours of data
Post by: Andreas on December 12, 2017, 08:54:04 pm
Question here is this.  Is the upward slope of this LM399 typical for >500 hours of operation?  I doubt the meter is drifting but it is possible I suppose.

Hello,
it really depends. See examples here:

https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829)

So you should wait for at least 5000 hours or 200 days before you can really judge a LM399.
But it could also be the ageing of your resistors. 25ppm/K is not really precision.
And your trimming scheme could also be better (see LM399 data sheet).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Insatman on December 13, 2017, 02:06:12 am
While the raw reference voltage is almost insensitive to changes in temperature, humidity and pressure, the amplified 10V output indicates a correlation. I'm currently not sure if the resistors haven't yet stabilized in their new environment or if it's the LT1001 which is a CDIP type. But as they are all within a Hammond aluminium die cast and covered with batting I can't find a simple and thus logical explanation. Any ideas on this?

Thermocouple effects?

Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.

I did not use the offset terminals in my build.   After looking at branadic's data, I should probably see if the amplifier is the source of the drift.
Title: Re: LM399 based 10 V reference - First 650 hours of data
Post by: Insatman on December 13, 2017, 07:48:45 am
Question here is this.  Is the upward slope of this LM399 typical for >500 hours of operation?  I doubt the meter is drifting but it is possible I suppose.

Hello,
it really depends. See examples here:

https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg796829/#msg796829)

So you should wait for at least 5000 hours or 200 days before you can really judge a LM399.
But it could also be the ageing of your resistors. 25ppm/K is not really precision.
And your trimming scheme could also be better (see LM399 data sheet).

with best regards

Andreas

I have no problem waiting but I thought the drift would settle down somewhat after about 200 hours like most other references.  I've also started monitoring the LM399 output directly, so I can isolate if the drift is the LM399 or op-amp.   I will take a month or so of data and see what comes out.

I measured the value of the 5k pot today and ordered a 15ppm resistor to replace it with.  Being in Philippines, I'm pretty much stuck with Digikey for good parts as they deliver here quickly.  Otherwise Ebay is my only other alternative but parts take 1-2 months to make through the postal system here.

 

Title: Re: LM399 based 10 V reference
Post by: David Hess on December 13, 2017, 12:39:40 pm
Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.

I did not use the offset terminals in my build.   After looking at branadic's data, I should probably see if the amplifier is the source of the drift.

In case it was not clear, the reason I brought the input offset terminals up is that using them to trim the input offset voltage to zero also trims the input offset voltage drift to zero.  Where people sometimes get into trouble is using the input offset terminals to trim offset from other areas of the circuit resulting in very high input offset voltage drift.

Title: Re: LM399 based 10 V reference
Post by: branadic on December 13, 2017, 07:47:44 pm
Quote
Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.

From where do you get those numbers? I do have LT1001ACJ8 with the following numbers for input offset voltage behavior given in the datasheet:

Input Offset Voltage: typ. 20µV, max. 60µV
Long Term Input Offset Voltage Stability: typ. 0.2µV/month, max. 1.0µV/month
Average Offset Voltage Drift: typ. 0.2µV/°C, max. 0.6µV°C

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 13, 2017, 08:02:41 pm
The ratio of 300 µV of offset producing 1 µV/K of drift is a good approximation, but not 100% accurate.

The main reason for offsets in BJT based amplifiers is usually a mismatch in current or effective size of a transistor pair. This results in an offset voltage proportional to absolute temperature (This is one way to make a PAT source)  and thus the approximate ratio.

With most BJT based OPs adjusting the offset to zero will also reduce the drift to a low value. However there can be small extra contributions to the drift. The relation between offset and drift is usually better at higher offset.

The min / max values in the datasheet are usually what is tested - so there ratio may not reflect the physics behind it.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 13, 2017, 08:25:09 pm

but I thought the drift would settle down somewhat after about 200 hours like most other references. 

Hello,

200 hours would be a typical time to stabilize for a plastic housing to environment humidity after soldering.
I guess the drift during this phase is much larger than the initial drift of a LM399.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: David Hess on December 14, 2017, 04:39:20 am
Quote
Did you use the offset voltage adjustment terminals of the LT1001?  They will produce an offset voltage drift of 1uV/C for every 300uV of offset away from zero.

From where do you get those numbers?

They on page 7 of the application section of the LT1001 datasheet:

Trimming to a value other than zero creates a drift of (Vos/300)µV/°C, e.g., if Vos is adjusted to 300 µV, the change in drift will be 1 µV/°C.

At one point when I was doing precision circuits, I had a test jig for measuring things like this including the ratio of current into a trim pin to input offset generated.  Unfortunately I left those notes with that job but they would be easy enough to recreate if I had a reason to.

The ratio of 300 µV of offset producing 1 µV/K of drift is a good approximation, but not 100% accurate.

Oh, of course it is not exact, but it is design dependent.  Not all operational amplifiers have their lowest offset voltage drift at their lowest trimmed offset voltage.  See below.

Quote
The main reason for offsets in BJT based amplifiers is usually a mismatch in current or effective size of a transistor pair. This results in an offset voltage proportional to absolute temperature (This is one way to make a PAT source) and thus the approximate ratio.

With most BJT based OPs adjusting the offset to zero will also reduce the drift to a low value. However there can be small extra contributions to the drift. The relation between offset and drift is usually better at higher offset.

The min / max values in the datasheet are usually what is tested - so there ratio may not reflect the physics behind it.

I remember an old precision operational amplifier datasheet or application note discussing this and thought it was the OP-07 but inspection shows that that was not it.  Maybe it was the Fairchild 725 but I do not see anything there either.  At some point this became a design feature in precision operational amplifiers which distinguished them from "741" type designs and is consistent with them using different offset trimming circuits.

I only brought it up on the outside chance that branadic had used the offset adjustment terminals to trim the output creating a large offset voltage drift.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on December 14, 2017, 12:52:28 pm
One additional contributor to drift in some bipolar OPAs is going to be internal input bias current compensation. The LT1001 has a bias current reduction circuit, so does the OP27 and so does its predecessor the OP07.

In most of these bias current compensation circuits there are two arms to the compensation circuit, one which measures the bias current on one side of the main differential pair (usually from a cascode bias) and one that supplies the 'make up' current to the inputs. They tend to look a awful lot like PTAT current sources.

For low drift applications it might be wise to avoid OPAs with input bias current compensation/reduction circuitry.
Title: Re: LM399 based 10 V reference
Post by: David Hess on December 15, 2017, 03:43:52 am
One additional contributor to drift in some bipolar OPAs is going to be internal input bias current compensation. The LT1001 has a bias current reduction circuit, so does the OP27 and so does its predecessor the OP07.

In most of these bias current compensation circuits there are two arms to the compensation circuit, one which measures the bias current on one side of the main differential pair (usually from a cascode bias) and one that supplies the 'make up' current to the inputs. They tend to look a awful lot like PTAT current sources.

For low drift applications it might be wise to avoid OPAs with input bias current compensation/reduction circuitry.

This is a deliberate design decision and not a factor.  I doubt there is any precision operational amplifier produced which does not operate this way.  The internal current biasing of the operational amplifier drives the differential input stage with a PTAT current so that the adjusted input bias current does not vary as much over temperature; it is a deliberate effort at temperature compensation.  The input bias current compensation circuit just goes along for the ride.

For those who might be interested in the history of this, the improvement made to the LM301A over the LM301 was to use a PTAT current source to bias the input differential stage instead of a constant current.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on December 15, 2017, 09:30:10 am
One additional contributor to drift in some bipolar OPAs is going to be internal input bias current compensation. The LT1001 has a bias current reduction circuit, so does the OP27 and so does its predecessor the OP07.

In most of these bias current compensation circuits there are two arms to the compensation circuit, one which measures the bias current on one side of the main differential pair (usually from a cascode bias) and one that supplies the 'make up' current to the inputs. They tend to look a awful lot like PTAT current sources.

For low drift applications it might be wise to avoid OPAs with input bias current compensation/reduction circuitry.

This is a deliberate design decision and not a factor.  I doubt there is any precision operational amplifier produced which does not operate this way.  The internal current biasing of the operational amplifier drives the differential input stage with a PTAT current so that the adjusted input bias current does not vary as much over temperature; it is a deliberate effort at temperature compensation.  The input bias current compensation circuit just goes along for the ride.

For those who might be interested in the history of this, the improvement made to the LM301A over the LM301 was to use a PTAT current source to bias the input differential stage instead of a constant current.

I think you misunderstand precisely because you quote the LM301A, which doesn't have input bias current reduction circuitry - the bases of the input transistors have no current sources connected to them.

We're not talking about the main current source that sits in the tail of the differential pair, but something that actively reduces the input bias current (i.e. base bias), typically by mirroring the base current of a cascode transistor and injecting that into the input. This is a completely different thing to trying to even out the fall of input bias current with temperature rise by compensating out the rise in beta of the main differential pair with rising temperature by sourcing/sinking a tail current that is PTAT, as the LM301A scheme tries to do. I believe that the OP-07 was the very first OPA to use active input bias current reduction.

There are plenty of precision OPAs that don't do active input bias reduction, the LT1013 for example.

The giveaway that you're looking at an OPA with active input bias reduction is that the input bias current on the datasheet will be (a) low for a bipolar OPA, (b) specified as +/- some current and be very close in value to the input offset current figure and (c) not a datasheet item, but they will work, albeit not at their best, without an external DC bias path to the inputs because they're already providing most of the base bias current internally. Conversely, the way to spot one without input bias current reduction is that the input offset current and input bias current differ by a factor of 10 or more, whereas with an active input current they are of the same order.

Again, I'm not saying that an input bias current reduction scheme is definitely a source of offset drift, but is something to consider as a possible source of drift.
Title: Re: LM399 based 10 V reference
Post by: montemcguire on December 15, 2017, 10:05:56 am
I doubt there is any precision operational amplifier produced which does not operate this way.  The internal current biasing of the operational amplifier drives the differential input stage with a PTAT current so that the adjusted input bias current does not vary as much over temperature; it is a deliberate effort at temperature compensation.  The input bias current compensation circuit just goes along for the ride.

This makes good sense, and points to the benefit of a non-bipolar input stage amplifier. There are modern chopper amplifiers such as the ADA4522, which I believe is a CMOS amplifier, so it lacks traditional input bias current. Since it's a chopper, the normally huge 1/f noise of any sort of FET input is eliminated. You do get chopper noise from the input chopper though, and one can think of that as "input bias", but it doesn't have to be all that large, and the temperature coefficient associated with it seems to be very low. So, it's worth considering, especially since its overall performance is pretty good, and its design eliminates a few of these PTAT factors that you mention.
Title: Re: LM399 based 10 V reference
Post by: Alex Nikitin on December 15, 2017, 10:22:52 am
Again, I'm not saying that an input bias current reduction scheme is definitely a source of offset drift, but is something to consider as a possible source of drift.

As the input bias current itself. A 15nA input current of the LT1013 would create a 150uV drop on a 10K source impedance and it is temperature dependent (about 25pA/C or 0.25uV/C on 10K). An opamp with a bias current cancellation would have a considerably lower current to start with and a lower bias current v temperature drift. For example the LT1097 has <100pA input current with about 1pA/C drift. One thing to watch for on the opamps with a compensated input bias current is the current noise, as it is much higher than that on opamps with a non-compensated input current of the same magnitude. Plus the polarity of the input current can be either positive or negative and can easily change over temperature.

Cheers

Alex
Title: Re: LM399 based 10 V reference
Post by: Cerebus on December 15, 2017, 11:15:00 am
One thing to watch for on the opamps with a compensated input bias current is the current noise, as it is much higher than that on opamps with a non-compensated input current of the same magnitude.

Added to that is that some datasheets get this wrong, and show input current noise specifications that are just the calculated shot noise in the quoted input current (which is less than the real total input current including the internal bias compensation). They can be one or more orders of magnitude out.
Title: Re: LM399 based 10 V reference
Post by: branadic on December 15, 2017, 08:15:59 pm
So what do you suggest as a replacement for LT1001?

BTW: Updated the graphs of the last measurement.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: David Hess on December 16, 2017, 07:27:32 am
So what do you suggest as a replacement for LT1001?

I would need to see the schematic again to make a recommendation but the LT1001 is one of the better options.  Unless there is a problem with high impedance, the input bias current and input noise current will not be a problem.

The LT1012 has 1/20th of the input bias current and twice the low frequency noise of the LT1001.  The LT1007 has 20 times the input bias current and 1/5th the low frequency noise of the LT1001.  The chopper stabilized LTC1150 will reduce all drift errors to essentially zero but has 6 times the low frequency noise of the LT1001 unless the measurement bandwidth is severely limited which may be the case anyway.

Before changing the amplifier, I would double check the impedances at the inputs to find the effect of the bias current and current noise.  I might also trim the offset voltage just to reduce the offset voltage drift.  Something else to consider is if the output drives a heavy load which includes the feedback network, then self heating will create other error terms.  This can be prevented by buffering the output with an emitter follower or other unity gain buffer.  The self heating from driving a heavy load results in lower open loop gain and longer settling time.
Title: Re: LM399 based 10 V reference
Post by: branadic on December 21, 2017, 07:31:55 pm
I was able to reduce the temperature dependency of the gain stage down to 1.41µV/K. How?

A long story short: During some industrial investigation we found, that printed resistors showed quite big humidity dependency. After covering the printed resistors with some self-adhesive aluminium foil this dependency vanished almost completely.

So this was the initial impuls for me to wrap the BMF resistors together with some copper foil for better temperature equality. Afterwards I wrapped around a few layers of self-adhesive aluminium foil as aluminium has a bigger thermal capacity. One side effect is, that humidity has less surface to get into the molding compound of the resistors, thus a much bigger time constant. Here are the current results:

Temperature coefficient: ~ 1.41 µV/K --> including 0.15 ppm of the reading plus 0.01 ppm of the range per degree c = 562nV/K of 3458A
Humidity coefficient: ~ -329 nV/%rH
Dew Point coefficient: ~ -371 nV/K
Pressure coefficent: ~ -1 nV/hPa

I think I'm now at a point with the need to measure inside a climate chamber.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: babysitter on December 23, 2017, 11:42:36 am
That means awesome14 was guided almost right by God and the ieee (imaginary electronic engel..angels), just misunderstood how to use the copper tape. :)

Title: Re: LM399 based 10 V reference
Post by: branadic on December 23, 2017, 11:53:18 am
Damn, no. This has nothing to do with kisses by some mysterious omnipresent creature. The resistors have been alomst close together, but I wanted to avoid thermal glue to tie them together, as this can give unpredictable mechanical stress. So I decided for copper tap to tie them together and bring them as good as possible to same temperature.

There is no free hanging copper tape bridge of special length between the opamp and the resistors to compensate for temperature  :-DD

-branadic-
Title: Re: LM399 based 10 V reference
Post by: 2N3055 on December 23, 2017, 12:18:39 pm

There is no free hanging copper tape bridge of special length between the opamp and the resistors to compensate for temperature  :-DD

-branadic-

Not only magical length.. Don't forget special twists and wrinkles according to divine rule of chaotic wrinklity.... :-DD

Sinisa
Title: Re: LM399 based 10 V reference
Post by: babysitter on December 23, 2017, 01:22:15 pm
Hehe, great.

At work I was recently trying to "clone" our best Thermometer 6 times, so I hooked 6 BC546 to a HP3488 switch and a 34401A.
When trying to get them at the same temperature, i first went to taping my sensors a block of cooper and later submerged it in a stirred Novec 7100 bath to adjust my setup to the reference temperature. My colleague who used it later re-verified with a similar setup using FC-40, setup was still stable and OK. So, to be really AWESOME, use a liquid filled chamber. :)

Title: Re: LM399 based 10 V reference
Post by: branadic on December 23, 2017, 01:59:47 pm
I have something similar on my to do list as you might remember, I do have some Vishay foil resistors on my desk, that want to be packaged in some golden copper package filled with ZT150/ZT180. Something I might manage after christmas or maybe next year. My wish to Santa Claus? More time for hobby :)

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on January 02, 2018, 07:12:27 pm
Unfortunately the measurement was interrupted during free xmas time by a reboot of the computer.  :--
However three days of the "copper tape modified" reference were aquired to get a raw idea about the improved 10V output of my LM399. It seems the reference is now stable enough for my purpose  :-// Connecting it to AD5971.  :clap:

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on January 15, 2018, 10:36:15 pm
Hello,

again a ageing chart of 2 LM399 references.
now after 2 years (more than 730 days) 24/7 operation.  8)

And probably the last time. Since I may need the multiplexer
channels for two brand new LTZ1000A references soon.


Hi Andreas,

another year has past. Do you have new data or have you skipped monitoring both LM399's? After your last update in 2016 it seemed, that the slottet LM399 needed longer, but had better performance afterwards.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 15, 2018, 10:55:37 pm
Hello,

I monitor my new LTZs per hand (only once per week).
So no change for the LM399 setup.

and yes. the CH6 seems to have some seasonal drifts. (or maybe popcorn noise),

CH6 with short legs and no slots
Ch7 with short legs and slots
on your "slot or not" PCB.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 16, 2018, 06:29:52 am
Thanks Andreas, nice to see latest data. It seem there is some lag between LM399 CH6 and LM399 CH7 or maybe this is just a miscorrelation or just an accident? We should have used a significant largner number of references for this experiment.

-branadic-
Title: Re: LM399 based 10 V reference Updated after mods.
Post by: Insatman on January 18, 2018, 07:55:45 am
Earlier I posted about a new LM399 reference that seemed to be drifting a bit too much.  I got a lot of great feedback and set about using it.   Eventually I modified the circuit to replace the LM1001CN8 with a OPA177FP.  Some experiments while building another circuit showed this op-amp superior to the LM1001.  The feedback resistors were also replaced and a smaller value trimpot used.   The unit's metal enclosure was also lined with thin foam and the LM399 enclosed inside the box by small hollow Styrofoam cube.   The new schematic and test results are attached.   Overall the unit has improved a lot.    In the meantime I've also built a LTZ1000 based reference but that's for another post.  I'd like to say Thank You for the advice. :)
Title: Re: LM399 based 10 V reference Updated after mods.
Post by: David Hess on January 18, 2018, 01:00:00 pm
Eventually I modified the circuit to replace the LM1001CN8 with a OPA177FP.  Some experiments while building another circuit showed this op-amp superior to the LM1001.

That is not surprising; the OPA177F is a premium grade and the LT1001C is not.  This is the kind of application where grading the operational amplifiers yourself will make a different in performance if only from weeding out the outliers.

The non-A (TI changed and reversed the suffixes, of course) OPA277 might be even better.
Title: Re: LM399 based 10 V reference
Post by: zhtoor on March 15, 2018, 04:37:19 pm
Hello Mickle,

your experiment orientations are that what I call orientation "0" and "5" where I also have only little differences (up to around 0.5ppm) in my setups.

The pin 1 marker directs to the right side on the photo.
So the critical directions would be the putting them on the left and on the right (connector) sides.
I guess that in this case there would be also about 3-4 ppm difference.

When looking at photos from older HP34401a or Keithley 2000 units the critical directions should be putting them on the left and on the right side. The pin 1 marker shows either to the left or the right side in this case.

With best regards

Andreas

hello andreas,

you might want to have a look at this.

https://www.youtube.com/watch?v=T07j1PVe16s (https://www.youtube.com/watch?v=T07j1PVe16s)

@ 1:03:50

best regards.

-zia
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 15, 2018, 07:41:43 pm
Hello,

what has the bad thermal design of the LM399 to do with that weird theory?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: The Soulman on March 15, 2018, 11:12:05 pm
Hello,

what has the bad thermal design of the LM399 to do with that weird theory?

with best regards

Andreas


Well, 
Quote
ZENER DIODES QUANTUM DETECTORS DETECTING SPACE DIRECTLY
obviously.  :palm:
He haven't even touched on the indirect detecting of space, go figure.  :horse:
Title: Re: LM399 based 10 V reference
Post by: chickenHeadKnob on March 16, 2018, 01:31:17 am
Hello,

what has the bad thermal design of the LM399 to do with that weird theory?

with best regards

Andreas


Well, 
Quote
ZENER DIODES QUANTUM DETECTORS DETECTING SPACE DIRECTLY
obviously.  :palm:
He haven't even touched on the indirect detecting of space, go figure.  :horse:


I think the point was the video claims the orientation of the zeners  within the detector and the orientation of the multiple detectors generating correlated noise are what measure the anisotropic "ether" in that presenters theory.
Title: Re: LM399 based 10 V reference
Post by: amspire on March 16, 2018, 02:51:04 am
Cahill's detectors were not operating in zener breakdown mode but were instead looking at the reverse leakage of zener diodes below the breakdown voltage. He uses 3.3V zeners with well under 1.5V applied. The details are here:

http://vixra.org/pdf/1403.0387v2.pdf (http://vixra.org/pdf/1403.0387v2.pdf)

At a guess, the leakage current would be around the level to put the voltage across the 10K resistor at about 50% of the battery voltage with a single diode. With multiple diodes, you could easily get over 1V across the resistor.

The effect is probably negligible on 6-7.5V reference zener+avalanche diodes at the breakdown voltage.

If Prof. Cahill's effect is real, it would be interesting. I will probably give it a go, just for fun. I found another paper that tried to verify Cahill's results but couldn't:

http://www.ptep-online.com/2017/PP-49-10.PDF (http://www.ptep-online.com/2017/PP-49-10.PDF)

I put it down to the fact that the Laws of Physics in California are probably different to the Laws of Physics in Australia.  :)
Title: Re: LM399 based 10 V reference
Post by: Cerebus on March 16, 2018, 03:02:04 pm
I think the point was the video claims the orientation of the zeners  within the detector and the orientation of the multiple detectors generating correlated noise are what measure the anisotropic "ether" in that presenters theory.

At this point I have to confess that I haven't watched the whole video just the indicated section.

We know that there is no Luminous Aether (Michelson–Morley experiment) but we do know that we live in a quantum soup of virtual particles that, in extreme gravitational fields, produces Hawking radiation. In theory some Hawking radiation ought to be created in any gravitational field and, in theory, that ought to produce some anisotropic effects in a gravitational field. Whether those effects are measurable in another question entirely - in everyday gravitational fields the particle production rate will be so vanishingly small as to be effectively zero.

Certainly, any effect would be smaller than any of the other effects you can measure in a gravitational field. You can certainly measure a difference in the frequency of crystal oscillators with respect to the orientation of a gravitational field (Dave demonstrated this in a video if I recall correctly). I see no reason that a small difference in noise production by semiconductor lattices with respect to gravity isn't a possible effect, and possibly measurable, but I can see no obvious mechanism by which it would yield any change in noise correlation but I wouldn't rule it out as a possibility without a lot of hard thinking and experimentation. However, any putative noise correlation effects wouldn't look like his purported ones with their neat 1 ns delay (did he check the electrical lengths of his cables with a TDR?).

So, gravitational fields, not space itself or the Aether are the likely source of any effects measured that aren't the result of experimental error or magical thinking.
Title: Re: LM399 based 10 V reference
Post by: The Soulman on March 16, 2018, 03:12:38 pm
Hello,

what has the bad thermal design of the LM399 to do with that weird theory?

with best regards

Andreas


Well, 
Quote
ZENER DIODES QUANTUM DETECTORS DETECTING SPACE DIRECTLY
obviously.  :palm:
He haven't even touched on the indirect detecting of space, go figure.  :horse:


I think the point was the video claims the orientation of the zeners  within the detector and the orientation of the multiple detectors generating correlated noise are what measure the anisotropic "ether" in that presenters theory.

Ok, i've  missed "correlated", interesting to see if there is some truth to this..

Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on March 16, 2018, 04:39:30 pm
That experiment conveniently ignores several effects - for instance resistor noise, additional leakage currents from 'scope protector front ends, EMI effects, etc.  I don't recall seeing that he described the resistors being used (I  didn't watch the entire show); it seems like he might assume all resistors are the same.  PWW? BMF? Film?

I can make two "detectors" and see all sorts of "correlated" effects just by having similar equipment switched on in two rooms.  Turn off the lights, move to a quieter testing location, and suddenly no noise correlation.  Or maybe I see a correlation, but I notice a lightning storm 5000km away, and both detectors are picking up the same storm EMI ripples - traveling around and through Earth.  Lightning storms come and go with night and day.  I did not see him check for that effect.

Plus the fact if you have 'scopes running at similar sample rates looking at any noise...look at the data long enough and the peaks and valleys start lining up.  Like old CRT analog TV's tuned to an unused channel:  Stare at the static long enough and you can see anything you want.

The fact is  any crystal lattice is affected by gravity, heat flows, etc.  Those effects are well known and not too mysterious using existing physics knowledge.  If two detectors are placed in the same orientation we might see similar noise - and by the same token if we flip one detector's orientation - do we see more or less or the same correlation?

I'm just not buying this at this point - not until the experiment removes more variables.  It would certainly be interesting to see if the experiment can be duplicated, but judging by the first attempt at duplication I'm not holding my breath.
Title: Re: LM399 based 10 V reference
Post by: zhtoor on March 16, 2018, 04:53:38 pm
Gravitational Wave Metrology - GWM ?  :-DD
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on March 16, 2018, 05:00:20 pm
...And by the way, look at his "Detector" circuit.  Add the 'scope input capacitance and wire loops.  Let me know what you see as a basic RF detector/demod. at the 'scope input.

Mysterious physics or just another radio/lightning stroke receiver?  That's what more refined experiments might look at.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on March 16, 2018, 05:03:52 pm
We know that there is no Luminous Aether (Michelson–Morley experiment) ...

Uhmmmm... no.  The Michelson-Morley experiment was flawed, but even so, it produced a very small (near the noise floor) positive result.  The experiment was later repeated by Dayton Miller (http://en.wikipedia.org/wiki/Dayton_Miller), and using a more sensitive instrument (http://www.orgonelab.org/miller.htm), he obtained very positive proof of an Aether (http://www.anti-relativity.com/daytonmiller.htm).  He then died before he could publish, and the university tried to bury his work, but his assistant was able to save enough of it that we now know that he was successful.  Shortly after this discovery (or rediscovery), Einstein started to change his attitude towards the existence of an Aether-- and you can read about this in transcripts of various conferences he spoke at.  Tesla was right, and Einstein was not exactly correct (http://www.newdawnmagazine.com/articles/tesla-vs-einstein-the-ether-the-birth-of-the-new-physics).

You're conveniently ignoring the current existence of laser interferometers (used to detect gravitational waves) that incidently repeat the Michelson-Morley experiment on a daily basis. LIGO is 360 times physically bigger than the Michelson-Morley interferometer and 144,000 times logically bigger (because it runs the light beams through the path multiple times). If Michelson and Morley had "produced a very small (near the noise floor) positive result", LIGO would produce a massive one.

I'm all for "question everything" but when there's overwhelming evidence that the orthodox position is correct (to current measurement limits) then alternative theories are best left on the new-age mysticism sites like the one you cited newdawnmagazine.com (http://www.newdawnmagazine.com). Here's their self-produced tag cloud, does that strike you as a likely reliable source for a physics story?

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=404255;image)
Title: Re: LM399 based 10 V reference
Post by: zhtoor on March 16, 2018, 05:43:13 pm
"Experimentum summus judex."  (Loosely translated: Experiment trumps theory.)

could'nt have said it better !

best regards.

-zia
Title: Re: LM399 based 10 V reference
Post by: Cerebus on March 16, 2018, 06:26:19 pm
Note that is was of Mr. Miller's opinion that the Aether is "entrained" by the Earth, and so he conducted his experiment at at high altitude.  I doubt that the LIGO experiment is being conducted at high altitude-- "apples and oranges" so they say.

What you just said is what I was eluding to-- that many people have closed minds to alternative theories-- often to the detriment of mankind.

In addition, when we are talking about a theory: "Even in the face of overwhelming positive evidence, if there is one piece of data that is not in agreement with that theory, then that theory is not correct, and you must develop a new theory that better explains the data." -- Richard Feynman

"Experimentum summus judex."  (Loosely translated: Experiment trumps theory.)

Can't disagree with that final part, and the LIGO experiment, with a sensitivity of 10-19m would seem to be the Ace of Spades. If altitude is a factor in the equation then I think that kind of sensitivity would be enough to still detect the effect of the aether at any altitude.

BTW-- I used to work at E.G.&G. "Special Projects" (Area-51).  I've "seen things" that would cause any physicist to doubt their assumptions-- and that's all I can say about that.  The Aether is real, it has properties, it's existence and properties are fundamentally important to physics; and the person that manages to figure it all out (and get published) will win a Nobel prize.

As to the rest, it isn't exactly well published experimental evidence if you can't talk about for fear that the FBI will be at your door in short order afterwards. I'm not saying that you are definitely confabulating but I hope you understand when I say that it does rather beggar belief when someone who claims to have worked at Area-51 and quotes physics articles from new-age sites claims that our current understanding of physics, which has good experimental evidence, is wrong in favour of a previously experimentally discredited theory.

The orthodox is rarely 100% correct, and all scientific knowledge is technically provisional,  but on this occasion I'm quite happy to trust the orthodox over the outré.


Title: Re: LM399 based 10 V reference
Post by: guenthert on March 16, 2018, 09:41:58 pm

I've already said enough.  I want to keep breathing...
What, YOU, worrying?
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on March 16, 2018, 10:21:11 pm
Current orthodoxy recognises space is expanding faster than the speed of light.

If nothing can move faster than the speed of light... ether ( fabric of space )  must be nothing.
Title: Re: LM399 based 10 V reference
Post by: Echo88 on March 16, 2018, 10:31:17 pm
While were still at hijacking this thread:

Im interested in detecting EMI peaks and noise, be it trough lightning storms or through the UFO-starting in Indiana Jones 4: What would be the best way to detect those spikes on a low budget? Simply set up a AM-radio or are there already suitable SDR-boards which also detect lower frequency spikes <1MHz and are sensitive enough?

Would be just a fun experiment to see if i can detect the elevator in my building and the switching of the flat-lighting. I already replaced the LED-bulbs in my measurement-room, use line-filters and changed every used SMPS against linear supplies.
Title: Re: LM399 based 10 V reference
Post by: MisterDiodes on March 16, 2018, 11:10:04 pm
Hijacked thread indeed - but since you asked -

At first glance has nothing to do with a 10V reference, but everything to do with studying "correlated noise events" on a grand scale.

http://en.blitzortung.org/cover_your_area.php (http://en.blitzortung.org/cover_your_area.php)

Check out the real time map and project schematics. 

In a nutshell - If you're looking for E-field ripples from closer T-storms, your antenna is a length of wire.  For Mag fields that travel through and around Earth, just about any circuit loop that encompasses a non-zero area is a potential lightning detector antenna.  Add some gain with an op-amp and some filtering.  Add a GPS to timestamp interesting noise pulse events that look most like a lightning stroke.

Project users from around the world send their data to a central cloud server and let it crunch the numbers - and then a few seconds later you have a near real time lightning map.  This is a Time of Arrival detection system that works best when lots of users are participating.

In other words - once you've "correlated"  your noise pulses with hundreds of other users on Blitzortung - the end result is a fairly accurate map of lightning strokes around the world.  Sometimes a stroke event will trigger a few detectors, sometimes you'll see a bunch of detectors contribute to the stroke position solution  Every detected stroke is logged for later analysis.

What does this have to do with measuring References?

Once you start using a lightning detector, then you start keeping an eye on that local signal output while you're taking sensitive measures in the lab....and then you start to realize that some of your noise you see on your Vref isn't exactly random. :)  Somewhere around then you also realize how far away that storm is that's tickling your circuit.  If you're timestamping your data carefully, then sometimes your buddy halfway around the world can easily be observing the same noise events on his data.  It happens, ESPECIALLY when you're chasing low ppm or very sensitive measures.

This is a type of worldwide interfering noise source that requires no mystery physics of any kind to complete the explanation. 

Title: Re: LM399 based 10 V reference
Post by: hugo on March 16, 2018, 11:24:23 pm
BTW-- I used to work at E.G.&G. "Special Projects" (Area-51).  I've "seen things" that would cause any physicist to doubt their assumptions-- and that's all I can say about that.  The Aether is real, it has properties, it's existence and properties are fundamentally important to physics; and the person that manages to figure it all out (and get published) will win a Nobel prize.

Oh!!  Do tell us more!!!

I've already said enough.  I want to keep breathing...

It's too late now, there are some men in black suits, waiting for you downstairs ...  :)
Title: Re: LM399 based 10 V reference
Post by: amspire on March 17, 2018, 01:27:28 am
BTW-- I used to work at E.G.&G. "Special Projects" (Area-51).  I've "seen things" that would cause any physicist to doubt their assumptions-- and that's all I can say about that.  The Aether is real, it has properties, it's existence and properties are fundamentally important to physics; and the person that manages to figure it all out (and get published) will win a Nobel prize.

Oh!!  Do tell us more!!!

I've already said enough.  I want to keep breathing...

It's too late now, there are some men in black suits, waiting for you downstairs ...  :)
When one of us is receiving our Nobel Prize for Physics for destroying Relativity, he/she will probably have to credit fmaimon for starting a thread on LM399 voltage references. Might have to call it the LM399 10V Reference Theory of Unrelativity.
Title: Re: LM399 based 10 V reference
Post by: Rafael on March 18, 2018, 01:59:21 pm
Hello,

The circuit shown here: http://www.linear.com/product/LM399 (http://www.linear.com/product/LM399) apears seems simple, some recommendation to build or "improve" stability? Is missing something?

(http://cds.linear.com/image/4908.png)

Thanks!!
Title: Re: LM399 based 10 V reference
Post by: eurofox on March 18, 2018, 02:15:19 pm
Available on ebay as well:

https://www.benl.ebay.be/itm/LM399-4-Channel-2-5v-7-5v-5v-10v-High-Precision-Voltage-Reference-Module/123014269610?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2055119.m1438.l2649 (https://www.benl.ebay.be/itm/LM399-4-Channel-2-5v-7-5v-5v-10v-High-Precision-Voltage-Reference-Module/123014269610?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2055119.m1438.l2649)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 18, 2018, 04:34:35 pm
Hello,

The circuit shown here: http://www.linear.com/product/LM399 (http://www.linear.com/product/LM399) apears seems simple, some recommendation to build or "improve" stability? Is missing something?

(http://cds.linear.com/image/4908.png)

Thanks!!

The circuit is very basic. It can be improved a bit by getting the current to drive the LM399 to a large part from the 10 V output instead of the 15 V supply. Another point worth adding is a RC filter to reduce the higher frequency noise of the LM399. The higher frequency noise is not that critical in some applications, but it is relatively easy to reduce it.  Stability against capacitive loading might also be an issue sometimes. From that point on it is mainly about having stable resistors.
Title: Re: LM399 based 10 V reference
Post by: Svgeesus on March 18, 2018, 06:17:39 pm
It can be improved a bit by getting the current to drive the LM399 to a large part from the 10 V output instead of the 15 V supply. Another point worth adding is a RC filter to reduce the higher frequency noise of the LM399.

I'm familiar with bootstrapping the zener current, but have a question about the low-pass filtering. Would that be best at the input to the op-amp (to avoid gaining up the noise), or after it (at the expense of giving the reference a higher output impedance) or between the op-amp and a second op-amp for drive? And if the third option is chosen, would a higher-order active filter be a better choice?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 18, 2018, 08:12:29 pm
It does not make a significant difference between filtering before or after amplification. Between the LM399 and the OP it is very easy to implement a simple RC filter: e.g. a resistor in the 100 Ohms to 1 K range and a 100nF-1µF capacitor. So this is the obvious place for some filtering.

The noise level of the LM399 is around 10 times the noise level of typical precision OPs. There is no practical way to filter out the really low frequency noise.  So the useful amount of filtering is limited to what is simple. As an alternative to complicated filtering there is the option to use a second LM399.

Usually one would avoid using a second OP in the path to reduce extra errors. So things like a more capacitive tolerant output would normally be done with the same OP as the 7 to 10 V step.
Title: Re: LM399 based 10 V reference
Post by: Rafael on March 18, 2018, 10:49:55 pm
Please, be gentle in comments!  :phew:

My knowledge is inversely proportional to my enthusiasm for accuracy. :)

And my first contact with KiCad, I learned this weekend watching videos on YouTube ...

I'm still going to put the tracks and I ask, would it work?  :palm:

Edit: Second attachment... DRC is okay! Noise, who knows? :(
Title: Re: LM399 based 10 V reference
Post by: TiN on March 19, 2018, 12:07:59 am
Heater must use separate wires that go to PSU.
You need kelvin connection to zener output too.
Title: Re: LM399 based 10 V reference
Post by: Rafael on March 19, 2018, 12:33:03 am
Heater must use separate wires that go to PSU.
You need kelvin connection to zener output too.

If I got the message:

 :palm:
Title: Re: LM399 based 10 V reference
Post by: TiN on March 19, 2018, 12:35:35 am
Yea, something like that. I'd go for additional tap (thin wires is ok) for voltage sense. So you have 6 wires going out of the LM399 ;).
Title: Re: LM399 based 10 V reference
Post by: Rafael on March 19, 2018, 01:06:03 am
Yea, something like that. I'd go for additional tap (thin wires is ok) for voltage sense. So you have 6 wires going out of the LM399 ;).

Thanks a lot for your feedback! I will try to improve the tracks distribution, a good design is important too!

I need to master Kicad yet...

Edit: Few track adjustments, more clean.
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on March 20, 2018, 11:30:53 pm
Great work.  Can you post your schematic?  You may want to add a couple of capacitors (one for the op amp's supply, and maybe one across a feedback resistor).
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on March 21, 2018, 03:38:58 am
Post the schematic as a PDF or jpg i dont have the cad package you are using to draw it
Title: Re: LM399 based 10 V reference
Post by: amspire on March 21, 2018, 04:16:05 am
Post the schematic as a PDF or jpg i dont have the cad package you are using to draw it
Just do a screen capture. Even with KiCad, you get this as you are using some custom schematic elements:
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on March 21, 2018, 04:52:51 am
Hi Rafael, right now, the LM399 zener is using the 15V rail and the 7k5 resistor as a current source.  As the current through a zener changes, its output voltage will also change (by a smaller amount).  So if your 15V rail drifts a bit, or if the 7k5 resistor drifts a bit, that can affect your output voltage.

There is a zener technique you can use (I believe it is called "bootstrapping"), where the regulated output of the zener is used to drive current through the zener, which means the current through the zener is as stable as the zener itself.

Here's the basic idea: (attached).  This is the same technique used on the circuit which is on the first page of this thread.


Edit: You may also be interested in some of the posts in this thread: https://www.eevblog.com/forum/metrology/influence-of-resistors-in-lm399-reference-circuit/ (https://www.eevblog.com/forum/metrology/influence-of-resistors-in-lm399-reference-circuit/)
Title: Re: LM399 based 10 V reference
Post by: Rafael on March 21, 2018, 10:40:28 am
My humble apologies.

Now attached properly.

Hi Rafael, right now, the LM399 zener is using the 15V rail and the 7k5 resistor as a current source.  As the current through a zener changes, its output voltage will also change (by a smaller amount).  So if your 15V rail drifts a bit, or if the 7k5 resistor drifts a bit, that can affect your output voltage.

Thanks for the explanations, the theory is clear now!! :)

There is a zener technique you can use (I believe it is called "bootstrapping"), where the regulated output of the zener is used to drive current through the zener, which means the current through the zener is as stable as the zener itself.

Here's the basic idea: (attached).  This is the same technique used on the circuit which is on the first page of this thread.


Edit: You may also be interested in some of the posts in this thread: https://www.eevblog.com/forum/metrology/influence-of-resistors-in-lm399-reference-circuit/ (https://www.eevblog.com/forum/metrology/influence-of-resistors-in-lm399-reference-circuit/)

I've read this post sometimes, my head and maybe the translation is not helping either. I accompany your discussion on this topic and in others for a long time, it is admirable the struggle of all here for precision and stability, so I am sure that I am in the right place. Often the difficulty in finding the perfect component (or combination of compenents) makes it a quest for the lost treasure.

Thank you for the patience.  :-+
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on March 21, 2018, 01:36:34 pm
No worries, sometimes the best way to understand is by doing. The LM399 circuit is cheap enough that you could just build both versions and then see how they behave.

I know that I didn’t really start to understand the LTZ circuit until I actually started building one.
Title: Re: LM399 based 10 V reference
Post by: Macbeth on March 21, 2018, 02:07:07 pm
Post the schematic as a PDF or jpg i dont have the cad package you are using to draw it
Just do a screen capture. Even with KiCad, you get this as you are using some custom schematic elements:
Best to use a PDF for schematics, or a PNG for graphics / charts you want to be seen embedded inline with your posts. JPG is great for photos but introduces artifacts due to its lossy compression which can be horrible for diagram type drawings.

If OP produced the PDF using his version of KiCad with his libraries then the missing symbols will turn out just fine. PDF output is available under the File -> Plot menu.
Title: Re: LM399 based 10 V reference
Post by: cellularmitosis on March 23, 2018, 07:17:53 am
Hello,

Last year I made an LM399 board which was intended to be a basic/cheap LM399 "utility" board (e.g., it accepts a single or dual op amp).  Useful for quick checks, like binning LM399's by voltage, initial 1000-hour ageing, checking 1/f noise, etc.

In minimal configuration, it can be populated with LM399, op amp, 3 resistors and 2 capacitors.

I recently sent a few copies of this board to a friend, and realized I hadn't ever published this board.

Here it is: https://github.com/pepaslabs/basic-lm399/blob/master/kicad/releases/v1/README.md

You can get 10 boards for $12 at dirtypcbs.com, or 3 boards for $19 oshpark.
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 24, 2018, 05:03:46 am
Mhm,

the 100nF on the output is not in the cirquit diagram.
And I think this is one essential component to imrove EMI.
(But should be mounted kelvin sensed between output and LM399).

With my (thyristor pre regulated) power supply I have also found
that a 100nF across the heater pins improves stability.

If you use the 10V output I would also
a) use a series resistor to improve stability against capacitive loading.
b) put a kelvin sensed 100nF between 10V output and this series resistor.

I would also put the 15V stabilisation on the board. (or at least within the housing of the cirquit).
Just to have the same output independant of the external supply.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: RandallMcRee on March 26, 2018, 11:04:00 pm

I've already said enough.  I want to keep breathing...
What, YOU, worrying?

Indeed, DiligentMinds is now gone--I hope he *is* still breathing.

DiligentMinds, wherever you are, we miss you.
Title: Re: LM399 based 10 V reference
Post by: zhtoor on March 28, 2018, 05:33:15 pm
hello friends,

here is an interesting variation on LM399 reference circuit as implemented in IWATSU 7510/11/12/13
LM399 represented by D1/R6

regards.

-zia
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 28, 2018, 06:24:41 pm
Mhm,

but the ratio stability of R1/R2 affects the +/-6V reference. (additionally  to R3/R4)
Why not use +/- 6.95V as reference and avoiding the additional error?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 28, 2018, 06:33:28 pm
The effect of R1/R2 is not very large (e.g. sensitivity should be around 1/7). Likely they want +-6 V and not +-7 V.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on March 28, 2018, 10:15:07 pm

Indeed, DiligentMinds is now gone--I hope he *is* still breathing.

DiligentMinds, wherever you are, we miss you.

I have some suspicions about what has happened to him.

His previous postings have been mostly considered and intelligent. Then recently he seemed to have some rather unorthodox opinions on the existence of the aether. OK, the odd eccentricity isn't that unusual and can easily be overlooked - we have one clearly knowledgable, experienced member who normally makes perfect sense but has clearly drunk some of the audiophool kool-aide and needs taking with a pinch of salt on that topic. But in short order after the aether came up,  DiligentMinds went on to tell us he previously worked at Area-51 and a few days later almost all his posts have gone, apparently at his own instigation.

Either he is playing one hell of a practical joke on us or I suspect he's had some sort of breakdown and is now hopefully being looked after and nursed back to full mental health.
Title: Re: LM399 based 10 V reference
Post by: hwj-d on March 29, 2018, 02:07:32 am
OT:

Or he is simply offended that his subjective life experiences have been made over-ridiculous.

That he obviously had all his posts deleted rather does not argue that he didn't know what he is doing. Psychologically, this is more of a kind of "punishment" of the forum, or some of its members.

Please also note that everything about area51 has a military relevance, even if it is no longer up-to-date.
Title: Re: LM399 based 10 V reference
Post by: Rafael on April 03, 2018, 06:31:06 pm
It's a great hobby to draw pcbs....  ;D
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 03, 2018, 06:43:54 pm
It's a great hobby to draw pcbs....  ;D

And one can do so many re-designs:

I fear that the heater current (change) will influence your output voltage.
Why do you refer the ground pin of the 20K resistor to the input and not to the output?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Rafael on April 03, 2018, 07:01:36 pm
It's a great hobby to draw pcbs....  ;D

And one can do so many re-designs:

I fear that the heater current (change) will influence your output voltage.
Why do you refer the ground pin of the 20K resistor to the input and not to the output?

with best regards

Andreas

Thanks for the tip...

I will update.
Title: Re: LM399 based 10 V reference
Post by: TiN on April 03, 2018, 07:45:02 pm
PCB fab will express great amounts of love to you for this one.  :=\
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on April 24, 2018, 02:58:30 am
around 6 month under +160 C;  1-st step done, will set a drifting measurement now ...
under board a thick thing are copper wire; board a green pcb, resistors blue  :)
interesting oxidation of LM's legs, it actually from ebay ( seen as very beginning of leg are OK )
(https://preview.ibb.co/bL4dYc/DSCN1862.jpg)

Title: Re: LM399 based 10 V reference
Post by: Cerebus on April 24, 2018, 11:26:56 am
interesting oxidation of LM's legs, it actually from ebay ( seen as very beginning of leg are OK )

Classic sign of salvaged parts, where the legs were re-plated or, in some cases, had additional length brazed/welded onto them.
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 01, 2018, 03:16:50 am
I'm wonder, if this circuit make sense ...
I'd like to combine 4 Vref, but worried about current flow, so adding buffer to each in assumption opamp will do it better ,
then probably not bad to try filtering a noise; so R10 + C6,C1 (electrolytic) appears, to decrease a current leakage.  And R2, R4 - in opamp loop for Vdrop compensation   ...
4  such units in parallel, ( not 2 )
(https://image.ibb.co/hvxPKS/noise.png)

then amplification module to +10V  ..
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 01, 2018, 06:55:08 am
Hello,

In this configuration with the feedback resistors R1+R3 at the summing point the both OP-Amps will fight against each other.
With the high output resistors it is likely that one of them will fall out of regulation due to saturation.
So I fear it makes no sense for a precision cirquit.

So either you can add additional summing resistors (I would reduce R2 + R4 in this case)
or you take R1 + R3 directly from the output of the OPs (in this case C2+C3 make no sense).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on May 01, 2018, 07:30:32 am
The usual way to do averaging with Lm399 or similar sources would be to do the averaging right from the LM399, before the 1st (and only) OP. In this step one can also add some filtering with a capacitor to ground in combination with the averaging resistors.

The LM399 output impedance is rather low and thus loading it with averaging resistors (e.g. 1-10K range) is not that relevant.

Even with 8 of the LM399 in parallel the noise is still rather high compared to the usual OPs (e.g. OP07, OP177), so no real need to use more OPs.

If the refs show different noise levels (and low noise is aimed for), one can use different resistors for averaging - the higher the noise the higher the resistor. If done right even a noisy reference can reduce the average noise, though from a certain point it may not be worth it, as the contribution gets too small.
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 01, 2018, 06:01:38 pm
that is my misunderstanding:     (Kirchhoff's circuit laws)
will LM  working as supposed,  when it current sink, not a source ...
further - changing voltage  output for a single LM,  do affect all LMs  due to current change for all, or it insignificant ?
(https://image.ibb.co/gc359S/Clip02.png)
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 01, 2018, 06:36:12 pm
Hello,

each LM399 has its pull-up resistor and will always sink about 1 mA.
So some of the LM399 will sink a bit more than 1mA others somewhat less.
The dynamic resistance of the LM399 zener is around 0.5 - 1 ohms.
so each change of current by 1uA will lead to a around 1uV change of the zener voltage.
But that is ok as long as the changes are small and/or remain constant. (above ~ 0.5mA)

best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 01, 2018, 06:47:26 pm
Great, thanks, I completely miss that point
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 06, 2018, 10:16:44 pm
ОК,  question:

I hook 5 of that burned thingi ( LM399), feed from precision V source , toss in thick Styrofoam box, and measure Vout precisely ( in use HP3456A;  temp env. +- 0.3C ) ...
that was APR 28,  now second measure today - MAY 06 ( 8x24  approx 200 hours ) - I got identical numbers, for all 5.

Normally, how long I should wait to get at least one last digit ( 5-th digits after comma) change for LM399.
Title: Re: LM399 based 10 V reference
Post by: pelule on May 06, 2018, 10:52:27 pm
I would simply calculate:
assuming you are measuring in the 10V range and LM399 volatge is ~7V, thus drift should be at least 2ppm change is 0,000014V
But only, if you assume, the 3456A does not dift at all.
LM399 drift over time is according to LT data sheet typ. 8 ppm / 1khr thus you need to wait typical 250 hours (~11 days)
/PeLuLe
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 06, 2018, 11:48:18 pm
Right, but:
Items seems not a new, so it has some runtime, in addition 6month under +160C add something,  so I'm thinking i found an answer:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg507527/#msg507527 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg507527/#msg507527)
and that lead to: I would never find a direction, in a reasonable time, where units are drifting to .....

Title: Re: LM399 based 10 V reference
Post by: zhtoor on May 07, 2018, 12:09:20 am
around 6 month under +160 C;  1-st step done, will set a drifting measurement now ...
under board a thick thing are copper wire; board a green pcb, resistors blue  :)
interesting oxidation of LM's legs, it actually from ebay ( seen as very beginning of leg are OK )
(https://preview.ibb.co/bL4dYc/DSCN1862.jpg)

Hello GigaJoe,

great work.
would you suggest "cooking" the LM399's in an inert gas atmosphere like nitrogen/argon to prevent oxidation of metal?

best regards.

-zia
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 07, 2018, 03:06:59 am
I had cooked resistors,  metal film, a couple days at 170 C - they like burned,  but very stable, immune to soldering and heating. as coating degraded need to recoat it. I use conformal coating a few layers.
Title: Re: LM399 based 10 V reference
Post by: e61_phil on May 07, 2018, 05:20:12 am
ОК,  question:

I hook 5 of that burned thingi ( LM399), feed from precision V source , toss in thick Styrofoam box, and measure Vout precisely ( in use HP3456A;  temp env. +- 0.3C ) ...
that was APR 28,  now second measure today - MAY 06 ( 8x24  approx 200 hours ) - I got identical numbers, for all 5.

Normally, how long I should wait to get at least one last digit ( 5-th digits after comma) change for LM399.

You should measure the LM399s against each other. This increases resolution and doesn't depend on the single LM399 in your 3456A.
Title: Re: LM399 based 10 V reference
Post by: neil t on May 07, 2018, 07:57:16 am
It's a great hobby to draw pcbs....  ;D

And one can do so many re-designs:

I fear that the heater current (change) will influence your output voltage.
Why do you refer the ground pin of the 20K resistor to the input and not to the output?

with best regards

Andreas

Thanks for the tip...

I will update.

Looks like you have been having fun in Kicad  :-+ my main issue is that the area and layout of the lm399 particularly the partially unshielded trace going from zener + to non inverting input looks remarkably like an E Field probe.
Title: Re: LM399 based 10 V reference
Post by: anymouse on May 07, 2018, 08:55:50 am
If you want to parallel LM399, have a look at the voltage reference of the Guildline 4400. This uses eight paralleled LM329 (that's like a LM399 without internal heater).
You can find the schematics of the reference module in an other thread: https://www.eevblog.com/forum/metrology/teardown-voltage-standards/ (https://www.eevblog.com/forum/metrology/teardown-voltage-standards/)
Title: Re: LM399 based 10 V reference
Post by: branadic on May 21, 2018, 09:53:26 pm
Did some repeating T.C. measurements (after almost one year) on my 16x LMx99 that are still waiting for further use, as well as low frequency noise measurements with heater on. I could confirm that their T.C. is still within the same region as one year ago. Andreas might forgive me that I use a picture he once sent me with the T.C. of my references included in his diagram.
I'm currently investigating the low noise frequency with heater off. Up to now I can't find any correlation on T.C. and low frequency noise within my measurements.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on May 24, 2018, 02:03:04 pm
just an update:
may 29 -ON june 24 - OFF for LM drifting measure:
in ppm, in the end of period:
1: -2
2: -47 (weird jump)
3: -3
4: -2
5: -5
6: -1
7: -2
8: i blew it up ...
it can be multi-meter drift, i use only one unit (3456) to measure periodically turn it on every week, and env temp very stable +18.7 (basement). at least 4 candidates ... As i assumed, do it in parallel doesn't increase the precision, as all of then will drift in the same direction with similar speed .....
 

Title: Re: LM399 based 10 V reference
Post by: Insatman on May 25, 2018, 03:08:48 am
just an update:
may 29 -ON june 24 - OFF for LM drifting measure:
in ppm, in the end of period:
1: -2
2: -47 (weird jump)
3: -3
4: -2
5: -5
6: -1
7: -2
8: i blew it up ...
it can be multi-meter drift, i use only one unit (3456) to measure periodically turn it on every week, and env temp very stable +18.7 (basement). at least 4 candidates ... As i assumed, do it in parallel doesn't increase the precision, as all of then will drift in the same direction with similar speed .....

My HP3456A takes about 48 hours to stabilize after being off for more than a couple of hours.   I suggest you just leave it on for several weeks and see if your data changes.  I leave mine on 24/7, but periodic power outages here in Philippines give me plenty of opportunity to see the effect.   After a couple of hours on time the unit appears stable but over the next two days it will drift by ~20uV compared to my other 4 meters (2x HP34401, HP3457A and FL8505).   
Title: Re: LM399 based 10 V reference
Post by: borghese on May 25, 2018, 09:41:17 pm
Four years ago Bob Dobkin said at "Diligent Mind":
"The LM399's do not age when they are turned off and have almost no hysteresis-- so keeping the reference *off* until a few hours before you need to use it [and/or calibrate it] is the best way to keep the long-term drift minimised."
Cheers
Title: Re: LM399 based 10 V reference
Post by: Insatman on May 27, 2018, 07:13:44 am
Four years ago Bob Dobkin said at "Diligent Mind":
"The LM399's do not age when they are turned off and have almost no hysteresis-- so keeping the reference *off* until a few hours before you need to use it [and/or calibrate it] is the best way to keep the long-term drift minimised."
Cheers

I'm not sure I believe it, but even if it is true there is the thermal equilibrium of the LM399 and any other elements to consider.   It may take more than a couple of hours. 

I also have my own observations which show a small but significant change after a couple of hours warmup. 
Title: Re: LM399 based 10 V reference
Post by: borghese on May 27, 2018, 11:52:25 am
I do not know the truth, but now I quote the whole sentence as from «Reply # 635 on: March 13, 2014, 08:02:10 AM»(now deleted):

6) After telling him that I wanted a voltage transfer device, he said a better way might be to use [at least 6] LM399's in parallel [like the Bob Pease idea].  He said that the burn-in procedure would be to operate these in an oven set to 125C for 2 weeks, which would be equivalent to 1000's of hours of normal operation.  Any LM399's that are drifting too much after that can be replaced [i.e., you burn-in more than you need, and select the best units for the array].  The LM399 is much more sensitive to board stress than the LTZ [because the LTZ has a special mechanical arrangement in the die mount]-- so the LM399 should be mounted off of the PCB a little bit to allow for this.  The long term stability of the array of LM399's will be directly related to the power required to run the heater-- and this can be minimized with insulation-- the more the better!  The LM399's do not age when they are turned off, and have almost no hysteresis-- so keeping the reference *off* until a few hours before you need to use it [and/or calibrate it] is the best way to keep the long-term drift minimized.

7) Note that for LM399-based designs, the slots in the PC board [plus a lot of insulation top and bottom] make sense-- Bob said that the less power the heater requires, then the more stable the output voltage will be.  So, in this case, the slots [plus insulation] are helping with this.  Oh-- and he also said that the LM399 should be run at about 1mA of Zener current for best stability.  The more stable you can make the Zener current, the more stable will be the output voltage.  He said that there is about 1uV of voltage change for 1uA of current change.
Title: Re: LM399 based 10 V reference
Post by: branadic on May 27, 2018, 12:44:22 pm
Using slots around LM399 and massive insulating of the reference at least works for my LM399 ff. (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg219527/#msg219527) quite well.
After a few years the reference is also pretty stable, beside the popcorn noise issue. Attached is a 150h measurement with LT1763 preregulation. Time to put the preregulator inside the case.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Insatman on May 28, 2018, 04:05:26 am
I do not know the truth, but now I quote the whole sentence as from «Reply # 635 on: March 13, 2014, 08:02:10 AM»(now deleted):

6) After telling him that I wanted a voltage transfer device, he said a better way might be to use [at least 6] LM399's in parallel [like the Bob Pease idea].  He said that the burn-in procedure would be to operate these in an oven set to 125C for 2 weeks, which would be equivalent to 1000's of hours of normal operation.  Any LM399's that are drifting too much after that can be replaced [i.e., you burn-in more than you need, and select the best units for the array].  The LM399 is much more sensitive to board stress than the LTZ [because the LTZ has a special mechanical arrangement in the die mount]-- so the LM399 should be mounted off of the PCB a little bit to allow for this.  The long term stability of the array of LM399's will be directly related to the power required to run the heater-- and this can be minimized with insulation-- the more the better!  The LM399's do not age when they are turned off, and have almost no hysteresis-- so keeping the reference *off* until a few hours before you need to use it [and/or calibrate it] is the best way to keep the long-term drift minimized.

7) Note that for LM399-based designs, the slots in the PC board [plus a lot of insulation top and bottom] make sense-- Bob said that the less power the heater requires, then the more stable the output voltage will be.  So, in this case, the slots [plus insulation] are helping with this.  Oh-- and he also said that the LM399 should be run at about 1mA of Zener current for best stability.  The more stable you can make the Zener current, the more stable will be the output voltage.  He said that there is about 1uV of voltage change for 1uA of current change.

Interesting.   The datasheet for the LM399 suggests a 7.5K resistor with a 15V supply, which gives 2mA rather than the one Bob suggested.   I have built several LM399 and LM199 references this way.  I use an LT3080 as a pre-regulator.  I have built 2 boards with 3x LM199 in parallel that were purchased used from Ebay.  These have performed well.  Some data from the first one has been submitted to this forum previously.
https://www.eevblog.com/forum/metrology/data-on-various-voltage-reference-ics-over-time/msg1479949/#msg1479949. (https://www.eevblog.com/forum/metrology/data-on-various-voltage-reference-ics-over-time/msg1479949/#msg1479949.)  These boards have an op-amp which scales the output to 10V.  This does likely increase the drift/tempco somewhat.

I have not used slots in my previous board designs, but I do use insulation.  My LM199/399 devices are surrounded by a small hollow styrofoam cube.  The entire board including the pre-regulator and output scaling op-amp (~2.5x2") are then typically housed in an insulated box. 

thanks for the feedback...I always enjoy learning.
Title: Re: LM399 based 10 V reference
Post by: AG7CK on May 28, 2018, 04:16:13 am
Quote: " ... datasheet for the LM399 suggests a 7.5K resistor with a 15V supply, which gives 2mA ..."

That's an old trap I guess we have all gone into here or there or somewhere else: 15 Volt over 7.5 kOhm is 2 mA, but 15 Volt minus the 7.x Volt over the LM399 gives - Yes, a tad more than 1 mA through 7.5 kOhm.
Title: Re: LM399 based 10 V reference
Post by: Insatman on May 28, 2018, 06:40:18 am
Quote: " ... datasheet for the LM399 suggests a 7.5K resistor with a 15V supply, which gives 2mA ..."

That's an old trap I guess we have all gone into here or there or somewhere else: 15 Volt over 7.5 kOhm is 2 mA, but 15 Volt minus the 7.x Volt over the LM399 gives - Yes, a tad more than 1 mA through 7.5 kOhm.

to quote Homer Simpson "Doahhhh..."  Of course you are correct :palm:
Title: Re: LM399 based 10 V reference
Post by: borghese on May 28, 2018, 09:05:29 am
The HP in the HP3456 multimeter uses a current of 1.3mA - 1.4mA.
Too bad that as isolation use only the cup made of valox and it is known that moving it varies the voltage probably because it varies the thermal gradient
Title: Re: LM399 based 10 V reference
Post by: Insatman on May 28, 2018, 09:53:06 am
The HP in the HP3456 multimeter uses a current of 1.3mA - 1.4mA.
Too bad that as isolation use only the cup made of valox and it is known that moving it varies the voltage probably because it varies the thermal gradient

I added some foam insulation around the voltage reference in my HP3456.  Not much, only about 1/8" as the clearance to the shield isn't much.  I did a full calibration afterward which includes adjusting the reference.  The unit is very stable once it's been on for a couple of days. 
Title: Re: LM399 based 10 V reference
Post by: JimmyJo on June 13, 2018, 02:39:20 am
Can I join the club now?
Single LM399H (non A), Vishay NOMCA 7V-10V gain resistor, LT1001.
Reference board is shielded from drafts, but the 3ft of pomona minigrabbers are not.


Questions and suggestions welcome.
(https://preview.ibb.co/egZOvJ/20180612_161344_1.jpg) (https://ibb.co/gtkk9d)
Title: Re: LM399 based 10 V reference
Post by: TiN on June 13, 2018, 04:26:20 am
Interesting that you used NOMCA's. Perhaps you can share board photo for us?
I got some NOMCA and NOMC networks too for tests, but did not get around to actually do it yet.
Title: Re: LM399 based 10 V reference
Post by: JimmyJo on June 13, 2018, 06:00:18 am
This is really a hack job compared to your work, TiN.
On paper, NOMCA just seemed like a relatively low cost way to obtain closely matched resistors.  Both in terms of tempco and aging.
Two 8x 1K NOMCA were glued back to back, then wires soldered, jumping between the two packages to connect all 16 resistors in series.  A section of 10 was picked at random to become the gain setting resistor.

A more cost effective design would use the leftover NOMCA resistors for the 3k current setting resistor.

(https://preview.ibb.co/gDDuKd/20180613_125555.jpg) (https://ibb.co/hzWgzd)
(https://preview.ibb.co/fOqkQJ/20180613_125656.jpg) (https://ibb.co/fbXzkJ)
(https://preview.ibb.co/m4VDCy/20180613_125735.jpg) (https://ibb.co/eOG9kJ)
(https://preview.ibb.co/jhS4Kd/20180613_125831.jpg) (https://ibb.co/dJOded)
(https://preview.ibb.co/mJYpkJ/20180613_130754.jpg) (https://ibb.co/fk2h5J)
(https://preview.ibb.co/dCaDCy/NOMCA.png) (https://ibb.co/fjBmXy)
Title: Re: LM399 based 10 V reference
Post by: TiN on June 13, 2018, 01:22:52 pm
Unlike my work, you have it up and running, so I would say you are way ahead of my LM399 projects.  :-+
And for prototype testing, your build is nice, tidy and clean. So you have my thumbs up right there.  :-DMM

You can build two of these, and run them back to back, using 34465 as difference meter between two REFs on 100mV range. That will give you some idea on the noise as well.
Then if surely few friends more local to you with access to calibrated 3458/reference could calibrate your ref and provide you pretty good accuracy for further tests.  :popcorn:

Worthy to put into metal box with low-thermal binding posts and battery, so you can use it as portable test source  ;)
Title: Re: LM399 based 10 V reference
Post by: Galaxyrise on June 13, 2018, 02:20:58 pm
Since temperature and resistance matching is going to best for the resistors in the same package, perhaps parallel two packages.  For example, for the top half of the ratio, use 1 from package A and 2 from package B.  Then for the bottom half of the ratio, use 7 from package A and 2 from package B.  That produces a 3/7 ratio such that only the ratio within each package needs to be maintained. The combination is 2.2 times the value of a single resistor, so maybe the 5k resistors makes more sense.  But with so many ratios available this way, one could choose something closer to what their LM399 actually needs to minimize trimming.



Title: Re: LM399 based 10 V reference
Post by: JimmyJo on June 14, 2018, 04:47:33 am
Thank you for the kind words, TiN. 

Since temperature and resistance matching is going to best for the resistors in the same package, perhaps parallel two packages.  For example, for the top half of the ratio, use 1 from package A and 2 from package B.  Then for the bottom half of the ratio, use 7 from package A and 2 from package B.  That produces a 3/7 ratio such that only the ratio within each package needs to be maintained. The combination is 2.2 times the value of a single resistor, so maybe the 5k resistors makes more sense.  But with so many ratios available this way, one could choose something closer to what their LM399 actually needs to minimize trimming.
That is an interesting arrangement of resistors, but im not sure if it works the way you described.  To simplify the mental math, lets grossly exaggerate the problem.  Lets say the resistance of one of the resistor networks increases by few orders or magnitudes, the overall ratio of the arrangement would then be dominated by the other resistor network.  A quick calculation in excel shows that a 5ppm change in one of the networks will lead to a ~1.2ppm change in my circuit, and ~2.2ppm change in the arrangement you've shown.  Can you confirm my math?
Title: Re: LM399 based 10 V reference
Post by: z01z on June 14, 2018, 07:38:26 am
It's possible to do the ratio with one package, as 1.5k/3.5k.
Title: Re: LM399 based 10 V reference
Post by: Svgeesus on June 15, 2018, 09:29:16 pm
I see six wires going to the LM399 so Kelvin connection. But I'm not sure where exactly those would go. One 0V wire to your output terminal and another to the 0V of the op-amp, I guess. And one 7V wire to the op-amp input and another to the resistor supplying the zener current?
Title: Re: LM399 based 10 V reference
Post by: JimmyJo on June 18, 2018, 09:26:45 am
I see six wires going to the LM399 so Kelvin connection. But I'm not sure where exactly those would go. One 0V wire to your output terminal and another to the 0V of the op-amp, I guess. And one 7V wire to the op-amp input and another to the resistor supplying the zener current?

That is my poor attempt at a Kelvin connection, I'm not 100% sure how it should be wired, especially the low side.  You/ve guessed the 7V side correctly, that's exactly the way i wired it.   As for the low side, one goes to power supply ground, the other one goes to output low.  But my 3k/7k return is also tied to the output node, so there is a constant 1mA there.
Title: Re: LM399 based 10 V reference
Post by: enut11 on July 24, 2018, 05:44:48 am
Hi All
In Reply #533 I purchased four LM399 chips and acquired two more since. The Refs have been 'cooking' now for about 18 months. At irregular intervals I spot tested the Vz and 10v outputs and occasionally reset the 10v to 10.00000v when it wandered too far, presumably due to aging components.
Initial tests were done with my HP3456A but lately on a near-new Keysight 34461A. The object of the exercise is to pick the most stable LM399 PCB for further experiments.

Results pages attached:
Title: Re: LM399 based 10 V reference
Post by: enut11 on July 24, 2018, 05:50:21 am
Expanded graph sections from the results above. All measurements are spot checks.
The blue line is Vz. Graph shows variation from test-to-test.
The red line is 10v output. Graph shows variation from 10v. I occasionally tweaked the 10v output to bring it back to 10.00000v.
Some of the graphs show off-chart excursions. These are simply missing data from when the Refs were traveling around Australia for the Aussie Cal Club.

I am zeroing in on Ref #5 as the best of this lot.
enut11
Title: Re: LM399 based 10 V reference
Post by: VK5RC on July 27, 2018, 08:25:04 am
Hi Enut11,
Did you use the same cables etc during testing? Any Temp data ?
Wth my 34461, I found if I used the Maths function, after 1hr warm up, connected everything, waited a while ie 2-3mins and then the display was often quite stable and would show down to uV levels.
My favourite cable is Pomona's twin banana to twin banana shield low EMF cable - mainly as I can't accidentally short out the reference - don't ask!!!!!!!!
Have you checked for any interference - I had a grounding issue - and switching some lights on/off would blow my reading by 30-100s of uV. Once I solved this - putting everything in a metal box - connecting electrically everything - it all settled down a lot.
Kind regards
Rob
Title: Re: LM399 based 10 V reference
Post by: enut11 on July 28, 2018, 06:01:48 am
Hi @VK5RC
I did use the same cables for the latter 12 months of testing as I made up a special 3-way set to measure 10v and Vz. I also simultaneously compared Vz on the HP3456A and the Keysight 34461A.

Before I make my final choice on the best one I will do a more thorough comparison.

At the moment I am working on Max's version of the insulated Ref oven and getting good results - around 0.1C max variation at 34C over 24hrs. Playing around with different insulators. It all takes time.
enut11
Title: Re: LM399 based 10 V reference
Post by: Wolfgang on July 28, 2018, 07:27:12 pm
IMHO the LM317 is a lot noisier than even more classic regulators like the LM723 with a PNP pass transistor.

DK7JB has created a paper (unfortunately in German) comparing a lot of voltage regulators for output noise (just look at the graphs). The LM723 won hands down.
https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern?download=115 (https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern?download=115)

Powering a LM399 should be done by a supply with minimum noise.
Some hints can be found in:

https://electronicprojectsforfun.wordpress.com/power-supplies/a-collection-of-proper-design-practices-using-the-lm723-ic-regulator/ (https://electronicprojectsforfun.wordpress.com/power-supplies/a-collection-of-proper-design-practices-using-the-lm723-ic-regulator/)



Title: Re: LM399 based 10 V reference
Post by: David Hess on July 28, 2018, 09:03:15 pm
IMHO the LM317 is a lot noisier than even more classic regulators like the LM723 with a PNP pass transistor.

DK7JB has created a paper (unfortunately in German) comparing a lot of voltage regulators for output noise (just look at the graphs). The LM723 won hands down.
https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern?download=115 (https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern?download=115)

All of the integrated regulators which use a bandgap reference are pretty poor compared to the zener reference on the 723.  They not only have have higher reference noise but must multiply it to get their output voltage.

I would like to see how using a 317 as the pass element of a 723 would perform.  Enclosing the noisy 317 inside the 723's feedback loop and low noise reference should work pretty well.  The advantage is integrated protection of the pass element.
Title: Re: LM399 based 10 V reference
Post by: Wolfgang on July 28, 2018, 09:31:30 pm
Hi Dave,

I have never seen this arrangement, but I could give it a try when I'm back at the bench. While protection is cool, I am afraid that the 317 produces more noise than the LM723 is able to kill (the error amp gain is only 33dB).

We will see  :)
Title: Re: LM399 based 10 V reference
Post by: Gyro on July 28, 2018, 09:59:09 pm
IMHO the LM317 is a lot noisier than even more classic regulators like the LM723 with a PNP pass transistor.

DK7JB has created a paper (unfortunately in German) comparing a lot of voltage regulators for output noise (just look at the graphs). The LM723 won hands down.
https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern?download=115 (https://www.bartelsos.de/dk7jb.php/rauschen-von-spannungsreglern?download=115)

All of the integrated regulators which use a bandgap reference are pretty poor compared to the zener reference on the 723.  They not only have have higher reference noise but must multiply it to get their output voltage.

I would like to see how using a 317 as the pass element of a 723 would perform.  Enclosing the noisy 317 inside the 723's feedback loop and low noise reference should work pretty well.  The advantage is integrated protection of the pass element.

If you want a protected pass element, then the LM395 would be a better choice. No bandgap reference at all.
Title: Re: LM399 based 10 V reference
Post by: Wolfgang on July 28, 2018, 10:06:55 pm
Hi,

the LM395 might not be a bad choice, but I missed any noise specs for it. My guess is that is way better than a 317, for a few €s more.
Anyway, also worth a try.  :)

Some Horowitz and Hill precision supplies as well as Linear Tech App Notes used a separate preregulator using a 317, but outside of the main regulators loop and with extra noise filtering before. Adding current limit to the precision regulating loop adds noise.
Title: Re: LM399 based 10 V reference
Post by: David Hess on July 29, 2018, 05:15:06 am
I have never seen this arrangement, but I could give it a try when I'm back at the bench. While protection is cool, I am afraid that the 317 produces more noise than the LM723 is able to kill (the error amp gain is only 33dB).

I have used it many times but with an external reference bootstrapped off of the output and a precision operational amplifier as shown in the example below.  I just never tried it with a 723.

If you want a protected pass element, then the LM395 would be a better choice. No bandgap reference at all.

I agree but they are way more expensive.

Adding current limit to the precision regulating loop adds noise.

Oddly enough there may be another good reason to add a current limit via a shunt between the pass element and output filtering capacitor.  A small amount of AC feedback can be taken from before the current shunt to stabilize the feedback loop with low ESR output capacitors.  Some integrated regulators use this trick but they have the advantage of sampling the AC feedback from the pass transistor structure itself without adding any series resistance.

Effectively this adds a controlled series resistance to the output capacitor for stability purposes while maintaining a low ESR for filtering purposes.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 29, 2018, 05:46:15 am
Powering a LM399 should be done by a supply with minimum noise.

Why that? in which application. (the LM399 dampens the supply noise at least by 3K / 1 Ohm).

And why do you look at the wideband noise and not the 1/f noise which is more important for integrating voltage measurement instruments?

For a HP34401A a LM317/LM337 seems to be enough.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Wolfgang on July 29, 2018, 08:41:48 am
As with all regulating circuitry, PSRR deteriorates with frequency. I may be conservative, but I try to kill noise or ripple in a signal or PSU chain as early as possible. This is not a solution for the problem of 1/f noise that is *generated* in (pre) regulator stages, I agree.
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on September 25, 2018, 10:22:17 pm
that it ....
(https://image.ibb.co/mioJ49/20180925_175156.jpg)

top 10V, then 1V then 0.1V ;  Right side variable output 0-20V, adjusted precision up to 1uV
it based on 2 LM399, that i grilled a while ago; self made 1:10 dividers. and other cheapo thingi ..  TC I measured +0.7 ppm\C (spent 2 weeks to do adjustment, finally give-up , but can be done better) on 10V, and not measurable on 1V.

variable based on 2DW232 and some 1:10 divider if need some millivolts output ... 

reference and variable has 2 independent transformers. so I had fun idea,  I set Variable output to compensate 1Vout,   and after 12 ours it was really stable , but after a day pass, it state -20uV - I guess due to env temperature up , as I didn't do anything for a thermal compensation on  variable output.

u use my K2001 to adjust as it suppose to,  consider a performance report on K2001 i did a year ago, soo will see around december how it will change ... if so ...

now I will think about current source now, a bit more challenge due to 1A output needs.
Title: Re: LM399 based 10 V reference
Post by: TiN on September 26, 2018, 10:12:36 am
BNC connectors are not good for precision microvolt stuff ;).
Title: Re: LM399 based 10 V reference
Post by: GigaJoe on September 26, 2018, 06:29:25 pm
but HP 3245A has it .... at least connect disconnect on 0.1V i had repeatable result .... anyway ....will see ...
Title: Re: LM399 based 10 V reference
Post by: branadic on November 10, 2018, 01:55:03 pm
While I was waiting for the parts for LTZ1000 references I had the time to draw circuit and board file for the "LM399 PWM DAC reference" based on the "ION" MickleT. presented over here as a representitive of the russian forum (radiokot.ru).
As far as I understood the original design came from Sergei and is available here:
The original design comes from adver (https://translate.google.de/translate?sl=auto&tl=en&js=y&prev=_t&hl=de&ie=UTF-8&u=https%3A%2F%2Fradiokot.ru%2Fforum%2Fviewtopic.php%3Fp%3D3113890%23p3113890&edit-text=&act=url), but sources can be found on Sergei's google drive (https://drive.google.com/drive/folders/1GnfWAyInh_M2gaa59xyV-j0-_jiwc0C1).
You can find the source code for two versions:
- 8bit + 14bit fractional
- 8bit + 16bit fractional
and lot's of pictures.

I made some minor changes and put some additional parts into the circuit:
    - LT1763 voltage regulator for voltage reference (12V) as well as microcontroller (5V)
    - additional resistor divider feedback (for testing purpose at first glance)
    - output buffer (LTC2057)
    - NTC temperature sensor on one of the ADC pins, to add a temperature compensation scheme
    - optical transmitter (TORX1952 or SFH752V) for galvanically insulated data transfer of temperature data and pwm setting to develope temperature compensation algorithm

The board does fit, what a miracle, into a Hammond 1455C801 case. :)
The boards have been ordered at Elecrow and will arrive soon. In case someone wants to duplicate it and contribute to this reference, attached are the design files (Gerber data). Eagle files on request.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: DC1MC on December 02, 2018, 08:40:42 pm
I'm chewing my fingernails, it's actually the first PWM implementation that I know off.
What range it's planned to have ?

 Cheers,
 DC1MC
Title: Re: LM399 based 10 V reference
Post by: branadic on December 02, 2018, 08:49:50 pm
Actually the PWM is only to boost the zener output voltage to 10V. I'm still waiting for the boards to arrive. Customs took quite long as the boards arrived in Frankfurt on 22th of november, so 10 days ago.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: DC1MC on December 02, 2018, 09:39:38 pm
Actually the PWM is only to boost the zener output voltage to 10V. I'm still waiting for the boards to arrive. Customs took quite long as the boards arrived in Frankfurt on 22th of november, so 10 days ago.

-branadic-

Ah, so it's not planned to be adjusted in a wide range, but just like a finely controlled amplifier. Still very interesting.

 Cheers,
 DC1MC
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 02, 2018, 09:51:32 pm
In this circuit PWM is used to set an amplifier gain. The very extreme PWM values can be problematic. So starting slightly below 7 V the lower limit would be a little over 7 V (maybe 7.1 V). The upper limit is likely set by the supply, so maybe 11.5 V.
The other limitation of the circuit is a relatively slow settling - not a problem with a fixed reference though.
The setting is also not necessarily very linear.

The circuit looks a little odd at two small details: one is the choice of OPs. My guess is it would work better with the OPs swapped. The AD8675 is a really good OP, but still not that suitable for a 300 K source. The AD8675 might than need a little more than 12 V to work at 10 V.

The filter at the output amplifier is only rather high frequency. I would expect at least a little more than 10 nF there.
Title: Re: LM399 based 10 V reference
Post by: branadic on December 03, 2018, 11:52:28 am
It's just a board for experiementing and for getting a gut feeling how to improve things, not a final schematic with all components defined forever. The board also includes some options to play with.
After waiting forever the boards arrived today.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: try on December 03, 2018, 10:30:48 pm
Hi enut11,

Expanded graph sections from the results above. All measurements are spot checks.
The blue line is Vz. Graph shows variation from test-to-test.
The red line is 10v output. Graph shows variation from 10v. I occasionally tweaked the 10v output to bring it back to 10.00000v.
Some of the graphs show off-chart excursions. These are simply missing data from when the Refs were traveling around Australia for the Aussie Cal Club.

I am zeroing in on Ref #5 as the best of this lot.
enut11

the variations of the 10V-output of your reference look unnecessary high. I have just one LM399-based 10V-reference which is permanently running and most of the time it is connected to a 34401A. I can't provide you with a fine documentation such as you prepared but I can tell you that my 10V output drifted 2ppm in the first year and 4,5ppm in the second year in absolute* terms.
The output drifts really slowly over time. During summertime with wild temperature swings in my appartment from 15 degrees Celsius to 30 degrees the 34401A shows variations of maybe 1,5 to 2ppm per day which provides a kind of overlay to the long-term trend.

*Absolute terms means, that the reference was 10V-2ppm at t=0 measured with a (calibrated!) 8508A from PTB's DC cal lab at the maker fair.
At t=1 year it was measured against a calibrated 3458A from PTB at the maker fair (10V flat)
At t=2 years it was measured against a calibrated 3458A from PTB at the PTB and at the maker fair (10V +4,5ppm).

Swings like 6ppm in one month only seem odd to me. As you wrote that you reset your reference after some time I assume your references contain a trimpot. Depending on the adjustement range and maker a trimpot could be a source of drift evil.
I would replace the trimpots by fixed resistors and would not modify the references anymore.

When you track your references and simply take notice of the adjusted value of 10,000 00 V at reset time you loose the drift information over the last period und you will be unable to compile a continuous drift chart.
You might consider recording the voltage BEFORE adjustment as well.  :)

There is nothing special in my reference. The resistors used are just Yageo MF in TK25.

Best regards
try
Title: Re: LM399 based 10 V reference
Post by: branadic on December 04, 2018, 07:02:04 pm
Most of the parts were in the junk box, so I could start to assemble the basic circuit. Still there are a few components missing, that I need to order.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: WillTurner on December 06, 2018, 06:40:44 am
JimmyJo built a reference with a 7V to 10V gain stage using a Vishay NOMCA resistor array
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1604887/#msg1604887 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1604887/#msg1604887).

These packages contain eight 25ppm/C resistors, tracking to 5ppm/C. Note that the datasheet specifications are given as "typical".
Individual resistor values of interest include 1k 2k, 5k, and 10k.

Then z01z suggested
Quote from: z01z on June 14, 2018, 05:38:26 pm (https://www.eevblog.com/forum/index.php?topic=15982.msg1607323#msg1607323)
It's possible to do the ratio with one package, as 1.5k/3.5k.


Yes indeed, and surprisingly (to me at least), the 10/7 ratio is independent of the individual resistor values.

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=589294)

Has anyone calculated an overall temperature coefficient for this circuit?
Title: Re: LM399 based 10 V reference
Post by: z01z on December 06, 2018, 08:19:38 am
These packages contain eight 25ppm/C resistors, tracking to 5ppm/C. Note that the datasheet specifications are given as "typical".
Individual resistor values of interest include 1k 2k, 5k, and 10k.
Diligentminds had an interesting post about these, saying that after initial burn-in, they are much more stable.

Quote
Then z01z suggested
Quote from: z01z on June 14, 2018, 05:38:26 pm (https://www.eevblog.com/forum/index.php?topic=15982.msg1607323#msg1607323)
It's possible to do the ratio with one package, as 1.5k/3.5k.

Yes, then I realized that for having a statistical divider, it is better to have more parts. For example, the single R in the upper resistor is the most sensitive to changes.

Quote
Yes indeed, and surprisingly (to me at least), the 10/7 ratio is independent of the individual resistor values.
This depends on how ideal your opamp is, you wouldn't get away with say, using a 1Ohm or 1GOhm resistor.
Title: Re: LM399 based 10 V reference
Post by: Vtile on December 06, 2018, 09:30:23 am
Edit. Silly me LM399 is drawn partially in the shcematic, not ref. zener type Y.

Not exactly this circuit, but I think what is better variations of that as a servo controller, if I have understood correctly. There the zener leg is connected to 10V output so it is self stabilizing (or oscillating if something goes wrong). Old basic design found at old datasheets of precision OPs etc.

No temp. co. calculations were done, but it shouldn't be too hard to put the values with error coefficients to some math environment.
Post #8,#19 shows picture of the results.
https://www.eevblog.com/forum/metrology/oh-no (https://www.eevblog.com/forum/metrology/oh-no)!-i-got-infection/msg1248065/#msg1248065

Conrad Hofmans original post, where I did get inspiration.
https://www.eevblog.com/forum/metrology/old-fashioned-zener-10v-reference/ (https://www.eevblog.com/forum/metrology/old-fashioned-zener-10v-reference/)


JimmyJo built a reference with a 7V to 10V gain stage using a Vishay NOMCA resistor array
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1604887/#msg1604887 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1604887/#msg1604887).

These packages contain eight 25ppm/C resistors, tracking to 5ppm/C. Note that the datasheet specifications are given as "typical".
Individual resistor values of interest include 1k 2k, 5k, and 10k.

Then z01z suggested
Quote from: z01z on June 14, 2018, 05:38:26 pm (https://www.eevblog.com/forum/index.php?topic=15982.msg1607323#msg1607323)
It's possible to do the ratio with one package, as 1.5k/3.5k.


Yes indeed, and surprisingly (to me at least), the 10/7 ratio is independent of the individual resistor values.

(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=589294)

Has anyone calculated an overall temperature coefficient for this circuit?
Title: Re: LM399 based 10 V reference
Post by: branadic on December 20, 2018, 10:49:40 am
After some progress the "ION" is up and running!
The missing parts arrived and were soldered into place. Silly me, I routed the power supply of the output buffer (LTC2057) to the wrong pins, but was able to botch some wires to correct for this. (corrected this in rev1 of the board files)
I then had some trouble with my AVR Dragon. Since much time has passed without using it, I didn't remember that the VTG pin is not able to supply the micro with the necessary power and as a result I wasn't able to program it. It took many hours and a lot of search on the web to get me to this point. And it took me another amount of time to get the latest version of AVR Studio (~25k/s download in 21th century is pretty slow, seems there was a bottleneck somewhere in between).
However, now that I solved all this issues I was able to program the Tiny this morning, to set the fuses and to power up the circuit. Hurray "Number 5 is alive" and pressing the buttons changes the ouput voltage. I was able to set it to 10,00000V on my meter.
So this is a good point to start further investigation and improvement on that reference, such as implementing a temperature compensation scheme in software (if necessary) and to find out if something similar is worth for a LTZ reference.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 23, 2018, 11:03:16 am
After some progress the "ION" is up and running!

2nd on this interesting theme within this forum  ;D

Since I use a external (well aged) LM399 for this tests and thus have larger wiring I had also to populate a additional 100nF output capacitor. (on the bottom side).

As population I use 14V for the VCC regulator (for my NiMH battery packs) and 2*LTC2057 for the OP-Amps.
Software is the 22 bit version for the PWM. (8 bit in Hardware + 14 bit dithering).
I have deactivated the key inputs since the NTC (at the moment without function) is using one of the inputs.

The EEPROM constant was adapted to my LM399#3 and so I should have exactly 10 V at the output.
But I got around 420-430 uV more than that.

As reason for the difference I found out the non-symmetrical delay times of the ADG419.
The rising 10V signal has 97.4 ns delay and the falling 10V edge 93.1 ns delay against the 5V PWM signal.
This 4.3 ns difference together with the nearly 10 kHz PWM signal (20MHz / 8 / 256) generates calculatory 42 ppm difference.
Bingo: 42 ppm are 420uV at 10V.

Hoping that the 4.3 ns remain constant over power supply variations and temperature ...

with best regards

Andreas





Title: Re: LM399 based 10 V reference
Post by: Andreas on December 23, 2018, 07:23:16 pm
Got a unexpected result on PSRR:

Supplied by a programmable linear power supply at the VCC node
(yes: the LT1763 can be back suplied/paralleled without damage).

from 12V to 18V  1.054 mV difference on 10V output.
This gives 176 uV/V or 17.6 ppm/V in average.
from 12 to 14 V the value is even 21.5 ppm/V,

compared to less than 1 ppm/V for the LM399 above 10V (mainly for the heater)
we have to take much more care on the supply voltage stability.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 23, 2018, 08:58:36 pm
The supply to the CMOS switch can have quite some influence on the switch. This can be in two ways: one is extra charge injection and the other is a change in the delay for turn on and turn off.  So a really stable supply to the DG419 would be important.

A 200 µV/V PSRR is not that bad.  It needs some care but should not be so hard.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 23, 2018, 10:28:55 pm
It needs some care but should not be so hard.

Lets do the math for the current cirquit.
LT1763 has up to 60mV change in bandgap voltage (1220 mV) over a 165 K temperature range.
so worst case up to 300 ppm/K.

Cheap 0805 1% resistors have 100 ppm/K
worst combindation will be 500 ppm/K for the 14 or 15V supply.
Over a temperature range of 30 deg C (which I want to have) this gives 1.5% possible supply voltage variation.
14V * 1.5% = 0.28V max change over temperature.
with 21.5 ppm/V ("typical", since only 1 sample measured) we can get 6 ppm change on the 10V output over the 30 deg C range.

For me that means that we have to check every single power supply cirquit for T.C.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: splin on December 24, 2018, 01:25:14 am
It needs some care but should not be so hard.

Lets do the math for the current cirquit.
LT1763 has up to 60mV change in bandgap voltage (1220 mV) over a 165 K temperature range.
so worst case up to 300 ppm/K.

Ironic considering it's used to power a < 2ppm reference! Those (two) LT1763s are very expensive - over $4.2 each from most distributers (but only $2.3 from Arrow which is great if you are in the States).

Personally I'd have used jellybean opamps and pnp transistors to achieve similar noise (why don't they specify <10 Hz noise for many/most voltage regulators?). TC and time drift would be way better and saving $9 in the process which would have been better spent on a second LM399. It would need a bit of care to ensure it starts up properly though.

[EDIT] Just to be clear, I'm sure the circuit met the designer's own requirements/trade-offs who might, for example, happened to have surplus LT1763s.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 24, 2018, 05:23:27 am
Hello,

by the way: a jelly bean LM317 has specced 100 mV worst case change over temperature range
of the reference voltage (1250 mV) resulting in nearly 500 ppm/K.

So the LT1763 is not too bad.
Besides other advantages (low power consumption,  low noise, low voltage drop).
So ideal for battery powered equipment like voltage references.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 24, 2018, 10:10:03 am
If the supply would be a real problem, there would be still the option to use the existing LM399 reference and use an OP as an amplifier from there.

The regulator may not be that bad: at least in the typical curve there is quite some curvature / slope at the upper end of the temperature range. In addition there are resistors that are better than 100 ppm/K and the calculation was worst case for the resistors.

I don't think the supply should be so critical that one would need to have a super low noise regulator. Keeping the emissions from the DG419 switch away from the reference and OP could be the more important point.

A totally agree that it would be nice to have data in the DS on the low frequency noise (e.g. 0.1 Hz to 10 Hz or even lower). However testing would take quite some time by definition. A band-gap reference typically used has quite some noise, but at least it should be reasonable predictable, so that typical data would already help.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 25, 2018, 08:17:32 pm
Hello,

another notice to the measurement of trise and tfall:

the loading by the 10:1 scope probe (10Meg / 15pF)
of the PWM signal at the processor changes the output voltage by -9 ppm.
The loaded PWM signal at the analog switch output is changed by +18 ppm.

So the results above for trise, tfall are already changed significantly by scope probe loading.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: MiDi on December 26, 2018, 08:26:19 pm
Personally I'd have used jellybean opamps and pnp transistors to achieve similar noise (why don't they specify <10 Hz noise for many/most voltage regulators?).

Think the noise <10Hz from vregs does not really matter for circuits incorporating e.g. opamps with their high PSSR at low frequencies.
Issues arise for higher frequencies where most circuits suffer from decreasing PSRR.
Title: Re: LM399 based 10 V reference
Post by: David Hess on December 27, 2018, 01:30:54 am
At low frequencies, power supply rejection is high so low frequency noise is not a problem.  At high frequencies, passive filtering is effective so high frequency noise is also not a problem.  The exceptions are circuits which have no power supply rejection like single ended logic and low noise oscillators where the noise in the power supply limits the ultimate performance.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 28, 2018, 10:49:44 am
Hello,

first T.C. measurement on branadics PCB.
This time the VCC is externally supplied (14V voltage regulator at room temperature).
Temperature sensor is at the ADG419 switch.

Box TC:
0.0366 mV max change over 22.43  deg C  = 0.1632 ppm/K

Linear regression curve:
-0.144 ppm/K

3rd order regression curve @25 deg C:
-0.124 ppm/K

the resulting error to the linear and 3rd order regression curves is not significant different.
so for a temperature compensation of the switch alone a linear correction would be sufficient.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 29, 2018, 11:22:25 pm
Hello,

2nd measurement
this time with 14V generated from LT1763 on PCB.
Temperature sensor is again at the ADG419 switch.

Box TC:
0.030 mV max change over 23.2 deg C = 0.131 ppm/K

linear regression curve:
-0.117 ppm/K

3rd order regression curve @25 deg C:
-0.117

so the result is slightly below that with external stabilized VCC.

when looking at the stabilized voltage over temperature
it gets clear that the negative TC of the voltage regulator (-65 ppm/K) together with the negative PSRR value of -22 ppm/V gives 0.02 ppm/K reduction of the negative T.C. of the switch.

Box T.C. of LT1763
0.021 V
23.2 deg C
-64.8 ppm/K

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on December 30, 2018, 05:22:29 pm
Thanks Andreas, that's not to bad. So as far as I understand it is possible to improve the output voltage by T.C. compensation in firmware. The question left is towards linearity of the PWM.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 30, 2018, 06:02:48 pm
Hello,

I do not think that linearity will be a large issue.
The different switching times may give a offset.

Another question for me will be noise.
Although with my 6.5 digit instruments I did not see anything unusual up to now.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 30, 2018, 06:21:22 pm
The -60 ppm/K for the regulator is a relative good case - it can be considerably higher, though normally not as bad the worst case.
Even with the worst case TC of the supply, the overall TC would not be much worse than the TC of the LM399.
Besides the effect via the supply, the DG419 could also have an TC from a change in the T_on and T_off.

Another point worth looking with this circuit is the relative high resistance of the filter. So bias current from the OP and maybe leakage could be a problem. Just a constant bias would cause an extra offset (e.g. 100 pA * 300 K = 30 µV).

For a reference circuit linearity of the PWM is not that important. It is more about long term stability and to a smaller extend the TC.
With the rather large resistance nonlinearity should not be that bad. As an crude estimate something like the difference in high / low resistance of the DG419 divided by 100 K could be a crude upper limit. So maybe 1 Ohms / 100 K or 10 ppm.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 30, 2018, 07:00:02 pm
the overall TC would not be much worse than the TC of the LM399.

Near room temperature the typical T.C. of a LM399 is not worse than that of a LTZ1000.
So a additional T.C. of >100 ppb/K matters.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 31, 2018, 12:13:36 pm
Wideband noise measurement of branadics PCB.

51uV RMS over 10 Hz - 100kHz
with only 210uV peak/peak. (so shurely no normal distribution).

in zoom we see that the 10kHz clock from the switch is dominant
-> 40uV RMS or -88dBV for the fundamental.
Frequencies above 10kHz are filtered by the switch/filter.

As comparison the input noise of the used LM399#3
59uV RMS with much higher peak to peak noise and significant parts above 10 kHz.

and finally the noise floor of the used low noise amplifier according to AN83 together with a 10.8V NiMH battery.
0.43uV RMS so enough distance to the measured values.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 31, 2018, 02:27:47 pm
The residual ripple looks quite high. The higher harmonics are also quite present. This suggest that this is not the normal filter function (a kind of soft roll of 3 rd order filter). It could be some other path than just the filtered PWM.

The one point, I don't like very much in the circuit is taking off the signal directly from the switch, where there can be quite some switching spikes.
It might be better to take the signal to the final buffer / filter from before the 100 Ohms resistor (R9), so directly from the 1. st OP. In the shown circuit one could test a little larger value for C22, to get more filtering for spikes coming from the switch.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 01, 2019, 11:46:10 am
@ Kleinstein

Feel free to draw your suggestions into to schematic below, so everyone can benefit from that.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 01, 2019, 12:24:23 pm
I would suggest trying getting the signal directly from the 1 st OP and make C22 a little larger - e.g. 100 nF (more if it fits) instead of 10 nF. The larger capacitor would also make sense in the original circuit.

I think Andreas is already using LTC2057 for both OPs. Otherwise the AD8675 with such a high resistance signal could be a weak point. It is more like a precision BJT based OP could work for the second OP.

Attached is a picture of the suggested change. The drawing is not very pretty, but I hope understandable.
Practically this would be likely removing R15 and add a bodged in resistor.

The hoped for advantage should be less ripple. However there is a slight disadvantage from possible leakage current from the DG419 in combination with the 100 Ohms  - still 1 nA of leakage would only result in 100 nV of error. So it is likely not that significant.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 01, 2019, 04:17:40 pm
Obviously a good idea.

I think we should take this modification into the re-design.
I had to look wether we have still a 10V output after the change ( yes ).

C22 is still 10 nF.
If I would change C22 I would also keep C21 at the same value.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 01, 2019, 05:36:15 pm
C21 is for the compensation of possible capacitance at the output. So this can stay the same, it is not helping much with filtering.

With C22 there is slightly the question where to take the ground from: this could be slightly better with the ground from the output. It's likely not a big difference anyway - just a different point in the layout.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 01, 2019, 06:16:54 pm
Thanks and by the way, happy new year to all of you.

Quote
Obviously a good idea.

I think we should take this modification into the re-design.

Already changed that in the schematic for the redesign, but need to test that too. The output spectrum looks for sure much cleaner now  :-+
Even though I currently have installed an AD8675 for the first opamp, it's no problem to replace it by another LTC2057 like the one I have populated for the output buffer.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 01, 2019, 07:55:30 pm
Another measurement on branadics PCB:

On Off-behaviour:

note that the off board LM399 #3 is always "on" during that experiment so a measurement with a on-board LM399 may differ.

Switching on is without any issues. On switching off we get a short over voltage of around 300 mV @ 14V VCC.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 02, 2019, 08:43:29 am
I just realized that the proposed change at R15 will likely have a negative effect on linearity. This is not so bad for just a reference as it would still be monotonous, but it should be good to know if one wants to use the circuit for other purpose.  The 100 Ohms (due to filtering maybe only a fraction of it) would kind of add to the switch resistance in a crude approximation.

So for a new layout I would keep the option to have the original R15 position.

Title: Re: LM399 based 10 V reference
Post by: Andreas on January 02, 2019, 09:36:01 am
Hello,

perhaps we could get better isolation of the switch by adding a additional buffer.

For the linearity: I see a 1/x function of the PWM signal.
100% PWM gives 7V
50% PWM gives 14V
25% PWM would give 28V

so linearity should be no real issue.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 02, 2019, 11:07:31 am
Linearity should not be that bad. With the original R15 I would expect something like 1-10 ppm INL (relative to the 1/N curve) with the changed R15 for lower ripple maybe 500 ppm INL.  So for the intended use as a reference this should be OK.

I don't think it would be worth an extra buffer for the switches. One might think about placing an inductor (e.g. 100 µH range)  in parallel to the 100 ohms, to lower the resistive drop and still get the transient and high frequency isolation.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 04, 2019, 07:38:14 pm
Some further results on branadics PCB.

voltage drop over R9 is around 2.4 mV (average current through resistor around 24uA)
so if I put R15 tap on the other side of R9 I get a 2.4 mV (240 ppm) higher output voltage.
Influence on T.C./PSRR has to be investigated if we want to do this.

Current consumption:
the 22 Bit software (with up/down Keys removed) uses 20.5 mA for the cirquit (without LM399).
As I always want to use batteries this cuts battery lifetime to 50% of the LM399 alone.
I tried sleep mode which reduces the current consumption to ~13 mA.
My target of staying below 5mA for the 10V transfer cirquit can only reached by reducing the processor and PWM clock frequency.
Reducing PWM frequency will also have a positive effect on PSRR.
But to compensate for this (for ripple) the filter capacities have to be increased.

further questions arise:
Is R2 (3K) on the "right" side of R9 (is the offset current of 0.1% stable? or is the ripple on the other side more critical).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 04, 2019, 08:14:26 pm
The "offset" due to R9 should be relatively stable. Most of the current should be flowing back to the OP through the 100 K resistors. It is more like added nonlinearity. With the position of R15 one kind of has the choice between good linearity, but some ripply, or relatively poor linearity (though in principle possible to calculate) but less ripple.  At high temperatures (e.g. > 40 C ?)  if the leakage of the DG419 really goes up, there would be additional drift with the tap on the quite side.

It might be possible to reduce the offset over R9 with an additional inductor in parallel.

From the pure calculation, one might get away with a slightly slower PWM without that much more ripple. Some of the ripple might be still from parasitic paths and not the ideal RC filter curve.

R2 is at the right side of R9 - it is not that critical anyway. However with the reference current depending on the PWM setting, the linearity and adjustment range is limited anyway. If one wants a larger adjustment range it would be more about an additional auxiliary less stable 10 V just for R2. This does not need to be so stable.

With a lower frequency for the µC one might be able to also reduce the voltage - no need to use 5 V, 4 V might be fine at 12 MHz. This also saves current about proportional to the voltage. There may be a more efficient use of the sleep mode - it depends on how much is used by the oscillator and how much by the actual CPU. I am afraid the lower power oscillator mode would add some jitter - though it might be acceptable.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 04, 2019, 10:24:13 pm
Hello,

why not go a step farther?
3.3V and 8MHz internal R/C clock (should have the lowest power consumption).

A pin-compatible 3.3V regulator (LT1763-3.3) is available.
I only do not know if the Toshlink device will work with 3.3 V.
So eventually only the alternative SFH756V is usable to read out temperature and actual PWM value.

A interesting test for the PWM-ripple would be to split the first 100K resistor into 3 * 47K resistors
one at the old place and the 2 new on the other 2 switch inputs (10V and Gnd) to isolate the switching noise from the output.

By the way: R8 is also a problem.
It uses 1.5 mA.  (5mA * 30% duty cycle)

It could be increased to 2K2 without functional impact.

The better way would use a pull-up resistor (100K) to VCC instead of a pull-down to GND.
(supporting the AVR internal pull-ups during reset instead of fighting against the AVR pull-ups).
But in this case the 2 inputs of the ADG switch would have to be exchanged.
And also the PWM output inverted.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Inverted18650 on January 05, 2019, 08:09:34 am
OEM with both revisions, may be easier for others to follow:

Thanks again to brains behind this project, bowing down "I am not worthy". :-DD
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 05, 2019, 10:00:14 am
I don't think splitting the 100 K to have some of the resistance before the switch is a good idea: this would increase the effect of charge injection quite a bit. In the current circuit most of the injected charge flows back trough R9 or directly to ground.  Charge injection depends on the supply and likely also temperature and this way adds possible drift.

I am not sure if the filter is optimized very well.  Maybe I try a few variations (e.g. different capacitor ratio, a capacitor to ground) in a simulation. So there may be a little room for improvement (better ripple rejection with same capacitance). Slightly larger caps may still be an option if lower clock is important. Also increasing C22 to e.g. 1 µF might be an option - adding another filter stage to the output.

Using the internal RC clock and lower voltage could be a good idea.

If no actively enabled the AVR internal pull ups should be off.
So I see no advantage in inverting the control and using pull-ups instead. I would expect even 50 K to be sufficient for R8. The main function should be during turn on, to prevent an overshoot. However the filter should be slow enough to not allow much overshoot - currently its more like sluggish and R8 may only be there for safety in case the crystal fails to start.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 05, 2019, 10:03:46 am
Changing power supply to 3,3V for safing energy is a good idea. If that means that TOSLINK for isolation does no longer work, fine, no problem whatsoever.
Lowering the PWM frequency means bigger caps and less space on the board. Need to check if that fits into the layout.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 05, 2019, 01:06:17 pm
I did a quick simulation of the filter, sill as a filter and not as an PWM controlled amplifier, but the effect on ripple should be essentially the same.
It looks like the capacitors are well large enough and might have room for a lower frequency. It seems to help with the higher frequency part if the C4 goes to ground instead of the OPs output. This is especially true if the OP is relatively slow (e.g. LTC2057).

The high frequency part (e.g. 4 kHz) the filtering is very effective. So if at all one would expect a problem from the low frequency end due to the modulation of the PWM signal. So there are lower frequencies present, but at a reduced amplitude, depending on how the PWM modulation is calculated.  Worst case one may have to reduce the resolution a little - more than 20 bit resolution only helps if one gets the adjustment at start that accurate.

I don't see a big different with C3 at 1.5 µF. For BOM simplification just 1 µF for the 3 caps is not much worse.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 05, 2019, 08:04:06 pm
If no actively enabled the AVR internal pull ups should be off.

Most processors that I am aware off have the pull ups active during Reset.
In this case the data sheet says "no".
So I have replaced R8 = 100K.
Power consumption decreased from 13.1 to 11.4 mA.

It seems that the switch off spike has increased somewhat.
But the relation is more to the overall power consumption
which previously was 20.5 mA. With a additional 1K pull down resistor
the change is minimal.

8 MHz internal oscillator: (still 5V for the processor).
further decrease from 11.4 to 6.6 mA @ PWM-Clock ~4kHz
So I guess we will be around 5.5 mA when reducing VDD to 3.3V so nearly on target.

Saw during setting the fuses there is a ATTINY15 compatible 6.4 MHz mode
With this clock we are at 5.5 mA at 5V already.
But something is wrong as the PWM clock drops to 800 Hz.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: branadic on January 07, 2019, 01:03:33 pm
Back on the bench I also modified my reference with the suggestions made by Kleinstein. I had minor issues with my first microcontroller, as the fuses where set wrong and it was no longer possible to get access to it.
However, after replacing the micro, programming the 22bit version of the firmware and setting the fuses more or less correct (Extended 0xFF, High 0xDC, Low 0xFF) I'm now able to change and safe the pwm settings.
My reference is currently running on one of my Prema 5017 SC. With the math function active (y=ax+b, with a=1 and b=-10) I can measure down to 8.5 digit resolution, even though this doesn't replace a 8.5 digit multimeter.
So far the reference looks to be stable, but I have to modify my Octave script to also log ambient parameters (TMP112 for measuring temperature at the switch, SHT25 and MS5611).
First impressions coming soon...

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 07, 2019, 10:13:44 pm
Hello,

now I fear I have done too many modifications in one step:

low noise modification according to Kleinstein (R15 tap)
low power modifications to 6.6 mA
- R8 = 100K
- 8 MHz internal R/C oscillator
- sleep mode active

T.C. increased by a factor of 10

so box-TC.
0.315 mV over 22.5 deg C = 1.4 ppm/K

average linear T.C. -1.35 ppm/K
3rd order TC @ 25 deg C -1.27 ppm/K

And I hoped that due to the PWM frequency reduction the T.C. would be better (lower number of edges over time).
Have to analyse why it got worse.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 07, 2019, 10:42:47 pm
Quote
Saw during setting the fuses there is a ATTINY15 compatible 6.4 MHz mode
With this clock we are at 5.5 mA at 5V already.
But something is wrong as the PWM clock drops to 800 Hz.

In this mode system clock drops to 1.6MHz (see page 25 for ATtiny15 compatible mode) --> 1.6MHz/8/256 = 781.25Hz ~= 800Hz.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on January 07, 2019, 10:58:50 pm
Quote
Hello,

now I fear I have done too many modifications in one step:

low noise modification according to Kleinstein (R15 tap)
low power modifications to 6.6 mA
- R8 = 100K
- 8 MHz internal R/C oscillator
- sleep mode active

T.C. increased by a factor of 10

Maybe this is due to the internal RC oscillator and its t.c.? Changing to an external 8MHz crystal could maybe improve things?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 08, 2019, 09:31:28 pm
Hello,

made a quick check by changeing the osccal register. (calibrated Value = 0x97)
Output voltages are after 2:1 divider into ADC25

0x97+00: 3987 Hz 5001.869 mV
0x97+10: 4165 Hz 5001.870 mV
0x97-10: 3803 Hz 5001.839 mV
so around 6 ppm change over 9.5% frequency change.

over a 23 deg C temperature change I would expect much less frequency change
(around 1% when reading from typical datasheet curve)
So this does not explain a 32 ppm change over temperature.

The whole measurement is also relative instable.
there are around 2 ppm changes without any large environment changes.
(perhaps some EMI related influence).

I decided to revert the low noise R15 tap for the next T.C. measurement.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 08, 2019, 10:26:18 pm
Quote
The whole measurement is also relative instable.
there are around 2 ppm changes without any large environment changes.
(perhaps some EMI related influence).

I decided to revert the low noise R15 tap for the next T.C. measurement.

Can't confirm the changes here, the readings look stable (10s integration time), beside ambient influences. I'm currently implementing my environmental sensors to correlate the readings.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 09, 2019, 09:49:35 pm
Hello,

the output voltage fluctuations seem to be correlated to using the internal R/C oscillator.

After having reverted R15-tap I recognized around 70uVpp on the output voltage measured with ADC25 and 2:1 divider.
I looked up the measurement protocol of prevous T.C. measurement ( with low noise connection of R15) and still 40 uVpp with R15-tap and R/C oscillator.

Now I changed only the fuses to XTAL oscillator with 20 MHz XTAL and:
Noise on ADC25 went down to around 10uVpp (which is the normal LTC2400 noise).
Readings are stable now. current consumption (with sleep mode active) is 11.6 mA

So now I think we cannot use the internal R/C oscillator.

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: branadic on January 09, 2019, 10:13:05 pm
Hi Andreas,

thanks for your efffort. I still think power consumption can be reduced by using a lower frequency crystal such as 8MHz. The question left is, what about TC with external crystal and R15 tap as suggested by Kleinstein?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 09, 2019, 10:38:03 pm
Hello branadic,

one step after the other.
my plan is to measure first the current setup with 20 MHz and sleep mode.
2nd step is to change back to low noise R15 tap.

And finally go to lower X-TAL frequency.
My preferred frequency would be 4.9152 MHz to get to a standard baud rate from within interrupt routine.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 12, 2019, 07:07:47 pm
Hello,

some further measurements with branadics PCB:

- 09.01.2019: removed R15 tap against Measurement of 07.01.2019

(RC-oscillator 8MHz + Sleep mode active ~ 6.6 mA no R15 tap)

box-TC. 1.3 ppm/K
average linear T.C. -1.34 ppm/K
3rd order TC @ 25 deg C -1.30 ppm/K

so no significant change in T.C. against R15 tap.

- 10.01.2019 changed back to 20 MHz XTAL oscillator

(XTAL-oscillator 20MHz + Sleep mode active ~ 11.6 mA no R15 tap)

box-TC. 0.12 ppm/K
average linear T.C. -0.10 ppm/K
3rd order TC @ 25 deg C -0.09 ppm/K

so comparable against first T.C. measurement with factor 10 lower T.C.

- 11.01.2019: re-installed R15 tap

(XTAL-oscillator 20MHz + Sleep mode active ~ 11.2 mA with R15 tap)

box-TC. 0.10 ppm/K
average linear T.C. -0.07 ppm/K
3rd order TC @ 25 deg C -0.04 ppm/K

So R15 tap seems to generage a slightly lower T.C.
(But I also guess that repeatability of measurement is around +/- 0.03 ppm/K)
But T.C. over temperature looks rather non - linear.

I have made a overview table over all T.C. measurements.

next measurement will be with 4.9152 MHz XTAL which looks promising with ~4.5 mA.

with best regards

Andreas

Edit: typo on XTAL frequency
Title: Re: LM399 based 10 V reference
Post by: branadic on January 16, 2019, 07:45:18 pm
Thanks for the results.

Quote
next measurement will be with 4.9152 MHz XTAL which looks promising with ~4.5 mA.

So you will run the pwm at 2.4kHz (with prescaler set to /8) or will you change the prescaler to 4 or 2?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 16, 2019, 08:00:12 pm
Hello branadic,

unfortunately there is only a ratio 1:1 or 1:8 as pre-scaler for the used PWM-unit.
The interrupt routine takes around 300 clock cycles (at the 22 bit version of the SW).
(so with 1:1 the interrupt routine would calculate longer than the PWM cycle)

So the PWM-frequency has to drop to 2400 Hz.

I have some strange effects with ~5 MHz now where I do not know if I have contact problems or if the ATTINY oscillator cirquit has generally problems with lower frequencies.  -> will have to repeat some tests.
Also the 8 MHz R/C oscillator shows instabilities (high T.C.) here as already shown.

Can you check wether you have the same instabilities with the R/C clock as I have here?
Could also be that I got a bad batch of ATTINies on my side.

with best regards

Andreas



Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 16, 2019, 08:39:32 pm
300 Clock cycles for the fractional divider sounds like quite a lot. So I assume this would be C code. If it is just die divider and not much interface via UART, one could write ASM code. For a different project I have done a fractional  before - AFAIR the ISR was more like 20 cycles. So the power savings due to sleep mode could get more efficient. However it is still the oscillator running.

I don't think one would need a prescaler at the timer - this does not make much sense, it would mainly give away some resolution with no  savings in current consumption.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 16, 2019, 08:54:35 pm
AFAIR the ISR was more like 20 cycles.

perhaps microseconds (at 16-20 MHz).
in C alone the saving/restoring of internal registers used by the Interrupt needs about 50 cycles.

And remember Its not only a fractional divider it is a fractional divider with dithering which takes a lot of the calculations.
(Have not analyzed yet if it could be optimized).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 16, 2019, 09:19:54 pm
Of cause it's not a fractional divider, but sigma delta type modulation of the PWM setting. However the code is very similar to a factional divider - about the same code length just updating different registers.

C code can be quite long because of saving the registers. This is where ASM can come handy, as one can get away without it, or at least much less savings. In ASM its about 20 cycles if I remember right. It essentially adding a 16 Bit value, and than moving one bit from the result to the PWM setting. So it's really short code and thus easy to write in ASM.

As the timer is 8 bit, it would only need some 2.5 MHz clock to get a 10 kHz PWM frequency.
The ISR would than handle something like 15 or 16 extra bits, though it might be better not to use all of them, but this would be more like choosing the suitable settings.

The ISR would be something like
ADD   sum_L ,  increment
ADC   sum_H,  increment
"Bit 7 of sum_H to  T"   (currently don't know the command)
"T reg. to bit 0 of PWM value"
out  compare1A,  PWM value


RTI  and go back to sleep mode in main loop

So this would be something like 6 cyles + interrupt start overhead and RTI, which should be around 10 cycles.

Using the UART could cause some minor glitches, but may be OK too.

The difficult part would be converting the incoming UART commands and such things.


Edit:
The end of the code suggestion is wrong. Right would be something like this:
mov temp, PMW_coarse
adc temp,  Reg=0
out  OC1A,temp
Title: Re: LM399 based 10 V reference
Post by: e61_phil on January 17, 2019, 07:42:55 am
I haven't in mind what is done over the UART, but if it is only setting new PWM values I wouldn't use an ISR for the PWM at all. One can save the time for pushing everything to the stack and popping it back afterwards. Just a loop which checks for a new UART byte and doing the PWM. If there are new PWM settings it shouldn't matter if the PWM glitches during the new setting.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 17, 2019, 08:30:04 am
To get the high resolution without too much amplitude at the lowest frequencies, it is not just simple PWM, but PWM modulated in a Sigma delta way. So the main PWM frequency is around 10 kHz, but some low amplitude components can extend down to 10 Hz.

The more I look at it, it might not even work so simple, starting with 8 bit PWM - so it may need more advanced software to emulate a higher PWM resolution from the start. Or one would have to start with an even higher PWM frequency.

I would normally prefer a µC that has 16 Bit (or at least 10 Bit) PWM in hardware. So it would be starting with 10 or 11 bit resolution PWM and get the finer part by modulation.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 17, 2019, 06:09:46 pm
Quote
Can you check wether you have the same instabilities with the R/C clock as I have here?
Could also be that I got a bad batch of ATTINies on my side.

Can do that next week, the reference is currently measured with 3458A.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: branadic on January 17, 2019, 09:41:40 pm
Since the source code is accessable via the former link (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1955149/#msg1955149) we can also quote it here, so it becomes obvious what is done and the base for a discussion:

Code: [Select]
/*******************************************
Original code: adver ([url]http://radiokot.ru/forum/viewtopic.php?p=3112844#p3112844[/url])
Modified & ported to AVR Studio: Mickle T.
*******************************************
Compiler                : WinAVR
System Clock            : 16-20 MHz
Fuses : Ceramic Resonator, NO prescaler
*******************************************/
#include <avr/io.h>
#include <util/delay.h>
#include <avr/interrupt.h>
#include <avr/eeprom.h>
//******************************************
#define BASIS 14            // 14 bits of fractional part (8+14bit total => 2.4uV resolution)
#define DIVISOR (1<<14)
#define Key_Up()   !(PINB & ~(1<<PB1)) // Is "UP" pressed?
#define Key_Down() !(PINB & ~(1<<PB2)) // Is "DOWN" pressed?
//******************************************
unsigned long Value = 2885765UL;         
// Value have a fixed point format [8.14]
// 256/Value = amplification factor of 7-10 V converter
// If LM399 output voltage = U_Zener, then
// Value = (U_Zener/10.000 V)*256*(1<<BASIS) = (U_Zener/10.000 V)*4194304
// Value = (6,877576V  / 10V) *256 * 16384 = 2884664,4527104 = 2884664
//******************************************
// MASH internal variables
unsigned int ValInt;                        // Integer part of Value
unsigned int ValFrac;                       // Fractional part of Value
unsigned int q[3]={1,0,0};                 // Latch outputs 16bit
char c[3][3]={{0,0,0}, {0,0,0}, {0,0,0}};   // Carry flip flops 1bit
unsigned int d[3];                          // Adder output / latch data input 16bit
//******************************************
// Calibration factor (Value), saved in EEPROM
EEMEM unsigned long CFG_CAL_VALUE = 2885765UL;
//******************************************
void Init(void)
{
    DDRB |= (1 << PB0);                 // PWM active
    DDRB &= ~((1<<PB1) | (1<<PB2));     // UP & DOWN key pins config
    PORTB |= (1 << PB1) | (1 << PB2);   // Pullup resistors

    ValInt = Value >> BASIS;
    ValFrac = Value % DIVISOR;
    OCR0A = ValInt-1;          // PWM compare match register initial value
    TCCR0A = (1<<COM0A1) | (1<<COM0A0) | (0<<COM0B1) | (0<<COM0B0) | (1<<WGM01) | (1<<WGM00);       
    // Set OC0A on Compare Match, BOTTOM, Normal mode, OC0B disconnected, Fast PWM Mode (WGM = 3)
    TCCR0B = (0<<WGM02) | (0<<CS02) | (1<<CS01) | (0<<CS00);    // Prescaler CK/8
    TIMSK = (1 << TOIE0);  // Enable overflow interrupt
    sei();
}
//******************************************
ISR (TIM0_OVF_vect)
{
    // Carry flip-flops
    c[1][1] = c[1][0];  c[2][2] = c[2][1]; c[2][1] = c[2][0];
    // Adders
    d[0] = (q[0]+ValFrac) % DIVISOR;
    d[1] = (q[1]+d[0]) % DIVISOR;
    d[2] = (q[2]+d[1]) % DIVISOR;
    // Carries
    c[0][0] = (q[0]+ValFrac) >> BASIS;
    c[1][0] = (q[1]+d[0]) >> BASIS;
    c[2][0] = (q[2]+d[1]) >> BASIS;
    // Latch
    for (int i=0; i<3; i++) q[i] = d[i];
    OCR0A = ValInt - 1
            + c[0][0]
            + c[1][0] -   c[1][1]
            + c[2][0] - 2*c[2][1] +   c[2][2];
}
//******************************************
// PWM configuration update
//******************************************
void Update(void)
{
    cli();
    ValInt = Value >> BASIS;
    ValFrac = Value % DIVISOR;
    OCR0A = ValInt-1;         
    sei();
}
//******************************************
// Loading the calibration factor (Value) from EEPROM
//******************************************
void ConfigLoad (void)
{
// If the both keys was pressed on the startup,
// only the failsafe initial value is loaded !!!
    if (!Key_Up() && !Key_Down()) {
      eeprom_read_block(&Value,&CFG_CAL_VALUE,4);
      _delay_ms(3000); };
}
//******************************************
// Update calibration factor in EEPROM
//******************************************
void ConfigUpdate (void)
{
    eeprom_update_block(&Value,&CFG_CAL_VALUE,4);
    _delay_ms(3000);
}
//******************************************
int main( void )
{
    ConfigLoad(); // If UP & DOWN keys was pressed at startup
// then load default config
    Init();

    while(1) {
      if (Key_Up() && Key_Down()) // If UP & DOWN keys was pressed
    ConfigUpdate(); // then save config to EEPROM
      if (Key_Up() && !Key_Down()) {
        Value ++;
        Update(); };
      if (!Key_Up() && Key_Down()) {
        Value --;
        Update(); };
      _delay_ms(300);
    };
}


-branadic-
Title: Re: LM399 based 10 V reference
Post by: fmaimon on January 17, 2019, 11:07:39 pm
Branadic,
Why not use something simpler like the ones below? It's much faster!

Sample code for ISR
https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg105509/#msg105509 (https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg105509/#msg105509)
Full code, including crude uart
https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg125081/#msg125081 (https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg125081/#msg125081)

Assembler optimized ISR https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg125776/#msg125776 (https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/msg125776/#msg125776)
Title: Re: LM399 based 10 V reference
Post by: iMo on January 19, 2019, 03:41:00 pm
An MCU is a generator of a broadband noise (almost continual spectra from mHz to hundreds MHz). For a cleaner PWM output I would recommend to use a buffer, ie those 5pin smd cmos gates, powered by a clean stable source.
For example: http://www.ti.com/product/sn74ahc1g08 (http://www.ti.com/product/sn74ahc1g08)
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 20, 2019, 03:56:17 pm
Hello,

some measurement results.
I think we have first to understand what is happening before we do changes to software/hardware.

The instabilities on 8 MHz R/C clock come together with a very high noise level on the 10V output.
My ADC25 measures around 80-90uVpp in the original cirquit and around 40uVpp with the R15 tap which filters away some noise.
You have to be aware that the 160 ms conversion time of my ADCs have 80 ms for self adjustment and 4*20 ms settling time for the 4rth order SD-Filter.
So effective it corresponds to a around 1 NPLC measurement on a normal instrument.

In the 20 MHz crystal oscillator measurements I did not find any excessive noise up to now. Only the normal ~10uVpp which belong to the LTC2400 of my ADCs.

with 4.9152 MHz (~5 MHz) crystals there is sometimes excessive noise. It seems to be temperature dependant. But sometimes at higher (with R15 Tap) and sometimes at lower temperatures (without R15 Tap). It is almost not visible with the R15 tap but always when the additional filtering is not active.
I also made different oscillator settings (3..8 MHz or > 8MHz) which seems to have some influence on noise and the resulting deviation/hysteresis from the LMS curve.

So I think the oscillator is critical in this application.


Another point:
Voltage across R15 is also dependant on clock (PWM) frequency. At 20 MHz we have 2.466 mV across R15 and At 4.9152 MHz it is 2.222 mV.
This gives 16.2 uV/MHz sensitivity if the (lower noise) R15 tap suggestion of Kleinstein is used.

I checked the output of the spread spectrum algorithm with a PC software:
Results are:
Repetition Period is 32768 PWM periods. (see Autocorrelation). So > 3 seconds with 10 kHz.
Nominal PWM output is exact the fractional part. (as claimed=2862050/2^22)
Noise is with gaussian distribution.
Allan deviation shows ideal stability of the nominal values.
the spread spectrum uses 8 different PWM values for the output, with the original program parameters.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 21, 2019, 10:31:21 pm
Hello,

some jitter measurements for the PWM output of the ATTINY85.
I used my scope for this since I do not have a frequency counter.

The spread spectrum of the PWM output is on the rising edge.
Unfortunately my scope does frequency measurements only on the rising edge.
But fortunately there is a math function to invert the signal.
So I did the frequency measurements on the inverted signal.

With the 20 MHz XTAL we get 0.002 Hz standard deviation.
The 8 MHz R/C oscillator shows 4.3 Hz. So that explains why I have such a instable 10V output with the R/C oscillator.
The 4.9152 MHz XTAL oscillator has 0.00044 Hz standard deviation. (so around a Factor 4 lower than the 20 MHz)
I guess that I had a temperature without excessive noise for the ~5 MHz XTAL.

When looking at the 10V output we see that there is a visible ripple on the PWM output.
(all measurements without R15 tap).

with 20 MHz around 0.25 mVpp
with 8 MHz 0.93 mVpp
and with ~5 MHz 1.7mVpp

so obviously we will have to improve the low pass filter.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Inverted18650 on January 22, 2019, 09:05:29 am

I used my scope for this since I do not have a frequency counter.


If you would, please Sir, allow me to send you an HP5342A, as a thank you. Please just PM me the address and I will send it out tomorrow. Please allow me to do this. It would mean a lot to me (and everyone here), because the work you do is amazing.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 22, 2019, 07:39:35 pm
Quote
With the 20 MHz XTAL we get 0.002 Hz standard deviation.
The 8 MHz R/C oscillator shows 4.3 Hz. So that explains why I have such a instable 10V output with the R/C oscillator.
The 4.9152 MHz XTAL oscillator has 0.00044 Hz standard deviation. (so around a Factor 4 lower than the 20 MHz)
I guess that I had a temperature without excessive noise for the ~5 MHz XTAL.

So relative to their frequency the jitter is the same for both crystals (20MHz: 0.1ppb, 4.9152MHz: 0.09ppb) while the R/C oscillator shows 0.5ppm jitter.
However, I don't really get the numbers for the standard deviation, what are they related to? If they are related to the PWM frequency the question is, what is the nominal mean PWM frequency depending on the clock source?
Attached the second rising edge of the inverted PWM signal in infinity mode with the 20MHz smd crystal assembled. I can see ~400ps noise + jitter with high resolution mode activated. The resolution of the integrated frequency counter of the DSO is not good enough to make a statement concerning mean pwm frequency.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 23, 2019, 10:15:42 pm
because the work you do is amazing.

Thank you liking my measurements.
But as a volt nut I would rarely use such a high end counter.


So relative to their frequency the jitter is the same for both crystals (20MHz: 0.1ppb, 4.9152MHz: 0.09ppb) while the R/C oscillator shows 0.5ppm jitter.


Mhm when we relate the jitter to the PWM frequency (9.7 kHz, 4kHz and 2.4 kHz) the numbers get more handy.
The standard deviation is in Hz so it is an absolute value.
for the voltage output change I would relate it to the PWM frequency.
But the jitter is also filtered like the PWM frequency so we have no direct relation to output noise.

0.4 ns / 100us (10kHz) gives around 4 ppm peak to peak jitter.
my 0.002 Hz in 10kHz standard deviation give 0.2 ppm rms jitter or with 6 sigma around 1.2ppm peak to peak jitter.
So both measurements are in the same ball park.
But I guess with a crystal oscillator we are both at the limit what can be done with a scope measurement.
(which has also a jitter, which is specced with 3 ps RMS in my case).
Would be interesting if the R/C oscillator shows a much more jitter also on your side.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 23, 2019, 10:25:06 pm
Hello,

and some further measurements on the 10V output.
This time with R15 tap according to Kleinstein.

The pwm ripple can now only be detected in the FFT.
In the time based mode the oscilloscope noise is dominant.

with best regards.

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 24, 2019, 06:22:24 pm
Quote
Would be interesting if the R/C oscillator shows a much more jitter also on your side.

For exact comparison, how did you set the fuses for use of internal R/C oscillator?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: nsrmagazin on January 24, 2019, 06:25:51 pm
You don't need 0.01% resistors, you can achieve the same with 1%.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 25, 2019, 06:54:28 am

For exact comparison, how did you set the fuses for use of internal R/C oscillator?


Attached:
Fuses settings (on RC I use minimum startup times but it should not make any difference)
with XTAL I use maximum startup times to let the crystal stabilize on power on.

my current software with sleep mode active and Keys deactivated.

Persistance measurement on my side to be comparable with your measurements with R/C 8 Mhz oscillator.
I used indefinite persistance in this shot to get maximum width because with 5 sec persistance you get only about the half value where the center travels from left to right and back. (drunken walk).

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: branadic on January 25, 2019, 07:35:52 pm
Hi Andreas,

thanks for your input. I tested jitter with your program with 20MHz crystal and internal 8MHz R/C oscillator. Even though I can't reproduce your exact numbers they seem to be in the same ball park.
In my measurements including the one above I used infinity persistance.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: iMo on January 26, 2019, 08:56:26 pm
In case you are chasing the best clock performance forget the xtal oscillator with the MCU. You have to use an external one, the best performance comes usually from a discrete solution. Ask RF people.
Title: Re: LM399 based 10 V reference
Post by: e61_phil on January 26, 2019, 09:46:12 pm
I like this research here very much. It shows that many things aren't that easy in pratice. Starting with the RC oscillator and internal xtal oscillator also shows a lot.

Great work! Thanks!
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 26, 2019, 10:17:16 pm
Hello,

my personal goal with this cirquit is not "best oscillator performance"
I want to have a 7->10V transfer that is:
- stable over temperature (target < 0.1 ppm/K with some temperature compensation)
- power consumption less than around 5 mA
- low noise compared to a LTZ1000 (so less than 0.4 uVpp for the cirquit)

So far as I have seen the oscillator has to be short term stable (within one PWM cycle only) and that over the whole 10-40 deg C temperature range.

What I have seen now:
- A 20 MHz oscillator which is stable enough but takes ~13 mA power.
- A R/C oscillator with 1% peak/peak change of frequency which gives some 10 ppm output variation on the 10V output.
- a 5 MHz oscillator which is in my 5mA target but shows larger instabilities over temperature.

Do you know a external stable Oscillator which takes less than around 2 mA power (since the analog section alone needs 2-3 mA).

Theoretically with lower PWM frequency the analog switch should deliver a more stable ratio (over temperature) at the output.
But what I have seen is more or less contrary. So perhaps we have some compensation effect.

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: e61_phil on January 26, 2019, 10:21:36 pm
Isn't it possible to go into a deep sleep mode during transportation and use the 20MHz oscillator only if you really need it?
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 26, 2019, 11:04:06 pm
I like this research here very much. It shows that many things aren't that easy in pratice. Starting with the RC oscillator and internal xtal oscillator also shows a lot.

Hello,

I already knew that it would not be easy.
Making the divider digital does not mean that all problems are solved against a resistive divider.
The problems are only different.
But what I really did not expect is that the clock stability (even with crystals) is a mayor issue.
I more suspected the rise time over temperature from the switch would be the largest problem.

Perhaps I should try a different processor or a larger crystal size ...

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on January 26, 2019, 11:04:12 pm
Andreas, let's reflect the observations:

the 20MHz Euroquartz crystal is specified

Frequency stability*
 -10° to +60°C from ±5ppm
 -20° to +70°C from ±10ppm
 -40° to +85°C from ±15ppm

Calibration Tolerance at 25ºC: ±30ppm
https://www.reichelt.de/keramik-smd-quarz-2-5x3-2x0-7mm-20-0-mhz-20-000000-mt-p101040.html (https://www.reichelt.de/keramik-smd-quarz-2-5x3-2x0-7mm-20-0-mhz-20-000000-mt-p101040.html)

while the 4,9152MHz crystal is specified
 -10° to +60°C from ±50ppm
https://www.reichelt.de/smd-quarz-grundton-4-915200-mhz-4-9152-hc49-smd-p72506.html (https://www.reichelt.de/smd-quarz-grundton-4-915200-mhz-4-9152-hc49-smd-p72506.html)

So to meet your requirements you need a crystal of low frequency and good temperature stability. What do you think of using Euroquartz 6,0 MHz MQ?
https://www.reichelt.de/keramik-smd-quarz-5x7x1-2mm-6-0-mhz-6-000000-mq-p84988.html (https://www.reichelt.de/keramik-smd-quarz-5x7x1-2mm-6-0-mhz-6-000000-mq-p84988.html)

Frequency stability
 -10° to +60°C from ±5ppm
 -20° to +70°C from ±10ppm
 -40° to +90°C from ±15ppm
 -55 to +125°C from ±20ppm

Calibration Tolerance at 25ºC: ±30ppm

This should fit your current budget as it is ~1/4 of the clock frequency for the µC but also low frequency for the pwm (~2.93 kHz).
The only trade off is, it's not the optimum frequency for communication via optocoupler and FT232RL-USB converter.

EDIT: You could also turn of 10V boost circuit during transport to save another few mili amps.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 27, 2019, 08:40:15 am
So to meet your requirements you need a crystal of low frequency and good temperature stability. What do you think of using Euroquartz 6,0 MHz MQ?

EDIT: You could also turn of 10V boost circuit during transport to save another few mili amps.

Hello,

I do not think that I have a problem with the tempco of the XTAL.
The output change is less than 1 ppm per 1% frequency change if the oscillator is stable.
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2103421/#msg2103421 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2103421/#msg2103421)

It is more the short term fluctuations of the combination of XTAL and Oscillator cirquit at certain temperatures.

And up to now I also have not recognized a difference between the 20 MHz SMD ceramic XTAL and this one
https://www.reichelt.de/standardquarz-grundton-20-000000-mhz-20-0000-hc49u-s-p32853.html? (https://www.reichelt.de/standardquarz-grundton-20-000000-mhz-20-0000-hc49u-s-p32853.html?)
that I am using since January 11th.

There are many options. Including using a resistor divider for a raw adjustment of the output voltage.
And doing a fine trimming of output voltage and T.C. per PWM only.
This would also reduce the requirements for filtering...

I will build a 2nd sample without the processor part.
So I also can check if a different processor will do the job and/or if the one that I picked is the only one which has a problem with the oscillator.
A PIC12F1840 consumes around 2 mA @ 20 MHz according to data sheet and should be able to do a 5 kHz 8 Bit PWM.
And with some code optimization also 19 kHz. (or 10 kHz with a 10 MHz XTAL).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on January 27, 2019, 08:53:54 am
I do not want to mess into your experiments, but my gut feeling is that this kind of pwm divider for your extreme precise low noise needs could work in following setup only:

Discrete Xtal oscillator (ie transistors) or a quality canned oscillator (low phase noise/ampl noise) powered by a low noise stable source ---> MCU or FPGA as the PWM generator -----> PWM signal buffer (ie. a single standalone CMOS gate powered by a low noise source with certain current capability) ------> RC filters -----> analog buffer.

Also mind the "jitter of the edges" is not only source of errors in your case - the overshoots/ringing of at the pwm edges may contribute to the output voltage levels (in case of certain unsymmetrical volume of charges in the pwm pulses).
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 27, 2019, 09:13:54 am
I don't think the oscillator is that critical. There is some nose from clock jitter, but the large noise source should be the lm399 reference. Much of the jitter is often low frequency and thus not effecting a 10 kHz PWM very much.  One may want some shielding so that mains hum does not modulate the oscillator very much.

The idea of the PWM gain setting in not so much a bout low noise, but more about a long time stable gain. So ringing / settling of the PWM steps and charge injection would not matter that much.

A possible effect could be from supply spikes origination from the CMOS switch (DG419) effecting the oscillator. The DG419 is expected to give quite some current spikes when switching and thus need good decoupling. Still the supply ripple would be constant in first approximation.
Title: Re: LM399 based 10 V reference
Post by: branadic on January 27, 2019, 10:24:51 am
Quote
It is more the short term fluctuations of the combination of XTAL and Oscillator cirquit at certain temperatures.

I don't get the reason for this short term fluctuations over temperature yet. Is it because the filter is not optimized for lower pwm frequency (2.4kHz) that comes with lower clock frequency (4.9152MHz)?
A simple test could be using the 20MHz crystal and the internal prescaler to check for that as an issue and compare 20MHz clock frequency vs. 20MHz/8 = 2.5 MHz to identify the problem.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: 2N3055 on January 27, 2019, 11:59:14 am
My 2 c...

If I'm correct, a Fast PWM mode is used, and PWM value is modulated to achieve more resolution.

Every time you change PWM register in Fast PWM mode, it will glitch. Unless you use Phase correct PWM mode on timers.

On 20 MHz clock, glitches are smaller, because of shorter time slices, but are same relative to clock...

I don't know if this is the reason but it might be worth looking into.

Regards,

Siniša
Title: Re: LM399 based 10 V reference
Post by: Echo88 on January 27, 2019, 12:17:35 pm
Interesting discussion!  :-+
But i dont understand the need for low current consumption, since the PWM-part isnt needed at all during transportation. Or do you aim for generally low current consumption Andreas? 
The Datron 49xx reference had a transit-mode which disabled all switching-circuits during transportation, to ensure a lasting supply during long shipping times.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 27, 2019, 01:23:47 pm
Is there really an glitch every time the fast PWM setting is changed ?
According to the DS there is a glitch with a PWM setting of 0 and the other PWM values are double buffered. The double buffering suggests there should be not glitches.

The linked software seems to use a higher order sigma delta modulation, at least it is a relatively complicated calculation, and with C codes it seems to be too slow to operate the timer at full speed and thus needs the 1/8 clock speed. I have not looked in detail at the calculations needed and how fast ASM code for the type of modulation could be. One could at least save quite some time from the IRQ overhead and by keeping the data in registers - so there is a chance it might fit inside the 256 clock window. With the half speed phase correct PWM mode one might have enough time to use the full timer clock and thus 4 times the PWM frequency or at 1/4 the clock speed.

Another option to consider would be using a µC with 16 bit timer (e.g. Tiny13A,Ttiny261).  The hardware PWM could use something like 11 bit resolution so that there is enough time to do the calculations even with the timer running at full speed. This would reduce the lowest frequency component by possibly a factor of 8.

With the LM399 it would be less about transport / storage in active mode. AFAIK the LM399 shows very little hysteresis on power on/off - so one could transport and store in cold state. It is more about using the reference in a floating application.
Title: Re: LM399 based 10 V reference
Post by: chekhov on January 27, 2019, 08:01:54 pm
Hi,

I'm not sure how far your compiler is able to go, but I think it is possible to reduce the job of microprocessor a bit.

Also, what about ARM (Cortex M0 for example) ? Personally I haven't used them before, but I believe this may be much more efficient with integer arithmetic and will definitely have several 16-bit PWM. For me these Tiny family looks attractive because they are small but easy to solder, but probably there might be a different power-effective solution.

Maybe something from my 2 cents is helpful.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 28, 2019, 05:50:55 pm
Hello branadic,

unfortunately there is only a ratio 1:1 or 1:8 as pre-scaler for the used PWM-unit.
The interrupt routine takes around 300 clock cycles (at the 22 bit version of the SW).
(so with 1:1 the interrupt routine would calculate longer than the PWM cycle)

So the PWM-frequency has to drop to 2400 Hz.

I have some strange effects with ~5 MHz now where I do not know if I have contact problems or if the ATTINY oscillator cirquit has generally problems with lower frequencies.  -> will have to repeat some tests.
Also the 8 MHz R/C oscillator shows instabilities (high T.C.) here as already shown.

Can you check wether you have the same instabilities with the R/C clock as I have here?
Could also be that I got a bad batch of ATTINies on my side.

with best regards

Andreas

If the code needs 300 Cycles wit C code, chances are that ASM could bring this down to less than 250 cycles, so that it could work within the 256 cycle window for 8 Bit PWM at full clock speed. It would already fit the 512 cycle window for phase correct PWM - though it might need some extra code to update only every 2nd time. Still this would mean a higher PWM frequency and thus more effect of charge injection or similar. So a µC with HW für 16 Bit PWM would be my preferred way. Starting with 10 Bit PWM at full clock would reduce the filter requirements as the lowest frequency components would be higher by a factor of 8. There area few AVRs (e.g. Tiny13A) with 16 bit PWM even in the small 8 Pin case - possibly even pin compatible. Part of advantage of starting without a divider could be used to use a lower clock.

I don't think the LM399 needs to run 24-7 and also during transport. One might still want battery operation for having a floating reference in some cases.  With a total consumption of some 50 mA with would still be practical with rechargeable cells.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 28, 2019, 08:10:51 pm
Hello,

very interesting discussions.
I have obviously a hardware problem (with the stability of the oscillator) and every body is discussing about software or "better 16-32 bit processor (at best with PLL)" or making wild assumptions on transportation modes which I never intended.

I think the next step is to check the oscillator itself e.g. according to following application note:
https://www.st.com/content/ccc/resource/technical/document/application_note/c6/eb/5e/11/e3/69/43/eb/CD00221665.pdf/files/CD00221665.pdf/jcr:content/translations/en.CD00221665.pdf (https://www.st.com/content/ccc/resource/technical/document/application_note/c6/eb/5e/11/e3/69/43/eb/CD00221665.pdf/files/CD00221665.pdf/jcr:content/translations/en.CD00221665.pdf)
(dont know if there is a better one around).

to the software:

according to data sheet there are no glitches on timer0 unit since all is double buffered and transfer to the registers is at "top" value.
And even on timer1 there would be only glitches if the update would take place "in between" the old and the new value.
Since we are doing only small changes around 70% whereas the calculation takes only 300/2048 cycles = 15% there can be no glitches anyway.

I have analyzed the software: if you are doing it right (by using the carry bit from the status register and correct alignment of the fractional part)
you can do the whole calculation within less than 100 cycles. The "C" implementation is rather long winded solely to calculate the carry bits.
I have made the check for a PIC implementation which generates exactly the same output with around 70 cycles.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: iMo on January 28, 2019, 08:36:05 pm
Also, what about ARM (Cortex M0 for example) ? Personally I haven't used them before, but I believe this may be much more efficient with integer arithmetic and will definitely have several 16-bit PWM.
FYI - these are the 16bit HW PWM module capabilities with STM32F103 @72MHz clock (PLL). That PWM performance is typical for all 16bit HW PWM you may find in those 32bitters.
Not all PWM freqs listed.

For example: 5000Hz PWM freq ---> 14400 max duty  (~13.5bits)

PS:  Try to avoid MCU PLL in your app.
Title: Re: LM399 based 10 V reference
Post by: chekhov on January 28, 2019, 08:38:52 pm
I have obviously a hardware problem (with the stability of the oscillator) and every body is discussing about software or "better 16-32 bit processor (at best with PLL)" or making wild assumptions on transportation modes which I never intended.

Hello,

Sorry for that, my point was to reduce MCU load and as result try to minimize power consumption. Also, in this case external oscillator might be used (with divider if possible), providing better frequency stability rather then internal one,  :blah:
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 28, 2019, 09:45:16 pm
Hello,

no need to apologize. Only wanted to say that software optimizations will come later...

a collegue at work told me that I also should keep an eye on the processor supply.
So I will try to improve the ripple by larger capacitors.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 30, 2019, 08:47:51 pm
Hello,

tried to improve VCC ripple on 5V supply of the processor on branadics 7->10V transfer board.
All measurements done with 4.9152 MHz Xtal.
The active phase and the sleep phase during PWM period time are clearly visible.
Measurements are done on the programming connector Pin 1+Pin 6.
(so the filter cap is on one side and the connector on the other side of the processor.)

Initially I used 100nF X7R + 4,7uF Tantal  -> 4.16mV RMS ripple
I stacked 2,2uF X7R over the 100nF -> 100nF + 2,2uF + 4.7uF -> 3.65mV RMS ripple
Checked 47uF Aluminium electrolytic instead of Tantal ->  3.7 mV RMS ripple
and finally 10uF Tantal -> 100nF + 2,2uF + 10uF -> 3.58 mV RMS

Not too much improvement but it shows that the 100nF+4,7uF is too small for a effective filtering of the ATTINY85.

I think its time to do some stability measurements with the new filter caps.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 30, 2019, 09:02:44 pm
If there is a slightly improved software used to get the ISR to less than 250 cycles, one could use a divider of 1 at the timer and a lower CPU clock (e.g. 4 MHz instead of 20 MHz, and maybe a divider at the oscillator - not sure the tiny85 supports this) . In this case there would be not much time left for the sleep mode and one might just skip it to avoid the power supply problems. So a improved software could be an alternative to more supply filtering.
Title: Re: LM399 based 10 V reference
Post by: chekhov on January 30, 2019, 09:06:23 pm
But what is the real problem with this ripple ?
Does it just influence other circuitry and adds noise through opamps even having second LT1763 ?
Or because of that frequency changes somehow ?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 30, 2019, 09:18:40 pm
With the RC oscillator the supply ripple can effect the frequency and this way add to nonlinearity and drift (as the effect can depend on temperature too). To a smaller extend this also applies to a crystal oscillator.

The ripple could also indirectly cause some charge injection.
Title: Re: LM399 based 10 V reference
Post by: chekhov on January 30, 2019, 09:22:51 pm
Thanks,
I was more interested in influence on crystal-based solution.
I wonder will that end up with opto-coupling of switch input from MCU and powering it from different source (separate 3.7V cell ?).
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 30, 2019, 09:34:25 pm
An opto-coupler would add extra delays (especially different turn on / of delays). So I don't think this would be a good idea. To get an accurate PWM across it would be  raw PWM plus an extra clock for synchronization.

The crystal based solution should not be that sensitive to supply ripply. If power is not critical not using a sleep mode and thus a near constant current consumption would be an option.
Title: Re: LM399 based 10 V reference
Post by: chekhov on February 01, 2019, 09:27:51 am
I think its time to do some stability measurements with the new filter caps.

with best regards

Andreas

Regarding the ripple, what if, as a test/temporary solution, instead of lowering MCU consumption you try a reverse way - don't allow MCU to rest ?
This way instead of sleep mode in the main loop, this will become busy loop, doing anything (maybe just the same detection of button press).
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 01, 2019, 06:03:17 pm
Regarding the ripple, what if, as a test/temporary solution, instead of lowering MCU consumption you try a reverse way - don't allow MCU to rest ?
This way instead of sleep mode in the main loop, this will become busy loop, doing anything (maybe just the same detection of button press).

Hello,

thats what Kleinstein already suggested.
But I do not like to patch hardware problems by software without knowing the reason.
Experience says that the problems will return. (then when you do not expect it).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on February 01, 2019, 10:12:08 pm
Andreas,

are you sure about the problems of the oscillator itself? A difference between your circuit and mine is, I use AD8675 for the first and LTC2057 for the second opamp. Could it be that you measure some interference? Just a thought, maybe it's abolutely on the wrong track.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 02, 2019, 06:45:00 am
Hello,

thats really a point worth checking.
On the other side: Up to now I have never seen the excessive noise with the 20 MHz XTAL
only with the 4.9152 MHz.

I see this as indication for the oscillator with low frequencies.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 05, 2019, 09:36:43 pm
Hello,

now it is clear to me what is going on with the 4.9152 MHz X-TAL.
It is not a instability of the Oscillator but simply a very marginal analog filter together with the short integration time of my ADCs.

To analyze this I monitored the 2400 Hz PWM signal with a mask test of the oscilloscope.
But all what I could find is that the 417 us period time changed around 2 ns over temperature.
(so around 5 ppm change over 20 deg C)

The oscilloscope probe 12-15 pF also changed the output voltage by 1-2 ppm.
So it is shurely worth to keep the parasitic capacitances constant around the PWM signal by shielding it on both sides.

For further analyzing I made a LTSPICE model of the complete cirquit.
Since there is no model for the ADG419 available on the ADI site I looked for a replacement.
I found a FAQ note that the switch in ADG411 is the same than for the ADG419 and so you can replace the ADG419 with a ADG413 which is in the same datasheet as the 411.

Then it got clear: the remaining ripple at the output of the filter is > 1mVpp with the 5 MHz XTAL. With a effective integration time of 20ms (1 NPLC) and if there is no exact integer number as ratio between the 2400 and 50 Hz we have only a dampening of a factor < 50 due to the integration time. So I see up to around 30 uVpp at the ADC depending on the LTC2400 internal oscillator frequency against the XTAL oscillator frequency.
That also explains why I have sometimes the excessive noise at high temperature and sometimes at low temperatures: The LTC2400 is not held at constant temperature.

Even at 20 MHz XTAL the ripple of the ION filter (without R15 tap) is > 100uVpp which is in my opinion too much for a reference cirquit.
So in any case we have to improve the filter.

with best regards

Andreas


Title: Re: LM399 based 10 V reference
Post by: Andreas on February 05, 2019, 10:03:25 pm
Hello,

I also tried to optimize the code for the ATTINY.
Here the first result (up to now only tested in simulator).

With this implementation the interrupt routine is around 120 cycles so the timer can be operated with a 1:1 divider instead of 8:1.
So a 2.45 MHz XTAL will give nearly the same PWM frequency like 20 MHz with 8:1.
But hopefully with much less current.
I think I will try this on week end.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 06, 2019, 05:31:54 pm
The new code looks odd - very much like the PIC code, but I don't see how the information flow from the q to the c values should work. I only see the shifts for the c's and no effect from the q's. So I guess there is something wrong in the code.

I very much like the idea of improved code to make it work inside the 250 cycle window and thus 1:1 timer clock. In this case one may not need a sleep mode anymore and still get low power consumption.

The oscillator has different settings for different frequency ranges. So one may have to adjust the fuse settings when going from a 20 MHz to a 4 MHz crystal.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 06, 2019, 06:14:26 pm
First: 2.5MHz crystals are available, so this could help to prevent 1/8 divider for the timer.
https://www.reichelt.de/standardquarz-grundton-2-500000-mhz-2-5000-hc18-p1852.html?&trstct=pol_1 (https://www.reichelt.de/standardquarz-grundton-2-500000-mhz-2-5000-hc18-p1852.html?&trstct=pol_1)

Second: Andreas, were you able to verify the simulated results by real world measurements in time and frequency domain? Up to know I couldn't measure such voltage swings.

Third: What if you change your simulation to R15 tap? Any improvements?

Instead of AD411 I used a switch model in LTSpice with parameters such as slew rate and resistance given from the datasheet.

I'm not an expert on software/microcontroller firmware so if anyone here with the necessary skills on AVR microcontrollers and Assembler programming would make a code suggestion we can test it. I can also setup one board for you to join the club.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 06, 2019, 08:11:32 pm
The new code looks odd - very much like the PIC code, but I don't see how the information flow from the q to the c values should work. I only see the shifts for the c's and no effect from the q's. So I guess there is something wrong in the code.

You have to think in assembler.
It is essentially the "Bresenham" code but 3rd order interleaved.

with the 8.14 code I have to shift all fractional values by 2 bits left (ValFrac4 = ValFrac * 4)
So I get an overflow into the C bit of the Status register when 65536 or above is reached.
The inline code shifts this Carry-Bit directly into the Bit 0 of the C0, C1 or C2 variable.
All other bits are shifted left (so I avoid the assignments from c[2] = c[1] ...)
In the disassembler the ROL command is translated as ADC Rx,Rx (which is the equivalent).

I have tested both codes in simulator in parallel and have compared the output values over
a simulation time of > 70 seconds real time. (so more than 10 cycles of the random code).
 the error counter of comparison of the two results was zero the whole time.
So dont tell me that it does not work.

With best regards

Andreas



Title: Re: LM399 based 10 V reference
Post by: Andreas on February 06, 2019, 08:23:19 pm

Second: Andreas, were you able to verify the simulated results by real world measurements in time and frequency domain? Up to know I couldn't measure such voltage swings.

Third: What if you change your simulation to R15 tap? Any improvements?


-branadic-

The measurement for the 20+5 MHz xtal is on the previous page see e.g.:

https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=630511 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=630511)
The results are not exact but similar.

Attached the LTSPICE cirquit and the ADG413 model so you can evaluate the R15 tap by your own.

Note that you will see the ripple only with either the scope or with a relative short integration time (1 NPLC).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 06, 2019, 08:55:47 pm
The new code looks odd - very much like the PIC code, but I don't see how the information flow from the q to the c values should work. I only see the shifts for the c's and no effect from the q's. So I guess there is something wrong in the code.

You have to think in assembler.
It is essentially the "Bresenham" code but 3rd order interleaved.

with the 8.14 code I have to shift all fractional values by 2 bits left (ValFrac4 = ValFrac * 4)
So I get an overflow into the C bit of the Status register when 65536 or above is reached.
The inline code shifts this Carry-Bit directly into the Bit 0 of the C0, C1 or C2 variable.
All other bits are shifted left (so I avoid the assignments from c[2] = c[1] ...)
In the disassembler the ROL command is translated as ADC Rx,Rx (which is the equivalent).

I have tested both codes in simulator in parallel and have compared the output values over
a simulation time of > 70 seconds real time. (so more than 10 cycles of the random code).
 the error counter of comparison of the two results was zero the whole time.
So dont tell me that it does not work.

With best regards

Andreas

OK I now see how it works:  It uses the carry flag from the C instruction before. This is a rather tricky way of combining C and ASM code - it works here, but depends on the code the compiler creates for the given C code. So it's slightly fragile code - though I won't expect a different compiler version to create much different code. But better not try -O3 here.
Title: Re: LM399 based 10 V reference
Post by: branadic on February 06, 2019, 09:32:11 pm
Andreas,

I've a different value for C6 in your LTSpice, which I changed to 100nF also on my board. I made a comparison with and without R15 tap on my spice model of the circuit. Attached is the circuit for cross checking. You can see that with R15 tap (vout2) there is no oscillation, while without R15 tap (vout1) it's clearly visible.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 06, 2019, 09:49:35 pm
The extra oscillation (which should better be called ripple) is likely due to charge injection from the CMOS switch. It presents quite some load and even with the capacitors to ground (C4 / C11) can cause the extra ripple. So the original circuit (vout1) shows more (especially high frequency) ripple. However the modified circuit should show extra nonlinearity (or better a more complicated function).

If one is after lower ripple, one could consider having the first filter cap (C10/ C3)  no towards the OPs output but to ground. At least in the simulation of the filter it showed better suppression of higher frequency ripple, where the OPs output impedance is not very low. How large the effect in real life is can depend on the OP model and layout.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 06, 2019, 10:47:42 pm
The extra oscillation (which should better be called ripple) is likely due to charge injection from the CMOS switch.

Charge injection is not modeled in the normal switch model.
And practically parasytic capacitances + charge injection are less than 10% in real world.
Most of the ripple in the output (without R15 tap) is from the voltage divider built from the
100R and the 100K resistor working against the 3V (charging + discharging)
of the feedback capacitor (C3).

So one measure to try (without R15 tap) would be reducing the 100R to 10-22R if the OP-Amp allows this (stability).
Usually the LTC2057 is stable with 22R and capacitive loads.

If one is after lower ripple, one could consider having the first filter cap (C10/ C3)  no towards the OPs output but to ground.

In my closed loop simulation C3 against ground showed no effect.

I've a different value for C6 in your LTSpice, which I changed to 100nF also on my board.

100nF gives a improvement of around factor 10 in simulation.  (corner frequency with 100nF is 340 Hz)
(On my board I still have the 10nF as I had to fix the ground pin of the output LTC2057 somewhere).

And your simulation works with 10kHz PWM frequency and not with 2400 Hz.

The R15 tap also gives further improvement. (never saw excessive noise with R15 tap and XTAL).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 07, 2019, 06:41:13 am
Attached is the circuit for cross checking. You can see that with R15 tap (vout2) there is no oscillation, while without R15 tap (vout1) it's clearly visible.

-branadic-

Edit.
Rise 21 nS. Fall 8.5 nS.

Edit 2
Delay

Edit 3
79 kHz
8+14 bit
source code
Title: Re: LM399 based 10 V reference
Post by: iMo on February 08, 2019, 12:13:06 pm
In order to run the above schematics of Andreas, you have to open the ADG413.cir, go to
.SUBCKT ADG413   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
right click on the "ADG413" and do "Create Symbol". It creates and saves the ADG413 symbol into your LTspice repo. Open the simulation again and you will see the yellow ADG413 symbol there.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 08, 2019, 02:37:04 pm
Another simulation..
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 08, 2019, 03:45:26 pm
Сomparison of different models, and test output.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 09, 2019, 06:24:05 am
Edit 3
79 kHz
8+14 bit
source code

I am still missing your source code implementation for the 79 kHz.

I have further optimized the code now being at ~100 cycles and verification is running at the moment.

The ideas:
C0 is not needed after being output so you can save the storage to a global variable and use a local register variable.
If you calculate the results for C1 and C2 directly 2 registers are not used (and do not need to be pushed and popped)
Further optimization is possible if not using the "+" operator together with and-masks. resulting in single bit operations.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on February 09, 2019, 10:13:01 am
Andreas,

I performed some simulations last night for verification and could reproduce all of your results. So for a redesign of this LM399 PWM DAC reference board we can conclude:

- R15 tap modification should be implemented in the board (did that already with bodge wires)
- C6 (simulation) / C22 (board files) should be changed from 10nF to 100nF (did that already on my board long time ago)
- replace SMD crystal by HC49 as this gives more flexibility in choosing the frequency value (e.g. from 2.5MHz up to 20MHz, while 2.5MHz has the advantage to surrender 1/8 prescaler and reduce power consuption of the circuit, while PWM frequency remains the same 10kHz)
- additional 100nF MKS2 at the output of second opamp

This board design could also be used for a reference based on 2DW233 by choosing an appropiate value for R2 in the board files.

To use the main circuit idea for a 7V --> 10V boost on LTZ1000 reference there are still some improvements to make, especially if the reference is used as a portable type to get power consumption down. However, the board design will look different as in this case the board doesn't need to fit into a Hammond 1455 case.
Time to order a redesign at Elecrow?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 09, 2019, 11:03:48 am
Getting the code down to well below 200 cycles for the ISR is a big step forward, as it allows to use the timer at full speed.  I see relatively little advantage in even fast code - with still about 50% of the CPU time needed It may not be worth using a sleep mode, unless really low power is aimed for.  In this case I would consider a different µC that supports 16 bit or similar PWM in hardware. One could than use something like 10 or 12 Bit HW PWM and have only some 5-10% CPU usage to allow more power savings. The Tiny13A seems to not support an external crystal - so not a good candidate.

A higher native PWM resolution might also be an advantage for a not power critical case. One could havehigher resolution,  less modulation or a lower PWM frequency without need for higher capacitance in the filter.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 09, 2019, 10:55:02 pm
- R15 tap modification should be implemented in the board (did that already with bodge wires)
- C6 (simulation) / C22 (board files) should be changed from 10nF to 100nF (did that already on my board long time ago)
- replace SMD crystal by HC49 as this gives more flexibility in choosing the frequency value (e.g. from 2.5MHz up to 20MHz, while 2.5MHz has the advantage to surrender 1/8 prescaler and reduce power consuption of the circuit, while PWM frequency remains the same 10kHz)
- additional 100nF MKS2 at the output of second opamp

Hello Branadic,

I see the R15 tap still as a option. The 2.4 mV (average) voltage drop on the 100R resistor is outside the regulation loop.
I would change C22 even to 1-10uF as Foil (MKS2) capacitor.
HC49: yes (since I have all usually used frequencies in HC18/HC49 in my drawer).
C25 at the output with 100nF is also ok (EMI reduction).

Other things not to forget:
C8: I want to populate 2.2uF (0805) is ok but do not forget either a 10uF Ta Size A or a 47uF 6.3 mm diameter in parallel for voltage regulator stability.
I want to increase C1-C3 to test lower PWM-frequencies. (The 100k resistors could be on the soldering side if space is a concern).
The PWM-clock from Tiny to switch should be close to a digital ground line. (or even surrounded by it).

As it is chinese new year until mid of next week I would not hurry too much with ordering.
I want to simulate the spitted 100k resistors this week end if there would be a advantage (or if we could leave it off).

with best regards

Andreas




Title: Re: LM399 based 10 V reference
Post by: Andreas on February 09, 2019, 11:05:24 pm
Hello,

attached the 2.45 MHz version with 9.6 kHz PWM and optimized Interrupt service routine.
I get occasionally jitter of 5-6ns on the 9.6kHz period time but seems to have no influence on the 10V output.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: pigrew on February 10, 2019, 02:55:14 am
Questions from an observer:

Frequency modulation is often good from an EMI standpoint for DC-DC power converters (many recent papers on random frequency variation). Also, multi-channel converters are often used.

Should these techniques be applied to a PWM-DAC? Multiple channels should get sqrt(N) extra bits of DAC resolution. The design in my head would have all channels be equal-weight, so the mismatch could be averaged out over time. I think that four-channels may be a reasonable.

But, I don't know if the frequency modulation would help or hurt the filtering. Though, it is certainly much more computationally intensive.

Also, how was the 9.6 kHz frequency obtained? Was it because the analog circuit is less ideal at higher frequencies, or due to computational concerns?

I'm tempted to put together a similar proof-of-concept circuit with 2DW232 and glow-cost eneric SMD parts, and see if I can get good performance and digital temperature correction.

Thoughts?

-Nathan
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 10, 2019, 06:13:15 am

But, I don't know if the frequency modulation would help or hurt the filtering. Though, it is certainly much more computationally intensive.

Also, how was the 9.6 kHz frequency obtained? Was it because the analog circuit is less ideal at higher frequencies, or due to computational concerns?


Hello Nathan,

I guess with frequency modulation (which is usually in the 1-2% range similar to the 1% frequency jitter what I had when trying the internal R/C oscillator) we get the same result as with the internal R/C-Oscillator where I had > 70uVpp noise and instable voltage readings.
So I would say it hurts.
But of cause it depends.
I personally would never use a CPU with PLL or spread spectrum in this application. (except if you can do the frequency change exactly at the full PWM period times).

The ~10 kHz PWM is a design decision from adver, see also
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1955149/#msg1955149 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1955149/#msg1955149)

I personally would go lower in PWM-frequency exactly for that reason.
On the other side: with my cirquit I had the lowest T.C. with the 20 MHz XTAL (so with 9.7 kHz PWM frequency) -> strange.
As the ripple measurements / simulations have shown: with a lower PWM frequency you will have also to improve the filter cirquit.

But there are limits: the 3*100k resistors are already at a value which is above that what you would want to have at the input of a LTC2057. (max around 70K). And good quality foil capacitors are rare to get above 10uF.

So I will make next some T.C. measurements at different frequencies to see if there is a "sweet spot" together with the ADG419.
The resulting errors are the lower the lower the initial T.C. is. (Because the temperature sensor is never at the place where it should be and thus measures not the exact temperature what you would need for the compensation).

By the way: with the 2.45 MHz X-TAL I am now at 4.7 mA without sleep mode and at 4.1 mA with sleep mode.
So the analog part of the cirquit draws obviously the most current.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 10, 2019, 10:25:05 am
If the 100 K resistors are a concern, one could consider making the resistors more towards the OP a little smaller (e.g. 47 k) and maybe use 2.2 µF caps. In my simulation I see no real advantage to have the middle capacitor larger (1.5 µF compared to the other 1 µF) - so 3 same size caps are likely OK. The initial 100 K resistor sets the values the R_on of the DG419 is measured when it comes to linearity - so here it's not that good to make the resistor much smaller.
For C6/C22 I would consider the same 1 µF value as for the filter, so it would already filter the 10 kHz main PWM frequency. Due to the smaller resistor it's not as effective as the cap at the main filter. It is more for some ripple that gets around the filter due to parasitic effects.

If the drop over R15 is an issue, one could still add an inductor (> 50µH as first guess) in parallel.

One might need an improved filter, if the µC clock gets to low. One could likely still use a lower PWM frequency if the µC supports higher PWM resolution. So it would than be 10 or 12 bit from the hardware PWM and less from modulation. The real difficulty for the filter should not be the PWM frequency, but the lower frequency contend due to the modulation - it's low amplitude, but can also contain rather low frequency parts.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 10, 2019, 02:12:41 pm
You may utilize the second 2057 as a 2nd order low pass (ie 10Hz below). And relax the first filter if necessary..
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 11, 2019, 08:26:12 pm
Hello,

@IMO: nice schematics. From where did you get the model of the ADG419?

I checked some variations of the cirquit by simulation:

- R1 100K splitted into 3 different 50K resistors
  I hoped that this would reduce the ripple after the 100 R resistor due to less influence from the charge injection.
  But obviously the effect is only a minor. not worth on getting different T.C.s for the 3 resistors.

- Reducing the 100R resistor. Same hope as above since I thougt that the 100R resistor together with the 100K build a voltage divider with the influence lower for lower 100R resistor.
  But obviously the ripple is only a thing between discharging C4 over the 100K Resistor R1.

- Finally exchanged C4 with 10uF: this gave the most reduction on ripple for the output voltage.

interestingly C3 does not play a large role for the filter. I have reduced C3 to 10nF without significant change on output ripple at VOUT2 in simulation. Now trying influence of C2.

@Branadic: can we change C4 up to 10uF in the layout?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on February 11, 2019, 08:44:13 pm
Quote
@Branadic: can we change C4 up to 10uF in the layout?

Currently C4 is a 1µF/63V MKS2 type (7.2mm x 5.0mm x 10mm). Changing it to 10µF means replace by 10µF/50V MKS2 type and requires quite some space (7.2mm x 11mm x 16mm). Have to check if I can make this space available.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 11, 2019, 08:59:09 pm
Leakage of C4 would only be effective through the 100 Ohms resistor and if the output signal is taken from C4, the leakage would not even matter. So it might be possible to use an electrolytic cap for C4, preferable low ESR and maybe with some 100 nF in parallel.

Splitting the 100 K to 50 K before and after the switch does not look like a good idea, I think we had that idea before.

For me C3 still has an effect. It may not be visible in the simple simulation with only 10 kHz. It can be important for the lower frequency components due to the modulation of the PWM signal. Even with 10 kHz main PWM there is a little amplitude down to the 100 Hz range.
For me it's more like C2+C3 are important, but C2 and C3 could be same value.
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 12, 2019, 05:38:15 pm
I am still missing your source code implementation for the 79 kHz.

My code is not optimized.
Enough only for PWM.

But in the source code there are errors.
1 - PWM start before memory initialization.
2 - load from memory before initialization of buttons.
3 - the length of the variables is more than necessary.

Just compare mine and your code.
Successes!   ;)

Title: Re: LM399 based 10 V reference
Post by: pigrew on February 12, 2019, 10:18:09 pm
I've started working on a PCB design for this project, but with few tweaks. First, I plan to use the STM32G071 CPU, which is fast and pretty low power. I've gotten some reasonable results (though too much HF switching noise due to breadboard construction) with a STM32F103  (PLL enabled), so I'd like to see how it performs. Also, I propose to build a 3-level bit DAC using two synchronized PWM. Since this is geared towards 10V, we can do better if we PWM between 5V and 10V instead of between 0V and 10V. For the 5V output, I will flip-flop between switches (use one for the "low" branch and use the other one for the "low" branch the next cycle). In my current implementation, I've also implemented random frequency variation which greatly spreads out the frequency response, but it may be bad due to offsets caused by unequal rise and fall times of the switch (though this does eventually average out).

I've drawn up an "analog" schematic and I've appreciate feedback.

I've marked all the capacitors as poly-film or tantalum, I figure they are less risky for piezo-electric badness, but I likely will just use X5R (and components I already have) on the first build.

I've tried to incorporate most of the features I've seen here. I was thinking about using Kelvin force/sense, or designing a BJT output stage, but I will add those after making a first board work decently.


EDIT: Adding the following specific questions, also replaced the schematic with an updated one.

Things in particular to decide before layout:

Title: Re: LM399 based 10 V reference
Post by: Andreas on February 13, 2019, 11:51:22 am
I've started working on a PCB design for this project, but with few tweaks. First, I plan to use the STM32G071 CPU, which is fast and pretty low power.

 Also, I propose to build a 3-level bit DAC using two synchronized PWM. Since this is geared towards 10V, we can do better if we PWM between 5V and 10V instead of between 0V and 10V. For the 5V output, I will flip-flop between switches (use one for the "low" branch and use the other one for the "low" branch the next cycle). In my current implementation, I've also implemented random frequency variation which greatly spreads out the frequency response, but it may be bad due to offsets caused by unequal rise and fall times of the switch (though this does eventually average out).

I've drawn up an "analog" schematic and I've appreciate feedback.

Hello,

Fine to see a different approach with different ideas. How much current is drawn by your CPU?
I have to admit that I do not understand how your PWM works. (also not familiar with the STM CPU).
Do you have something like a timing diagram?
When combining 2 different PWMs I would have expected a e.g. 100:1 ratio (coarse/fine) between the mixing resistors.
How many bits are used from the PWM?

I´m trying different PWM frequencies at the moment.
The frequency has (as expected) a influence on the output voltage in the mV range.
Interestingly the frequency also has a influence on the T.C. of the whole cirquit.
The sweet spot (at least for my sample of ADG419 seems to be between 9.6 and 19.2 kHz (more near the upper frequency)).
Of course this may also depend on if you use 3.3V or 5V input levels.

Things in particular to decide before layout:

  • ADG419, ADG1419, or ADG5419. I'm leaning towards ADG1419 due to lower on resistance, but it has larger parasitic capacitances though much lower charge injection
  • Switch to LTC2058 (at least for prototyping) since it is cheaper (per op-amp), nearly as good?
  • Use ceramic X7R 0.1uF caps for decoupling?
  • Add ferrite at switch output (next to R1)??? Perhaps need to build the circuit and try with/without?
  • SMD resistors have fewer parasitics, but perhaps we want the series inductance of THT resistors?
  • Looking back into the thread, I screwed up the positioning of the 10V node and 100 ohm R4. Fixing that, will upload 20190212b soon.
  • Switch to Sallen-Key LPF on the output?
  • I want to use the STM32G0 because it's low power, but it seems that the 64-pin package is the best choice. Ugh.

For the switch: we started with an existing design (ION) so no idea which is really the best. (we need more experience so perhaps it would be good to use a different one).

LTC2058 is the dual version of LTC2057. So should give no differences. But usually single OPs make the PCB layout easier.

I see no problem with ceramic CAPs on low impedant power supply lines. In the analog path I would use Film capacitors in the final layout.

for the ferrite: as R1 is rather high impedant 150K in parallel with 0.x pF a ferrite at this point will not have too much effect.
I also had the idea of using a ferrite/inductor but more from SA input of the switch to the C16 (VRef10V) side.
But LTSPICE simulation had up to now not the effect that I have expected. I guess I still have a problem with my Switch model.

What we did not up to now is a series resistor between digital PWM output and switch for dampening digital noise.
but of course this may give additional influences on rise/fall time symmetry.

I would not rely on the inductive behaviour of resistors. As this may vary from batch to batch.
I use through hole only if I need long term stability.

To the schematics:
C11 C12 could be larger in value to reduce ripple on the output. Also C16 plays a large role for the remaining ripple.
I would place C10 between PIN6 and PIN2 of the OP-Amp to allow capacitive loads on the output.
If you actually place a series resistor between the voltage regulator and heater of the LM399 you might need a additional electrolytic capacitor for the LM399 heater (see LM399 data sheet).

with best regards

Andreas






Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 13, 2019, 04:24:55 pm
The ARM based circuit is not that much different. The PWM is still in the same way, just 2 channels in parallel. As shown the two channels have the same weight and one would get about the average PWM value that count. With 2 equal PWM settings this reduces the ripple to start with a little, as in most case the PWM would be more like out of phase.

Going from 7 V to 10 V has a PWM setting of about 70% and thus a considerable reduction in ripple to start with. However the loading problem to the first OP is still there. A lower resistance switch would help a little with "linearity", but could cause more load on the input side. It could help to have switches that a specified for only 16 V or so, as these might have low R_on with still low spikes.

The STM32 µC should be able to use 16 bit or similar PWM, which can be a slight advantage with a higher base clock. So less low frequency component. to filter.

Still I am not so sure using 2 switches is worth the extra effort (e.g. 2 x DG419 or similar). Especially with a slightly higher PWM base frequency the low frequency ripple should not be that critical. It is more about parasitic effects and  higher frequency spikes that go around the filter.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 13, 2019, 07:18:20 pm
@IMO: nice schematics. From where did you get the model of the ADG419?
Bordodynov library. It includes various switches.
https://forum.allaboutcircuits.com/threads/the-libraries-and-components-models-of-ltspice-free-download.133690/
Title: Re: LM399 based 10 V reference
Post by: iMo on February 13, 2019, 07:34:09 pm
The STM32 µC should be able to use 16 bit or similar PWM, which can be a slight advantage with a higher base clock. So less low frequency component. to filter.
Above are the tables with stm32 @72MHz and 16bit hw pwm ranges. When not using PLL and targeting "low power" a realistic clock could be something like 24MHz. With 24MHz clock you may get 10kHz PWM with ~11bit resolution.
PS: a 10kHz PWM with full 16bit resolution would require 655MHz clock :)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 13, 2019, 08:07:01 pm
The point was not so much really using 16 Bit PWM, but having HW that supports at least 12 Bits, so that something like 11 bits can be used. So it's about 3 more bits's of PWM resolution and still 10 KHz. So something like an amplitude for the modulated lower frequency part that is only 1/8 and does not extend all that far down as the original solution.

So the higher PWM clock is already more efficient than using 2 channels, which kind of halves the amplitude.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 13, 2019, 08:07:12 pm
STM32x HW PWM. The "max duty" is the max value (100%) you may load into the pwm register for a specific pwm frequency.

For example at 10kHz pwm frequency the max duty is "2400" that is something like 11.x bits.

You may have many 16bit pwm channels active, btw (up to four independent 16bit pwm channels for each single 16bit Timer_x used).

Theoretically you may design a 44+bit DAC with four pwm channels at 10kHz pwm frequency.. :)
Title: Re: LM399 based 10 V reference
Post by: Magnificent Bastard on February 13, 2019, 08:26:30 pm
Microchip has a new zero-drift op-amp: MCP6V51 (https://www.microchip.com/wwwproducts/en/MCP6V51)

It has very low input current noise, which allows for a very high input impedance without spoiling the rather good input voltage noise spec.  Of course, all of the horrors of using a chopper need to be addressed-- such as clock feed-through, etc.-- but with proper circuit design, this is a real contender to the offerings out there right now.

-MB
Title: Re: LM399 based 10 V reference
Post by: iMo on February 13, 2019, 09:10:54 pm
Switch to Sallen-Key LPF on the output?
There is around 800uVpp (@10KHz pwm) ripple at the output of the ADG switch (in my simulation above). You have to decide what is the targeted ripple amplitude at the ION's output. That gives you an idea how to filter it out..
PS: Getting a little bit different results when the V- of the DG switch is connected to -15V instead to the GND.
Title: Re: LM399 based 10 V reference
Post by: pigrew on February 13, 2019, 10:00:53 pm
I´m trying different PWM frequencies at the moment.
The frequency has (as expected) a influence on the output voltage in the mV range.
Interestingly the frequency also has a influence on the T.C. of the whole cirquit.
The sweet spot (at least for my sample of ADG419 seems to be between 9.6 and 19.2 kHz (more near the upper frequency)).
Of course this may also depend on if you use 3.3V or 5V input levels.

In my simulations (and in real life with the ADG1404), my major issue has been charge injection into the "common" terminal of the switch, causing the common node's voltage to swing widely for a few hundred ns. I figure that this is due to charge injection coupled with the "slow" switching time. The Vishay DG419LE may give somewhat better performance as it is much much faster (it even has a few body snatchers in it, see datasheet.). The Renesas ISL43210A also looks good. The V_IN,H of both of these are a bit high (3.0V), but I think it's workable. I think the turn_on/turn_off times are why the voltage is so frequency (and temperature) dependent.

This transient issue may also drive me to use a lower switching frequency (with a larger number of channels.) Another option would be to reduce the value of the resistors at the switch's common.

- Reducing the 100R resistor. Same hope as above since I thougt that the 100R resistor together with the 100K build a voltage divider with the influence lower for lower 100R resistor.
  But obviously the ripple is only a thing between discharging C4 over the 100K Resistor R1.

If the 100R is reduced, perhaps we would run into stability issues due to the large capacitors to ground?


Fine to see a different approach with different ideas. How much current is drawn by your CPU?
I have to admit that I do not understand how your PWM works. (also not familiar with the STM CPU).
Do you have something like a timing diagram?
When combining 2 different PWMs I would have expected a e.g. 100:1 ratio (coarse/fine) between the mixing resistors.
How many bits are used from the PWM?


The STM32F103 I used is about 15 mA. The prototype only has a single channel (unlike what I'm proposing). My final design will use the STM32G071 which is much lower power, probably ~3 mA. One can actually run the timer clocks faster than the core CPU which saves power, too. The PWM has shadow registers which apply at the start of each cycle, so one has the entire clock cycle of time to calculate and update the HW registers and there is no edge case where you have to apply the PWM settings after the pulse. Each timer has a few PWM generators, so it could easily support 4 or 8 channels.

The disadvantage of mixing resistors is that the resistor ratio must remain stable. In my proposal, we don't get additional resolution, but the resistor ratio can drift significantly.

I am attaching a LTSPICE simulation of the dual-PWM scheme. Notice that large spikes on most of the voltage nodes. Not good. As I mentioned, I think they are a combination of charge injection and slow switching times.

ADDENDUM: The MCP6V51 looks reasonable, but it has a larger max offset voltage (15uv vs the 4 uV of the LTC2057), and double the offset drift. Other than that, the parts look very similar, even sharing the same chopper frequency (100 kHz).
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 13, 2019, 10:58:50 pm
Add 10 more phases and get russian V1-18  :-DD

https://www.qrz.ru/schemes/download/4445 (https://www.qrz.ru/schemes/download/4445)
Title: Re: LM399 based 10 V reference
Post by: bsw_m on February 14, 2019, 12:39:51 am
With the only difference that in V1-18, 12-phase voltage pulse divider (as originally named PWM DAC in documentation) used for divide ref output (about 12.92V) to the output range 0-12v with steps in 1uV.  And this is not used in feedback, but what's the difference :)
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 14, 2019, 09:23:55 am
There is around 800uVpp (@10KHz pwm) ripple at the output of the ADG switch (in my simulation above).

PS: Getting a little bit different results when the V- of the DG switch is connected to -15V instead to the GND.

Hello,

you mean the node A2 in your simulation above?
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=648510 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=648510)

This ripple mainly depends on the 1uF capacitor together with the 100K at the common of the switch (R4/C4) versus PWM frequency.
Thats why I asked branadic to increase the cap to 10uF in our test-board.

of course you will have to filter the remaining ripple anyway to reach something which is below the noise of a LM399 or a LTZ1000.

If the 100R is reduced, perhaps we would run into stability issues due to the large capacitors to ground?

Depends on the Op-Amp which is used. For the LTC2057 I have tested that 22R are stable up to 10uF foil capacitor if there is a large enough (10-100nF) feedback capacitor (here the 1uF directly at the LTC2057).

I am attaching a LTSPICE simulation of the dual-PWM scheme. Notice that large spikes on most of the voltage nodes. Not good. As I mentioned, I think they are a combination of charge injection and slow switching times.

Thanks for that.
Now I understand what you want to do.
So you are reducing the PWM amplitude at the switch from 10 to 5V by interleaving the PWM signals.
But I think you will have to take care that the ratio of R1a+R1b is stable over temperature and time.
(besides the R,on in high/low state of the switches).

For the spikes: I am still not shure if they are real or a simulation artefact.
I am not shure wether the ADG413 model supports break before make feature.

A STM32G071 with 3 mA would also be in my target spec.
(although a SO-8 package is more easy to solder).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 14, 2019, 10:24:11 am
Hello,

attached diagram from measurements of VOut (@25 deg C) and
resulting average (linear) T.C. of my ADG419 switch over PWM frequency. (without R15 tap).

so around + 40.5 uV/kHz near linear change in Vout
and  +12.7 ppb/khz change in T.C.

so interpolating: Zero T.C. somewhere near 18.5 kHz

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: pigrew on February 14, 2019, 03:43:34 pm
attached diagram from measurements of VOut (@25 deg C) and
resulting average (linear) T.C. of my ADG419 switch over PWM frequency. (without R15 tap).

so around + 40.5 uV/kHz near linear change in Vout
and  +12.7 ppb/khz change in T.C.

so interpolating: Zero T.C. somewhere near 18.5 kHz
Andreas

Interesting, I hadn't considered using the frequency as the TC-adjustment knob.

I just got some parts in the mail today. I'll compare my existing switch (ADG1404) with the DG419LE to see if I see less variation of DCV with switching frequency, on the breadboard.

Thanks!
Title: Re: LM399 based 10 V reference
Post by: iMo on February 14, 2019, 05:37:43 pm
There is around 800uVpp (@10KHz pwm) ripple at the output of the ADG switch (in my simulation above).
..you mean the node A2 in your simulation above?
This ripple mainly depends on the 1uF capacitor together with the 100K at the common of the switch (R4/C4) versus PWM frequency.
Thats why I asked branadic to increase the cap to 10uF in our test-board.
I think a better solution would be a good higher order low pass filter. Even the 10Hz 2nd order one in my schematics should filter the 10kHz+ perfectly out.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 14, 2019, 05:44:21 pm
Add 10 more phases and get russian V1-18  :-DD

https://www.qrz.ru/schemes/download/4445 (https://www.qrz.ru/schemes/download/4445)
Shots made by Max Otto von Stierlitz :)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 14, 2019, 05:55:41 pm
The filter at the first OP is already 3 rd order and there is another 1st order filter before the 2nd OP, that could be changes to 2nd order. For the filtering the 10 kHz or so main PWM frequency should ideally not be a problem anymore. However in real life there may be limited gain / output impedance of the first OP that can limit the effectiveness of the real filter.  In a similar way the 2 nd order Sallen-key filter is in theory better than just a passive 2nd order low pass with buffer, but with real OPs there may be a little higher frequency signal loading the OPs output and this way bypassing the filter. So not sure if the Sallen key circuit is an improvement in reality.

Due to the modulation, there are also much lower frequencies present, though at a lower amplitude. These could also be a problem, not that visible in the simple simulations.

With the 2 PWM stages it depends on the settings, whether the ration of the two 150 K resistors really matters. If the PWM settings are the same, there would be essentially the same voltage from both sides and than taking the average. Different resistors would not really matter than.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 14, 2019, 07:38:14 pm
I still see 700uVpp ripple (vref10 node) in the pigrew's simulation.
How should your dual PWM work? What is the trick?

FYI - the best PWM equipped STM32 is the STM32F334 series. It includes 217ps resolution timers.

For example you can generate a 100kHz PWM with ~15.5bit resolution:

144Mx32/100k = 46080  (100% duty)
46080*0.7 = 32256  (70% duty)
Title: Re: LM399 based 10 V reference
Post by: pigrew on February 14, 2019, 11:05:09 pm
I made some oscilloscope captures of the ADP1404, ISL43210A, and DG636EEQ. My below plots used Vdd=15V, Vpulse=3.2V, S1=9V 10V, S2=0V. Blue in input, Yellow is output.

The ADP1404 is clearly the worst, with quite visible charge injection during the dead time (60us) and very slow (150ns). The delay of the ADG1404 is a strong function of Vdd, with the "best" response with Vdd=20V.

The other switches were much faster (~15ns delay, <10ns rise/fall time), and have minimal charge injection. The DG633EEQ looks best (less ringing), but that could be just because the switch is too fast for my probing techniques.  |O

For those of you using the ADG419, I'd be interested if you could measure the voltage (versus frequency, constant duty cycle) with the ADG419 vs. the Vishay DG419LE. I've not tried the DG419LE (forgot to order it), but I think it'd be an improvement (unless the supply current goes up too much?).

EDIT: Added photo of breadboard.

-Nathan
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 15, 2019, 04:35:50 pm
Modeled a little  ;)

Many thanks to Kleinstein for invaluable help.  :-+
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 15, 2019, 04:37:50 pm
Part 2
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 15, 2019, 04:39:03 pm
Part 3
Title: Re: LM399 based 10 V reference
Post by: serg-el on February 15, 2019, 04:40:08 pm
Part 4
Title: Re: LM399 based 10 V reference
Post by: pigrew on February 15, 2019, 11:43:42 pm
I still see 700uVpp ripple (vref10 node) in the pigrew's simulation.
How should your dual PWM work? What is the trick?

FYI - the best PWM equipped STM32 is the STM32F334 series. It includes 217ps resolution timers.

For example you can generate a 100kHz PWM with ~15.5bit resolution:

144Mx32/100k = 46080  (100% duty)
46080*0.7 = 32256  (70% duty)

You are absolutely correct, my simulation increases the ripple. Darn.  :palm:

I've realized that the R4 100 ohm resistor is the problem. The switches are loading the output C4 capacitor, causing extra ripple. So, either that capacitor value needs increase, the series R4 needs a decrease, or the R1 values need to be larger.

I'm running into stability issues with reducing R4. 75 ohm seems OK.

Increasing R1 makes the initial transient longer, but seems OK. It may also help somewhat with charge injection.

Increasing C4 to 15u or 22u, along with increasing R1 to 200k seems to be a winner. The ripple is so small I can't zoom in that far in LTSpice. I think I need to build it and measure it at this point. However, the startup time is now a good 15 seconds for stabilization.

But, there are HF spikes on vx,x1,x2,x3 at the pulse edges. Adding a C5 from the switch drain to ground helps a little bit.

I'm still undecided between using the passive LPF and a Sallen-Key on the output. And now I've found Linear Tech Design note 9 (https://www.analog.com/media/en/reference-design-documentation/design-notes/dn009f.pdf)... But it also adds to the power requirements (another 6 mA).

Try the attached simulation? I'm sure it's not perfect, but it seems close. I don't know if the 2nd switch is a good idea or not. I can't see much of a difference in simulation. I added a  (commented out) parameter step in the simulation to change the PWM phase. Having the pulses back-to-back seems optimal. However, it seems good enough without the 2nd switch, and there would be fewer of the big spikes to add offset, as I expect the spikes to be temperature/frequency dependent.

Do we have a target spec for the turn-on transient time?



The STM32F334 HR timer does look interesting, but using 10 mA makes it less than ideal. I looked around and I think that the MSP430F51 has a similar type of circuit running on a dedicated high frequency oscillator, but I _think_ has a lower power footprint. I just ordered a few free samples, and I'll see if I can get a HR PWM going (it'll take a good week for them to be delivered, so no results on that sooner. The analog stuff is more interesting, anyway.).
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 16, 2019, 12:01:08 pm
For those of you using the ADG419, I'd be interested if you could measure the voltage (versus frequency, constant duty cycle) with the ADG419 vs. the Vishay DG419LE.

which voltage? 10V output voltage?
versus PWM frequency?

for the ADG419 I recorded output voltage over PWM frequency above:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2200098/#msg2200098 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2200098/#msg2200098)

The DG419 seems to be better in R,On but worse in charge injection and leakage currents.
But I can order some for comparison.

Modeled a little  ;)

And what are your conclusions on this?


I've realized that the R4 100 ohm resistor is the problem.
Not only, also the output impedance of the OP-Amp ... (so reducing the 100R gives no proportional effect)
The real problem is the C4 * R1 product.

I for my part want to set C4 = 10uF which gives around 80uVpp ripple.

The ripple is so small I can't zoom in that far in LTSpice.
Try V(out) - 10 as formula for the output.

But, there are HF spikes on vx,x1,x2,x3 at the pulse edges.
Fortunately they are filtered out easyly by the final low pass filter.
I first thought they are artefacts from the simulation.
But I also have them on my scope. (even larger).
If I limit the bandwidth of the scope to 10 MHz (mainly to reduce the noise) the spikes are similar as in simulation.

Do we have a target spec for the turn-on transient time?

I for my part: should be below 10-20 minutes. As the zener also will need some minutes to warm up.
This cirquit is not intended to be used as calibrator.

with best regards

Andreas

edit: attached screen shots of spikes
Title: Re: LM399 based 10 V reference
Post by: iMo on February 16, 2019, 12:31:54 pm
edit: attached screen shots of spikes
That exactly is the 700-800uVpp ripple I've been referring to (with 1uF cap)..
And you have to filter it out with a quality low-pass filter, to a level which is acceptable by the metrology folk, say, 7nVpp :).
The second order low-pass (ie the 10Hz one in my simulation) has 40dB/decade roll-off rate, thus the 10kHz+ will be down by 120dB+ (sure, the chopper adds something to it, so it will be not such great in reality).

PS: the STM32F334 HiRes PWM does not use ~GHz clock, but it works such it has got a 32 taps delay line (temp and voltage compensated) which together with 144MHz PLL clock can increase the PWM resolution as described above.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 16, 2019, 07:37:03 pm
And you have to filter it out with a quality low-pass filter, to a level which is acceptable by the metrology folk, say, 7nVpp :).

Hello,

I for my part will be satisfied with a noise level factor 3 below the LTZ1000 so 0.4 uVpp (= 400 nVpp) would be sufficient.
From the last change with 80 uVpp that is a factor of 200 which should be obtainable with a passive low pass filter of 1uF instead of the 10 nF at the moment.

By the way:
by increasing C4 to 10 uF the remaining ripple in simulation on the non R15 tap (VOut2) is lower as on the R15 tap output (VOut3).
And also the ripple on Vout2 has less high frequency parts.
So for me the filter design of the output stage will be as follows:

PWM-Frequency 9600 Hz (because of current consumption and fair T.C. of -0.1 ppm/K which can be compensated in software).
C4 = 10uF
C6 >= 1uF (alternatively a 2nd order passive filter with smaller caps).
no R15 tap (less high frequency parts in output)

I am still looking for a method how to check the influence of C1-C3 and if they can be reduced without too large influence on the low frequency dithering.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: RandallMcRee on February 16, 2019, 11:52:51 pm
I have been experimenting with low-pass filters to reduce noise. It is not as easy as it might seem. First, chopper amplifiers all fail to give expected results. Empirically it seems that the filter "undoes" the chopper's attempts to lower noise and a significant amount of noise appears at the output, defeating the filter.

A low-bias current precision opamp does work, however. In particular the ADA4625-1 is exemplary--it has the triple threat: low bias current, low input current noise and low flicker noise which are all necessary for best results. (Thanks to chuckb for recommending this part).

Attached is a circuit. The AD filter wizard gives values that are high so I scaled them down and built the circuit on the left (C1=46uF). The settling time is still only 21 seconds to 0.1 ppm according to the simulation. I think C1=25uF would also work but I went for overkill. Probably NPO ceramics are worth a try, although I used very large MKP types.

The results speak for themselves: 0.32uVp-p noise is reduced to ~ .17uVp-p (right at my measurement noise floor).

Attached are 1second/div and 5second/div scope graphs showing filter noise output (last two attachments).
First shows input to filter, its at 10volts dc.

Should be fine for PWM purposes.
Edited to get attachments into post!
Title: Re: LM399 based 10 V reference
Post by: iMo on February 17, 2019, 11:02:12 am
..A low-bias current precision opamp does work, however. In particular the ADA4625-1 is exemplary--it has the triple threat: low bias current, low input current noise and low flicker noise which are all necessary for best results..
The guys here are pretty keen on the low voltage offset tempcos, +/-1.2 μV/degC is on the higher side, however.
Title: Re: LM399 based 10 V reference
Post by: RandallMcRee on February 17, 2019, 03:47:33 pm
..A low-bias current precision opamp does work, however. In particular the ADA4625-1 is exemplary--it has the triple threat: low bias current, low input current noise and low flicker noise which are all necessary for best results..
The guys here are pretty keen on the low voltage offset tempcos, +/-1.2 μV/degC is on the higher side, however.

Its true.  Here's a thought: the LM399 has its own heater and the ADA4625 an exposed thermal pad....just need to mechanically couple the two and problem solved. Unless the ADA4625 has too much bias current at the LM399 setpoint? Someone would need to try it. I'll give it a go if someone sends me a 399...

Like I said all of the choppers I have tried were unsuitable.  For example, same circuit with the OPA189 shows 0.6uV (twice the input).
Randall
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 17, 2019, 04:22:22 pm
Relative to the 7 v of the LM399 the 1.2µV/K are something like 0.15 ppm/K this is not that much.

The LM399 runs rather hot - so it's not such a good idea to use it's temperature. The insulating cap around also makes it difficult.

There are a few good non chopper OPs that may be an alternative. Besides the ADA4625, the ADA4077 and OPA140 would be other candidates. There is no real need to be much better than the reference. With BJT based OPs it might still be an option to reduce the resistors a little and use slightly larger caps.  2.2 µF MKS caps are still affordable and not that large.

With a suitable circuit chopper stabilized OPs should still work.
Title: Re: LM399 based 10 V reference
Post by: RandallMcRee on February 17, 2019, 07:19:27 pm
Here is the OPA140 and the ADA4522-1 in that same Sallen-Key circuit. C1=47uf, C2=4.7uF, R1=15K, R2=470K. R1 current noise will dominate as you make C1 smaller.

OPA140: 0.22uV
ADA4522-1: 0.384uV

Perhaps the 4522-1 meets your goals.


Title: Re: LM399 based 10 V reference
Post by: pigrew on February 17, 2019, 11:40:18 pm
I'm getting closer to finalizing a schematic. So, here is my draft 2, with the following features:


Thanks in advance for any critiques.

I intend this board to be used from an 17 to 24 V power supply. The board is intended to be more of a experimentation platform than a final product, and I expect a second revision once I get good results with this one. I have the footprints complete for the selected components but have not done any routing/placement yet.

Depending on the complexity (via count), I may mill a board here, or perhaps order a 2-layer board from China. In the first case, I'll get measurements in a week or so. In the second case, it'll be about three weeks.

-Nathan


Planned changes:
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 18, 2019, 08:02:06 am
Hello,

I would use 100 nF across the zener of U5. (removing EMI influences eg. from the chopper).
At least the ADG419 showed a large power supply voltage dependancy. I do not know if I would use a capacitance multiplier.
C11 and C25: I would go up to 1uF here (to have the space on the PCB).

After my experiences the LM399 has much less temperature dependancy than the 1 ppm/K of the data sheet near room temperature.
So I would place the thermometer sensor near the analog switch.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on February 18, 2019, 10:06:39 am
Where your values for the low-pass come from? I did 10Hz and 1Hz low pass with flat pass-band with 16k+16k and 1u+1u (10u+10u for 1Hz).
Your values (0.1Hz lowpass) seem to have "overshoots" and "oscillation" (?) - at least the tool reports it.
The simulation shows no ripple in the pass band, though.
Title: Re: LM399 based 10 V reference
Post by: RandallMcRee on February 18, 2019, 05:38:47 pm
Hey imo,
The original filter came from Analog's filter designer, 2nd order bessel, fc=0.1Hz with scaling to the noise min (slider all the way to the left).
I then scaled it a bit more, another factor of 4.7.

The http://sim.okawa-denshi.jp (http://sim.okawa-denshi.jp) tool always seems to show an oscillation frequency for many of these low-pass filters, but Filter Wizard does not. The overshoots are pretty clear: if you ask for step response text data and click on the file you will see that these so-called overshoots are at the 1ppm level, so not a concern in reality. Any transition from 0.9999... to 1.0000... is labeled an overshoot in that tool, I guess.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 18, 2019, 06:07:55 pm
The filter resistors are kind of a compromise between needing large caps and problems from large resistors like noise and sensitivity to bias / leakage. For OPs like TLC2057 the resistors should ideally not go much above above 50 K. So the 3 x 100 K in the initial filter circuit is already quite high, especially for the originally proposed BJT based OP.  The first resistor additionally is important for the effect of the switch resistance. So going much below 100 K is not a good option.

I personally would keep about 100 K for the 1 st resistor behind the DG419 and more tend towards less (e.g. 47 K) for the next 2 resistors. As a compensation the capacitors would need to be a little larger - usually something like 2.2 µF is still reasonable as MKS type.  In my simulations in AC mode there is a slight improvement when the first cap goes towards GND instead of the OPs output - this keeps away most of the AC current load to the OP.

The final filter could use similar caps - possibly as a 3 rd order to filter out the fastest spikes before the active filter with a small cap first.
It could be that just a 1st order filter is good enough, especially with the STM32 based solution with less low frequency component.

I don't think one would need PP-caps as I don't see a need for fast settling in this circuit. For the capacitor at the switch one might even consider an electrolytic cap if one has time for settling of DA. With only 100 Ohms (and possibly a little less) leakage is not that critical.

The choice of OPs is not that critical, as the LM399 still limits the performance: it has quite some low frequency noise and still often more drift than conventional precision OPs. So neither a 0.4 µV_pp .1-10 Hz noise from the chopper or 1 µV/K from a normal precision OP would be that bad.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 22, 2019, 10:27:36 am
I'm getting closer to finalizing a schematic. So, here is my draft 2, with the following features:
  • Ability to use my two-phase PWM, or a single phase, ganging together the two SPDT to get a lower on-resistance.
  • STM32F334-based PWM, for 250 ps (or so) resolution.
Would the switches work fine at say 70-100kHz PWM?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 22, 2019, 10:46:53 am
The PWM frequency is a compromise: at high frequency there is more effect from switching transients like charge injection or nonlinearity during transients. This extra part can depend parameters like temperature, supply voltage and similar - so a tendency to get more drift. There is also more higher frequency background creeping around the filter.

A lower frequency gives more ripple or requires a better filter.

Even when using just simple PWM, one may not need 2 switches in parallel. The R_on effects the linearity, but with little direct effect on the stability. 2 switches add extra charge injection and more load - so it may get worse with 2 switches in parallel.

With the higher PWM resolution, there would be less low frequency part anyway. So I don't think it would need such a high PWM frequency - I would more like expect the best frequency to be lower.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 22, 2019, 11:24:31 am
Another point to consider - the HiRes PWM mode of the F334 requires the PLL to be involved. The 217ps and alike resolutions work with 144MHz timer clock only, afaik (doublecheck). With some stm32s (ie the BluePill's F103) the PLL slowly drifts, something like 60Hz @72MHz in a sawtooth manner (not related to sawtooth with GPS). No idea whether that is the case with F334, however. Also the jitter with PLL has to be considered somehow. May be it is not too relevant with 399, but do consider when moving to LTZ1000 :)

..With the higher PWM resolution, there would be less low frequency part anyway. So I don't think it would need such a high PWM frequency - I would more like expect the best frequency to be lower.
Considering the F334's HiRes options the useful PWM freqs are the 8.8kHz, 17.6kHz or 35.1kHz, all w/ 16bits resolution.. 8)
Title: Re: LM399 based 10 V reference
Post by: pigrew on February 22, 2019, 03:43:14 pm
Another point to consider - the HiRes PWM mode of the F334 requires the PLL to be involved. The 217ps and alike resolutions work with 144MHz timer clock only, afaik (doublecheck). With some stm32s (ie the BluePill's F103) the PLL slowly drifts, something like 60Hz @72MHz in a sawtooth manner (not related to sawtooth with GPS). No idea whether that is the case with F334, however. Also the jitter with PLL has to be considered somehow. May be it is not too relevant with 399, but do consider when moving to LTZ1000 :)

..With the higher PWM resolution, there would be less low frequency part anyway. So I don't think it would need such a high PWM frequency - I would more like expect the best frequency to be lower.
Considering the F334's HiRes options the useful PWM freqs are the 8.8kHz, 17.6kHz or 35.1kHz, all w/ 16bits resolution.. 8)

I just ordered some F334s, but have not gotten them yet. I can do some measurements to see how stable the frequency is.


 I hope that the output voltage's dependence is more reduced with the switch I'm using, so even if the frequency drifts, it won't have a big impact on the output. With the ADG419, there were large spikes at the switching time, along with somewhat slow switching, which I believe make the output much more dependent on the frequency (the number of spikes per second adds an offset to the output since the negative spikes have different area than the positive spikes).

Because of the transients, lower frequency should be more linear. But, the lower the frequency, the larger the filter has to be.

The F334 does uses some sort of PLL or FLL, but I don't know how it'll behave:

Quote
HRTIM1 timer is made of a digital kernel clocked at 144 MHz followed by delay lines. Delay
lines with closed loop control guarantee a 217 ps resolution whatever the voltage,
temperature or chip-to-chip manufacturing process deviation. (from the datasheet).

I've ordered components to put together my boards. I split the design into a digital board and an analog board so that different microcontrollers can be compared using a single digital board. I'll make the board on a PCB mill. It uses the back as a ground plane with nearly all the routing on the top. It incorporates most of the feedback I've received. The SMD components are on the PCB's bottom while the THT components are on the other side. The LM399 routing is a bit funky, to be fixed in rev2.

I'm attaching an updated schematic and 3D rendition of the analog board.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 22, 2019, 05:37:25 pm
Would the switches work fine at say 70-100kHz PWM?

My ADG419: no
see T.C. over PWM frequency:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2200098/#msg2200098 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2200098/#msg2200098)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on February 22, 2019, 06:25:15 pm
Would the switches work fine at say 70-100kHz PWM?
My ADG419: no
see T.C. over PWM frequency:
How did you measure that? Did you set the PWM duty to 68% and changed the PWM frequency from 2kHz to 100kHz?
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 22, 2019, 08:34:17 pm
Hello,

I did 5 T.C. measurements (each lasting one complete day)
every measurement with a different XTAL for the ATTINY85.
(2.45 MHz, 4.91 MHz, 10 MHz, 20 MHz and one with 8:1 prescaler)

The voltage for the overview is measured at 25 deg C.
The T.C. is the linear approximation (slope of the green curves) over ~15 to 40 deg C.
PWM is kept constant. at the precalculated value which should give exactly 10V at the output for the given LM399#3.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: iMo on February 23, 2019, 04:06:50 pm
Hello,

I did 5 T.C. measurements (each lasting one complete day)
every measurement with a different XTAL for the ATTINY85.
(2.45 MHz, 4.91 MHz, 10 MHz, 20 MHz and one with 8:1 prescaler)

The voltage for the overview is measured at 25 deg C.
The T.C. is the linear approximation (slope of the green curves) over ~15 to 40 deg C.
PWM is kept constant. at the precalculated value which should give exactly 10V at the output for the given LM399#3.

with best regards

Andreas
OK, you kept "PWM duty" constant (hard-coded in your fw) and you changed the "PWM frequency" (also hard-coded in your fw) by replacing the 4 Xtals, and the TC is the lin interpolation between the 4+1 Xtal measurements.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 23, 2019, 04:13:35 pm
Hello,

exactly. The "dots" are the measurement points. The rest is imagination  8)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on February 26, 2019, 03:15:39 pm
Preparing to try NCC-1701 Vishay MPM-divider...
It's not exactly 10V output, but cheep and i hope is pretty good stable for my 6.5 DMMs.
FYI: Maxim does the similar:

1:1-1:30 30k end to end divider with 2ppm/degC resistance-ratio tempco (for <3.5 ratios):

https://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/MAX5491.html (https://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/MAX5491.html)

PS:

1:1-1:10 10k divider, up to 1.5ppm/degC ratio tempco
https://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/MAX5492.html (https://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/MAX5492.html)

1:1-1:100 100k divider, up to 1ppm/degC ratio tempco
https://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/MAX5490.html (https://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/MAX5490.html)

From DS:
Quote
..and custom ratios are also available upon request..

And the Vishay https://www.vishay.com/docs/60001/mpm.pdf (https://www.vishay.com/docs/60001/mpm.pdf) to compare https://www.maximintegrated.com/en/alternatives.cfm/part/SOT23/pk/437 (https://www.maximintegrated.com/en/alternatives.cfm/part/SOT23/pk/437)
Title: Re: LM399 based 10 V reference
Post by: branadic on February 26, 2019, 04:34:35 pm
Hi Andreas,

I didn't forget the T.C. measurement, but was of the bench for a while because of illness.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 03, 2019, 08:34:52 pm
Hello,

I made some LTSPICE-simulations for the ION board of Branadic.
Reason was to see if the integrator filter around the first OP-Amp can be reduced in size after we found out that C4 should be rather 10uF instead of 1uF.

Starting from 1uF, 1.5uF and 1uF for C1,C2,C3 and reducing to 100nF for all
I also tried to simulate the influence of the low frequency part from the dithering.
First I tried to put a sine of 1kHz and 40 mV Amplitude (10V/256) in series to the input (on top of the LM399 voltage).
But I soon decided to put the 40 mV sine in series to the ADG419 output.
To get all details I had to put the maximum simulation distance to 2ns. (So each calculation several hours).
Simulations were all done with a 10uF and 10nF output filter so that changes in the integrator can be seen more easily.

Here some results:
If all 3 capacitors are used there seems to be no difference if 1uF, 10uF or 100nF are populated.
Except for Startup time after power up which scales exactly to the 1uF or 100 nF value.
I will have to do further simulations with lower Sine frequency and reduced capacitors to see where the limits are.
Especially as the full period time is several seconds for the dithering.

Although there is no difference on the VOut2 node it makes a large difference on the VUC1 node wether C1 is connected to Gnd around (100uV high frequency spikes) or the UC1 node (8 mV high frequency spikes).
As I also can see high frequency spikes on my scope I take this serious.
So the suggestion of Kleinstein for C1 to GND should ease the filtering.

Interestingly I can omit C2 with only minor changes on the output signal or UC1.

But for now I only would put C1 against GND instead of UC1.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 03, 2019, 09:47:11 pm
The resistance of R1+R2+R3 = 300 K is rather high, even for the LTZ2057.  So if the filter is not that critical, I would prefer a reduced resistance instead of much smaller caps. R1 may have to stay at a relatively high value to keep the R_on effect and the current flowing back to C4 small. However I see no big problem reducing R2 and R3, maybe to something like 10-20 K. These 2 are just filtering a low amplitude signal.

AFAIK the lowest frequency from the modulation should not be so super low, more like the same as if one would use plain PWM for the same resolution. So with some 22 Bit resolution and 4 MHz clock the period should be near 1 Hz. At least this happens with 1st. order digital filtering.
Higher order digital filtering is supposed to reduce the amplitude quite a lot, so that not much analog filtering is needed anymore, even of there is a small very low frequency part.
Even if the code allows for 24 Bit or even 32 Bit resolution (to simply the code and use the CPU carry), there should be not lower frequency part if the last bits in the set value are 0.
Title: Re: LM399 based 10 V reference
Post by: pigrew on March 05, 2019, 02:09:24 pm
Thanks for the simulation results.


I just ordered my "Analog" board, along with STM32F334 and MSP430F5172 digital boards. They should arrive towards the end of next week. I'm leaning more towards the MSP430 timer_D, as it is generated internally with a ring oscillator, apparently with a length directly controllable by registers (so no wacky PLL glitches). Though it will drift with temperature, I hope it will be fairly stable.

I'm happy to share my layout files upon request (PM me). The boards were cheap enough (2-layer, 1mm thick) that it doesn't make sense to mail spares ($10 for everything, including shipping from HK).
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 08, 2019, 08:15:47 pm
Hello,

did some further simulations (now with my "final" population).
The 1uF, 1.5uF + 1uF (Gnd) for the integrator part.
10uF (foil) to reduce the ripple + spikes.
And a further 4K7 + 1uF low pass.

Simulated several sine frequencies as "low frequency modulation"
At low frequencies (50 Hz and below to 0.1Hz) the filter could be better
But on the other side the 40mV amplitude is shurely more than that
what we will get as change of +/-4 counts in the existing 10kHz triangle signal of some 10 uVpp.

So some results and the overview table (which was wrong regarding the 10uF in the previous post).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on March 09, 2019, 12:49:19 pm
Is that an ".io" or the ".ic" (the correct one) LTspice directive in your simulation??
Title: Re: LM399 based 10 V reference
Post by: iMo on March 09, 2019, 02:20:12 pm
I've been experimenting with the decreasing the opamp's output current and boosting the output current with 399 classic circuit. Is there any experience with that transistor at the output? More noise or drift?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 09, 2019, 02:32:34 pm
The OP in the filter does not see much DC load. So just adding the transistor would not work . With some extra DC load the extra transistor would not change much. If at all the extra transistor could allow to use a slightly smaller resistor for R4 - however it looks like the capacitor is more important than the resistor.

For the output stage the extra transistor could be useful to reduce heating of the OP under heavy load. It would not change much with noise and drift, as the transistor is in the loop.
Title: Re: LM399 based 10 V reference
Post by: pigrew on March 09, 2019, 02:33:33 pm
I also thought about doing that, but it seemed unnecessary. The BJT would add a bit of noise. Another technique would be a resistor between 15V and the op-amp output (to provide the current), 1mA and 7V 3V would make it a 7k 3k resistor.

ADDENDUM: It's also possible that the pull-up would reduce the op-amp's switching noise, since the switched caps would be injecting less current. (I also changed the 7V figure to 3V)
Title: Re: LM399 based 10 V reference
Post by: iMo on March 09, 2019, 02:51:24 pm
..For the output stage the extra transistor could be useful to reduce heating of the OP under heavy load. It would not change much with noise and drift, as the transistor is in the loop.
Yes, we talking here the output stage only (399 classic). I don't see any diff against the naked opamp's output in tempco drift. No idea about the added noise, imho it should be by orders lower than the 399. Except the way higher current capability the boosted opamp's output gets some emi/esd protection as well..
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 09, 2019, 03:50:33 pm
There is no extra noise expected. However there is also essentially not added ESD / EMI protection as the he feedback path is still there, though one could have a small series resistor here for some protection (but than added noise from the resistor and possibly drift). 
Title: Re: LM399 based 10 V reference
Post by: iMo on March 09, 2019, 03:59:28 pm
The fb path is resistive - 399 cathode resistor and divider resistor.
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 09, 2019, 06:52:10 pm
Is that an ".io" or the ".ic" (the correct one) LTspice directive in your simulation??

Attached the latest simulation file
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 10, 2019, 10:19:05 pm
Hello,

I did a 2nd sample of branadics ION PCB.

This time I wanted to be more flexible with the processor so it is built on a Sub-Board connected to the programming connector.
The capacitors are according to the latest simulation except for C3 which is connected to UC1 instead of GND.
Since I ran out of LT1763-5 I simply used a LP2950 for the 5V regulator in TO-92 (SMD mounted)
The 4K7/1uF filter at the output is built with a 0805 1uF X7R capacitor. (so no foil at the moment).
The 10uF WIMA is mounted on the "2nd floor" above some other components.

On the sub-board I use a PIC12F1840 together with a 10 MHz XTAL to get ~10kHz PWM.
Power consumption is 3.7mA for the whole cirquit compared to 4.1 mA of the ATTINY85 with 2.45 MHz and sleep mode.
The sleep mode on the PIC cannot be used since it would also stop the PWM output.

With the new filter population the 10kHz PWM clock is no longer visible (below the noise floor)
when being measured with a LNA 10Hz - 100kHz according to AN83 (LT).

The diagram to be compared with (with the previous filter) is described here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2080246/#msg2080246 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2080246/#msg2080246)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: matches on March 15, 2019, 08:13:43 pm
Hi.

Earlier in this thread the use of the AD107 was recommended when building a self biased version.
HP did so for the 34401A, I think Keithley for the 2000, 2015, ...

LT suggests the LT1001.

Can someone help me understand why to prefer the AD107 over the LT1001?

Regards
Title: Re: LM399 based 10 V reference
Post by: Andreas on March 15, 2019, 08:20:59 pm
Hello,

some T.C. measurements over PWM-frequency with the 2nd sample on branadics PCB.
The PIC-based solution shows this time a much larger T.C. at 9.765 as sample #1. (0.56 ppm/K)
Whereas voltage over frequency increases again.

see post:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2218482/#msg2218482 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2218482/#msg2218482)

And T.C. over frequency shows a negative slope instead of a positive slope.
"Zero T.C." would be at negative frequencies.

we have:  - 18.8 ppb/kHz
and :        +0.104 mV/kHz

now the big question:
- is the stray from ADG419 from device to device for the T.C. so large
- or is the different processor (PIC) the guilty. (e.g. different rise time at input)

-> we need much more samples/measurements

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on March 15, 2019, 08:53:43 pm
The higher PWM freq the higher temperature of the ADG switch, more stray capacitance leakage, more losses in the 10uF capacitor (a nonlinearity?).
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 15, 2019, 09:40:21 pm
I don't think the power used by the ADG419 is a significant factor.  Also the loss in the capacitors should be not such a big effect - to a large part dielectric absorption is still linear.

The effects I see that could contribute significant to the TC are the change in R_on of the DG419.  The R_on contributes a little to nonlinearity as current in both phases is different.  This effect should get larger if the 100 K resistor at the filter is reduced. However this effect should be largely independent of the frequency. So it's more like part of the TC extrapolated to zero PWM.

Another temperature dependent effect is from a change in the switching time of the AD633. Besides the internal switching time also the level for detecting the logic signal could change with temperature and this way, depending on the slew rate of the control signal effect the timing.  The ADG619 is relatively slow - a faster switching chip might be better.

I don't see a large effect related directly to the µC or the slightly larger stray capacity from the adapter board. I theory there could be an effect of the PWM signal coupling to the µCs clock circuit and this way leading to some phase shift there, that would effect the PWM accuracy. However this should be more like a transient effect and linear in the PWM frequency.

Capacitive coupling from the control signal to the filter input should be relative to the rather low switch resistance. It could still contribute somewhat, and since the switch resistance increased with time would also be temperature dependent. Assuming something like 1 pF, this could contribute something like 5 pC of charge and at a 40 Ohms switch resistance would be some 2 µV at 10 kHz PWM frequency. However a similar effect would happen for the other side and this would to a large part compensate. So only the R_on difference would be effective e.g. some 5 Ohms from 0 V to 10 V. So I don't see a significant contribution.

AFIAK the rise time of the AVR and PIC is not that different. In addition the rise time is way shorter than the DG419 switch speed.

So may guess would be more with the different ADG419 sample than the µC.
Title: Re: LM399 based 10 V reference
Post by: Grandchuck on March 18, 2019, 04:15:54 pm
Here is my fun box enclosed in Styrofoam and sealed with duct tape.  To get it to a useful state, I soldered the appropriate voltage selection pins together.  It is powered on 24/7 and has been for months.  It has settled into a useful state at 9.817495 volts and wanders no more than 2 ppm.  My home lab is rather constant for temp and humidity.
Title: Re: LM399 based 10 V reference
Post by: pigrew on March 18, 2019, 09:15:09 pm
Small updates on my PWM implementation:

I've received the boards I've ordered, and have built up a MSP430 MCU board, finding out I've made a few mistakes along the way (but nothing unfixable)....

The MCU can operate in two modes: free running at up to about 1 GHz, or XO-disciplined at up to about 400 MHz.

Running with a 8MHz XTAL, multiplied to 24 MHz, used to discipline a 384 MHz timer clock, outputting 10 kHz, uses about 2.7 mA @ 3.3V. stdev(period) ~ 18.3ns, p-p(period) ~ 120. Enabling "clock error accumulation" correction increases the jitter slightly, but probably should be left disabled in this application.

Free-running with a 256 MHz clock uses about 1.2 mA, though with a much higher temperature-coefficient. stdev(period) ~ 21ns, p-p(period) ~ 135.

These measurements were performed with a MSO-X 4154A oscilloscope, having it measure the period of a single output cycle.

I think my preference would be XO-disciplined because of its much lower TC. The higher current consumption can be mitigated by using a switching DC-DC converter (though not the one that I had originally chosen).

Next up is populating my analog board... I like the MSP430 enough that I won't bother with the STM32F334.

One interesting thing was that when setting the period, the lowest 3 bits are ignored in 8x clock mode, and the lowest 4 bits are ignored in 16x clock mode. The pulse width parameter is controllable with full 16-bit resolution.

(Some of the mistakes: The pins I used for TD0 have higher channel->channel skew than the other available pins. Also, TD0.0 can only output the full period, not PWM (unless you set the period to a power of 2 (2^12, 2^14, 2^16, etc). Next revision will use TD0.1 and TD0.2 as the outputs. The micro's FLL only multiplies, but doesn't divide, so it is not very adjustable if you use a high-frequency XTAL. Oh, and I mixed up the pins of the LDO regulator on the PCB.)

Once I get the whole thing running well, my plan is to integrate the boards back onto a single PCB.
Title: Re: LM399 based 10 V reference
Post by: pigrew on March 25, 2019, 01:25:10 am
I've developed firmware for the MSP430 to run the PWM using sigma-delta modulation, plus built up enough of the analog PCB to be able to output a somewhat stable 10V. Yay.

Source code is posted athttps://github.com/pigrew/msp399pwm (https://github.com/pigrew/msp399pwm).

In terms of performance, it (or the Keithley 2000) is reasonably stable (staying within 30 uV of 10V over a few minutes) and quite adjustable (32-bit resolution). The PWM adjustment ISR takes about 3.5us, allowing operation at over 50 kHz without glitches (faster in the future with further optimization). The circuits draws about 20 mA at 18V and 18C, most of which is the LM399's heater current. The MCU board uses a DC-DC converter while all of the analog circuitry is driven at 15 V with a LDO. I have a simple RC filter at the output (10u/4.7k), but will populate the output amplifier soon.

A plot of a quick frequency sweep is attached, showing a gain of 0.069 mV/kHz.
Title: Re: LM399 based 10 V reference
Post by: iMo on March 25, 2019, 12:03:45 pm
While curious about how your 32bit DAC works and looking at the code and SD DAC whitepaper you refer to, let me kindly ask you - does it mean you modulate (within a periodic ISR) a 16bit PWM duty with the sigma delta modulator's 1bit output?
Title: Re: LM399 based 10 V reference
Post by: pigrew on March 25, 2019, 01:33:53 pm
While curious about how your 32bit DAC works and looking at the code and SD DAC whitepaper you refer to, let me kindly ask you - does it mean you modulate (within a periodic ISR) a 16bit PWM duty with the sigma delta modulator's 1bit output?
Yes, that's exactly what I do.

At 10 kHz, the base PWM is only using 15.2 of the 16 bits, but I specify the resolution with 32-bit precision over the serial terminal.  It uses the left-over bits as input to the modulator
Title: Re: LM399 based 10 V reference
Post by: iMo on March 25, 2019, 05:53:58 pm
Ok, the LSB of the 15.2bit 10kHz PWM "duty" is resolved by the 16.8bit SD modulation, basically. What is the SD's isr() frequency (or period)? [I can see there 3.13us in the comment but it does not fit into my current understanding of the process..]
Title: Re: LM399 based 10 V reference
Post by: pigrew on March 25, 2019, 08:15:12 pm
Ok, the LSB of the 15.2bit 10kHz PWM "duty" is resolved by the 16.8bit SD modulation, basically. What is the SD's isr() frequency (or period)? [I can see there 3.13us in the comment but it does not fit into my current understanding of the process..]

The ISR is called once per PWM period and the ISR takes about 3.1us to complete. With a period of 100us, this leave plenty of time for other tasks to happen in the MCU.

The ISR is called at the timeout of Timer_D (which is the CCR0 value in up mode). At this point, the "CCR1" is latched into the "CL1" register and the ISR is triggered. The ISR takes about 3us to compute the next sigma-delta bit, which it'll use to update the CCR1 register. The interrupt latency plus the processing time (about 3us) must be shorter than the duty cycle in order to ensure the duty cycle is updated before the next period starts.

I also have the UART and systick using interrupts. While the TD0CCR0 has a quite high priority, I have not enabled interrupt nesting which would have allowed it to preempt the lower priority interrupts. So, I must have the sum of the the durations of the TD0CCR0 and UART ISRs to be shorter than the PWM period so that CCR1 can be updated in time for the new value to be latched when TD0 times out.

My UART ISR is very non-optimized currently such that some D-S computations will take too long if the PWM freq is set to >50kHz.

See the attached diagram.

EDIT: (If anyone wants to build a MSP430 PCB, let me know and I can post updated board files).
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 01, 2019, 02:33:04 pm
My firmware work continues on my LM399 implementation. Notable changes include:


Jitter continues to be an issue. The period of the 100us cycle has a P-P jitter of about 450ns. This seems to be caused by the Timer_D oscillator, itself. Fortunately, it seems to be random and does eventually average itself out.

I've also started thinking about what criteria to use for design optimization. My current plan is to focus on two measurements: voltage across temperature and the short-term standard deviation of DMM measurements. The tempco will be measured by a 6.5 digit meter, from about 15 C to 40 C, with the reference placed in an oven (while the DMM is outside the oven). The short-term DMM noise measurement will analyze 50 or 100 NPLC DMM samples (10V range, auto-zero on). I will try to minimize the standard deviation of these.

For the "noise" measurement, I'm getting standard deviations of about 5 uV on a 34401A with 50 samples with NPLC=100 (60Hz). Is this good or bad? I'm not sure. I'm guessing this is close to the noise floor of the instrument, so I may need to start using a second 10V reference, and measuring (10Vref vs MSP399PWM) differentially in order to get the DMM into a lower range.

Are there other measurements I should make?

Another idea I had was if it's a good idea to use the LM399 without its heater. The die temperature could be measured by looking at the diode forward-bias voltage between the heater and the reference circuit, and programming a correction table into the PWM generator. The advantage is a greatly reduced power consumption. Disadvantage is probably quite awful temperature coefficient. One bad thing would be that the two reference circuits voltages would no longer be equal. Why not just use a 2DW232? Because the Zener reference amplifier in the LM399 provides a lower Zener resistance, making the pull-up resistor less critical.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 01, 2019, 04:03:53 pm
A 5 µV noise sounds bad too me. However I don't know how much is normal for the LM399 on the board and inside the meter. One could check this by measuring the LM399 voltage directly - this would give a kind of background noise (multiply by 10/7 to get the noise at 10 V level).
450 ns jitter sounds like awful, even if it over time might average out. From a different circuit with an AVR I think the jitter should be well better than 1 ns, likely better than 100 ps.

The 2DW232 reference usually has a lower noise, but higher TC and an unknown long time stability. To get a good TC it would need something like a temperature stabilization. With a well adjusted current a relatively crude stabilization (e.g. +-0.5 K) could be sufficient.
The idea of the PWM scaling to get 10 V is more like getting a long time stable voltage with a low TC without worrying about resistor aging.
Because of the TC I would keep the heater on. With some extra insulation the power consumption is not that high. AFAIK the LM399 does not mind that much about getting turned on and off. Due to the high temperature there should be little hysteresis, as one is always coming from below.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 01, 2019, 08:13:11 pm
For the "noise" measurement, I'm getting standard deviations of about 5 uV on a 34401A with 50 samples with NPLC=100 (60Hz). Is this good or bad?

Another idea I had was if it's a good idea to use the LM399 without its heater.

Hello,

5 uV standard deviation is rather high (I have this as Peak/Peak value on a HP34401A).
It also depends if your measurement is made slightly above or slightly below 10.000000V
The resolution switches from 100nV to 1uV above 10V on the interface.

So I have 1.1uV standard deviation for the output of the ION (ATTINY version)
and 0.76uV standard deviation with a much higher resolution at 6.8V.
Around 2000 values each measured after warm up with 100 NPLC (50 Hz).

I personally would not use a LM399 without heater. I would use a AD587 10V reference when I want to save energy. (see the AD587LW project).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: e61_phil on April 01, 2019, 08:25:08 pm
If you connect the 34401A to a computer anyhow, I would use 10NPLC with 10 samples  instead of 100NPLC. The 34401A can't do more than 10NPLC and the math doesn't work correct if the meter does the averaging. And you don't loose resolution above 10V.

I also see far below 1ppm pp with my 34401As
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 01, 2019, 08:37:08 pm
If it is about getting the noise introduced by the PWM, one could measure the difference from the 7 V LM399 to the 10 V output. This would reduce the reference (both the external and DMM internal) noise by about a factor of 3.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 02, 2019, 05:53:19 am
Jitter continues to be an issue. The period of the 100us cycle has a P-P jitter of about 450ns.

Hello,

that was about the relation when I measured from rising edge to rising edge (which is modulated).
When measuring from falling to falling edge I have less than 1 ns (which is the scope resolution).
So typical I have 3-4 mHz standard deviation in 10 kHz. (shurely also limited by the scope resolution).

But it could also be that the PLL is not stable enough.
(Similarly to the R/C oscillator in ATTINY which made visible jumps by some counts on the 34401A)
That could be seen when switching off the modulation of the fractional part of the ION.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on April 02, 2019, 06:02:28 pm
His 100us PWM period is with 15.2bit resolution, afaik. That is 2.6ns LSB and thus the 1bit SD modulation jitter is 2.6ns.
When reading TI's forum I see there people complaining on the DCO FLL jitter which could be several % short term.
Most probably those 450ns come from there..
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 03, 2019, 04:33:45 am
Thanks for the suggestions. I did some more measurements and tweaks on the MSP430 and got it down to 100 ns p-p jitter.... still pretty bad.

I quickly ported my code to the STM32F334 (running with a 128MHz*8=1.024GHz clock) and the oscilloscope was giving figures on the order of 1 ns for the p-p jitter. Much better. The downside is that the HRTimer uses about 40 mA! This may still be within the power budget when using a switching converter. I'll do some more careful measurements in the coming days. I guess the question is if the faster (1GHz) timer with its higher clock jitter and power consumption is worth it compared to a slower (64 to 128 MHz) lower power, lower-jitter PWM generator. I'm becoming pessimistic about the usefulness of the MSP430.

Code is in the same  github repository (https://github.com/pigrew/msp399pwm).
Title: Re: LM399 based 10 V reference
Post by: iMo on April 03, 2019, 06:30:21 am
There is not such thing like ~GHz clock in '334. As I wrote earlier, there is a 144MHz Xtal derived internal clock, used together with a 32taps delay line (voltage and temperature compensated) which together create the "virtual" up to 144Mx32 PWM clock. Thus you may use 144Mx32 as well as the power consumption should be the same, imho. Would be good to try whether those 40mA change with the highest PWM "clock".
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 03, 2019, 02:45:27 pm
Juts for a stable 10 V source, there is no need for the super high resolution PWM. Just having some 10-12 Bit PWM with some 10 MHz base frequency would be good enough with the additional SD like modulation.

As the LM399 is only stable (over the 2 nd year) to maybe 5 ppm, there is not much sense in an adjustment much finer than 20 Bits. So the 22 Bits of the original project is fine.

AFAIK the high resolution PWM in the STM32 need the high µC clock and this causes quite some power consumption. It could get better with heavy power saving modes, but this could again cause jitter or ripple from a modulated supply current.

One point possibly worth a look would be the ADG419 switch. It is relatively slow and might cause some jitter. One might consider a faster switch, maybe even 74HC4053 - it's faster but would need a kind of +-5 V supply.
 
Title: Re: LM399 based 10 V reference
Post by: Magnificent Bastard on April 03, 2019, 04:48:47 pm
There are a number of microcontrollers and digital signal controllers that offer a "High Resolution Timer" in the form of a 16-bit PWM followed by a 32-stage selectable delay line.  Some of these control the accuracy of the delay line transparently in hardware, while most of them require firmware support to do this.

A straight delta-sigma approach with a smaller number of primary bits can have code-dependent low frequency artifacts that are near impossible to filter.  The HR-PWM approach either eliminates this problem, or reduces it to a minimum if you still want to "dither" the lowest bit for even greater resolution.  For example, a 16-bit timer, with an additional 5 bits (through the delay line) results in a LSB that represents only 10V/2^21 (~4.77uV).  In this case, if you are "dithering" the LSB, the possible code-dependent artifacts are limited to 4.77uVpp (and will be much lower after the filter).

The "trick" is to find a part (an MCU or DSC) that has a full hardware controlled HR-PWM, that doesn't require a nuclear reactor to operate, and the filter can be something that is reasonable and low cost.
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 03, 2019, 05:20:23 pm
There are a number of microcontrollers and digital signal controllers that offer a "High Resolution Timer" in the form of a 16-bit PWM followed by a 32-stage selectable delay line.  Some of these control the accuracy of the delay line transparently in hardware, while most of them require firmware support to do this.

Would you mind giving some examples of the microcontrollers/ICs you mention with the selectable delay lines? (I've only found the STM32F334 and MSP430F5172)

EDIT: Ah, I see, something like NXP MC56F827xxx (NXP's "nanoedge").



One issue is that the rising edge has to be delayed separate from the falling edge. Do the parts you mention support that? Or one could use digital logic circuits (AND) to combine a delayed and non-delayed output to generate the different fractions.



Once I get a chance, I'll measure the STM32F334 current consumption with the various dividers enabled, and at 128 + 144 MHz (the two allowable clock frequencies allowed with the DLL). Would a spectrum analyzer (with a DC blocking cap) be a reasonable way to measure the phase noise? Or one-shot period measurements with a Racal 1992 (with 1 ns resolution)? (measured with delta-sigma modulation off)
Title: Re: LM399 based 10 V reference
Post by: iMo on April 03, 2019, 06:20:36 pm
The '334 hires pwm is with 16bits resolution only. I would be happy to know more chips with 16bit pwm at 4.6GHz pwm clock, indeed.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 03, 2019, 08:28:13 pm
As the LM399 is only stable (over the 2 nd year) to maybe 5 ppm, there is not much sense in an adjustment much finer than 20 Bits. So the 22 Bits of the original project is fine.

Not my opinion:

20 bits = 1ppm steps at the tap would give 14uV steps (like popcorn noise) if you change the pwm by one count.
I want to compensate the temperature of the AD419 by changeing the PWM.
So to not get any "popcorn noise" by temperature compensation one step has to be much smaller than around 0.5-1 uV so that it is within the LM399 1/f noise level. So I need at least 16 times more than 20 bits for the final cirquit.
Of course one also could compensate for the ageing too. (which is usually 1-2 ppm/year after some run in phase).


One point possibly worth a look would be the ADG419 switch. It is relatively slow and might cause some jitter. One might consider a faster switch, maybe even 74HC4053 - it's faster but would need a kind of +-5 V supply.
You are welcome to set up a cirquit with the 74HC4053 and make some comparisons.
If you have the cirquit you could also replace the 74HC4053 by a MAX4053A (which is better in charge injection and leakage currents).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Magnificent Bastard on April 03, 2019, 08:57:01 pm
The '334 hires pwm is with 16bits resolution only. I would be happy to know more chips with 16bit pwm at 4.6GHz pwm clock, indeed.

It's not very clear how this works in the datasheet, so you have to go to the reference manual for this part.  Yes, the PWM DIGITAL timer is 16-bits.  BUT--- that PWM output *can* be followed by a fractional-clock delay.  One 144MHz clock period is divided by a 32-stage delay line, which is delay-locked to the 144MHz master clock.  So, you can delay the PWM by 0..31 fractions of a single master clock.  This provides the equivalent of 5 extra bits, without having to crank the clock up to 4.4GHz (which the technology would not handle anyway).

So, what we are saying, is that with a 144MHz clock with a 16-bit PWM delayed extended by a 32-step fractional-clock delay line, you get the equivalent of a 21-bit PWM driven by a 4.4GHz clock.

You can connect an external AND-OR gate (SN74LVC1G0832) to the PWM output and two extra GPIO pins.  You set up an interrupt on the positive edge of the PWM and another on the negative edge.  These two interrupt routines control a 'PWM-FORCE' signal (which goes to the OR gate input) and a 'PWM-ENABLE' signal (which goes to one of the AND inputs)-- with the original PWM going to the remaining input of the AND gate.  This can provide an extra 4-bits of resolution, where the PWM controls both the positive and negative edges (and the force and enable signals are not that critical).  This can be (practically) a 25-bit DAC without any dithering at all.  If you want even more resolution, you can add (maybe) 3 more bits on the bottom, that dither the 25th bit, providing 28-bits total-- and the code dependent low frequency artifacts will be very small.

Get it now?
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 04, 2019, 12:10:02 am

You can connect an external AND-OR gate (SN74LVC1G0832) to the PWM output and two extra GPIO pins.  You set up an interrupt on the positive edge of the PWM and another on the negative edge.  These two interrupt routines control a 'PWM-FORCE' signal (which goes to the OR gate input) and a 'PWM-ENABLE' signal (which goes to one of the AND inputs)-- with the original PWM going to the remaining input of the AND gate.  This can provide an extra 4-bits of resolution, where the PWM controls both the positive and negative edges (and the force and enable signals are not that critical).  This can be (practically) a 25-bit DAC without any dithering at all.  If you want even more resolution, you can add (maybe) 3 more bits on the bottom, that dither the 25th bit, providing 28-bits total-- and the code dependent low frequency artifacts will be very small.

Get it now?

I think I see what you're suggesting.

I think we need to run the final PWM at between 5 and 30 kHz to allow for a reasonable analog filter. The extra bits you output let you construct a few extra MSB that are controlled in software. This also allows the internal timer clock/DLL/whatever to be run at a speed (providing a better time-resolution).

However, I don't see how you could add many more bits without going to much lower resultant PWM frequencies. For example with the STM32F334, the internal PWM period could be up to about 62 kHz (with f_HRTIM=128 MHz*32, 16 bit), and add three extra bits to get a final PWM rate of 7.75 kHz (without dithering). I'm not sure that it's a good idea to go any slower than that.



During my testing of the 'F334, I'm now reading about 53 mA. No matter what I do (other than turning off the HRTIM unit), I can't get the current draw to significantly decrease. I've tried, for example, disabling the unused subtimers and setting their clock dividers to the most divided.

I did also order another dev board today, a TI launchxl-f28027. I'll report on how its PWM generator performs (and current consumption).
Title: Re: LM399 based 10 V reference
Post by: Magnificent Bastard on April 04, 2019, 12:39:30 am

You can connect an external AND-OR gate (SN74LVC1G0832) to the PWM output and two extra GPIO pins.  You set up an interrupt on the positive edge of the PWM and another on the negative edge.  These two interrupt routines control a 'PWM-FORCE' signal (which goes to the OR gate input) and a 'PWM-ENABLE' signal (which goes to one of the AND inputs)-- with the original PWM going to the remaining input of the AND gate.  This can provide an extra 4-bits of resolution, where the PWM controls both the positive and negative edges (and the force and enable signals are not that critical).  This can be (practically) a 25-bit DAC without any dithering at all.  If you want even more resolution, you can add (maybe) 3 more bits on the bottom, that dither the 25th bit, providing 28-bits total-- and the code dependent low frequency artifacts will be very small.

Get it now?

I think I see what you're suggesting.

I think we need to run the final PWM at between 5 and 30 kHz to allow for a reasonable analog filter. The extra bits you output let you construct a few extra MSB that are controlled in software. This also allows the internal timer clock/DLL/whatever to be run at a speed (providing a better time-resolution).

However, I don't see how you could add many more bits without going to much lower resultant PWM frequencies. For example with the STM32F334, the internal PWM period could be up to about 62 kHz (with f_HRTIM=128 MHz*32, 16 bit), and add three extra bits to get a final PWM rate of 7.75 kHz (without dithering). I'm not sure that it's a good idea to go any slower than that.



During my testing of the 'F334, I'm now reading about 53 mA. No matter what I do (other than turning off the HRTIM unit), I can't get the current draw to significantly decrease. I've tried, for example, disabling the unused subtimers and setting their clock dividers to the most divided.

I did also order another dev board today, a TI launchxl-f28027. I'll report on how its PWM generator performs (and current consumption).

The 334 can have a 144MHz clock, which is what you want to run it at to take best advantage of the HR-PWM over temperature.  144MHz/65536 = 2.197KHz (the 'natural' hardware PWM output frequency).  2.197KHz / 16 = ~137Hz (the PWM frequency after the firmware division, as above).

A filter for 137Hz is EASY, since you would be filtering the output of the Delta-Sigma integrator, which will have a rather small triangle wave with some spikes due to the finite slew rate of the op-amp.  Just a 3-pole filter would do it, and the 137Hz remaining would be well below the noise floor.
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 06, 2019, 07:52:11 pm

A filter for 137Hz is EASY, since you would be filtering the output of the Delta-Sigma integrator, which will have a rather small triangle wave with some spikes due to the finite slew rate of the op-amp.  Just a 3-pole filter would do it, and the 137Hz remaining would be well below the noise floor.

Your comment encouraged me to do a frequency sweep using a signal generator for the PWM instead of using a micro. In this case, it's a 33520B driving two 71% PWM signals (out of phase) to drive my dual-PWM analog board (with a Sallen-Key output filter). There ended up being a nice region around 700 Hz where the output voltage was less dependent on frequency, which would be great to capitalize on.

Investigating microprocessors, I have the following comments:
Title: Re: LM399 based 10 V reference
Post by: iMo on April 06, 2019, 09:21:52 pm
What is the -3dB of your 2pole SK filter?
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 06, 2019, 09:49:48 pm
What is the -3dB of your 2pole SK filter?

It's built from 10 uF polyfilm and generic 0805 2.2k resistors, yielding a fc of 7.2 Hz. I do have a slight RC filter beforehand (R4,C4 = 100,10u).

Also note that the switch's outputs have 150k series resistance (so 75k in parallel). Another line of experimentation would be reducing their values.

See the attached schematic.

EDIT: Pleases note that I screwed up the pin numbers of the LM399 and the analog switch.... Will be fixed in the next board.

EDIT: Also that I used MCP6V51 op-amps...
Title: Re: LM399 based 10 V reference
Post by: iMo on April 06, 2019, 10:17:21 pm
Is the value of R7 ok?
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 06, 2019, 10:27:23 pm
Is the value of R7 ok?

My idea (not sure if good or bad) is to balance the current output by the U7 op-amp such that the op-amp has near-zero current output. In a traditional push-pull stage, this could create crossover distortion, but I don't think that applies to chopper op-amps?

The LM399 draws ~1 mA. VDDA is set to 15V by the regulator, feedback sets Vx to 10V. R7 would be 5V @ 1 mA: hence 5 kOhm.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 07, 2019, 07:03:46 pm
Using Vx to supply the current to the LM399 will only work if the output is a an approximately constant level. If the voltage is adjusted to other levels, it may be better to use a different source (e.g. 15 V supply or maybe another OP for an approximately 9 V).
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 07, 2019, 11:44:42 pm
Using Vx to supply the current to the LM399 will only work if the output is a an approximately constant level. If the voltage is adjusted to other levels, it may be better to use a different source (e.g. 15 V supply or maybe another OP for an approximately 9 V).

Good point. This design is very much optimized for +10V (plus or minus a volt or so). The circuit's output is limited by 7V (LM399 Vref), the analog switches (+16V max), and the op amp's output headroom, so perhaps 7 to 14 V). The 15V supply should be good enough to use to bias the LM399.

Speaking of which.... I'm currently biasing the heater with the 15V supply, which seems a bit wasteful of power. How sensitive is the Zener voltage to the heater voltage? Should I bias the heater with the board's input voltage?

I don't see using it outside of between 9.5 and 10.5 V.  I'm not even sure how linear it would be (due to the analog switches).



I'm starting to redo the digital board based on the C2000. Are there any features that I should add? (Is there any interest from others in building one if it works out? I'd be happy to share the layout files.)

Schematics are attached. I'm planning to order boards tomorrow, maybe a second set of analog boards (with a few bugs fixed. I'm currently contemplating adding guard tracks around the high impedance nodes, and if I want a solid copper GND plane on the back.)
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 08, 2019, 05:08:48 am

Speaking of which.... I'm currently biasing the heater with the 15V supply, which seems a bit wasteful of power. How sensitive is the Zener voltage to the heater voltage? Should I bias the heater with the board's input voltage?

Hello,

Questions like this are usually already answered earlier in this thread:

e.g. here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg441913/?topicseen#msg441913 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg441913/?topicseen#msg441913)

and here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg443181/?topicseen#msg443181 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg443181/?topicseen#msg443181)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on April 08, 2019, 08:01:33 am
FYI - there is not such thing like a zener diode wired to the "zener" pins of the LM399 package. The two terminals are wired internally as below (from DS).

@Andreas: as we have PMed re LM399 model - I've run through all your great measurements and graphs in this thread in order to assemble some data. Despite the fact you frequently do interpolate the data in your graphs (ie linear) you do not display the "equation" which could be used in modelling, a pity..  :)
Title: Re: LM399 based 10 V reference
Post by: WillTurner on April 16, 2019, 04:49:23 am
May we briefly return to the bootstrap non-inverting operational amplifier circuit of fmaimon in the first post of this thread?

(https://www.eevblog.com/forum/projects/lm399-based-10-v-reference/?action=dlattach;attach=44356;image)

Consider the simplified op amp 7V -> 10V stage :
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=706791)
The transfer function is :
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=706797)
Since the transfer function has both a pole and zero which are very close together, I conclude that the circuit doesn't operate as a low pass filter as intended, and that the capacitor in the feedback leg is redundant. The pdf below (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=706806 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=706806) ) further develops this analysis. I appreciate the intention of rolling off noise above, say 10Hz, however I don't believe that it is possible in the non-inverting op amp circuit without incorporating resistance (perhaps a simple RC filter) between the zener and the positive input.

I would appreciate someone more knowledgeable corroborating my thoughts.
Title: Re: LM399 based 10 V reference
Post by: iMo on April 16, 2019, 06:03:45 am
FYI
UPDATE - OPA177 and R values as in above schematics.
Title: Re: LM399 based 10 V reference
Post by: WillTurner on April 16, 2019, 06:21:13 am
So your simulation uses a split (+15V/-5V) supply (op amp not grounded), with a 1k load, and you roll off only 3dB, and then your response flattens out (0dB)? Not the 20dB per decade that you would want from a LPF? Doesn't that support the case I made?
Title: Re: LM399 based 10 V reference
Post by: iMo on April 16, 2019, 06:32:08 am
I updated above with the OPA177 schematics you have mentioned in your post..
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 16, 2019, 06:58:03 am
The filtering action only applies to the added voltage from the amplifiers gain. In the PWM version this is mainly filtering the PWM part.  Filtering of the reference noise itself is only done by the extra filter stage at the output and possibly by some filtering between the LM399 and the 1 st. OP. AS the input to the OP is high impedance, one could add some RC filtering (e.g. 1 K +  100 nF) here with little extra effort.  However this would be more something for a non PWM version.
Title: Re: LM399 based 10 V reference
Post by: iMo on April 16, 2019, 07:03:27 am
For example
Title: Re: LM399 based 10 V reference
Post by: WillTurner on April 16, 2019, 07:10:31 am
@imo : please look carefully at your first and second simulations. I believe that they both show you are achieving -3dB gain. I suggest you try a single pole RC filter, and look at the difference  :) . Your third circuit is a completely different kettle of fish.

@kleinstein : Sorry, I haven't been following the recent PWM design. To be clear, I have gone back to the first post of the thread, and am considering the most basic LM399 circuit with a 7V -> 10V gain stage. Probably not that good without very good low TC matched resistors. Nevertheless, my conjecture is that the feedback capacitor in that circuit doesn't do much.

@everyone : I realize that what I have posted may look like heresy. Please think carefully before jumping in  :-[.

Edits:
  1. Corrected 3dB to -3dB.
  2. Clarified "precise resistors" to "good low TC matched resistors"
Title: Re: LM399 based 10 V reference
Post by: iMo on April 16, 2019, 11:58:54 am
I've been messing with an "LM399" LTspice model for a while, trying to simulate the behavior of a real LM399.
So far I've added TCs, dependency on Iz, and white and pink noises.
The params are "derived" from various sources in this thread.
Not useful for metrology purposes, imho, but rather a tool for optimizing the circuits around the 399.
The white and especially the pink noise p-p values are my rough estimates, such I get nice pictures :)
Long term drifts and heater power effects not included so far.
Would be great to have some more real data on the pink noise from field.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 16, 2019, 07:40:50 pm
Would be great to have some more real data on the pink noise from field.

Hello,

there are relative large differences in 1/f noise for the LM399 from sample to sample.
Good devices have around 2-4 uVpp @7V
others i have measured > 6uVpp @ 7V

So if you need low noise you have to sort out the bad ones.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Magnificent Bastard on April 16, 2019, 08:59:42 pm
Would be great to have some more real data on the pink noise from field.

Hello,

there are relative large differences in 1/f noise for the LM399 from sample to sample.
Good devices have around 2-4 uVpp @7V
others i have measured > 6uVpp @ 7V

So if you need low noise you have to sort out the bad ones.

with best regards

Andreas

Hi Andreas,

Have you noticed anything special about the devices with low 1/f noise  (Such as lower TempCo, or lower time drift, etc.) ?

-MB
Title: Re: LM399 based 10 V reference
Post by: branadic on April 16, 2019, 09:13:16 pm
Quote
Hi Andreas,

Have you noticed anything special about the devices with low 1/f noise  (Such as lower TempCo, or lower time drift, etc.) ?

-MB

I'm not Andreas, but can share a table measuring on 16x LMx99. Maybe you can see something obvious?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on April 16, 2019, 10:54:07 pm
At first blush it seems devices which show a change in voltage between heated and unheated greater than a few milivolts  should be carefully scrutinised.
Title: Re: LM399 based 10 V reference
Post by: iMo on April 17, 2019, 08:07:52 am
For example the white and the 1/f model.
Scale on right in 1uV/div.
100Hz 1/f corner.
The amplitude of the FFT spectra seems to be in "V_rms".
Title: Re: LM399 based 10 V reference
Post by: iMo on April 17, 2019, 09:54:13 am
And the model v2 source, for people who want to play with it :)
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 28, 2019, 04:03:18 pm
In the last week or two, I've tried out the C2000 HRPWM, and while it is slightly better that previous attempts, its strategy for generating the high resolution isn't consistent enough that I want to use it for this project.

The C2000 adds a variable delay element (using a string of inverters) on each PWM output. Thankfully, the C2000 has an extra register to configure delay allowing ~23bit resolution, unlike the STM32F334 and MSP430 which reuse the compare bits, limiting the PWM counter to 16-bit. The datasheet specifies the delay very loosely, but my board has a step size (MEP) of about 150 ps. A MEP count of up to 255 can be added to a falling or a rising edge. Incrementing my dual PWM board by one MEP on one channel increased the scaled output voltage by 0.7 uV. (0.07-ppm resolution). Because the MEP delay changes with temperature, one much continually run a calibration routine (using a separate delay loop) to keep track of the scaling factor. With a 60 MHz clock, I have about 140 MEPs per timer tick. The TI provided calibration routine outputs a bunch of noise (on the order of 5 MEP p-p). At this point I should have given up, but instead I learned C2000 assembly and rewrote the routine to add averaging.

The averaging code helps, but there is still a non-linearity at the high-end of the MEP count. I'm realizing that inverter-based delays have relatively bad jitter performance. Another wisdom I've learned in to run the PLL as fast as possible in order to reduce their output jitter. The C2000 with a 60 MHz sysclk using a 20 MHz XTAL, outputting 1 kHz, creates a jitter of about sigma=330ps. (I've realized that I can make fairly accurate jitter measurements with a TDS784D oscilloscope in histogram mode using a delayed timebase. Great!)

So, what next?

I want about 25 bits of resolution, a 1 kHz PWM frequency, and a delta-sigma loop running at 10 Hz or faster. This implies I want a >167 MHz timer clock (10 Hz * 2^(25-1)). The STM32G01 can nearly achieve that, but I'm predicting its jitter performance isn't so great.

I'm concluding that a FPGA may be the proper solution. The hard-IP PWM/counters I've seen do not have the speed or resolution I want. Restricting my search to non-BGA packages also wipes out much of the competition. The popular ICE40 specifies bad jitter and isn't that fast (I've not tried it). The Lattice MachXO2 seems suitable. Using a "DDR" output, I expect it to be able to achieve an up to 750 MHz equivalent timer clock (twice the internal frequency due to DDR) with the timer logic running at (f/4 = ~90 MHz) (which it can according to static timing analysis). With a slower model in the same line (LCMXO2-7000HE-6 dev board) and an external 80 MHz oscillator, a 500 MHz equivalent timer used to generate a 1 kHz PWM had a jitter of sigma=28 ps, and I suspect this is approaching the noise floor of the oscilloscope (which samples at 4 GSa/s).

I had hopes to use a QFN32 FPGA package, but it doesn't seem to support high-speed DDR outputs, so it looks like I have to use a 100-pin TQFP. I only need about ten pins, what do I do with the other 80?
Title: Re: LM399 based 10 V reference
Post by: iMo on April 29, 2019, 07:33:35 am
Quote
I only need about ten pins, what do I do with the other 80?
With the LM399 and the PWM calibrator on your board, you may add 2 opamps (A creating -10V, B integrator), 1 comparator and 3 ADG switches (-10V, 10V, integr cap discharge) and create a 6.5+ digits ADC with all the FPGA's LUTs and IOs available :)
A complete multislope 6.5d ADC (sending ADC results via serial) fits in ~400LUTs, afaik.
Title: Re: LM399 based 10 V reference
Post by: guenthert on April 30, 2019, 04:04:00 pm
[..]
I want about 25 bits of resolution, a 1 kHz PWM frequency, and a delta-sigma loop running at 10 Hz or faster. This implies I want a >167 MHz timer clock (10 Hz * 2^(25-1)). The STM32G01 can nearly achieve that, but I'm predicting its jitter performance isn't so great.

I'm concluding that a FPGA may be the proper solution. The hard-IP PWM/counters I've seen do not have the speed or resolution I want. Restricting my search to non-BGA packages also wipes out much of the competition. The popular ICE40 specifies bad jitter and isn't that fast (I've not tried it). The Lattice MachXO2 seems suitable. Using a "DDR" output, I expect it to be able to achieve an up to 750 MHz equivalent timer clock (twice the internal frequency due to DDR) with the timer logic running at (f/4 = ~90 MHz) (which it can according to static timing analysis). With a slower model in the same line (LCMXO2-7000HE-6 dev board) and an external 80 MHz oscillator, a 500 MHz equivalent timer used to generate a 1 kHz PWM had a jitter of sigma=28 ps, and I suspect this is approaching the noise floor of the oscilloscope (which samples at 4 GSa/s).
[..]
25bits at 1kHz will be a bit of a challenge, but if you could ease the requirements a bit, Parallax' Propeller might be an alternative.  That MCU has a quite odd and by now fairly old design, requires support chips (external I2C connected EEPROM), draws quite a bit of power and has other drawbacks.  It does however allow for arbitrary length PWM via bit banging (since it has eight 'cogs' -- think of them as cores -- and no interrupts, bit banging can be done jitter-free easily).   It's routinely clocked up to 100MHz, some reported success with slightly higher clock rates.  Sadly, despite its age, the MCU is notoriously under-specified.  It derives its clock from an external crystal via an internal PLL circuit.  It's unclear to me how much jitter that introduces (on small time scales -- within a second, I measured the error to be less than 2 parts in 80e6 using a 10MHz reference signal from a GPSDO).

FPGA seems to be the more promising choice, the Propeller might just be easier to program.

I meant to try this for a long time now.  You could beat me to it.  ;)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 30, 2019, 04:20:57 pm
I don't think there is any need to get 25 bits with 1 kHz speed. The LM399 has a much higher noise. So there would be much more noise from the LM399 than any SD modulation amplitude from using "only" some 20 Bit true PWM and than SD modulation on top.
Title: Re: LM399 based 10 V reference
Post by: pigrew on April 30, 2019, 07:09:30 pm

I don't think there is any need to get 25 bits with 1 kHz speed. The LM399 has a much higher noise. So there would be much more noise from the LM399 than any SD modulation amplitude from using "only" some 20 Bit true PWM and than SD modulation on top.

Hmm, Maybe I need to rethink things. Looking at the LM399 datasheet again, it shows ~10 uV (~1.5ppm) of p-p <1Hz noise (Low Frequency Noise Voltage). Looking at the other plot, "Zener Noise Voltage (stabilized)", shows all the noise. I also need to figure out what I'm trying to achieve (My thought has been a 10 V transfer-standard with long-ish term stability, but not necessarily low noise.).

I was intending to get the resolution down to be similar to what I see as the S.D. of voltmeter readings when measuring my source (about 0.5 uV of 10V), so 0.05 ppm, even though this is wasted due to the accuracy of the reference.....

Do you have any suggestions? Is there something you'd like to see done with the LM399?

Connecting a few LM399 in parallel would reduce the noise, but burn a lot of power.

Disciplining a lower-noise source (like LTC6655) with a LM399 (Stack two LTC6655-5 to get 10V, use a 8 to 12 bit DAC to tweak the output to be 10V, divide down to ~7.1V with PWM and compare to the LM399 to use as a feedback loop to control the DAC, using a micro for temperature compensation?)? I guess this is just another name for auto-cal. The disciplining can be disabled, and the PWM divider can be used as a fairly linear high-resolution calibrator output.

25bits at 1kHz will be a bit of a challenge, but if you could ease the requirements a bit, Parallax' Propeller might be an alternative.  That MCU has a quite odd and by now fairly old design, requires support chips (external I2C connected EEPROM), draws quite a bit of power and has other drawbacks.  It does however allow for arbitrary length PWM via bit banging (since it has eight 'cogs' -- think of them as cores -- and no interrupts, bit banging can be done jitter-free easily).   It's routinely clocked up to 100MHz, some reported success with slightly higher clock rates.  Sadly, despite its age, the MCU is notoriously under-specified.  It derives its clock from an external crystal via an internal PLL circuit.  It's unclear to me how much jitter that introduces (on small time scales -- within a second, I measured the error to be less than 2 parts in 80e6 using a 10MHz reference signal from a GPSDO).

FPGA seems to be the more promising choice, the Propeller might just be easier to program.

I meant to try this for a long time now.  You could beat me to it.  ;)

The Propeller looks interesting, but probably can't do what I want because of its "only" 80 MHz clock speed (without using an external serializer IC). I'm also worried the low-ish PLL frequency of 128 MHz implies significant clock jitter. Another similar (but more modern?) design are the processors by XMOS. They quite likely could do the wizardry I'm trying to get, with "tile" clock frequencies of 500 MHz on their lower-end hardware (XS1-L6A-64-TQ48-C5).

I already have a proof-of-concept PWM generator running on a FPGA dev board, so my plan is to further explore it.

So, the question is if I should finish up this 10V LM399-based source. I'm leaning towards yes, just to see how good of a PWM generator I can make. But, I won't make a dedicated PCB this time. I'll just use the machXO2 breakout board, as-is, with a small external MCU to do temperature readings, and package it in a box to avoid drafts. I need to get my design into an oven soon for doing tempco checks.
Title: Re: LM399 based 10 V reference
Post by: branadic on April 30, 2019, 07:40:48 pm
Even though RMS noise (10Hz - 10kHz) is specified to be typ. 7µV and max. 50µV for LM399 it does make sense to use higher DAC resolution, as you can average values to get more of an idea of the real value, as long as you don't have a jumper (popcorn noise). On the other hand you find specimen with low noise, but you have to select them (same is true for drift...).
Don't trust a single picture on noise given in the datasheet, measure the device you have at hand to get a picture of what you are dealing with.

As far as I remember noone found a low noise LTC6655 by now. The datasheet values given are misleading as the LNA used to specify noise on them doesn't match the defacto standard bandwidth?
However, the approach of building a disciplined ADC or DAC by using a low noise reference combined with a reference of good long-term stability is nice but not new. But it puts you in the rabbit hole to find a very low noise reference.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: iMo on April 30, 2019, 08:26:23 pm
I think you have gained pretty good momentum in the "High Resolution PWM Divider" direction (the name for a new topic?).
The actual zener (399 or LTZ, etc) is a different topic, imho. And not extremely important in the situation where you have been spending a lot of effort with the PWM divider.
Forget the 399 for a while and do create a 25bit 10kHz PWM divider  :-+
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 30, 2019, 10:23:48 pm
Looking at the LM399 datasheet again, it shows ~10 uV (~1.5ppm) of p-p <1Hz noise (Low Frequency Noise Voltage).

Many LM399 have around 4 uVpp 0.1 - 10 Hz noise @ 7V so ~ 0.6 ppm (like most integrated buried zener references).
But there is also some stray from sample to sample (My LM399#2 has around 9 uVpp) so I would select for noise.
If you have large integration times then the DMM will already filter some of that noise.

As far as I remember noone found a low noise LTC6655 by now. The datasheet values given are misleading as the LNA used to specify noise on them doesn't match the defacto standard bandwidth?
The too small bandwidth of AN124 is one reason. The other sad truth is that the LTC6655 noise is very dependant on the input power supply voltage.
To get minimal noise you have to reduce the input voltage. At 5.7V for the 5V version I measured factor 1.3 lower than at 10.2V. But with 2uVpp average @ 5V still more than the promised 0.25ppm p_p.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: pigrew on May 08, 2019, 02:55:40 pm
Over the last few days, I've driven my LM399 board by a PWM generated by a MachXO2 FPGA (FPGA code is posted on GitHub (https://github.com/pigrew/msp399pwm/tree/master/mach399pwm).) I estimate a TC of ~0.25 ppm/C based on about 70 hours left in a room without air conditioning. In addition, I've seen about a ppm shift in the output voltage, along with RTN-like events with a magnitude of 1 ppm.

Note the setup is far from optimal. I placed the circuit in a cardboard box with no electrical shielding, used long cables, etc, but it is able to demonstrate the FPGA-based PWM. I'm able to control a 1 kHz PWM with 19-bit control, and 8 bits of D-S modulation, with a PWM resolution of 2ns (250 MHz clock with IO using DDR outputs).

Another small investigation I made was a comparison of various 25MHz and 50 MHz 3.3V surfacemount XO. Using a 4 GSa/s oscilloscope, I could not draw any conclusions about which had lower jitter. Over 1 millisecond, I saw about 40 ps p-p of jitter. The FPGA's PLL did add some jitter, boosting my measurements up to about 80ps p-p period jitter when generating a 1 kHz PWM. The PLL's VCO was running at 480 to 500 MHz, divided by 2 to clock the DDR IO cells at ~250 MHz (higher FPGA speed grades can go slightly faster).

The ASFL1-25.000MHz-EC-T XO used ~2mA of current while the ASV-50.000MHZ-LR-T uses 4 mA. Some of the other models used higher currents.

Where is the RTN coming from? My guess is the reference, but I'm not sure. I have not done any measurements of Vref. The plot showing RTN is a residual plot, after subtracting out the estimated output voltage based on a linear model of time and temperature.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on May 08, 2019, 03:02:02 pm
The jumps in the residual curves could very well be popcorn noise from the reference.

Using a FPGA just for the PWM looks like a lot of effort and power consumption.
For the drift / TC  I would expect the switches to be a possible source. Some units may behave better than others, as the R_on ratio could be a factor.
Title: Re: LM399 based 10 V reference
Post by: iMo on May 09, 2019, 10:06:33 am
Where is the RTN coming from? My guess is the reference, but I'm not sure. I have not done any measurements of Vref. The plot showing RTN is a residual plot, after subtracting out the estimated output voltage based on a linear model of time and temperature.
You may try with a "reference" made of a battery cells, like 4x1.5V or 2x3.6V.
The batteries have pretty low noise usually.
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 09, 2019, 07:05:30 pm
The batteries have pretty low noise usually.

... as long as you keep the temperature constant.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: pigrew on May 10, 2019, 01:09:36 am
Using a FPGA just for the PWM looks like a lot of effort and power consumption.
For the drift / TC  I would expect the switches to be a possible source. Some units may behave better than others, as the R_on ratio could be a factor.

Yeah, it's a bit of effort, but I've always been curious about the Lattice parts.

The power consumption isn't that bad. The 7000HE uses about 28 mA at 1.2 V, plus about 5 mA at 3.3 V for the XO+IO. The dev board itself is reading 100 mA total, but I think a lot of that is the FTDI chip and the other accessories on it. Most of the microcontrollers I looked at use about the same, so this really isn't that bad (assuming the use of a switching regulator to drop down from the input voltage). Given everything being powered by 3.3V, this is about 115 mW compared to the ~350 of the LM399 heater. Not insignificant, but not a dealbreaker. Most of the FPGA's power is burnt by the PLL. With additional programming, a low-power mode could be created where the PLL is disabled and the circuit is clocked by the internal RC oscillator. Maybe I should go to the other thread to implement a multi-slope converter with the 7000HE? I wonder if the PLL's jitter is low enough to get a bunch of digits...

Since I can't use an oven at the moment, my plan is to do ratio measurements of the output voltage vs the reference voltage for a few days using a Solartron 7081 in its 8-digit mode (3072 NPLC, ~=51 seconds) to check how stable the circuit is (apart from the reference).
Title: Re: LM399 based 10 V reference (Ratio of PWM generated output to LM399 ref)
Post by: pigrew on May 13, 2019, 04:40:09 pm
Over the last few days, I performed a ratio measurement of my reference using a Solartron 7081 in ratio mode (and 50 second acquisition periods). For each measurement, the DMM acquires the input voltage and then the reference voltage (on a separate reference input). My 7081 is fairly noisy as is the LM399, but a few days of logging should provide enough data to average things out.

I calculated the ratio in PPM units as as (ratio - ratio[0])/ratio[0]*(10^6)). Since I'm making ratio-mode measurements, the measured ratio should be quite insensitive to the DMM's temperature (which also changed during the logging). Next time, I'll use an external relay to switch inputs instead of the DMM ratio mode, so that I'll be able to log both voltages instead of just the ratio.

Based on a linear fit, the ratio varies by 0.018 ppm/(C-°). I conclude that most of the thermal drift I have is due to the LM399 reference itself and not the PWM divider circuit. After the addition of temperature calibration, I expect the circuit I constructed will have a short-term (on the order of a day) accuracy of <1ppm, with a TC of <0.1 ppm/C. No, I didn't do any math, and I have not yet implemented temperature corrections, but this is my feeling based on the logged data so far.

The step-changes in the temperature are where I was convinced to turn on the house's furnace.... it shouldn't be this cold in May.


EDIT: Another small note is that I looked at the MachXO3 vs MachXO2. The power consumption was estimated as being very similar, but the MachXO3 does have smaller BGA packages which have the high-speed pins. Of all the FPGAs I've looked at, the MachXO2 seems best.
Title: Re: LM399 based 10 V reference
Post by: guenthert on June 23, 2019, 10:42:26 pm
[..]
The Propeller looks interesting, but probably can't do what I want because of its "only" 80 MHz clock speed (without using an external serializer IC). I'm also worried the low-ish PLL frequency of 128 MHz implies significant clock jitter. [..]
I was wondering about that.  I don't think max. clock jitter is specified for the P8X32A.  Don't really know how to measure it, but recently got a new hammer (LeCroy Wave Ace 2032 which allows fairly fine resolution timing measurements using ETS) ...
(http://)
Here I attached passive probes to the output pins 2 and 3 of a propeller on a PiHat board (6MHz crystal -- 96MHz clock frequency.  Unfortunately on this board it's difficult to access the external clock), which I toggle as fast as possible w/o using the counters (16 cycles).  Resolution here is 40ps (25 GS/s), variation in impulse width (including uncertainty in trigger point) peak-to-peak seems to be 160ps, i.e. about 2% of the width.  That sounds terrible at first, but over several cycles it averages out to the stability of the crystal (can't measure that with the oscilloscope).  Now it could be that all early cycles in such a meta-cycle are short and all late ones are long, then that would wreak havoc on a PWM scheme ...
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 24, 2019, 05:11:50 am
Hello,

I simply used the statistics on the pwm output frequency.
(I had to do it on the inverted PWM signal because the rising edge was the one with modulation).

see here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2138395/#msg2138395 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2138395/#msg2138395)

the x-tal frequencies had a standard deviation of about 5*10e6 lower than the PWM frequency.
the R/C frequency was around 10e3 lower in standard deviation than the PWM frequency.
The R/C frequency was unusable with some 80-90uVpp noise on the 10V output.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: guenthert on June 24, 2019, 06:04:21 am
Thanks, I need to ponder this a bit.

Further, the 160ps pk-pk variation in the width of the pulse need to be taken with a grain of salt, as the Wave Ace 2032 spefices:
Trigger and Interpolator Jitter: 0.4ns pk-pk
Title: Re: LM399 based 10 V reference
Post by: Andreas on June 24, 2019, 07:53:23 pm
Hello,

the spec of my scope is: Sample jitter 3ps (rms), typical

But how to compare to a "trigger jitter"?

perhaps it is better to sample a long sequence and do the statistics over the samples.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: David Hess on June 24, 2019, 08:45:48 pm
Measure the trigger jitter with a crystal oscillator.
Title: Re: LM399 based 10 V reference
Post by: pigrew on June 24, 2019, 09:34:47 pm
Thanks, I need to ponder this a bit.

Further, the 160ps pk-pk variation in the width of the pulse need to be taken with a grain of salt, as the Wave Ace 2032 specifices:
Trigger and Interpolator Jitter: 0.4ns pk-pk

Be careful with equivalent-time sampling, as each period has a different jitter and it'll be combining them all together. That's fine for looking at histograms, but not for interpolating periods from the acquired data. I've been disabling ET sampling. In single-shot mode, I've been interpolating between acquired point in order to determine the timing of the rise, to a resolution better than the sample rate, though this relies on a fast rise time of the input signal (since the oscilloscope's amplitude usually has significant noise). I use a P6243 1 GHz active FET probe for this.

In your case, a period of 166ps, and I think that you were looking at the single-period jitter? You have enough memory to acquire an entire period of the clock, so can sample at 2 GSa/s (500ps resolution). If you calculate the period based on the rise-to-rise of what's on the screen, the trigger jitter won't matter. On the other hand, it will matter if you enable ET sampling since it'll mess up the interpolation.

How the manufacturer likely measures trigger jitter is by acquiring a very fast edge, and comparing the t=0 mark on the screen with where the interpolated acquired data actually passes the trigger voltage. Do this enough times and you'll find the pk-pk and the RMS jitter.

In my measurement, I often was looking at signals with periods of about 1 ms. I used a delayed trigger mode (setting the delay to the signal period), and plotted a histogram of the acquired data around the trigger voltage on the screen. This measurement is very much influenced by the trigger jitter. If I knew the trigger jitter, I could have subtracted it out, but didn't.  The Tek TDS784D has a datasheet specified jitter of 7ps (unsure of if it is RMS or pk-pk). Its max sampling rate is 4 GSa/s.

I've been distracted by work recently, but also looking into building something like the TAPR TICC in order to better characterize the jitter. I breadboarded a TDC7200, but was getting about 80 ps RMS jitter  (twice the datasheet value of about 45ps RMS). I then got sidetracked by looking at FPGA-based TDC which made me run into some bugs in the Lattice MachXO2 compiler... The MachXO2 should be able to be at least as accurate as the TDC7200, probably much better, with the use of an external PLL.
Title: Re: LM399 based 10 V reference
Post by: David Hess on June 25, 2019, 01:57:35 am
I agree that normal ETS mode will not work properly but that is not how ETS is used for jitter measurements.  Instead it is combined with persistence and then the horizontal mean, standard deviation, and peak-to-peak values at the second trigger point are calculated.  Combining ETS with persistence may or may not be the same thing as combining real time sampling and persistence depending on the DSO.

I think newer DSOs can make these measurements using their statistics and segmented memory capability but I would verify any method used to be sure it is producing correct results.
Title: Re: LM399 based 10 V reference
Post by: guenthert on June 29, 2019, 05:44:04 pm
Well, I like to understand this better.  I'm afraid in my scope 'persistence' is just a display feature, aiding visual inspection, but of course, one could export a series of waveforms to a PC and do the calculations there.  Need to read up on the subject. 

  Clearly the stated 0.4ns jitter must be a maximum value across all time scales (they considered relevant).  If it would apply to short time scales, ETS couldn't work at 20ps resolution.  Wished the manufacturer would publish more elaborate specifications.

  In any case, if the observed jitter indeed stems from the MCU (or at least a good share of it) and a better clock were available, then one could use an external shift register, I suppose.  I dimly recall that such has already been suggested.
Title: Re: LM399 based 10 V reference
Post by: David Hess on June 29, 2019, 07:28:37 pm
If ETS is supported, then the timing resolution can be calculated by just taking the reciprocal of the specified ETS sampling rate.  Whether the DSO achieves that without added noise is another matter.
Title: Re: LM399 based 10 V reference
Post by: iMo on August 19, 2019, 05:31:55 pm
If you connect the 34401A to a computer anyhow, I would use 10NPLC with 10 samples  instead of 100NPLC. The 34401A can't do more than 10NPLC and the math doesn't work correct if the meter does the averaging. And you don't loose resolution above 10V.

I also see far below 1ppm pp with my 34401As

Could you elaborate on the 34401A math problem with 100NPLC averaging, plz?  ???

PS: found a post https://www.eevblog.com/forum/metrology/real-integration-time-of-hp-34401a/msg1197640/#msg1197640 (https://www.eevblog.com/forum/metrology/real-integration-time-of-hp-34401a/msg1197640/#msg1197640)
perhaps you are referencing to..
Title: Re: LM399 based 10 V reference
Post by: iMo on August 28, 2019, 05:43:35 pm
Burn-in of LM399 - do I need to wire the zener reference part as well?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on August 28, 2019, 06:14:40 pm
It helps to also have the Zener part working during burn in. This way one can observe the initial drift phase and get an idea how fast it gets slower.  Powering the zener part is not that difficult: the current should be reasonable constant, but is not that critical. So it can be a normal OP (like LM358, 741, OP07 is already fancy) and 3 reasonable stable (e.g. 100 ppm/k, maybe 25 ppm/K) resistors.
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 28, 2019, 07:16:48 pm
Burn-in of LM399 - do I need to wire the zener reference part as well?
Hello,

I would do that to monitor the ageing from time to time.

In my LM399 ageing box I use 6K8 (1% metal film) to a 14V stabilized supply for the zener.
(ok in the 2nd row when the 6K8 ran out I used 10K || 22K)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on August 28, 2019, 08:10:02 pm
The aging of wine.. Does the aging help with the popcorn too?
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 28, 2019, 08:27:53 pm
Does the aging help with the popcorn too?

Popcorn = impurities of the silicon.

Does the silicon get cleaner during ageing?
So I fear: no.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on August 28, 2019, 08:35:38 pm
I've been bending the leads a little bit (using 2.54mm pitch socket). Hopefully that stress does not contribute to the popcorn too..
Title: Re: LM399 based 10 V reference
Post by: Andreas on August 29, 2019, 08:53:30 am
Hello,

I think that is no issue as long as you do not damage the glass seals.
I have here one LM399 which is affected by humidity.
(Seasonal changes of the output voltage).
So I think the hermetical seal is broken.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on August 29, 2019, 11:29:43 am
Interesting.. Pretty hot the 399 package and the humidity has such an effect..  :phew:
Title: Re: LM399 based 10 V reference
Post by: FriedLogic on September 03, 2019, 12:11:07 am
I've been bending the leads a little bit (using 2.54mm pitch socket). Hopefully that stress does not contribute to the popcorn too..

I got some 2mm pitch sockets (2x2 Preci-Dip 833 Series) for an LM399 ageing box since the spacing is closer to correct and the insertion force looked reasonable. Although unlikely to be a problem, I thought that the less mechanical stress there is the better.
It's not built yet so I can't comment on how well they work (has anyone else tried them for this type of application?)

Even in the few LM399's that I have running I've seen occasional popcorn noise appear or disappear over months, so I'd expect that ageing has some effect, but it could be good or bad.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 03, 2019, 09:45:46 am
I was using the 2.54mm rounded precision ones and my observation had been the touching the 399 while in the socket changed the output voltage by dozens of uVs.
Title: Re: LM399 based 10 V reference
Post by: Robert763 on September 05, 2019, 09:09:29 am
There is a proper socket available that the 399 will fit (0.1" 4 pin square. It's a TE connectivity 8060-1G4
Mouser part No. 571-8060-1G4  Farnell order code 1608447
https://www.te.com/usa-en/product-1-1437504-7.html (https://www.te.com/usa-en/product-1-1437504-7.html)
Drawing here:
https://www.te.com/commerce/DocumentDelivery/DDEController?Action=showdoc&DocId=Customer+Drawing%7F1437504-3%7F99%7Fpdf%7FEnglish%7FENG_CD_1437504-3_99.pdf%7F1-1437504-7 (https://www.te.com/commerce/DocumentDelivery/DDEController?Action=showdoc&DocId=Customer+Drawing%7F1437504-3%7F99%7Fpdf%7FEnglish%7FENG_CD_1437504-3_99.pdf%7F1-1437504-7)

Robert.

 
Title: Re: LM399 based 10 V reference
Post by: free_electron on September 06, 2019, 04:52:01 am
if anyone needs real lm399's ( not from china ) . i just found about 20 or so in my garage. brand new , NOs , nat semi (old logo )
10$ a piece.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 28, 2019, 11:03:38 am
What would be the "optimal" value for the RS1..RS4 and CLP1 in case I want to average 4x LM399s?

I plan to use an OPA277 (bip) or OPA2210 (low noise bip), or OPA187 (AZ) there (all are with drift <=0.1uV/C).

PS:

OPA2210  0.4pA/sqrtHz      2.5nV/sqrtHz
OPA277    0.2pA/sqrtHz    12nV/sqrtHz   (fyi - OP07  0.35pA/sqrtHz   11nV/sqrtHz)
OPA187    0.16pA/sqrtHz   20nV/sqrtHz

PPS:

added some calcs, it seems 4x RSx=47k + OP create around 1uVpp (CLP1 not connected).

OPA2120 below should be OPA2210.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 28, 2019, 03:00:25 pm
The LM399 has relatively high noise. So it is relatively easy to get get resistors and an OP with lower noise. If filtering is used, it is more the filter function that sets the suitable resistance. To high resistance, bias current drift, leakage and resistor noise set an upper limit. I would consider the 47 K about the upper useful limit, maybe better only 10 K to be conservative with high bias OP.
Without a filter one can go lower - like 1 K if one wants, as there is still not much voltage across the resistors. In most cases I would add at least some filtering, as it's low effort in this position.

The choice of OP also depends on the load to the OPs output - usually no need for something really fancy and the OP07 may be good enough.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 28, 2019, 03:18:15 pm
When using RSx=100k and CLP1=1uF the BW(-3dB) is 1.6Hz. That lowers the resistor's noise, and let the OPamp noise BW stays the same (??) (say <10Hz due to DMM's sampling period). Sure it is at the edge.

Re noise the OP-07 is similar to the OPA277, the only issue I see with the OP-07 is up to 10x higher TC drift.

PS: fixed RS_CLP1 BW.
Title: Re: LM399 based 10 V reference
Post by: antintedo on September 28, 2019, 04:03:49 pm
When using RSx=100k and CLP1=1uF the BW(-3dB) is 1.6Hz.
1-pole filter is far from a brick wall filter, so the effective noise bandwidth is 1.57 * 1.6Hz. Not very important here but good to remember. Also watch for 1/f current and voltage noise of the opamps - parameters are specified at 1kHz but you are most concerned about 0-10Hz region. AD app note MT-048 has some relevant information. To be able to ignore resistor 1/f noise make sure to use good quality metal film or wirewound resistors, crappy thick film ones can be problematic.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 28, 2019, 04:15:25 pm
The averaging resistors see rather low voltage. Thus resistor excess noise is not really relevant for these resistors. This is different for the resistor(s) in the 7 to 10 V stage, that see some 7 V.

The 1/f noise can in deed be relevant. Especially the current noise of the BJT based OPs tends to be higher at low frequency.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 28, 2019, 04:25:19 pm
Therefore the OPA2210 - from DS
Quote
The OPA2210 precision operational amplifier is built on TI's precision superbeta complementary bipolar semiconductor process, which offers ultra-low flicker noise.. SiGe bipolar process..

The Q is how is the current noise at 0.1-10Hz.
A: aprox 22pA p-p (3.3pA rms).
Title: Re: LM399 based 10 V reference
Post by: antintedo on September 28, 2019, 04:58:18 pm
Current noise is specified in the datasheet: page 7, figure 4. With 1/f corner frequency and white noise density it is possible to approximately calculate the noise. The app note I mentioned does exactly that.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 29, 2019, 10:03:11 am
FYI - the noise analysis of an OP07 and an OPA2210 in a traditional Vref circuit.
The noise of the LM399 not included here.

PS: interestingly, the built-in noise model of the OP07 is pretty optimistic, at least when comparing the OP07 and OPA2210 datasheets :)

PS: added a few popular opamps
Title: Re: LM399 based 10 V reference
Post by: iMo on September 29, 2019, 01:03:52 pm
A summary of the above noise simulations..

PS: added Vn_out with RS=10k

PPS:
BTW - the chopping frequencies:
                  Datasheet           Simulation peaking at
LTC1050     2.5kHz                2.5kHz
LTC1052     330Hz                 140Hz (model without external 100n caps)
LTC1250     5kHz                   none
LTC2057     30kHz                 30kHz
ADA4522    4.8MHz/800kHz   none (sim up to 10MHz)
Title: Re: LM399 based 10 V reference
Post by: iMo on December 23, 2019, 12:32:38 pm
Fyi - what happens when you add resistance of pcb traces (35um copper, l=length, w=width, includes TC) into the LTspice simulation :)
An example where gnd and power traces are 50mm long with 0.8mm width.
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 27, 2019, 10:50:58 pm
Hi,

Maybe I'll start my own topic, but I'm working on this circuit right now.
The circuit will be placed in a small oven that will be kept at about 42 degrees.

The heaters of the LM399AH will get their own "floating" power supply.
And these power supplies will be connected at one point, that is in the schematic by means of R1.

The output buffer is not in the oven but on a separate circuit board.
If there is enough interest, I can create a new topic about this 10V Reference.

The schematic
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-08.png)

Shoot @ it!  ;)

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: guenthert on December 28, 2019, 12:28:00 am
[..]
Shoot @ it!  ;)
[..]
  I'd appreciate some stated goal, some specifications to be met, to ground the discussion and to evaluate, whether such specifications can (and once realized) have been met.  This is even more needed, if a not-so-obvious design is chosen (here: why put the LM399, which has its own heater, in an oven?  The LM399 was developed to make ovens or temperature compensation unnecessary).

  Said that, even w/o such stated goals, I can say that you're in hot water there as from the schematics it looks like low on the heater (pin 4 on the LM399) might be on a higher potential than low on the "zener" (pin 2 on the LM399).  The data-sheet says that's a no-no.
Title: Re: LM399 based 10 V reference
Post by: Andreas on December 28, 2019, 09:08:36 am
the heater (pin 4 on the LM399) might be on a higher potential than low on the "zener" (pin 2 on the LM399).
Not if the heater supply is really "floating".

(here: why put the LM399, which has its own heater, in an oven?  The LM399 was developed to make ovens or temperature compensation unnecessary).
I guess the oven is only for the 7->10V transfer resistors.
And of course it also reduces the Seebeck effect on the Kovar leads of the LM399.

Shoot @ it!  ;)
Id like to hear some of the design decisions / reasons.
Why use a LT1010 (150mA) as low ohmic buffer if you use 50 Ohms series resistors.
(The ADA4522 with  ~80 mA would be sufficient).

- I would use decoupling capacitors parallel to the heater. (47uF + 100nF see also data sheet).
- also 100 nF across the zeners (to reduce EMI).
- D7 (12V) is not clear to me together with 15V supply

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 28, 2019, 10:04:41 am
There is some sense in using an extra buffer, to avoid the heating inside the AD4522 from the load current. However the load current should normally be low, so a normal OP (e.g. good old NE5534) would be good enough.

The zener D7 also confused me at first  it is just for start up. Later on it would see some 5 V and thus not conduct.

Having the LM399 also in the oven is a two sided thing. It helps a little with the Kovar leads - though not very much as the LM399 is internally still much hotter and the average temperature at the other side does not matter that much as the wires are likely reasonably well matched.
The down-side of the LM399 in the oven is that this is a significant heat source. So the oven temperature needs to be higher or more conduction to the outside. This would also need more power to the heater. The parts that could really profit from a stable temperature are the resistors.

The heavy filtering after the reference can have a tricky side: To get the high capacitance one would likely need to use polyester caps and these show quite some DA. I have not done the math and don't know the exact DA numbers for the caps, but from some point on a low frequency filter will need considerable settling time. Not because of the filter RC time constant but the DA settling over the time of several hours. The typical time scale for the popcorn noise of the LM399 is still longer than the filter time constant - so it would not really help with the popcorn noise.
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 28, 2019, 11:14:36 am
Hi,

I'm going to explain why the schematic is structured like this.

But first this...
I still had a handful of good parts lying around and another one of my component test ovens, so I thought about building a 10V reference with it.
I just want to do the best I can with the parts I have here.
.
The setup shown here will probably only be built once, so I don't take into account the parts costs and labour as this is unimportant to me.

About a month is here in a test box 2x LM399AH to age and from these then two are selected for low noise and absence of popcorn noise.
Andreas, thanks, i forgot the heater capacitor in the next version of the schematic is show it wil be there.

R4 and R5 mix the reference voltage of the two LM399AH IC's and these resistors are also used to create the low pass filter together with C1 to C6.
R6 helps to protect the +input of the opamp a little with the charge of the capacitors when switching off.

The opamp + buffer is besides a 1.4x amplifier also a low pass filter.
This is mainly meant to keep the noise of the amplifier and the buffer low.

The measurement of the noise without already being optimized was about 1uV RMS measured with my Audio Precision measuring set at a 22KHz bandwidth.
Of course this was using the 400Hz high pass filter.
This was necessary because despite the decent build on a breadboard en some shielding the 230V 50Hz field is still present.

This measurement was done in the test setup with 1x LM329 as zener.
But the noise tests were meant to find out which opamp works best with the precision resistors I have at my disposal.

Yesterday I tested three opamps, the LT1007AC, the OPA140 and the ADA4522 which is currently doing the best in my test setup.

Andreas, you're asking for the extra buffer if the ADA4522 can deliver peak currents of up to 80mA?
I don't like that this precision opamp is directly connected to the output.
My starting point was also, that the 10V output can be abused.  :box:

D7 and R20 is used for startup, this was not necessary for an ADA4522, but it was sometimes necessary for the OPA140 and LT1007.
When power on this circuit, it pulls the input of the LT1010 buffer positive for a moment.
In normal operation D7and R20 do nothing, a little leakage current if there is one, is controlled by the opamp loopgain, normaly that wil be no problem.
I'm going to do some more tests on these parts, to see if these two components can be connected directly to the +15V output, then they're completely out of the loop.

Q1 limits the output current, and in my test circuit this is a little less than 35mA short circuit current.

The problem I run into at the moment, is that I am against the noise floor of my Audio Precision Analyser...
This is about 0.8uV RMS in 22KHz bandwith.

A simple extra preamplifier is not possible.
This should be a symetrical preamplifier with low noise, high common mode suppression and a not no too low input impedance.
I am thinking about how I am going to build it.

So, this schematic isn't ready yet, but I'm satisfied with the output buffer.
It's solid and relieves the opamp of eny load and overload.

I also tested some of my transistor buffers after the opamp, but that wasn't a good plan as far as the noise behavior of the circuit is concerned.

This needs to be tested further:
R10 + C8 and C10 with a small series resistor of say 10-Ohm and a 1uF capacitor directly over the output.
This to keep enough bandwidth to control the noise behavior of the LT1010 buffer.

The 2nd order filtering of the opamp + buffer is only meant to keep the noise of this circuit as low as possible, it is not meant to filter the noise out of the 2x LM399.
R4 +R5 and the 6x 10uF have a much lowerfilter corner.

So, this is not a circuit to become a better than an LTZ1000 reference, that was not the starting point.
But to build as good a circuit as possible with the present parts.
By building the temperature sensitive parts in a oven, I have almost no more problems with the Seebeck effect.

These are de components for R4 and R5, R7, R8 and R13, R14 and one of the 10uF capacitors.

(http://www.bramcam.nl/Diversen/Chopper-Opamps/Componenten-01.png)

When I've done some more tests, I'll show you the results here.

Thanks for your remarks!

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 28, 2019, 01:00:31 pm
Hi Kleinstein,  ;)

About the buffer, yes you can use a NE5534a as a buffer if its compensated well.
But i think i have more then 100 pieces of all kind of buffer IC's, so why not use some of them...
And the LT1010 is a great buffer.

The oven is not initially intended to help the LM399, but of course it does help to make the effects of ambient temperature even smaller.
As far as my knowledge of the documentation on the LM199 zener is concerned, a constant ambient temperature helps to increase stability.
Also some extra shielding helps the stability because this limits the power of the heater, so if the space allows it in my oven,
there will be some insulation around the two LM399 ICs.
The difference between the LM399 oven temperature and de main oven temperature is big, normaly > 40C.

The main reason for the oven is to keep the opamp circuit as stable as possible.

The difference between the oven in the LM399 and the main oven is big enough for a good control of the oven in the LM399.
Because I keep the buffer IC and the output load resistor outside the oven, I have enough control capacity left for the main oven at about 30C LAB temperature.
With some simple insulation, the main oven consumes about 1.5-Watt.

Filtering
Concerning the filtering of the zener noise, only the passive filter and putting the two LM399 ICs in parallel really helps by the low corner of 0.3Hz.
I wanted to do something different this time, several times I used a bootstrap electronic capacitor filter.
Such a filter(bootstrap) is also in my main LAB reference with 4x an LT1021.

I'm going to do another test with the 6x 10uF capacitor, but the current test has been done with the same series of capacitors only of a higher voltage and high capacity, these only don't fit in the oven.

Starting up the oven naturally takes time to reach the equlibrium.
The slow reaching of the maximum charge of the capacitors is therefore not such a problem.

Popcorn!
By selecting the LM399 IC's I hope to have as little popcorn as possible.

Kind regards,
Bram

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 28, 2019, 02:56:10 pm
Compared to a relatively large extra oven the capacitors may in deed be fast, though slower than one may think.
With a dielectric absorption in the 0.5% range the equivalent circuit for a 100 µF capacitor has something like 0.5 µF with a series resistor in the GOhms range in parallel. With a low RC cross over this extra part can become a problem, as 1 Gohms and 10 K from the references give an effect in the 10 ppm order of magnitude. So settling may take a few of the Gohm * 0.5 µF (  = 500 second range) time constants. So it could take up to about 1 hours for the DA to finally settle to less than ppm level of effect. The resistance value in the equivalent circuit depends on the DA and may change with temperature.

Chances are one could still be just OK with the 60 µF, but one should keep this effect in mind if the filter cross over gets lower than 0.3 Hz. At some point (maybe from 0.1 Hz) on one would have to wait for the DA (and thus more like hours) and not just the filter RC (minutes).
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 28, 2019, 03:29:49 pm
Hi Kleinstein,

I measured one of the 10uF WIMA MKS 2 capacitors on my Philips PM6303 Bridge.
F = 1KHz

C   = 9,9uF
Q   = 240
D   = 0.004
RP = 3.8K
RS = 0.07-Ohm
Z    = 16.08-Ohm
Cp = 9.898K
Cs = 9.898K

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: dgminala on December 30, 2019, 10:09:34 pm
Hi,

Maybe I'll start my own topic, but I'm working on this circuit right now.
The circuit will be placed in a small oven that will be kept at about 42 degrees.

The heaters of the LM399AH will get their own "floating" power supply.
And these power supplies will be connected at one point, that is in the schematic by means of R1.

The output buffer is not in the oven but on a separate circuit board.
If there is enough interest, I can create a new topic about this 10V Reference.

The schematic
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-08.png)

Shoot @ it!  ;)

Kind regards,
Bram

Hi blackdog,
I just found this topic a while back, and have been perusing some of the more intense noise-related posts, many of which are above my pay grade (I'm retired).  Your post caught my attention, because I've been working on a similar concept for a reasonably accurate, stable and quiet 10V reference.  My approach was going to be very similar to yours - sum 3-4 LM399s into a scaling amp/filter, ovenize the circuitry following the LM399s to minimize environmental effects, and wrap the LM399s with insulation. 

For the record, I'm a retired engineering tech, with FINALLY enough free time to build some of the stuff that I had wanted to build when I was employed. So, I'm not on the level of most of the guys & gals who have participated in this thread, but I have a pretty good grasp on the electronics being discussed here.

I'd like to see this topic continued and maybe I can contribute in some way.  I'd certainly like to see a working product for my own use at some point. 

Thanks for offering your idea for critique!
Dave M
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 31, 2019, 02:57:20 pm
Hi dgminala,  :)

Thanks for your comments, I'm almost at the age of retirement for my Pension.
I'd like to see how you've built your circuit and what your considerations are.

And now something else that I think fits in this LM199/399 topic.
Because remarks were made about the D.A. of the capacitors I used, I thought it was time to make a box to test this.
This is my first trial and input from the forum users is appreciated as well.

The starting point was to be able to measure the DA and the leakage current of not too large capacitors.
I take leakage currents into account, therefore the function switching is done in a somewhat different way than what is probably expected.

Measuring the D.A.
What I know so far is that a capacitor is charged to a certain value for one hour, then short-circuited (discharged) for 10-second and then after 15 minutes measure how much voltage is back in the capacitor connections.
I do the charging in this first version from a 10V reference IC with a 10K resistor. <= Why use a 10V reference? I've got a lot of them lying around and the low noise...
The same 10K resistor is used to measure the leakage current and/or to see how far the capacitor is charged.

My DVM's which have between 30 and 70-pA input current at the high input impedance setting for the 1 or 10V DC range.
The error caused by this will be small enough for this simple measuring set.

This is the first version schematic.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-01.png)

.
This is the second Bimbox where I try to find a good mechanical setup, the first box was too small.
From the middle to the right where the large 100uF capacitors are located, is meant for the capacitors to be tested.
From the middle to the left side is for the two 9V batteries the 10V reference, the switches and LED.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-02.png)
.

Here you can see how the measurement function is switched with the connecting bridge,
I have two more beautiful versions of this connecting bridge made by the brand "MC" with cold contacts, but I couldn't find them when I took the pictures.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-03.png)

Shoot @ It!

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 31, 2019, 03:52:57 pm
There usually is no real need to measure the current during charging. A 10 K resistor would not be sensitive enough to see leakage current of many caps anyway. The external link may be more prone to leakage than a suitable relay or switch, as one has to somehow handle it and also just moving the hand around could cause some (though small) charge transfer.

If one is also interested in shorter time scales than the standard 10 seconds / 15 min case, a relay could help for accurate timing for the discharge time.

If the DMM is only connected after some 15 minutes, one should have the DMM shorted before, to avoid charging the input capacitance.
If the DMM leakage is really low or the capacitor not that good and large, one could also measure during recharging and thus get more data from one run. In this case it would help to test both polarities to suppress leakage. The on the fly measurement is more suitable for the faster DA part (e.g. up to a few seconds), not so much for several minutes.
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 31, 2019, 05:06:57 pm
Hi Kleinstein,

Thank you for your comments, it's made me think more about it and make some more adjustments. !

1e
The reason that I also made it possible to measure the voltage across the 10K resistor is this, I also want to use it to select electrolytic capacitors on leakage current.

2e
This test box is intended for capacitors from say 1uF.

3e
I am aware of the charge my body has when switching the function.
But by using the isolated bridge of the "MC" Brand and touching the metal of the box with one hand, I can switch " low of charge" for that matter.

4e
My first thoughts were a rotary switch and/or relay's to switch the functions.
But I don't have a good rotary switch that will do, and with relays, my batteries will run out quickly.
It can be done with relays by using bistable versions to save power. but I will have to order those and I also want to build this measuring instrument with the parts that are now available here.

5e
Leaks due to dirt can of course always be dare with the connecting bridge as suggested here.
But measuring instruments of mine, that are already about ten years old still look like new.
I regularly clean the connectors and make sure my hands are clean when operating my instruments.
So i have to be shure this box is clean before i use it!

6e
Regarding the charge of the DMM input on the high impedance setting, that is correct, one of my KeySight 34461A goes on the 10V range in overrange, if you wait long enough.
For those who don't know yet, the source of the problem is the leakage current of the DMM running out of the input and which charges the capacitor and the paracietic capacitors in the DMM.

So short-circuiting the DVM input before you start measuring might be a good plan.
But I'm still thinking about maybe doing it a bit differently, I will show you in the 2nd version schematic.

7e
I measured two of my DVM's in terms of input capacitance, that is for the KeySight 34461A ~300pF and for the Tektronics DM4050 about 210pF, both in the 10V DC range.
The charge in these capacities together with the capacity of the short measuring cables will, as I think about it, give little disturbance to the measuring data if you measure capacitors larger than 1uF.

Kind regards,
Bram


Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 31, 2019, 05:41:21 pm
I just did some DA measurements on 2.2 nF low loss caps at shorter time scale (e.g. 100 µs to 50 ms). So I may be thinking about slightly different effects that are less of a problem here, though the smaller cap and shorter time scale require similar insulation resistance.

One could have the circuit with only one switch that is critical - the one to connect the DUT to the circuit. The rest of the circuit could than use less critical switches to select charging, discharging and measuring. The meter could be always connected, to also measure charging voltage. So it would be charge, discharge and "open" for the 2 nd switch. 

In most cases the charge in the DMM input capacitance is not a problem, at least with 1 µF or similar caps. It could be with smaller caps.
So limiting the initial voltage is usually good enough.  In a few seconds the input voltage will not rise very much, the full 15 minutes could be a little long.

Measuring the leakage current of the cap under voltage may requite a very stable reference. Just the DA measurement does not need such a stable voltage (just a 9 V block cold be good enough), as it is not to know the DA to 1% of its value. It is usually more the question if the DA is in the 1%, 0.1 % or maybe 0.01% range.  Looking at different time scales may be more informative than a very accurate scale factor.
DA can be quite temperature dependent.
Title: Re: LM399 based 10 V reference
Post by: maat on December 31, 2019, 06:58:49 pm
@Blackdog: You might want to put a temperature sensor on that board, close to the capacitor, because leakage current is strongly dependant on temperature.

Judging from your circuit the power consumption will be pretty much negligible, therefore  the temperature will be around room temperature, but if you put those caps into real circuit you will probably some 10 K above ambient. Depending on the capacitor, this might double or quadruple your leakage current. So maybe you want to add some kind of heater to your box.

If you are interested, I do have leakage current values at different temperatures for the WIMA MKS2 10 µF, 50V cap you intended to use. I'll will dig them out.

To get some reasonable values from the WIMAs I had to wait about 8h. Nichicon UKLs needed about a day or two to settle. Kemet T491 about 8h. Regarding the Tantalum capacitors, after eight hours, there was still significant DA, but in comparison to the leakage in absolute numbers it could be considered settled.

I usually measure across a 1 MOhm resistor using a good DMM. Their input current is on the order of a few tens of pA or lower so having them connected in parallel can be neglected when measuring capacitors 1 µF or up.
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 31, 2019, 08:11:33 pm
Hi, maat,

All information is welcome, I think also for other forum users.
There is not much information available about capacitors except some generalities.

Currently I am testing a 100uF MKS4 from WIMA with a 10V reference module and a 10K resistor,
after 8 minutes of charging the difference between the reference output and the voltage across the capacitor is 10uV.
The current that was flowing at that time was 1nA and it is still dropping.

Of course this is not 6x the WIMA MKS2 10 µF, 50V capacitor, but I will do a test with that tonight.

So all information is welcome and I will also do a test with the capacitors in the 42C oven to see if the leak is significantly higher than at 25C.

Thanks!

And have all a nice last day of the year, here still 3 hours to go before the end of the year!

Bram
Title: Re: LM399 based 10 V reference
Post by: blackdog on December 31, 2019, 09:13:48 pm
Hi,

I had a little time before the lady of the house called me.  :-DD

Here some measurement data done on the philips PM6303 measuring bridge and the 100uF WIMA at 10V leakage current test setup.

This is the offset of the meter when I short-circuit the meter (10K resistor).
If the MC short-circuit bridge applies, you have to wait about 5 minutes until Mr. Seebeck has left.  :)
When the display is stable on the meter, I use the "Null" function on the meter to adjust this offset away.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-10.png)
.
This is the value after using the "Null" function.
2uV across the resistor means a current of 200pA, there is about 50pA bias current from the meter.
I forgot to see if the current is off the meter right now or running out of the DVM.
Take these measurements with a grain of salt, they are difficult to perform reliably, just see that it indicates a direction.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-11.png)

These are the six WIMA 10uF 50V capacitors in parallel, total capacity 58.31uF.
This is already difficult to do a good 4 wire measurement because of the low impedance.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-05.png)

With well connected sense wires, the meter shows a relatively high Q.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-06.png)

This is the "D" of the capacitor which is also 1/Q.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-07.png)

And this is the impedance of the six capacitors at the measuring frequency of 1KHz.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-08.png)

Kind regards,
Bram



Title: Re: LM399 based 10 V reference
Post by: Kleinstein on December 31, 2019, 09:21:16 pm
If the current is still drop, it may very well be part of the DA and not real leakage. DA is not just happening as charge recovery, but also as delayed charging of the capacitor. With the usual 0.5% DA number for polyester caps one can about expect 0.5% of the charge and thus some 5 µC to go to the delayed part. The number is only approximate, as it is for a given time scale. If the effect is slower, there could be even more and the fast part is already done.
Still 5 µC are good for 1 nA over 5000 s, so this could be most of the current seen.
1 nA and 10 K source resistance and thus some 10 mV may still be noticeable for the LM399 reference. 200 pA may be below the level where it is a real problem.

With high DA like in MKS caps it is really difficult to measure leakage, as it is difficult to separate it from DA effects. One could see the difference comparing the charging and charge recovery part. The DA part in principle comes back as charge recovery, the leakage does not. However it is a slow process and the difference may not be visible over short time (e.g. less than hours).
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 01, 2020, 02:27:08 pm
Hi,

You all have a nice 2020  :-+

Yesterday I connected the 6x 10uF 50V WIMA to get an impression of the leakage current and DZA of these capacitors.
The large WIMA 100uF 63V MKS4 had become almost immeasurable, so low was the voltage across the 10K series resistor.
It was close to the drift, input leakage and noise of the reference and the measuring instrument used.

This 100V 63V WIMA MKS-4 capacitor could well be used in my LM399 reference circuit,
the leakage is very low and the DA can be used for a reference that is in a oven because of the heating time to get stable.
But it is much too big for my oven.

Great, let's go back to the 6x 10:00 WIMA in the 50V version...
It's always good to put your components to an extra test.

I have already used these capacitors in several filters and had no problems with that.
But with the filter application of this LM399 reference design, many more properties of the capacitor are important.
Of course I looked in the datasheet of this type of WIMA MKS2 and hoped that the insulation value would be better than mentioned in the datasheet.
And I think this is also the case but I use six capacitors in parallel, that doesn't really help.
And when I think of what forum user "maat" explained, that the leakage current of this type of capacitor is quite temperature sensitive, it is better not to use this type of capacitor in this project.

But then what?
Back to my old filter which I also used in other voltage references including my main LAB 10V reference which is also in an 42C oven.
That is a bootstrap electrolitic capacitor filter, placed here in the latest version of the schematic.

(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-09.png)

So now we are going to perform the next test with 330uF capacitors.
There are already a few types here, which I have already bought for low leakage applications in this kind of filters.

So now we are going to perform the next test with 330uF capacitors.
There are already a few types here, which I have already bought for low leakage applications in this kind of filters.

For the first test, I attached two 470uF 35V Nichion and two 470 Nichion 16V to my 10V reference PCB in a "real" professional way!  :-DD
This test is with new capacitors, no formatting in place.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-11.png)

And here I am formatting 13 capacitors at 15.5V.
I'll leave this on for about 36 hours.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-12.png)

I still use my series HP and Harrison power supplies for general testing in the LAB.
Here I use the Harrison 6823A for the capacitor test, on this power supply I also mounted a BNC so it's easier to connect it to one of my function generators.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-13.png)

The test with these capacitors on the 10V reference is now on for 0.5 hours and the leakage current is already lower than with the 6x WIMA 10uF 50V capacitors!
And that is with completely new capacitors without having formatted them on 15V first.
I now measure 18uV over the 10K resistor and the 6x 10uF was 45uV after 12 hours.
Looking good for the first test...

Kind regards,
Bram

PS
I forgot to add the Spice simulation of the low pass filter.
Does anyone know how I can get the graph in LTSpice so that there are more lines so that it is easy to read?
I don't use spice much so I don't have much experience with it yet.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/Spice-LP-01.png)

Title: Re: LM399 based 10 V reference
Post by: iMo on January 01, 2020, 02:35:04 pm
Your R6/7 (or R2 in your sim) should be something like 100k-200k, imho..

Quote
Does anyone know how I can get the graph in LTSpice so that there are more lines so that it is easy to read?

What do you mean by "more lines"?
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 01, 2020, 02:40:21 pm
Hi imo,

R5 and R6 are fo charging the capacitor C3 to almost the full reverence voltage, soo over C2 wil than be only a verry low voltage, therefore also very low leakage current.

Kind regarts,
Bram
Title: Re: LM399 based 10 V reference
Post by: iMo on January 01, 2020, 02:52:24 pm
R5 and R6 are fo charging the capacitor C3 to almost the full reverence voltage, soo over C2 wil than be only a verry low voltage, therefore also very low leakage current.

But the bottom R value impacts the corner freq of the low pass.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 01, 2020, 02:53:39 pm
One can use parameter sweeps to get more curves in one plot. With a command like
.step param par LIST 100 1K 10 K
The parameter par  (use as {par} for values)  gets assigned the 3 values and one gets a curve each. However it is sometimes tricky to see whick curve is which.

With the bootstrapped electrolytic cap, one may reduce the leakage problem, but could still get DA effects. So in the simulations it may be worth to include some DA equivalent circuit: e.g. a capacitor of some 10% the size and series resistor in the high MOhms range. Settling in electrolytic caps may take several days.
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 01, 2020, 03:05:06 pm
imo,

I know that the turn-over point and the steepness influences it is of course always a consideration of what you want and what comes out for your application.
Most of the applications I've seen from others was a resistance ratio of about 10x.
Thanks for showing me the step parameter in LTSpice!

Kleinstein
This picture made by forum user RR shows what I mean, all small steps are also visible in the frequency and amplitude scale.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/Spice-LP-02.png)

Thank you both for the help.

Bram
Title: Re: LM399 based 10 V reference
Post by: iMo on January 01, 2020, 03:17:07 pm
Plot settings -> Grid tick
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 01, 2020, 03:27:50 pm
imo,

YES!

(http://www.bramcam.nl/Diversen/Chopper-Opamps/Spice-LP-03.png)

Thanks
Bram
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 01, 2020, 07:49:12 pm
As imo already commented: For a meaningful application the lower capacitor in this circuit should be much bigger. Unless its RC time constant is larger than that of the filter itself, the circuit won't work as intended. Can't you keep your 60 uF MKS cap for the upper part?

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 01, 2020, 09:39:20 pm
Hi Dieter1 and 3roomlab, :-)

My time for this project is up now, I'll probably have time again coming weekend.

I already played in LTSpice with a larger value of the lower capacitor, the last version the lower capacitor was 1500uF and the top capacitor was 470uF.
That looks good in LTSpice, but I still have to think about whether or not other effects will play a role.

This evening I also tuned the amplifier stage in LTSpice in terms of AC behavior.
I'm already satisfied with that, but I still have to do some checking of the DC settings, if the output from my the last schematic is able to deliver the 10mA I always want to beavailable.
I haven't shown this schematic here yet, that will come at the end of the week, afther I've done some real noise tests with the Audio Precision measuring set.

Thank you guys for al the input.  ;)

Kind regards,
Bram

PS
The 2x 470uF capacitors now only give 5.5uV over the 10 K resistor, so that goes very well for these unformatted parts.
Probably this is a mix of the DA and the formatting.
Title: Re: LM399 based 10 V reference
Post by: iMo on January 01, 2020, 09:49:31 pm
Here is a simple simulation with pink noise (1/f) from a file.
Download pink.wav (various sources) and assign the file as the voltage value of a voltage source (wavefile="C:\Pink.wav" chan=0).
I've added a popcorn too..
PS: the green and red signals look weird, the amplitude is around 1uV, needs a better setting somewhere.
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 02, 2020, 06:52:18 am
Hi,

A small update about the capacitors under test (2x 470uF) de voltage over the 10k resistor dropt to <2uV, that is less than 1ppm deviation on the 10V output voltage.

Later today I will use the DVM "Null" again to see if it has been changed to get a more accurate measurement.

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: maat on January 02, 2020, 02:01:43 pm
Here are the numbers promised, but first the setup. I used a lead acid battery (12.84V) in series with the capacitor and a 2.2 MOhm (2.192 MOhm measured, 1%, 50 ppm/K, Vishay MRS25000C2204FRP00). I *continuously* measured the voltage across the resistor using an HP3458A (~1.5 pA input bias current).

Ambient temperature in the lab is stable to within 0.5 K, DMM temperature to within 1 K, DUT 0.1 K over the course of a day.

Temperature range: 25 °C to 45 °C in 5 K steps. I am currently limited to 45 °C, because of the sensor setup. I might change that in the near future. I could add temperatures down to  15 °C and, if I replaced the sensor, up to 65 °C if there is interest. The temperature of the DUT is usually a little 'off', because the sensor used for the set-point is different from the one attached to the DUT and there is a thermal gradient inside the box.

WIMA MKS2B051001N00JSSD, 10 µF, 50V, 5%, Sample size: 5
Temperarture of DUT in °CLeakage @12.84V in nAInsulation Resistance in GOhm
44.74.123.11
39.52.455.22
34.31.498.59
29.10.8914.38
24.00.5523.27

And when I think of what forum user "maat" explained, that the leakage current of this type of capacitor is quite temperature sensitive, it is better not to use this type of capacitor in this project.

I did by no means imply that these capacitors are bad in any way. They are fantastic and I would say WIMA is one the best brands you can find. They are a few orders of magnitude better than their datasheet says (which is likely due to the timescales involved in measuring significant sample sizes during production).

The problem is, that you are 'magnifying' the issue by a factor of 10k. Using that much gain, will make anything look big in comparison to microvolts.

So lets do a quick back-of-the-envelope calculation:
This will yield you a net tempco of 2.63 ppm/K using 6 capacitors (likely a bit better)!

Regarding the use of a bootstrapped electrolytic cap. This is not a good idea either, because not only does the leakage current change significantly more, but also the  capacitance. The latter will severely diminish the effect of the bootstrapping and get you horrible 1/f noise. I did try that at one point before using a 249/2490 Ohm + 330 µF Nichicon UKL bootstrapped setup and it already got me all kinds of weird drift.

I don't know your application, but in my opinion I would keep the the rc filter somewhere in the lower Hz (only use one capacitor for example), because any DMM you are going to use to compare your ref will low pass filter the result anyway (NPLC=100), so you only need to push the noise down until the DMM filter does the rest. If you want to improve the tempco beyond a single MKS2 and you have the space, use a polypropylene cap (as suggested by 3roomlab). I can recommend the MKP4 series. They come in 10 µF at a space (and price) premium ;).
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 02, 2020, 03:58:26 pm
Usually a DMM will do a lot filtering. The main part to filter is something like the 2.5 Hz range, as a DMM may use 10 PLC sampling the signal and than 10 PLC sampling zero. This would make the DMM slightly sensitive to noise around 2.5 Hz that can be avoided. Ideally the DMM would include such a filter, but only a few meters do, and the filter in the Datron1281 does not look like low enough to suppress 2.5 Hz (if they really use 10 PLC).

Instead of spending a lot on filtering, there would be the alternative of using a 3rd or maybe even 4 th LM399. Of cause the return of more references goes down, so does the gain from a little more filtering.

For multiple references it may be interesting to also get access to the individual voltages and differences, to see if some of them are drifting more than others.
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 02, 2020, 04:29:50 pm
Hi, maat,

Thank you for the information about the WIMA MKS-2 capacitors, it is appreciated.
I am the last one to say that these are bad capacitors, I think they have made a nice product as far as the 10uF 50V capacitors are concerned, I used them many times.

But as always, use the components that fit your application.
The 10uF 50V MKS-2 is just not the right capacitor for this application.

Bootstrap filter
My 10V LAB reference which contains 4x an obsolete LT1021, in an oven of around 42C, drifts extremely little in the last four years.
According to my calibrated 3458A it is still within 2PPM.

The 3458A I have, drift according to the Keysight reports, not more than 1PPM per year.
This LT1021 10V LAB reference also contains a bootstrap filter, but I do not experience any strange behavior.
Maybe that's because I use select capacitors...

I tried to buy this capacitor from Farnell: 330 µF Nichicon UKL, but they don't have this type in stock.
Then I have to wait or see if I have a list for another electronics wholesaler, this because otherwise the order costs will be too high.

At this moment, the test of de new 470uF 16V capacitor is goin wel and it looks like its hit the bottum at 2uV over de 10K resistor.
The other 470uF capacitor are stil forming @ 15,5V, late this evening or tomorrow I wil place in series with every capacitor a 10K resistor to test de Leakage current.

I'm also happy with all the input I'm getting here.
Yesterday I even put the opamp circuit in LTSpice and tuned it and tuned the filter responce.
Without examples and help on this forum, I would never have done this.

I am especially someone from the "Old School" do real measurements because Spice often fools you.
But I've known for a long time that Spice can also take a lot of work out of your hands.
But I always do real measurements, most of what I do with Spice is calculate filters and that almost always goes well.

Sorry for the crooked sentences, I suffer from dyslexia and English is not my native language, that's a lose, lose situation  :-DD

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: iMo on January 02, 2020, 05:27:00 pm
I think the KSFs here are a) to make the oven with say <0.15Kp-p stability, b) to make <1ppm/K "ratio TC" 7->10V divider, and c) to find popcorn free LM399s (difficult). A lot of challenges, indeed  ;)
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 02, 2020, 06:02:23 pm
I don't think there is much hope to find LM399 really free of popcorn noise. It is more about finding some where is popcorn noise is not too high.

The  7 to 10 V divider can be the real challenge: the TC matching may the easier part. The really hard part is to have the ratio long time stable.
Especially with an oven one may not even need good TC matching.
For a possible check one could also consider having the option to get the raw 7 V out, not just the final 10 V.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 03, 2020, 01:55:13 pm
Hello,

fortunately most of the LM399s cannot read the papers (nor data sheets).
Usually the ageing is only the first few weeks and turns to a "random walk" afterwards.
This first ageing is most probably equalizing the tensions of the cement between chip and housing.

attached a ageing curve LM399#2 and #3 are pre-aged devices.
LM399#CH6 and #CH7 are from day zero for the first 2000 days.
As reference 2 LTZ devices (unbuffered) with some "accidents"
(see the sudden shifts of the short cirquits).

All measured against a LTC2400 with temperature compensated AD586.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on January 03, 2020, 04:34:37 pm
@Andreas: my conspiracy theory would be AD/LT is selecting somehow (before a burn-in) the LM399s based on popcorn or 1/f noise level. They sell parts with "LM399" markings to the general public, and the parts marked with "weird numbers" to the selected customers. Guess who gets what..
Title: Re: LM399 based 10 V reference
Post by: guenthert on January 03, 2020, 05:17:11 pm
@Andreas: my conspiracy theory would be AD/LT is selecting somehow (before a burn-in) the LM399s based on popcorn or 1/f noise level (or Vref<7V?). They sell parts with "LM399" markings to the general public, and the parts marked with "weird numbers" to the selected customers. Guess who gets what..
  Not quite and reality is not quite as nefarious.  There are companies who offer a selection service and it stands to reason that AD/LT is one of those.  HP (erm, Keysight) is one of those major customers who pay extra to get units with tighter/guaranteed specifications (and a customized label).  Those units are selected from the general pool.  The open market gets those which don't quite meet those tighter limits, but also - and that might be the overwhelming majority - all those units which haven't been tested at all.  You're free to purchase a set and select the best and sell the rest on Fleebay (which I think some do).
Title: Re: LM399 based 10 V reference
Post by: iMo on January 03, 2020, 05:44:41 pm
@Andreas: my conspiracy theory would be AD/LT is selecting somehow (before a burn-in) the LM399s based on popcorn or 1/f noise level (or Vref<7V?). They sell parts with "LM399" markings to the general public, and the parts marked with "weird numbers" to the selected customers. Guess who gets what..
  Not quite and reality is not quite as nefarious.  There are companies who offer a selection service and it stands to reason that AD/LT is one of those.  HP (erm, Keysight) is one of those major customers who pay extra to get units with tighter/guaranteed specifications (and a customized label).  Those units are selected from the general pool.  The open market gets those which don't quite meet those tighter limits, but also - and that might be the overwhelming majority - all those units which haven't been tested at all.  You're free to purchase a set and select the best and sell the rest on Fleebay (which I think some do).
An lawyer would say "KS is not using an LM399H/AH in their DMMs".. :)
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 03, 2020, 07:54:53 pm
@Andreas: my conspiracy theory would be AD/LT is selecting somehow (before a burn-in) the LM399s based on popcorn or 1/f noise level. They sell parts with "LM399" markings to the general public, and the parts marked with "weird numbers" to the selected customers. Guess who gets what..

sounds familiar to me see:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg854683/#msg854683 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg854683/#msg854683)

so it must be true somehow.   >:D

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 03, 2020, 08:24:53 pm
Not quite and reality is not quite as nefarious.  There are companies who offer a selection service and it stands to reason that AD/LT is one of those.  HP (erm, Keysight) is one of those major customers who pay extra to get units with tighter/guaranteed specifications (and a customized label).  Those units are selected from the general pool.  The open market gets those which don't quite meet those tighter limits, but also - and that might be the overwhelming majority - all those units which haven't been tested at all.  You're free to purchase a set and select the best and sell the rest on Fleebay (which I think some do).

Exactly because the noise and drift tests are very expensive because of the time they take.  So most common devices only had the short functional test.
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 03, 2020, 09:15:28 pm
Hi,

I want to show the oven that I'm going to use for this LM399 reference, this because I already have quite a few test hours in it and that the "gain" is above 150x.
So 15C temperature increase gives less than 0.1C change in the oven.
How good it really gets depends on how well the insulation is done for the oven.
By placing the furnace in a simple polystyrene box, I got more than 150x gain.

I showed this oven before on this forum, and the controller is equipped with a LM723 IC.
And the NTC is located in the 7mm thick bottom under one of the power MOSFets that heat the box.
The extra TO220 IC is a LM35DT, I thought it would be easy to measure the temperature of the oven, but its a bit of a drama temperaturesensor, noise as hell.  :-DD

The oven is made from an already used HF box from which I removed the connectors.
It is a beautiful box for an oven because, of the very thick walls, the bottom plate on which the MOSFets are mounted is 7.5mm thick.

The battery is for the impression of the size of this oven.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-14.png)

.
Here you can see the extra 5K NTC measuring the temperature in the oven, this NTC was connected to a 34461A DMM.
It is clearly visible that I made sure that the wiring is mounted against the oven body.
This has been done to prevent you from making wrong measurements by leaking energy from the connection wiring.
During measuring, the lid was closed and the holes covered with tape.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-15.png)

.
This picture is for the impression how much space there is left for the Reference circuit, the parts around the LT1010 buffer do not come in this box.
That leaves some extra space to place the parts in an optimal way.
This because I'm still thinking about the remarks to leave out the bootstrap capacitor filter and only use the noise reduction of the 4x LM399 parallel.
I can then add more insulation to the LM399 IC in the oven, because no filter capacitors are needed.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-16.png)

.
I also thought about not connecting the oven in the LM399 IC's, but this oven is not good enough for that.
It requires an oven with a gain well above 1000x.
That's why this oven should be placed in a flask, or you could opt for a double oven.
As far as I'm concerned, both fall outside my project objective.

My first tests with the two types of Nichion VR(M) 470uF capacitors are good.
Eventually the 35V versions look good after first forming the capacitors at 15V and after that I started measuring leakage current around the 7V voltage which is the reference level.
One of these 35V types is at less than 1uV at 7V across the 10K resistor series.
So this is not the bootstrap filter but 1 single capacitor!

I've been thinking about "maat" his remarks about the strange drift of the bootstrap filter he tested.
My experience is that it works well but I also came across some weird effects today.
That wasn't the filter, but the cause was the power sources I used for the tests.

Those were two different power supplies, an old Harrison and the Rigol DP832.
Both can't be used properly because of the 1/F noise of these power supplies.

Today I picked up some different types of batteries I ordered from the wholesaler.
I first had to get these batteries to LAB temperature otherwise the drift of these batteries also caused too many errors...

Now the 4x penlite are quieter than my 10V reference PCB with the LT1021 on it.

This is a picture of the measurement on the 16V 470uF bootstrap filter with my old Philips PM3424 microvolt meter.
The DC value over the 10K resistor is less than 2uV.
Oh yeah, of course I regularly check if the "null" level of this meter hasn't changed.

(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-17.png)

And this is for the impression of my 10V Quad LT1021 LAB using a capacitor bootstrap filter.
The PostIt note is from June 2018 and the measured value on the 3458A is from this evening with a well warmed up 3458A and after an Autocal.
(http://www.bramcam.nl/Diversen/Chopper-Opamps/LM399AH-10V-Ref-18.png)

And below a link to the website of Rubicon with clear information about the properties of capacitors.

If you look at the leakage current between 22C and 42C, that difference is only small.
You have much more advantage in properly forming the capacitor and selecting the one with the lowest leakage currents than to worry about maybe 20% extra leakage current at 42C.
(http://www.rubycon.co.jp/en/products/alumi/technote.html#1)

But, uh... still a lot to think about.
For the sake of completeness, I'm not looking for the perfect 10V reference, but try to do the best I can with the stuff I've got.

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: iMo on January 03, 2020, 09:32:12 pm
Does an electrolytic (or tantalum) capacitor need a minimum DC bias in order to work properly (EDIT: to be properly "polarized")? I ask while the top capacitor in the bootstrapped combo sees "0V" best case.
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 03, 2020, 09:37:20 pm
Hi imo,

Actually, I don't know...
I don't know what manufacturers say about this.
Maybe some forum members have more information on this?

Kind regards,
Bram
Title: Re: LM399 based 10 V reference
Post by: David Hess on January 03, 2020, 10:35:27 pm
Does an electrolytic (or tantalum) capacitor need a minimum DC bias in order to work properly (EDIT: to be properly "polarized")? I ask while the top capacitor in the bootstrapped combo sees "0V" best case.

No, electrolytic capacitors work fine with 0 volts of bias.  Noise and leakage depend on the capacitor so they may need to be tested for it.
Title: Re: LM399 based 10 V reference
Post by: blackdog on January 04, 2020, 11:57:25 am
Hi 3roomlab,

Do you want to tell us where the picture of the drift of the LM399 at different temperatures comes from.
Thanks

Bram
Title: Re: LM399 based 10 V reference
Post by: FransW on January 05, 2020, 08:54:12 pm
The drift rate at that temperature of 95°C is nowhere specified, but it is an Arrhenius calculation, according to e.g. P J Spreadbury: 'The Ultra-Zener.. is it a portable replacement for the Weston cell?' , Meas. Si. Technol. 1 (1990).

They demonstrated, that the drift of the LTZ1000 doubles with each 10°C increase of the oven temperature. At 55°C they measure typically -2ppm/year. 95°C would yield a 16 times higher drift.
....
So this mean one can expect about a 2 times higher drift rate for the LTZ1000 A due to the about 10 K higher set point needed. This could be a bit less if good insulation is used around the non A version.

Due to the square law for the heater, half the power needed means about 70% of the current needed. So there is some power saving for the A version possible, but not that much if external insulation is added.  A first point to save power and keep the drift low is choosing an oven temperature that is not too high for the planed use.

Just as a crazy idea: The heat lost from the transistor to drive the heater could in theory also be used to heat the reference - so maybe have the transistor on top of the LTZ instead of somewhere far away. This may need a modified compensation however, to really work at the low power end.

From: https://www.eevblog.com/forum/metrology/ltz1000ch-or-ltz100ach/msg2341581/#msg2341581 (https://www.eevblog.com/forum/metrology/ltz1000ch-or-ltz100ach/msg2341581/#msg2341581)
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 06, 2020, 08:50:46 am
The measurements presented are not that conclusive to use them for extrapolation below 25 °C. Aging at 25 °C may be small because the parts used had been on shelf for years at room temperature (20 .. 30 °C). When you go below room temperature, aging may start anew.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 06, 2020, 09:31:15 am
The normal theory suggests that aging is slower at low temperature. However the data shown for the LTZ refs are not very conclusive.
Another difficulty with aging is that there are different processes / parts that show aging. At a high temperature the faster processes may have reached the final values and one only sees the slow processes, while at low low temperature one would see the effect of the faster processes with the slow part essentially at a stand still. So this can be more of less independent effects, especially after some burn in process.

The LM399 is made with a fixed temperature - if not using the temperature control, one could as well use a different more conventional zener with higher current, like the old 1N829A or the Chinese 2DW232. The main point of the LM399 is that it makes it easy to use a heated zener. That lower temperature slows down aging is no a new finding. So when designing the LM399 they probably had a reason to choose some 95 C and not 70 C or 50 C. If it would be so easy to improve, I would expect a slightly different version with lower temperature setting to be available too.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 06, 2020, 09:35:37 am
Ease of use is also the reason for choosing a temperature near 100 °C, otherwise you need humidity control.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on January 06, 2020, 10:08:41 am
The LM399 and LTZ1000 have different internal wiring. While LTZ is basically a naked zener, the LM399's zener is embedded into a circuit with many transistors and resistors. That may influence the temp vs. aging too.  The zener's current in LM399 is ~0.25mA, afaik.

PS: the LM329 datasheet shows the Ref's schematics identical as the LM399's one.
PPS: Dobkin's paper (https://www.eevblog.com/forum/metrology/poor-man_s-reference/msg1237259/#msg1237259) shows the 30k and 10k resistors are trimmable, it could be they can set the zero TC for 329 and 399 accordingly.
Title: Re: LM399 based 10 V reference
Post by: stijena1973 on January 09, 2020, 04:19:26 pm
I doubt that one can calculate individual noise from pairs.

Indeed you can :) Click :

www.intersil.com/data/an/an177.pdf (http://www.intersil.com/data/an/an177.pdf)

The calculation is not difficult!

The measurement is a pairwise sample DUT noise density over frequency. Because the assumption is that the samples are statistical independend the noise power at sa specific frequency is simply the addition of the noise power of the DUT pair.

If you have n=4 samples you can make n*(n-1)/2=6 pairwise measurements. I will call the reference sample spectral noise power at a specific frequency simply a,b,c,d.

What you measure with the pairs is the sum of the noise powers

a+b
a+c
a+d
b+c
b+d
c+d

You get the noise power of the first sample reference by the following expression:

a=(1/3)*(((a+b)+(a+c)+(a+d))-(1/2)*((b+c)+(b+d)+(c+d)))

a=(1/3)*((3*a+b+c+d)-(1/2)*(2*b+2*c+2*d))

a=(1/3)*((3*a+b+c+d)-(b+c+d))

a=(1/3)*(3*a)

a=a

For the general case n the expression is:

a=1/(n-1)*(sumwith(a)-1/(n-2)*sumnot(a))



The application note from LT does not give a calculation example. It is only stated that you have to use a much better reference as a DUT partner.


This method is called "three-cornered hat" method. Of course it can be four-cornered etc.

 But minimum is three, and pre-condition is that they are statistically indepedent.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 09, 2020, 09:30:18 pm
Calculating the noise of individual reference from measurements in pairs is rather difficult, as the noise is not constant in time  :-DD. So the individual noise amplitudes will scatter quite a bit. So the cornered hat version is not really practical, as the individual measurement will contain quite scattering.

There are usually 2 practical method to check the noise:
One is to amplify the noise with an AC coupled amplifier. So there will be lower frequency cut off, at maybe some 0.1 Hz.

The other method is using a known lower noise reference to subtract. A first approximation could be using the average of several (e.g. 5) references as the point to compare too. Once really bad units are out, one can assume the average is lower than the individual units.
Title: Re: LM399 based 10 V reference
Post by: ch_scr on February 14, 2020, 08:50:57 pm
I have finally built my east-german-themed burn-in rig for my czech LM399-clones. Some western parts sprinkled in as per tradition.
[attach=5]
It is based on a B589(AD589 clone) Ref / OP07 / 1k shunt current sink and one quad NPN array / three six-fold p-fet arrays used as current mirrors to supply 15 current sources tied to one reference.
[attach=4]
The B341 and U105 IC's were made in east germany but there are still some on ebay. I tried my best to have star ground / supply where applicable. Kicad project zip attatched.
Maybe the best thing that has come from this folly is that I have found a socket that really grabs the references. And I mean grab, no voltage fluctuation because the reference does not move after insertion. I believe they are "precision sockets" (like the turned ones but with additional inserts). As they are from the parts bin, I have no name or number, but the pictures of precision sockets look like it. As I tried to match the pitch of the references on the pcb the sockets have to be broken apart, soldered in one-by-one and insulated with heatshrink...
How would you set the measurement cycle on a 34970A? I thought to dwell on one reference(call it #1) for like 8 hours, then switch through all other ref's (#2 to #15), measure voltage there for short amount of time and stop on #15 for the next 8 hours. Now measure #1 - #13 for a short amount and dwell on #14; and so on, to minimize relay actuation while having long and short term data on all references.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 14, 2020, 09:09:59 pm
The current mirrors from discrete transistors usually don't work very well. even for less precision needed, one should add some resistors at the emitter / drain.  Even with transistor arrays the mirrors tend to be not the precision wanted to test references.

The more normal circuit would be to wire one of the LM399 with an OP to create a reasonable stable 10 V and than use resistors from that 10 V level. Even if just an TL431 is used for the 10 V level, this may be more accurate than the current mirrors.
The heater part usually likes to have more than 10 V anyway.

For the reading cycle, I would consider just switching to the next reference at a regular rate, like every 10 or 100 seconds.
It kind of depends on the effort spend in separating the data later and the MUX used. The obvious choice here would be CMOS MUX chips like DG408 (or even old  CD4051), that are good enough for a 7 V level.
Title: Re: LM399 based 10 V reference
Post by: ch_scr on February 14, 2020, 09:24:47 pm
The heaters are supplied by two 7815 . Good point about the emitter / drain resistors. I have some CD4067, that would allow fast cycling without stressing the relays in the 34970. Did not think of that, just took the relays as given.
Edit: Just realised, U105 has the drains connected inside the ic - no resistors there.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 14, 2020, 10:06:15 pm
Hello,

I am using some MAX4051A (with Kelvin sensing) in my LM399 ageing box:

https://www.eevblog.com/forum/metrology/measurements-on-emf-error-of-switches/msg1254204/#msg1254204 (https://www.eevblog.com/forum/metrology/measurements-on-emf-error-of-switches/msg1254204/#msg1254204)

edit: the MAX4051(A) has the advantage that I do not need any level shifters to the 5V logic which would be needed with CD4051/CD4067.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: notfaded1 on February 15, 2020, 02:04:29 am
Calculating the noise of individual reference from measurements in pairs is rather difficult, as the noise is not constant in time  :-DD. So the individual noise amplitudes will scatter quite a bit. So the cornered hat version is not really practical, as the individual measurement will contain quite scattering.

There are usually 2 practical method to check the noise:
One is to amplify the noise with an AC coupled amplifier. So there will be lower frequency cut off, at maybe some 0.1 Hz.

The other method is using a known lower noise reference to subtract. A first approximation could be using the average of several (e.g. 5) references as the point to compare too. Once really bad units are out, one can assume the average is lower than the individual units.
I seem to remember seeing this somewhere on here where someone had 5 LTZ1000 references that were averaged (must be nice).  Is there a multiple LTZ1000 (or even two) reference design pcb on eevblog?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 15, 2020, 10:10:11 am
TiN has a board with 4 x LTZ1000. The circuit is pretty straight forward: separate references and than average with resistors and have a buffer / amplifier after that.  So there is limited need for a special board - just use 2 identical boards and a single buffer / amplifier  (e.g. have one board partially populated).

For the LM399 references one could use a single source for the driving current, so only one amplifier to 10 V needed, that can supply the current to all refs. 

Even with extra emitter resistors, the current mirror type current source will likely no be very accurate or stable. A resistor from a higher voltage (even if just the 15 V from an 7815) is  more stable.
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 08, 2020, 07:28:55 am
Made a few updates to my LM399 design, not tried this version yet
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 17, 2020, 04:57:41 am
That was terrible :) I forgot everything about LM399, much better now, 3 VCC planes, and 3 GND planes

I'm using the connector that the 3458A reference is using (Molex 22172052), the board size is 50mm x 26mm. I'll be using VPG VHD200 for the gain stage.
Title: Re: LM399 based 10 V reference
Post by: KK6IL on May 24, 2020, 11:36:40 pm
R5, 200K looks like too high a value. the LM399 operating current is 0.5-10 mA, so with a 24V supply, R5 should be between 34K  and 1.73K.

The schematic does not show the connection between Z- and GND.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on May 25, 2020, 12:25:12 am
R5, 200K looks like too high a value. the LM399 operating current is 0.5-10 mA, so with a 24V supply, R5 should be between 34K  and 1.73K.

The schematic does not show the connection between Z- and GND.

Look again. R5 is there just to guarantee start-up. R4 supplies the zener's operating current.
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 26, 2020, 01:22:08 am
R5, 200K looks like too high a value. the LM399 operating current is 0.5-10 mA, so with a 24V supply, R5 should be between 34K  and 1.73K.

The schematic does not show the connection between Z- and GND.
Here you go; and an update on the PCB layout
Title: Re: LM399 based 10 V reference
Post by: Jay_Diddy_B on May 26, 2020, 02:52:11 am
Hi niner_007,

Are you aiming to get a 10.0000V reference or around 10V?

The datasheet shows:

[attachimg=3]

and

[attachimg=1]


with your 10K 22K divider you have a nominal gain of 1.4545

So your 10V output could be between

9.81 and 10.6V

Are you going to include a trim circuit?

Here is a picture of Datron 1281 connected to a Fluke 732A with Pomona 5291A low thermal emf patch cords.

They disagree by 4.69ppm  :-DMM

[attachimg=2]

Regards,
Jay_Diddy_B

Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 26, 2020, 03:11:22 am
No, I don’t care about the exact value, I’m not aiming for a perfect 10V
Title: Re: LM399 based 10 V reference
Post by: Jay_Diddy_B on May 26, 2020, 03:31:05 am
Do you have the VPG VHD200 10K / 22K resistor pack?

Jay_Diddy_B
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 26, 2020, 03:46:26 am
I have not made the order with Vishay yet, I'm planning to this week

The final value not yet decided, probably more like 21.3K and 10K
Title: Re: LM399 based 10 V reference
Post by: Jay_Diddy_B on May 26, 2020, 03:52:50 am
Think about LT5400BIMS8E-1#PBF.

You can build a 10K 20K divider with this part. It will give you a gain of 1.5 instead of 1.45

Should be in stock and a lot cheaper.

Regards,
Jay_Diddy_B
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 26, 2020, 04:06:03 am
I have made a design with LT5400 and LM399 before, but I decided against it this time, opting in for through hole parts
Title: Re: LM399 based 10 V reference
Post by: branadic on May 26, 2020, 05:56:59 am
Quote
but I decided against it this time, opting in for through hole parts

Go for TDP1603 10k network then or use a NOMCA1603 network the same way and use a SMD adapter board, to make it throughhole. Still much cheaper.
10k:25k gives you the right 0.4 ratio you want and all resistors within the package can be used to form that ratio.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: The Soulman on May 26, 2020, 08:08:33 am
Quote
but I decided against it this time, opting in for through hole parts

Go for TDP1603 10k network then or use a NOMCA1603 network the same way and use a SMD adapter board, to make it throughhole. Still much cheaper.
10k:25k gives you the right 0.4 ratio you want and all resistors within the package can be used to form that ratio.

-branadic-

Looking at the data sheet these nomca networks may not be good enough?

Quote
TCR: Tracking ± 5 ppm/°C (typical) -55 °C to +125 °C
Stability: RatioΔR ± 0.015 %1000 h at +125 °C
Shelf Life Stability: RatioΔR ± 0.002 %1 year at +25 °C
Title: Re: LM399 based 10 V reference
Post by: iMo on May 26, 2020, 08:48:11 am
I've been using LT5400 as the 1.5 multiplier. The LT5400 is 0.2ppm/K (AD told me it is rather <1ppm/K) TC of ratio tracking. With +/-5ppm/K with NOMCA you may get similar tracking when lucky (more resistors in the divider may average out the TC of the divider's tracking ratio).
Title: Re: LM399 based 10 V reference
Post by: branadic on May 26, 2020, 09:19:29 am
Quote
Looking at the data sheet these nomca networks may not be good enough?

Please refer to the thread Statistical arrays (https://www.eevblog.com/forum/metrology/statistical-arrays/msg2026729/#msg2026729), they behave much better than the datasheets indicates.

There is a lot of knowledge on this board, that is worth checking.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 27, 2020, 03:42:08 pm
I ordered the PCB with oshpark, I will share the project and gerbers after I tested it.

Just to validate my understanding of the resistors in the LM399 circuit, I understand the ratio setting ones very well. My understanding is that the stability of the 200K resistor is irrelevant, and that the 5K current setting resistor has very little impact. Based on the LM399 resistor thread. Is this correct?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on May 27, 2020, 03:55:23 pm
I ordered the PCB with oshpark, I will share the project and gerbers after I tested it.

Just to validate my understanding of the resistors in the LM399 circuit, I understand the ratio setting ones very well. My understanding is that the stability of the 200K resistor is irrelevant, and that the 5K current setting resistor has very little impact. Based on the LM399 resistor thread. Is this correct?
That is correct. The actual zener current is set inside the LM399 and due to the regulation loop inside the output impedance is quite low. So the current setting resistor is less important (some factor of 10, maybe more) than with a zener direct.
Title: Re: LM399 based 10 V reference
Post by: niner_007 on May 28, 2020, 08:55:30 pm
Quote
Looking at the data sheet these nomca networks may not be good enough?

Please refer to the thread Statistical arrays (https://www.eevblog.com/forum/metrology/statistical-arrays/msg2026729/#msg2026729), they behave much better than the datasheets indicates.

There is a lot of knowledge on this board, that is worth checking.

-branadic-
Price is of no issue, that's why I was leaning towards VHD200, I'm not planning on selling any of these, they are for me, and me alone
Title: Re: LM399 based 10 V reference
Post by: KK6IL on July 08, 2020, 07:00:16 am

I just bought a 34401 from Craigslist and the seller threw in a free 3468. They both agree with my pair of 3456's that my DIY LM399 standard is outputting around 9.998. Measured 9.9991 on a trio of good but long out of cal meters a year ago. So the time has come for me to build a proper standard using what I've learned on this thread.

Attached is my design for everyone to shoot holes in.  Component selection somewhat driven by contents of my junkbox. 3K 0.01% divider resistors likely from same production batch but I should be able to verify TC tracking by putting 4 at a time as a whetstone bridge and watching null as I vary temp.

I'm considering placing the LM399's upside down on the vectorboard with RTV and wiring all 4 leads with 30 ga wire to avoid mechanical stress, to slot or not slot, etc. Would that likely help or just add more thermocouples?

Main use would be to calibrate my meters, and to drive a KVD, but no reason not to make it as noise free as practical.

Bought a 3455 years ago for $20 if I promised not to return it if it didn't work. Well it didn't, most front panel switches bad as well as unknown problems after that. Saw a thread on the 3455 reference module, so removing that and making it into a calibrator also tempting.

John
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on July 08, 2020, 07:31:16 am
The capacitor C3 is wrong - just leave it out, or much smaller (e.g. 10 pF range). If more filtering is wanted, increase C10 instead and maybe add one more RC stage there.

The LH0033 is kind of overkill there.
Title: Re: LM399 based 10 V reference
Post by: branadic on July 08, 2020, 08:04:04 am
I'm missing some caps on the +22V rail. I would also add some 100nF foil capacitors across the zener diodes.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: KK6IL on July 08, 2020, 10:39:40 am

The LH0033 is indeed overkill but I have 1 or 2 dozen of them. Interesting part with their 10^10 input impedance and high speed. Remains to be seen if I can put a 10 Hz low pass filter in the loop without creating a high power oscillator. If it doesn't work, I'll remove it. I can't get a lot of noise reduction from C10 without getting into electrolytics and their leakage and DA. I'm a digital/FPGA designer, not so much an analog guy. Never did understand Bode plots.
Title: Re: LM399 based 10 V reference
Post by: Andreas on July 08, 2020, 07:15:27 pm

I'm considering placing the LM399's upside down

but no reason not to make it as noise free as practical.

Hello John,

Upside down for a LM399 means: maximum power consumption. (around +10% against normal)
And most likely: higher thermocouple voltage at the soldering joints.

I also hope that you select the LM399 for minimum noise. There is a large stray from device to device.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Inari on January 10, 2021, 05:58:40 pm
Hello Mickle T.

What is that "вольтметр постоянного тока" in one of your pictures? I know what it is but seems very interresting device - handmade? Do you share any information about this nice looking instrument?
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 14, 2021, 06:20:15 am
Just looking through the schematic posted by John/KK6IL, i noticed the OpAmp power supplies pin 4 resp. 5/6 are connected to reference Gnd, while they should be on supply Gnd. The LH0033 datasheet does not contain the term "noise", maybe it doesn't have any..

Concerning the remark of Branadic about "capacitors parallel to the zeners": Those aren't exactly zener diodes, but composite circuits (schematic in datasheet). Don't know if there is experience here about phase margins to observe.

What was the result about the filter loop stability? I think it might be stable since the time constants of the divider with C4 and R13 with C3 are completely different. The advantage of the circuit as drawn is that the filter error vanishes inside the reference control loop. If C10 causes oscillation during power-on one could drive R1..R3 from the output of U4.

Also i'd like to ask for the actual voltage differences between the three references. I think datasheet says like +/- 100 mV or so.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on January 19, 2021, 11:21:17 pm
@KK6IL: Afaik we had a discussion on 723's noise and stability. Probably not important here too much, but especially the "new parts" drifts..
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 20, 2021, 12:13:32 pm
Last week i made a board with 5x LM399 fresh from rscomponents and with averaging. Individual  voltages are within +/- 15 mV from the average. I used 274R resistors for averaging. Then i needed a day or so for making and testing a proper bipolar power supply with guard and the like. Initial heater currents sum up.. Once hot, the whole board runs at 50 to 60 mA at +/- 10 V = 1 W, mainly for the heaters.

After that i started a log using one of our HP3456A. This test is clearly limited by the 10 uV resolution of the HP3456A. Anyway i observed that the average reference voltage started at 7.01531 V, went to 7.01532 V after 3 minutes, to 7.01533 after 10 minutes. 7.01534 V appeared for the first time after 80 minutes. Since then it seems to sit at 7.015336 - with a large majority of samples at 7.01534. So initial drift after startup has been some 3 or 4 ppm. As we know there may be more to come during the first year or so.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on January 21, 2021, 11:27:45 am
@dietert1: a pcb or proto?
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 21, 2021, 01:59:16 pm
It's a small raster board. I wanted something low noise for characterizing a JVR. May serve as an exercise for making a 5x LTZ1000 board if the LM399s are still to noisy. Or i can draw a board and make some more of them. If i remember right one LTZ1000 is about the price of 5x LM399s,

Regards, Dieter

PS: After another 24 hours the log contains 30 000x 100PLC measurements.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 21, 2021, 03:27:06 pm
The LM399 can be relatively noisy, though the noise level can vary quite a bit be tween units.  Not sure how typical the 2 units are that I have tested.

Attached is a reading of a LM399 relative to another LM399 at the ADC. So the curve is essentially the noise of 2 x LM399 as a ratio or difference. For comparison the blue curve is a JRV type reference measured in fast sequence (1 PLC with JRV, 1 PLC LM399A, 1 PLC zero, 1 PLC ADC ref=LM399) the points are average over 20 such cycles.  One can see quite a bit of popcon noise for the green curve (2xLM399) and much less for the blue one (LM399+JRV).  So the large jumps all come from one of the LM399 (the other one still shows a few similar ones, but very rare)  The JRV reference is quite a lower noise, but more drift (likely from temperature).

So to test the JRV (2N4391 + resistor+ bootstrapped drainvoltage), the LM399 is OK for temperature drift and may be good enough for the long term drift (the data from Alex suggest very low dirft, buit this was just 1 unit), but not good enough for short time ( < 1 hour) noise. Even 5 x LM399 would still be relatively high noise.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 21, 2021, 03:45:13 pm
Yes, appears we had the same idea: For the beginning some LM399 should be good enough to study the JVR. In your diagram the LM399 comparison shows about 2 uV p2p noise aside from the jumps. That would mean a LM399 could be 1.4 uV p2p and the average of 5x LM399 would be about 0.6 uV p2p.
Except i still have to make a good 7 -> 10 V gain stage for the LM399 board and some OPA189 preamplifier to get better resolution. In fact i wanted to make more preamplifiers with a multiplexer to get high resolution differences between the LM399s as well.

In this thread i just wanted to provide my numbers about initial drift of LM399.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on January 22, 2021, 02:58:04 pm
Well, how to discipline the jvr reference with help of the lm399?
Something like "399DJVR" ultimate reference?
The relationship between a GPS and an OCXO is almost the same as the relationship of the lm399 to the jvr..
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 22, 2021, 04:31:54 pm
There is a similarity in the stability on different time scales. However I very much hope the drift seen with the JRV ref. is mainly temperature. So with an oven to stabilize the temperature, there is the hope that one does not need another ref. for the long term stability. This is a bit off topic here and more goes to the JRV thread.

The main point in the graph is that the noise of the LM399 can be quite different between units. The 2 LM399 involved both show jumps of some 4 µV, but one unit with about 1 jump per minute and the other unit with about 1 jump in 1-2 hours (no jump visible in the graph, but a few in other similar curves).  It is still a question which of the 2 refs is better:  The more jumpy one looks more noisy, but after averaging over a longer time this would average out, or if one could identify the levls would even takt the 2 levels separate. At least for this sample there were only the 2 levels visible, no obvious 3rd level. The one with only few jumps still has the 4 µV uncertainty, with sometimes very long between jumps, so not much averaging in a reasonable time.

There are a few DMM (e.g. Keithley 2010, 2182, 2001) that use 2 references: a LM399 for the long term stability and a likely lower noise and higher current zener for the shorter time directly at the ADC. The corss over is digital and may be mainly because of ADC gain, not referece noice. It is in software, with not many details known.

It could be rather difficult to do the corss over / combination in the pure analog domain - this would be similar to low pass filtering and this is not easy and could get away without the 2nd ref.. Digital filtering would need a 2 nd ref. and may be possible also for a longer time constant (e.g. hours). I would build it in a way to correct the long time drift of the drifty ref. by looking at the amplied difference. So the ADC / DAC resolution would not need to be very high, if the 2 refs are similar voltage. 
Title: Re: LM399 based 10 V reference
Post by: qosch on January 29, 2021, 10:04:01 pm
After reading in this thread a few evenings I now registered because I still have a few questions.
First I am interested what people refer to when talking of the "Branadic PCB". I've read plenty of very informative posts by Branadic but could not find schematics or a layout posted by him.
Also I have an idea that I am surprised I have not read about yet: putting two LM399 in series instead of in parallel. Maybe I am missing something obvious but to me it seems that this solution is simpler because it just needs one buffer instead of N+1 (N=number of LM399), the relative noise and offset of the output buffer is lower (because the signal is larger) and a DAC could be connected to generate 10V without any resistive divider (that adds more drift and noise). Of course the supply voltages would need to be higher and the current source becomes more critical because it is a source of correlated noise and drift.
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 30, 2021, 07:49:13 am
Hello,

if its in the context of the 10V PWM DAC the cirquit is here:
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2082496/#msg2082496 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2082496/#msg2082496)

Of course you can series connect (to a limited number of zeners) the LM399 zeners.

There are 2 constraints since the heater and the zener are not independant of each other.
- the maximum voltage rating between the pins
- you should never forward bias the substrate diodes.

But many DMMs have their best accuracy range around 10V and only some up to 20V.
So a higher reference voltage will be of limited use.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 30, 2021, 08:18:56 am
Can use a series connection if the heater voltage is in the right range with 2 x lm399 this is usually OK.

However scaling from 14 to 10 V is about as difficult as scaling from 7 V to 10 V. It would be only with 3 in series that going from 21 V to 10 V may be a little easier in some ways or assuming an advantage from 2 identical resistor.

A DAC (e.g. PWM type) can be in the forward path (than 14 V is the better starting point), or in the fedback path (than 7 V is the better starting point). So it depends.

Noise wise the LM399 is normally quite a bit higher noise than a low cost precision OP like OP07. So there is not much difference between series and parallel connection.
Title: Re: LM399 based 10 V reference
Post by: r6502 on February 12, 2021, 11:52:03 pm
Hello all,

I just designed a adjustable 10V reference as a replacement for the reference circuit inside of my "DIAL - A - SOURCE" "model "DAS-56A", based on a LM399H reference and some Vishay precision resistors. The DAS-56A is a calibrator is based on a 6 digit KV divider, that is supplied by a 10V reference voltage.

The original reference circuit of the DAS-56tA is based on 2 zener reference diodes, that have normally around 6.2V each. They are in series and the 10V ouput voltage is set by several precision resistors and a trim pot to 10V. This 10V is used as input voltage for the KV divider.

In my device one of the reference diodes has now only around 5V, so there is the need to do something. Instead of replacing the defect z-diode, i decided to build a new 10V reference circuit based on the LM399H. The resistors have been chosen based on availability at the distributors (actual: Mouser and Bürkin).
 
So I saw this thread, and looked a bit, not at all posts but at a lot and that I've drawn 1st the schematic and than the PCB, that should fit in the original shielded compartment for the reference voltage circuit. The dimensions of the original PCB are about 43.2mm x 40.6mm. I increased the size a bit to 45mm x 45mm.

I'm not 100% sure, what OP to use. In the schematic I pointed out to an good old LT1001 that will do the job I think.

There is also a thread for the DAS56A:
https://www.eevblog.com/forum/metrology/general-resistance-das56a/ (https://www.eevblog.com/forum/metrology/general-resistance-das56a/)


Please take a look  and place your comments on my post. I atached Photos and the schematic ...



 
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 13, 2021, 09:45:16 am
Chances are the LM399 may be a little more noisy than the 2 old zeners. However the actual noise varies a lot between individual units.  Reference zener diodes like 1N827 are still available (though relatively expensive) - easier to get than a LM399 these days.

The LT1001 would be good enough, so would be the good old OP07.

The footprint for the resistors looks a little odd. The trim circuit can be build in a way that only 2 really critical resistors are needed - the other resistor (e.g. those to set the zener current) are far less critical. Different from the old zeners there is no need to accurately trim the reference current to a specific value.  So a single resistor should be good enough.

The vertical mounting of the resistors like on the old board is normally not a good idea for a precision circuit: The ends of the resistors see different temperatures and one can thus get some thermal EMF. In addition the distance from the solder joint to the resistor is quite short at one end.

Depending on the connections to the rest of the circuit, it may be a good idea to have a separate ground connection for the LM399 heater, as the current can vary.

The capacitor directly in parallel to the zener output is a little dubious. It is sometimes seen in other circuits, but a capacitor at the output does not really help much with a regulator circuit.  The LM399 has a shunt regulator circuit, not just a bare zener. Some filtering may be good, but more with a little series resistance. A capacitor with added series resistance can help a voltage regulator. For some odd reason  :-// this point is still not very well investigated for the LM399 - ideally I would have expected a note on the capacitive load, like with the TL431 or similar.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on February 13, 2021, 02:12:40 pm
In my board i have a 51R resistor in series to C2 and the amplifier input. Plus some protection diodes for the LM399 against reverse voltage and against overvoltage. I left the 4 legs as long as possible and carefully formed some bends into them in order to reduce stress on the hermetic seals. Landing pattern isn't a 1/10 " but a 3/10" square.
Yes R1/R2 are not critical, i am using one good metal film (2.87K 0.1 % 15 ppm/K). While a solution with a potentiometer is simple and correct, i would prefer tuning with fixed resistors and omit R4, R5 and R6 but double up R7 instead.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on February 13, 2021, 02:29:58 pm
You may try to simulate the dynamic response of the 399 to the output capacitance. Below a model I put in LTspice (the TC characteristic is an example only as you need to know the actual parameters of the zener and other components on the chip, the schematics comes from the datasheet).
Title: Re: LM399 based 10 V reference
Post by: mawyatt on February 13, 2021, 03:15:05 pm
You may try to simulate the dynamic response of the 399 to the output capacitance. Below a model I put in LTspice (the TC characteristic is an example only as you need to know the actual parameters of the zener and other components on the chip, the schematics comes from the datasheet).

Interesting!! Here's link to some earlier LM399 LTSpice simulations we did back in 5/1/2020 to get a "feel" for how the LM399 behaved.

https://groups.io/g/LTspice/topic/73959871#120404

https://groups.io/g/LTspice/files/z_groups.io/Files%20sorted%20by%20message%20number/msg_120195/Voltage%20Ref%20V1.asc

https://groups.io/g/LTspice/album?id=246490

Best,

Edit: Forgot to add the schematic from June 2020 :-[

Here are 3 references in the schematic to allow comparisons, one is simple Zener with 2N3904 as a diode powered from VCC (+15V), other is LM399 model powered from VCC, and one with LM399 model & Op-Amp to produce +10.000V Output. Also a couple test circuits with the LM399 and other references (unpopulated LTZ1000).

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 13, 2021, 04:34:30 pm
Thanks for the spice circuit of the LM399.  To make it work with my installation I changed the transistors to 2N3906 and used a different zener. In addition R6 should be larger, more like the original 2 K. Otherwise much of the circuit is inactive or it needs a way higher current.

The AC impedance than looks quite similar to the data-sheet. Not very surprisingly the 100 nF cap is not really helping the regulator, more like adding a resonance at some 50 kHz.  50 Ohms and 100 nF in series on the other side look good.  For EMI suppresion less than 10 nF would be sufficient.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on February 13, 2021, 04:55:26 pm
The 50 ohm resistor and the capacitor should not be in the supply path from R1/R2. C2 moves towards the input of the amplifier.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on February 13, 2021, 07:14:31 pm
@Kleinstein: The model v5 with my own zener diode.. The current via zener set to aprox 250uA (as it is in reality afaik)..
The first line in the zener's model are the temperature coefficients you may play with..
PS: could you show me your model?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 13, 2021, 09:25:36 pm
Attached is my model variation with the setup to simulate the dynamic impedance with a parallel capacitance. The main changes are the zener ( just a type that comes as standard), R6 = 2 K as in the data-sheet and R2 reduced to 10 K to get an AC impedance closer to the DS curve. The the PNPs are 2N3906 ( instead of the less common 3905).
The zener details have quite some effect on the temperature curve, but this is not the relevant part here.


With an AC current source of strength 1 the resulting ouput voltage directly corresponds to the impedance. 0 dB corresponds to 1 Ohm and 20 dB corresponds to 10 Ohms.
Title: Re: LM399 based 10 V reference
Post by: r6502 on February 14, 2021, 10:40:56 pm
Hello All,

When I designed the schematic, my focus was, to get a configuration, that will be adjustable to 10V, regardless of the tolerances of the LM399 zener voltage and the tolerance of the available resistors. The zener voltage of the LM399 is specified to 6,95V +/- 2%. I made a excel sheet, where I calculated several resistor values, that will be available on stock at a distributor. Than I came to the conclusion, that with the values, represented in my schematic, I will be able to adjust the reference to 10.000V. I made an annotation of the schematic, that's why the parts may have different IDs compared of my last posing.

In the beginning I had thought, to bye a new set of 1N827 zener diodes, but when I looked at the schematic of the original reference circuit, there are several resistors marked with "SAF" (Set At Factory) to be able to settle the tolerances of the of the Zener diodes and to set the output voltage to really 10.0000V. Due to the fact, that the resistors with low TK are relatively expensive and I do not have a stock of those, especial many different values,  I decided to replace the original reference with one based on LM399.

The fixed resistors I used for my schematic are Vishay 102 series resistors with TK of 2ppm°C. The trimpot is also a Vishay (series 1202) part with a low TK of 10ppm and based on metal foil. The cosponsoring TK of the resulting resistor 2x5k parallel to the trim pot will be lower then the TK of the trim Pot but higher the the TK of the two fixed resistors.

The simulation with the C parallel to to the LM399 zener is very interesting. I did not really thought about this parallel C, it is just like something, that I always place a C in parallel to a zener diode or similar circuits. I'm using Altium - is it possible, to use the Model you posted  also in Altium, or do I have to install LT spice for this?

@Kleinstein:
I made a photo of the used resistors from Vishay - see below. Here is also a link to this series of foil resistors:
http://www.vishaypg.com/foil-resistors/list/product-63001/ (http://www.vishaypg.com/foil-resistors/list/product-63001/). The spacing between the pads is 3,81mm ( 0,15inch) . The heater of the LM399 gets its own tacks of course - have a look at the PCB screen shot.

So have a nice week Guido
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 15, 2021, 08:44:22 am
The simulation of the LM399 internal circuit is rather basic, using kind of normal transistor models and also the zener model is just a random pic of what is available. For the AC simulation this should be reasonable OK, for the temperature dependence this could be off quite a bit. As no special features are used a different spice variant should be OK for such a simulation too. 

The main point from the AC simulation is that just a 100 nF cap does not help at low frequencies, more like makes things worse at a round 50 kHz and only help above some 100 kHz.

For filtering the solution from Dietert looks good and reasonable.
The amplifier part could also have the usual capacitor in the feedback path (e.g. about parallel to R3).

In the  7 to 10 V scaling circuit not all the resistors are equally critical, one could get away with less of the good BMF resistors. Other resistors usually have a different footprint. Especially R1 and R2 are not that critical with the LM399. The adjustment range can be quite a bit smaller, so a lower value for R5, R6 and than R5,R6 are less critical so that a single more normal resistor (e.g. good thin film like PTF56 series) would be sufficient. If a 2nd optional resistor is preferred for coarse adjustment is wanted this would be something like an optional one (likely 100-400 K range) in parallel to R7 and / or R3.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 16, 2021, 07:58:49 am
.. Interesting!! Here's link to some earlier LM399 LTSpice simulations we did back in 5/1/2020 to get a "feel" for how the LM399 behaved.
FYI - here are the LM399 and MAC199 die shots with an analysis what is there on the chip. A good simulation model would be nice to have! :)
https://www.richis-lab.de/REF02.htm (https://www.richis-lab.de/REF02.htm)
https://www.richis-lab.de/REF02a.htm (https://www.richis-lab.de/REF02a.htm)
Title: Re: LM399 based 10 V reference
Post by: mawyatt on February 16, 2021, 02:53:35 pm
.. Interesting!! Here's link to some earlier LM399 LTSpice simulations we did back in 5/1/2020 to get a "feel" for how the LM399 behaved.
FYI - here are the LM399 and MAC199 die shots with an analysis what is there on the chip. A good simulation model would be nice to have! :)
https://www.richis-lab.de/REF02.htm (https://www.richis-lab.de/REF02.htm)
https://www.richis-lab.de/REF02a.htm (https://www.richis-lab.de/REF02a.htm)

Thanks, we had seen this work by Noopy and it helped develop the spice models we created. It's interesting to see the differences in these chips and trying to understand how they behave.

Best,
Title: Re: LM399 based 10 V reference
Post by: dietert1 on May 10, 2021, 07:30:51 pm
Meanwhile i have been running the 5x LM399 board with a PWM gain stage 7 => 10 V for more than 2000 hours and initial drift stopped. From the data at 1500 h and 2250 h less than 1 ppm/year is expected. Recently i used some epoxy glue to fix the five thermal protectors, didn't make a big difference.
This is a ratio measurement where my DIY reference runs against the built in reference of a HP 3456A. Soon i hope to finish a setup with a low thermal scanner and difference measurements (better resolution).

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: branadic on May 11, 2021, 06:44:45 pm
Would be nice to share not only results, but schematic and the like, a good base for a valuable discussion.  :-+
What about (switching) noise, t.c. of your solution?

-branadic-
Title: Re: LM399 based 10 V reference
Post by: dietert1 on May 11, 2021, 08:19:19 pm
Since i learned a lot from the previous PWM discussion in this thread, here are some schematics i made when working on this. Hope this info serves the experts, since i won't have time to contribute much more.
The first schematic represents the circuit i am still using (log above). You can see a second order filter and ripple cancellation. Other details are a resynch Flip-Flop, separate power and supply grounds, a 7 V guard for the filter circuitry and circuitry for operating the filter caps near zero voltage. This is based on biasing large electrolytic caps. This circuit is very slow and useless for a programmable calibrator.
I built ripple cancellation as a mod that would allow me to reduce PWM frequency to 1 KHz, yet in fact i did not modify the MSP430, so it still runs at 10 KHz. The second schematic shows a study of a low resistance output stage that would serve another factor 10 or more of precision. It is much better than any integrated switch i know of: Switching within some nsec and less than 1 Ohm residual resistance. One day in the future that could replace the main PWM output stage. Ripple cancellation can remain with the IC switch.
Don't have a more detailed schematic of the reference board, but an image. The LM399 thermal protection caps are glued to the yellow 100 nF film caps.
The log i showed is with the reference board and the PWM on a table, wrapped in a towel, no metal enclosure, no additional oven. Since we have a 2 or 3 °C day/night temperature variation the log demonstrates that the TC of the whole circuit has less than 0.5 ppm/°C. This isn't a determination, just an upper limit. Each blue data point in the log represents 500 measurements with NPLC=100 and it looks like those are quiet on a fractional ppm scale. The 1 ppm steps are definitely present.

Regards, Dieter

PS: My PWM prototype has an offset of 11 ppm (max of log). The PWM ratio remained unchanged after implementation. In my opinion a more perfect PWM divider would be more accurate, i mean without tuning or "calibrating" it. Accuracy depends on the symmetry of the PWM switches: resistance and/or timing difference. At 10 KHz a 1 nsec timing asymmetry makes 10 ppm. IC muxes have delays of about 50 nsec that can easily contribute an asymmetry of 1 nsec.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on May 19, 2021, 06:28:49 pm
Today i took an image of the PWM board. At the top left corner it has a n-channel JFET to start up the references, J2 as shown here https://www.eevblog.com/forum/metrology/building-your-own-voltage-reference-the-jvr/msg1343014/#msg1343014 (https://www.eevblog.com/forum/metrology/building-your-own-voltage-reference-the-jvr/msg1343014/#msg1343014). This was missing in my schematic.
In the lower right corner there is a 3.3 V regulator for the MSP430 module. The mezzanine on that module is the sync Flip-Flop. The second PWM switch i added later for ripple compensation is hidden under the processor module. The large filter capacitors are what i found (spares from speaker xover revision). Maybe 100 V film capacitors are completely sufficient for low leakage and cleaning the board would be more important.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: dietert1 on July 09, 2021, 04:26:01 pm
Looking at this image i noticed a wiring error. The second PWM DAC (ripple cancellation) got wired to flip flop pin 10 instead of pin 8. Pin 8 remained unconnected. In effect ripple cancellation was implemented but remained inactive. Some days ago i corrected that mistake. PWM output shifted by -2 +/- 0.5 ppm and noise is lower now. Right now the HP 3456A used for logging delivered more than 18000 consecutive readings with no other results than 10.00000 and 10.00001 V. Need to improve logging for better resolution.

Some months ago i got a batch of 10x LM399AH and made two more 5x LM399 modules like the one shown above. When i started fitting them into an enclosure i got the idea to change the circuit into 5x 14 V instead of 10x 7 V. Then the PWM DAC divides the 14 V average down to 10 V. The DAC will no longer be in a control loop, but just feed an output buffer. And the reference will have programmable output voltage proportional to PWM ratio. I already got the ten heaters and the 5x 14 V pairs working. Power consumption of the ten heaters is about 1.8 W but there will be more insulation. This concept can be very cost efficient, since a TEC controller isn't needed.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on July 09, 2021, 05:01:12 pm
Starting from some 14 V (2 reference in series) and go down with a PWM divider makes some sense, especially with a variable PWM ratio, to also get voltages different from 10 V.  A similar configuration is used with some of the fluke calibrators.
Driving the 14 V reference may need an auxiliary voltage a little higher than 15 V, so likely more than 15 V supply needed.

With multiple LM399 in parallel it may be worth having a way to measure the individual reference too, to identify more drifty units and also the higher / lower noise ones. The noise can be quite different between units and a single noisy unit would effect the average quite a bit. Ideally a more noisy unit would get a larger resistor to contribut (R ~ En²) to get the lowest noise for the average.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on July 09, 2021, 05:43:34 pm
Yes, the supply is completely different now. The heaters are running from 19 V. The same voltage is used to drive the dual n-channel mosfet for the DAC output stage. And there is a 1.2x amplifier to derive 17 V from the 14 V reference voltage. This amplifier isn't very critical, deviations get attenuated by about a factor 3000 (2 * 0,5 Ohm / 3 KOhm).
Then there is a buffer between reference voltage and PWM output stage. So, together with output buffer of the PWM filter this time there are two opamps involved with their errors. The output buffer may need bootstrapping for perfect linearity, don't know yet.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: TiN on July 10, 2021, 12:16:38 am
Was the ref powered all the time with no interruptions? Drift can't be estimated easily with just few points for not permanently on zeners.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on July 10, 2021, 05:48:27 am
Yes, after there was some interest in this PWM reference i kept it logging continuously. Will show results later.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: TiN on August 02, 2021, 06:04:25 am
Tested LM399 out of Keithley 2001 for operation at sub-zero temperatures. Perhaps this is interesting for someone other than me.  :-DMM

Device is SL40057 IC-257 (custom Keithley P/N for selected LM399) date code 30 week 1991. Heater part of LM399 was not connected to anything, and "zener" part biased with +24V from Keysight E36312A and 11.1kOhm BMF resistor (static at ambient temperature). Output (kelvin-connected) from zener circuit was monitored by trusty Keithley 2002 DMM.

Run took about 1 hour, and temperature of GM cooler cold-tip with attached LM399 (without plastic casing) was reduced from ambient +27°C to -262°C.

(https://xdevs.com/doc/xDevs.com/cryotest/day3/gmcryo_lm399_cooldown.png)

Reference stopped working at around -185°C or so, with current drawn from PSU reduced down to 770uA (vs 1.45mA nominal) and output voltage going over 8V, all the way until 15.4V.  :bullshit:

Crude logging equipment setup:

(https://xdevs.com/doc/xDevs.com/cryotest/day3/lm399_cryotest.jpg)

I guess there is too much going on inside LM399 to operate this much far out of specified 0°C  :-DD
Next specimen is obvious LTZ1000...
Title: Re: LM399 based 10 V reference
Post by: dietert1 on August 02, 2021, 08:09:57 am
The curves don't show a sudden failure. Can you say to which degree this test was destructive for the poor LM399?
Title: Re: LM399 based 10 V reference
Post by: saturnin on August 02, 2021, 07:21:34 pm
LM399's output voltage is out of specs at room temperature already (the datasheet says its max reference voltage is 7.3 V). Has it been tortured in another experiment?

I also miss what the purpose of the whole exercise is. It is well known fact that silicon devices generally do not work well at deep cryogenic temperatures (BJTs are worse than MOSes in this respect). In presented case extremely low temperatures were reached (~ 10 K). No surprise LM399 ceased to work at some point...
Title: Re: LM399 based 10 V reference
Post by: ramon on August 03, 2021, 02:33:39 am
Just for testing, I guess.

His eyes looked around the table to find which creature can be sacrificed for science. And he found that LM399. Testing a resistor noise could had been more interesting .. or boring (who knows!). But just for testing the setup I think that it doesn't matter which device you choose.

The final goal might be to find the ultimate VOLT reference.
Title: Re: LM399 based 10 V reference
Post by: FransW on August 03, 2021, 11:17:19 am
A reminder:

Absolute measurements do NOT exist.

Frans
Title: Re: LM399 based 10 V reference
Post by: razvan784 on August 03, 2021, 02:41:41 pm
Absolute measurements do NOT exist.
Number of people inside a building.
Number of molecules inside a vessel.
The imbalance in electric charge between the plates of a capacitor, expressed as a number of electrons.
Should I go on?
Title: Re: LM399 based 10 V reference
Post by: donlisms on August 05, 2021, 03:11:23 am
Are counting and measuring the same thing?
Title: Re: LM399 based 10 V reference
Post by: IanJ on August 06, 2021, 10:47:43 am
Jings!......thats a chart you don't see very often!.......I can feel the cold from here!

Ian.
Title: Re: LM399 based 10 V reference
Post by: MegaVolt on September 10, 2021, 02:10:28 pm
This is the first version schematic.
(http://www.bramcam.nl/NA/Cap-DA-Leak-Tester/Cap-DA-Leak-Tester-01.png)
What program allows you to draw such beautiful circuit?
Title: Re: LM399 based 10 V reference
Post by: snik on September 10, 2021, 02:21:08 pm
It could be sPlan.

https://www.electronic-software-shop.com/lng/en/elektronik-software/splan-70.html?type=N&opt=%7B11%7D22&language=en (https://www.electronic-software-shop.com/lng/en/elektronik-software/splan-70.html?type=N&opt=%7B11%7D22&language=en)
Title: Re: LM399 based 10 V reference
Post by: voltsandjolts on September 10, 2021, 02:47:28 pm
Yes, could well be sPlan.
But inkscape can also do similar nice diagrams. Once you have a few basic parts/blocks, it's surprisingly quick.

How about a +200C run for the LM399? >:D
Title: Re: LM399 based 10 V reference
Post by: iMo on September 24, 2021, 11:55:34 am
After I collected and burn-in a bunch of 399s I am thinking to design a small 10x10cm large board for 4xLM399, LT5400 + OP07/2xNPN.
Here is the schematics - it is a classic design, what would be great is a feedback from experts here
a) on resistor's values (and types, TC)
b) capacitor's values and types, and
c) the layout - here the rectangles depict the star points, every wire will be a separate track on the pcb.
The power supply will be off this board.

PS: on the first glance the 100nF at the 399 zener should be modified as the ADR1399 has got some specific reqs for the stability..
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 24, 2021, 12:54:58 pm
C14 and L2 don't combine that well. I would tend to leave out C14. For the supply I would consider a common mode choke, if really externally. Some Protection diodes would also be nice.

I am not a big fan of the foild caps directly parallel to the LM399. This may make the life for the internal stabilzer more tricky. Chances are some 10 Ohms in series each would be a good idea. This is still relatively little compared to the higher frequency output impedance. It kind of sets a realistic traget for the regulator circuit.

Chances are C1-C4 can be smaller, like 1µF local and than the common C18. 10 µF MLCC tend to be low voltage and show much capacitance drop at some 15 V.

C9 makes the regulation rather slow, allowing for a lot of capacitance at the output. I think some 1-10 nF should be enough.

P.S.:  with the relatively large C9 the supply hum suppression would be not so great. So maybe split R9 and add some local filtering there.



Title: Re: LM399 based 10 V reference
Post by: iMo on September 24, 2021, 01:37:57 pm
There are 2 common mode chokes (small FT toroids) now.
Not sure where to ground the new C14 (R9 split)..

PS: Version 3 - Added 78(L)15 for analog only..

What about R1, R2, R3, R4 values??
Title: Re: LM399 based 10 V reference
Post by: dietert1 on September 24, 2021, 09:43:08 pm
Your circuit looks perfect.
In the last LM399 array i made i used an opamp to derive the power supply from the avaraged reference voltage as well, so it's also stable to 10 or 20 uV. Of course a 7815 with low TC serves as well.
The output circuit stability/transient response can be optimized with a pulsed load.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: Andreas on September 24, 2021, 10:13:45 pm
There are 2 common mode chokes (small FT toroids) now.

PS: Version 3 - Added 78(L)15 for analog only..

What about R1, R2, R3, R4 values??
Mhm,

10 uH  is rather small for a common mode choke. I would take at least 50 uH.
Typically you need a PI-Filter arrangement so at the output a capacitor is missing.

What are the 10R resistors doing in series with the 100nF?
The 100 nF are useless in this case because the LM399 has 0.5 - 1 Ohms dynamic resistance.
The 100 nF are not functional (not reducing noise) but are a EMI-Filter.
https://www.eevblog.com/forum/metrology/emi-measurements-of-a-volt-nut/msg2684100/#msg2684100 (https://www.eevblog.com/forum/metrology/emi-measurements-of-a-volt-nut/msg2684100/#msg2684100)

There is also a T.C. dependant on heater voltage at the LM399. So why only stabilizing the analog part?

The R1-R4 values have to limit the cross currents well below 1 mA. With typical +/-100mV.
So the 560 R look about right to me.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: dietert1 on September 25, 2021, 05:37:24 am
I am using 274R resistors that we had for averaging, as the LM399s i got agree to within +/- 15 mV. As the average is near the middle, max currents are 15 mV/274 Ohm = 0,055 mA or so. The idea of the 100 nF caps close to each LM399 is to protect the zener regulator from external EMI. The averaging resistors can run from the caps, leaving the LM399 output pins with very short traces (< 5 mm). One can also try and place the averaging resistors close to the LM399s and omit the additional filters 10R+100n.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on September 25, 2021, 06:19:58 am
The 100nF ser 10ohm - EricM indicated the new ADR1399 is more sensitive to capacitive loads than the LM399..
The closest components around the LM399 will be aprox 20mm off its pads (there will be the "ring")..
Needs to be investigated (EMI vs. stability aspects)..
The heaters "power" voltage will be stabilized too (say 18-19V). It will be a separate PCB.
All components TH except opamp and LT5400 (provided it fits the PCB)..
PI filter - the output cap could be added (ie 1nF cer).
Btw. - I do not like ceramics in the signal path..
I've increased the spread of the 399's voltages - you may see the currents through the averaging resistors.

The cathode's and averaging resistors will be 25ppm/K one (?).
Below a sim of TC (ppm) with those resistors between +10 to +25ppm/K.. (refs and opamp and LT5400 are 0ppm/K here).
The copper traces will add up some TC too (could be simulated as well).
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 25, 2021, 07:25:38 am
The LT5400 is a bit difficult to solder (especially the thermal pad, that is not absolutely needed). A ORN type network in an SO8 case would be a possible alternative, though with not as tight specs.

The averaging resistors can also be larger. The added noise is still small compared to the LM399 noise, and the filtering action with the 1 µF cap would get a little more effective.

For EMI prodection it may have to also include some filtering (e.g. Ferrite beeds or sereis resistors) with the individual reference ouputs.

Even with some series resistance the capacitors would still be daming high frequency interference.  In the relevant frequency range where the capacitor has any effect the output impedance of the LM399 is no longer at the 0.5-1 Ohms level valid for low frequency. A 100 kHz it is more like 10-50 ohms and quite inductive (some 10 µH). So a pure capacitance can cause a resonance and some series resistance helps to dampen the resonance.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on September 25, 2021, 07:41:55 am
Apropos "ring": One problem to think about with LM399 arrays is board temperature. The larger the array, the closer the board temperature rises to oven temperature, possibly causing one of the ovens to shut off. Since that risk depends on ambient temperature, one should watch oven currents. It's a pitty one cannot adjust each oven temperature.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: miro123 on September 26, 2021, 03:02:14 pm
The 100nF ser 10ohm - EricM indicated the new ADR1399 is more sensitive to capacitive loads than the LM399..

Where can I find the ADR1399 datasheet?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 26, 2021, 03:15:25 pm
AFAIK there is not yet a datasheet for the ADR1399. There was a more or less unoffcial anouncment that there will be one coming out late this year. Suggestest changes to the LM399 were:
4 x the zener current  (so likely more like 2 mA to run instead of some 1 mA as the usual LM399 current)
1/2 the noise
lower differential output impedance  (so less critical resistor for the current, is already not that critical)
more picky about capacitance at the ouput, but no details given.
Likely similar case, so it may work as a drop in with minimal changes (more current and maybe ESR for the cap if present).

A voltage regualtor is always a bit tricky with a low ESR cap at the output. A voltage regulating loop tends to have an ouput impedance that goes up with frequency and thus means an inductive looking impedance. Capacitance with ESR is usually easier to regulate. We don't know how much ESR is needed, but for the PCB layout it is definitely a good idea to have a resistor planed. One can still change to 10 Ohms to 1 Ohms or 100 Ohms if needed. I doublt they would need a very low ESR - this usually needs extra effort to get stable.
Title: Re: LM399 based 10 V reference
Post by: miro123 on September 26, 2021, 03:51:13 pm

4 x the zener current  (so likely more like 2 mA to run instead of some 1 mA as the usual LM399 current)
1/2 the noise
lower differential output impedance  (so less critical resistor for the current, is already not that critical)

Thanks  for sharing.  I have another question
Twice the current means 50% reduction of bootstrap resistor. e.g. from 3k -> 1,5k
According to me - twice the current and half the Zenner dynamic resistance requires the same resistor tolerance. Am I right?
E.g LM399 has 1 Ohm dyn resistance and Rb=3k . Hypothetic ADR1399 has 0,5 Ohm & Rb=1,5k. Both circuit will require the same resistor ppms.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 26, 2021, 04:08:21 pm
There is none free zener in the 399. The zener is included in a regulator, consisting of several transistors, resistors, caps etc.
Afaik the zener current in LM399 is set to 250uA.
Title: Re: LM399 based 10 V reference
Post by: Cerebus on September 26, 2021, 04:16:22 pm
Afaik the zener current in LM399 is set to 250uA.

Yup.

Quote from: SNOA589C–June 1976–Revised May 2013
Current through D3 is held constant at 250 μA by a 2k resistor across the emitter base of Q13 while the emitter-base voltage of Q13 nominally temperature compensates the reference voltage.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 30, 2021, 11:30:38 am
I collected several MAC199 (LM399 clone in TO18 case) with broken heater. I've tried to utilize the parasitic diode for chip temperature measurement.

Basically it works, with say 80uA fw current the voltage at the heater's ground changes like (aprox)

31C (fingers)  620mV
26C (amb)     637mV
IPA large drop  677mV

There is an issue with it, however - while the 80uA current is injected into the heater's gnd (parasitic diode anode) the Vref (standard 10V Vref like the above) increases by +3000ppm. That is way more than a drop on the bond wire - thus it messes up something on the chip, it seems..
PS: or chip temperature increase?? That is 50uW loss at the diode..

The "cold zener" itself shows something like 30-40ppm/K.

It seems it could be useful for measuring the chip temperature before the Vref measurement, then the diode current has to be switched off (a mechanical switch for example) and you may measure the Vref.
The loss at the diode during the measurement is aprox 13uW with 20uA.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 30, 2021, 11:44:51 am
AFAIK the neg. side of the heat part is the substrate and should be the most negative voltage in the whole system. A current troub a parasitic diode can injsct substarte current and this can upset the rest of the circuit electrical  (current from the substrate to parts of the circuit).

The MAX199 with a broken heater is essentially like an LM129, so a metal case but no heater. If the TC is reasonably low, one may get away with an external oven. There is no absolute need for a sensor directly at the chip. With a constant power at the reference it is OK to measure outside.  An external oven is slower to stabilize, but not that bad.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 30, 2021, 11:52:15 am
Sure, that oven is the standard approach..
Back to the roots - what would be the max resistor value in the RC filter (from the 399 zener to the opmap's input) for OPxxx, or, LTC2057?
The Johnson noise is 0.12uV rms with 80k at 33C, dF 10Hz..
A single 399 has something like 1-2uV rms.
We should add opamp's current and voltage noise, right?

PS: below LTC2057 (89nV rms) and OP177 (387nV rms) noise in 0.1 to 10Hz, 10k resistor. Not sure it shows all contributions..
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 30, 2021, 12:51:36 pm
With a RC filter you have the OPs input current (bias and curent noise) and the Johnson noise of the resistor.
With a reasonable size capacitor (e.g. 10 µF) the upper frequency limit would be well below 10 Hz. So the relevant noise is not 0.1 to 10 Hz, more like 0.01 to 1 Hz or even lower. The main parts there are the 1/f noise of the LM399 (or similar) and the 1/f noise of the OP. Especially the current noise may have quite some 1/f part.
So the OP07 would not be a good choice for high resistance (e.g. > 30 K). More like OPA202 as a BJT based one with low current noise, or an AZ OP like LTC2057, AD8628 or maybe max4238. The specs for the current noise of AZ OPs seem to be not very relable - 1 would not count on all the numbers there, especially when very low.  I would consider some 100 K a reasonable upper limit as than one starts to have to worry about leakage curents in the pA range that may change. The main noise source would still be the LM399, even if using 4 or so.  Still filtering the reference is limited. The filter will also need quite some settling. It is not only the RC time constant, but there are also slow DA effects that need to settle when it comes to settling to something like 1 ppm of the final value. This part starts smaller but can be 100 or 1000 times slower.

The capacitor may also need a reasonable stable temperature (e.g. thermal shield): I can see a small thermal effect even with only some 5 K and 4.7 µF (MKS with 3 x muliplication), it would get more with higher resistance.

If it is about a more pulsed operation (e.g. 99 ms zener and 1 ms temperature measurement) the more suitable way would be a switch to turn off the connection when in temperature mode.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on September 30, 2021, 03:05:05 pm
Those considerations are subject to experimental confirmation. Caps to be used for filtering references need to be qualified to the same level of 1 ppm or below. Don't assume anything.

I made a PWM 10V stage for a 5x LM399 array (schematic above). In my case the filter is the PWM filter and it is second order with a bandwidth less than 1 Hz. Resistors used are 22K +47K. From the very start i implemented a bootstrap scheme with electrolytic caps to keep the voltage over the filter caps near zero. But the electrolytic caps are very temperature sensitive. When i ran the unit in an incubator, noise due to forced air flow went up to 2.5 ppm p2p, while average stability was near perfect. When wrapped in tissue, noise is down to about 0.3 ppm p2p, but the unit exhibits notable TC, as it partially decouples from incubator temperature. The next revision will be without bootstrap.

Also i remember a leakage test with a 650 uF foil capacitor (AVX FFLI6L0657K). I could see temperature variations as voltage changes - over several hours. I think when capacitance changes with temperature, the constant charge in the cap will result in changing voltages. It was difficult to model as the capacitor has temperature gradients inside.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on September 30, 2021, 03:08:46 pm
..
If it is about a more pulsed operation (e.g. 99 ms zener and 1 ms temperature measurement) the more suitable way would be a switch to turn off the connection when in temperature mode.
You can even use the substrate diode as a "heater", imho. Say 20mA*0.65V=130mW. TO18 thermal resistance is something like 300C/Watt, 0.13*300 = 39C, with Tamb=25 you'll get Tpackage=25+39=64C.
With 50mA diode current you may anneal all defects off the silicon  :-DD
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 30, 2021, 03:31:57 pm
Current through the substrate diode can interact with the rest of the circuit. Usually this is not in a good way. So once there is substarte current all bets are off.



For filtering I would not consider electrolytic caps practical, because of the temperature effect and leakage that can change with time and temperature. The DA will also requite extremely long settling - maybe not as long as the LM399, but it would start all over after power off.
The question is more if polyester caps are sufficient or if it needs polyporpylene.

Filtering is mainly an option if the instruments to measure or use the reference are especially sensitive to a certain not too low frequency band. The main slow part would be someting like the DMM auto zero cycle with maybe 2-4 readings of 200 ms each. So this would be frequencies of some 1-4 Hz.
Instead of the effort for filtering a lower noise reference could be the easier way.
Title: Re: LM399 based 10 V reference
Post by: guenthert on September 30, 2021, 04:40:50 pm
[..]
For filtering I would not consider electrolytic caps practical, because of the temperature effect and leakage that can change with time and temperature. The DA will also requite extremely long settling - maybe not as long as the LM399, but it would start all over after power off.
[..]
      Is DA really an issue, if one uses two capacitors in series and charges the mid point from the output of the zener using a 2nd resistor (as described in AoE3)?
Title: Re: LM399 based 10 V reference
Post by: Cerebus on September 30, 2021, 05:05:14 pm
That little trick is designed to deal with leakage currents, not dielectric adsorption.
Title: Re: LM399 based 10 V reference
Post by: guenthert on September 30, 2021, 05:50:30 pm
That little trick is designed to deal with leakage currents, not dielectric adsorption.
      Well, yes, but what is the ill-effect of DA, if not voltage drop over the resistor due to the charging current?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on September 30, 2021, 06:15:04 pm
The trick with 2 capacitors would also suppress the effect of DA, at least to a large part. There is still the DA for the lower capacitor (that still sees the voltage). Because of the usually smaller resistor there, the extra current from DA would have less effect, but could still be an issue for some time. Still more like minutes or hours instead of weeks. Some of the temperature effect can still be there, even with zero voltage at the capacitor.
Title: Re: LM399 based 10 V reference
Post by: iMo on September 30, 2021, 07:57:38 pm
The substrate polarization - is the wiring with H- and Anode on the same potential (ground) comparable (in relation to achievable Vref parameters) with the wiring where H- is somewhere at -15V, for example?
Title: Re: LM399 based 10 V reference
Post by: eplpwr on November 08, 2021, 08:26:17 pm
Where can I find the ADR1399 datasheet?

It is now released on the AD website, complete with datasheet Rev 0.

For those of us that have a VAT# the ADR1399KHZ is orderable at US$ 12.88 ea, MOQ 18 pcs. I also saw the "Request samples" button was available, though didn't try it.
Title: Re: LM399 based 10 V reference
Post by: maat on November 09, 2021, 07:44:18 am
I also saw the "Request samples" button was available, though didn't try it.

I works  ;D. Delivery date pending...
Title: Re: LM399 based 10 V reference
Post by: Cerebus on November 09, 2021, 04:35:58 pm
I also saw the "Request samples" button was available, though didn't try it.

I works  ;D. Delivery date pending...

Yup, it does.  :) I've got "Backorder Allow 48 hours to be scheduled" as the current status.
Title: Re: LM399 based 10 V reference
Post by: magic on November 12, 2021, 06:30:58 pm
I collected several MAC199 (LM399 clone in TO18 case) with broken heater. I've tried to utilize the parasitic diode for chip temperature measurement.

Basically it works, with say 80uA fw current the voltage at the heater's ground changes like (aprox)

31C (fingers)  620mV
26C (amb)     637mV
IPA large drop  677mV

There is an issue with it, however - while the 80uA current is injected into the heater's gnd (parasitic diode anode) the Vref (standard 10V Vref like the above) increases by +3000ppm. That is way more than a drop on the bond wire - thus it messes up something on the chip, it seems..
Bringing the substrate above circuit ground may forward bias parasitic diodes from the substrate to circuit elements like NPN collectors.
Try biasing the substrate normally and using the diode going from the substrate to H+ instead. (If the heater failure is not due to H+ being open circuit).
Title: Re: LM399 based 10 V reference
Post by: r6502 on November 17, 2021, 12:09:00 pm
Hello all,

I've had supid idea:

I do not own a 7.5 or 8.5 digit multimeter, only 2 PREMAs and one HP 34401a, so all 6.5 digit types. These devices are using the same reference as my reference board that I'd like to characterize.

My thought was based on better thermal stability of LTZ1000  based reference design. I thought I will replace the LM399 in on of my PREMAs and put a LTZ1000 based reference in. The 7.2V could be direct used in the PREMA. Due to the higher stability of the LTZ100 I will now be able to see the drift of the LM399 based 10V reference, if the drift is > 10µV.

What do you think about this Idea?

Edit: may be I will use one of the new AD ADR1399 references for the 10V reference board, when I get some ...
Title: Re: LM399 based 10 V reference
Post by: Echo88 on November 17, 2021, 12:56:14 pm
Improve the stupid idea: Use the LTZ1000 as a standalone stable reference and take a 34401A to measure the voltage difference between a DUT like the LM399 and the LTZ1000 in opposition mode (ref-minus connected to ref-minus).
Dont really know if a LTZ1000-modded 34401A would show better stability or if its limited fundamentally by its ADC/Frontend.
Title: Re: LM399 based 10 V reference
Post by: branadic on November 17, 2021, 04:58:45 pm
I totally agree with Echo88, at least the Prema DMMs are not limited by the noise of the reference, but from the ADC. Using them in the 100 mV range to measure the difference between two references makes much more sense.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on November 17, 2021, 06:42:23 pm
Hello,

most LM399 references that I have have less than 2 ppm voltage change near room temperature over a 30 deg C span (10-40 deg C)
2ppm/30 deg C = 67 ppb/K.
A LTZ1000 is usually not better if it is not trimmed for T.C.
So you have a hen and egg problem.

Attached a measurement between 2 LM399 (one at room temperature and one in my car cooler box)

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 09, 2022, 11:53:46 am
i have been reading a million hrs in this thread :-) thanks to you all for the high level of skillsharing

one thing keep comming up, the opamt to use, i think i get the point,
the only thing that really matter is the drift pr temperature,
here are the 4 opamps all folks talk about, the best one is obsolete, darn it..
so.. what is the best, that i can get ? if cost is not an issue ?
anything else beat the mentioned types ?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 09, 2022, 01:25:53 pm
A few more modern classic OPs to consider are OPA277, OPA207, OPA2205 (dual) and ADA4077.  One could also consider an auto zero OP.

The accuracy of the LM399 is anyway somewhat limited, so even slightly lower grades are not that bad. It is not only the drift with temperature that matters. Also drift with time and humidity can be an issue.
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 09, 2022, 04:14:50 pm
DANKE !
the mentionened opamps was a great help, wow some are really amazing,
another thing to consider, we dont see so many people use a temperature regulator on the opamp, and resistors,
it could be another case, round the LM399, like +40C regulated, and then isolated,
if done correctly, temperature ultra stable, the opamp and resistors used could be much less hiper-performing.
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 09, 2022, 07:11:46 pm
I am impressed, darn stable at first impression
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 09, 2022, 09:59:01 pm
to make 7V from the LM399 into 10V
i need two resistors on my opamp
in the ratio 3/7
prefered a two in one temperature compensated ratio resistor chip,
scrolling thru this thread, i dont see this magic ratio mentionened ?
and a solution that is best possible, no matter what cost.
all seems to use seperate resistors, 5k and 20k are mentionened quite a lot,
but i just dont see how they make this gain ratio ?
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 09, 2022, 10:24:23 pm
in the ratio 3/7
Hmm,

what about 1.5 / 3.5 ?
could be made from a 8 Resistor chip like TDP1603

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 10, 2022, 09:18:10 am
thanks alot for the input andreas

i am unsure how you make the 3/7 or 1.5/3.5 ratio with 8 equal value resistors ?
please explain ?

the suggested 8 in one, 5ppm tracking
https://www.vishay.com/docs/60045/tdp.pdf (https://www.vishay.com/docs/60045/tdp.pdf)

this is a 4 in one (i got many of them stocked)  4 x 1k type, also 5ppm tracking
https://www.vishay.com/docs/60005/orn.pdf (https://www.vishay.com/docs/60005/orn.pdf)

so i hope i can use them, i am a bit worried about build size in this feedback circuit, i prefer it to be smallest possible
less noise pickup

next question : capacitors for opamp feedback, and 7v filtering, they say 1uF, but i am very sure the type is important,
noise ? leak, temperate problems, and what not from capacitors must be investigated.
recommendations ?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 10, 2022, 09:52:53 am
Using seprate resistor networks reduces the matching effectiveness. It can still be acceptable though not perfect.
In the 1.5 and 3.5 combination this would be 2 mA off current for the divider and thus a bit on the high side. The attchment shows the circuit, using 1.01 for the resistors of the 2nd chip. The sharing between the 2 chips is not perfect, but still good with only some 10 % unbalances part  (1/3 for the top and 1.25/3.5 for the lower part).

A reference source is essentially static and there is thus no really need to look for low DA in the capacitors. So polyester film should be good enough.
Directly at the ADR1300 reference the impedance of the reference is quite low and leakage is thus not as critical. So there even X7R ceramic or a tantalum electrolytic can be good enough. With X7R ceramic I would plan for more than 1 µF (e.g. 2 or 3 in parallel) because the capacitance under DC bias goes down. From the response curve for the ADR1300 the 1 µF already look like they are on the low side - enough to make it stable, but a little more would not hurt.

Edit:
Just to improve stability at the OPs one gets away with way less than 1 µF and 1-10 nF should be enough there. This can than be C0G ceramics.
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 10, 2022, 04:38:00 pm
I try a mockup with the ORN resistors i got plenty off, like this..

i arranged the connections, so the ic is symetric powered,
 dont you think they TC match best this way ?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 10, 2022, 05:10:20 pm
With 2 chips one would get the best maching in the way I have drawn it. So array no 1 for the 2 parallel resistors in the upper part and one single and one in the parallel pair for the lower resistance. With quite some current in the divider one needs to make sure not to have to much copper trace and a suitable ground point.

The idea it to mix the resistors from both chips so they contribure about equal to both sides of the divider:  In my version the right chip provides 1/3 of the upper resisistor and about 35%  (1.25/3.5) to the lower resistor. So not ideal, but pretty close. So matching between the 2 chips is less critical (only about 1/10).
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 10, 2022, 05:30:29 pm
here, i made it, as i drawn it, it seems to work, no temperature cycle yet, but i get 10V out :-)

i challenge your drawing: see mine again,

the two parallel resistors are in EACH ic,
and this way same power is lost in each ic = best TC match, correct ?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 10, 2022, 05:58:23 pm
The matching is not about getting the same power to both chips, but about getting a similar fraction in contributing to the upper and lower resistance.
So one wants to use the resistors from 1 chip in both the upper and low part and not have the upper part from 1 chip only.

In your version the help chip has 100% of the upper resistor and 1/3.5 for the lower resistor.  So about 70 % of the matching error is left over.  This is more like a worst case solution.

In my suggestion the richt chip contributes 33% and 35%, so relatively colse to equal. This gives around 7% of the matching error remaining or about a 10 fold reduction in the sensiticvity.

With chips from the some batch the matching between chips may be quite good anyway. So one can be lucky and still get a stable output - it ust needs a little more luck.
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on February 10, 2022, 06:36:39 pm
thanks a million times, a very good answer,
since i got two brand new LM399 running in parallel, on the same mock up board,
i hereby make the other two resistor chip wirings exactly like you suggest,
and for the final challenge, temperature cycle it,
then it is going to be interesting if we are able to detect or reveal what perform the best :-)

here is your suggestion, redrawn a bit, for easier soldering,
but i am sure it is exactly the same as you explained, right ?

EDIT : your suggestion works, and i got 10V out from that one too
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 08, 2022, 08:12:45 pm
Here i have some measurements on the 5x LM399 array with PWM gain stage that i made last year (see above). The device is still in prototype stage. I put it into a NH2 incubator controlled by an Arroyo TecSource at 22.1 °C. The 10 V reference output was monitored by an Advantest R6581T sitting in a temperature chamber at 23.25 °C.

First diagram shows a six months log of voltage measurements in units of uV, so the diagram vertical axis spans 1.2 ppm of 10 V. I added two temperature logs, the pink one inside the temperature chamber, the orange one ambient temperature (room temperature). I added 14.5 °C to room temperature values to make both logs fit onto the right y scale. That's about how much the R6581T heats above chamber air temperature.
There is a negative correlation between voltage and ambient temperature. Drift is estimated as -12 nV/day, about -0.4 ppm/year. A two parameter fit for temperature and time wasn't yet conclusive.

The red circle indicates a TC determination shown in the second diagram. The incubator temperature with the LM399 array & PWM got an increase of 1 °C. TC was determined as -0.3 ppm/K. It means the PWM can compete with precision resistor dividers.

As the incubator isn't a precision oven, it likely causes the negative correlation observed in the log.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: try on March 08, 2022, 08:35:46 pm
thanks alot for the input andreas

i am unsure how you make the 3/7 or 1.5/3.5 ratio with 8 equal value resistors ?
please explain ?

naked chip, value does not matter, it is all about ratios:
1 1 1 1 1 1 1 1

use two of them in parallel
1 1||1 1 1  1 1||1

gives
1 0,5 1 1 1  0,5 

switch them in series:

1-0,5   1-1-1-0,5

gives
1,5 and 3,5

total is 5
3,5/5=0,7

7V / 0,7 = 10V.

But honestly, what improvement do you expect over a simply solution with two resistors?
Are you able to estimate that in quantitative terms or even measure it?

Using that TDP1603 Andreas mentioned could be considered overkill (but it is an already much cheaper solution compared to what was promoted years ago here...).
You might even try out 8x 2,2kOhm resistors in metal film quality.

Real voltnuts obviously only use hermetically sealed custom-built Vishay VHP dividers at the cost of the GDP of Nicaragua.
Another possibility is to use custom-built wirewound precision resistors, wound by virgins at full-moon in leap years.

But if it makes you feel good throw in all the expensive stuff. ;)


Regards
try

P.S.: You can try out an arrangement that takes into account heat distribution because the two paralleled ones are running with half the current: P= (I/2)^2*R = 0,25*I^2*R compared to I^2*R for one single resistor. But with larger resistor values it won't matter anyway.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 08, 2022, 10:29:11 pm
The resistor networks tend to be quite stable and the ORN ones are not that expensive (some $4). The specs only call for 5 ppm/K TC matching, but so far my measurements with 10 K / 50 K sets were more in the 1 ppm/K range for the relative TC.  Chances are good that the relative drift can also be quite good, though here only time will tell.

In a 10 V reference the 7 V to 10 V step is still one of the more critical parts. Chances are that when starting with a LM399 / ADR1399 the resistors for the gain may still be the larger TC part and also be a part of the long term drift. So it makes sense to use rather good resistors there. Compared to a $10-$20 reference the price for the 2 resistor arrays is not that bad, and they can still be in stock.

The self heating should not be that crictical, as in a voltage reference the power is essentially constant in time. The slightly complicated circuit comes from distributing the resistors of the 2 sets of 4 so that they contribute about similar and get away with an easy layout / soldering.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 03, 2022, 01:44:32 pm
For one of my LM399 refs I got the popcorn noise on a very long time scale, but still about the same amplitude / step size.  This makes the curve for the short time look well, but a for a real world use the long times that are usually important. I may be good to look at the noise also for a little longer and maybe in the time domain.

What is the vertical scale, is it volts (rel. to 10 V or 7 V ?) or PPM of voltage.
Another point is that the reference in the meter may contribute, both at the fast and slow end.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 03, 2022, 02:53:18 pm
In the initial phase all references are expected to drift a little more. So there is a good chance the drift will go down after some 100 hours of operation.
The Ref102 data show some drift and popcon noise with about the 0.3 ppm step size. This is comparable to the step size typical for a LM399. So some of the noise may as well come from the reference in meter and also some drift could in principle come from the meter as well.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 03, 2022, 03:16:13 pm
It is not just the time after warm up, but usually the time after soldering or production that matters for aging. The turn off / turn on cycle can introduce some new drift cycles too, as part of the hysteresis. At least for the LM399 the hysteresis part is usually small and no long warm up needed for the reference. For the 10 V level of the LM399 there can also be some drift of the resistors.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 03, 2022, 06:28:00 pm
Hello,

it is a unfair comparison in my opinion:
The plastic package of the REF102CU will also be sensitive to relative humidity.
(up to around 0.5ppm / %rH and with a time constant of several days).
So you will have also seasonal changes depending on region.

see eg. here for a REF102AP
https://www.eevblog.com/forum/metrology/best-out-of-the-box-10v-reference/msg1242608/#msg1242608 (https://www.eevblog.com/forum/metrology/best-out-of-the-box-10v-reference/msg1242608/#msg1242608)

https://www.eevblog.com/forum/metrology/tales-from-the-mouth-of-a-voltage-reference-ageing-box/msg3618917/#msg3618917 (https://www.eevblog.com/forum/metrology/tales-from-the-mouth-of-a-voltage-reference-ageing-box/msg3618917/#msg3618917)

The statistical average will not help much for humidity.

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 03, 2022, 09:37:08 pm
The higher frequency noise, like above some 10 Hz can be filtered with not that much effort. So that is usually not the noise to worry about. The weak point of the LM399 is more the popcorn type noise, that can extend to rather low frequencies / happen at a long time scale. I have one unit with jumps about once per hour - at first this looks good, but can also be trouble as the noise extends down to very low frequencies.
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 04, 2022, 05:05:25 am
The weak point of the LM399 is more the popcorn type noise,

Sorry no:
Popcorn noise is not related to LM399.
From theory popcorn noise is the result of impurities in the silicon chip.
You can see popcorn noise on all kinds of voltage references if you look long enough.
Here a example of a AD587 with a LNA 0.1-10 Hz on scope.
Or a MAX6350 here as reference of a ADC which had popcorn noise at certain temperatures only.

Of course I would do a screening for popcorn noise for a stable statistical reference and sort out the stinkers.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Andreas on April 07, 2022, 05:47:12 pm
6500 is noisy toy, but very stable.
34410A is very low noise equipment... but poor stability.
Hello,

How many NPLC for each instrument. And how is the measurement rate?
~450 ms for one measurement looks a bit strange to me.

my K2000 at 10 NPLC and 50 Hz needs around 600 ms. (is the 6500 faster?)
The 34401A needs 400 ms at 10 NPLC.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on April 07, 2022, 07:49:02 pm
With timing from the RP and some extra delay for data transfer and processing the 440 ms make absolute sense.

The data for the DMM6500 look indeed a bit noisy.
The 34410 drifty part / low frequency noise is also a bit unusual. I have expected better performance for both meters.
For comparison it may be interesting to compare to the noise of both meters with a short at the input.

Much of the noise from the references could be on a shorter time scale (e.g. minutes) and may be a bit hard to see at this scale. So it may help to look at a shorter part, like 0.5 or 1 hour, to see if there are correlations on the shorter time scale.
The slight problem here is that there is noise from the DUT, the meters references and also the ADCs are of comparable size. So it can be a bit tricky to tell those noise sources appart. Things are hard enough when the meters are lower noise.

It may be better to test the references by having 2 x  LM399 based 10 V reference (or the 7 V level) and directly take the difference there and than only look at the difference with one of the meters in the 100 mV range. This would reduce the noise sources to 2 x LM399 and only a little noise from the DMM.
Title: Re: LM399 based 10 V reference
Post by: guenthert on April 08, 2022, 12:04:27 am
Some time ago Ill try to explain about "time delay between points" to Mr.Kleinstein, but my English skills is poor, my explain looks bad.

No worries, your English is much better than the Russian of most readers here  ;)
Title: Re: LM399-PWM-DAC
Post by: branadic on April 30, 2022, 09:24:21 pm
In preparation of some lab intercomparisons I did some measurements on the LM399-PWM-DAC.
First a 1h measurement using an uncalibrated 2182A @ 1NPLC, second a t.c. measurement using the integrated NTC. It turned out the t.c. is -0.275 ppm/K. Further stability measurements and taking some more readings with different calibrated meters are both in progress.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 01, 2022, 04:23:47 am
Hello branadic,

just to get shure:
it is the schematic mentioned here?
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1955149/#msg1955149 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1955149/#msg1955149)

which analog Switch + PWM frequency did you use?
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2200098/#msg2200098 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2200098/#msg2200098)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on May 01, 2022, 05:35:50 am
Hi Andreas,

it's the version with the modification suggested by Kleinstein, but with the original 8 + 16 bit resolution firmware, 20 MHz crystal and ADG419.

https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2082544/#msg2082544 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2082544/#msg2082544)

-branadic-
Title: Re: LM399 based 10 V reference
Post by: Andreas on May 01, 2022, 02:48:21 pm
Hello branadic,

so you have the  ~10 kHz PWM frequency with prescaler = 8:1.

with the optimized Code (for 8.14 Bit PWM) from here
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2189460/#msg2189460 (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg2189460/#msg2189460)
Prescaler 1:1 and XTAL frequency 4.5 .. 5 MHz you could get near zero T.C.

With best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: branadic on May 01, 2022, 04:11:29 pm
Thanks Andreas, but I stay with what I have at the moment as I don't have the required components in my drawer.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on May 31, 2022, 11:56:40 am
shodan i dont understand your circuit ?
in your post 1216 you show a schematic with 5K and 10K in the feedback
how can this deliver 10.000V out, when the LM399 is about 7V ?

Title: Re: LM399 based 10 V reference
Post by: oz2cpu on June 08, 2022, 11:58:59 am
thanks a lot, this helps : since it is very easy to get 1:2 resistors in the crasy matched type kind of temp match
then it is possible to add a "normal type of ressistor to trim the value to at least 10.0xx
this minor trim will not affect temperature match so much
Title: Re: LM399 based 10 V reference
Post by: free_electron on June 08, 2022, 11:21:04 pm
TL;DR . Bob pease or jim williams had a circuit where they used like 10 of the lm199's. anyone know where to find that schematic ?
Title: Re: LM399 based 10 V reference
Post by: guenthert on June 09, 2022, 12:04:11 am
how can this deliver 10.000V out, when the LM399 is about 7V ?
You can see exact value on legend into post #1227
No one could prove to me "REF should be 10.00000000000V exactly", therefore, I use a voltage around 10.6V.
10.6V Its fine value, because most DMM accept overload to 10% on 10V range, without any problems.  ;)

Probably the main problem is that I'm not a voltnuts, so 10.6V works fine for me. >:D I can calculate ppm drift with 10.6 base w/o problem.

Moreover: Into my last toy, i use direct connection LM399 to 20 bit DAC with OPA189 buffer and mirror... so... as result DAC output is +6.9V...-6.9V, no problem here... works very precise!  ;)
   10.6V is fine as a voltage reference, as is 6.9V.  The reason to chose a 'standard' voltage is so that it can be compared directly (difference method) with other 'standard' voltage sources (using e.g. nanovoltmeters with high resolution, but a limited range or a null-voltmeter with an even smaller range).

   Incidentally, I do however have a 16bit PCI ADC adapter with a 10V range, which doesn't allow for *any* overrange.  Even the 40mV offset on my reference are too much  |O
So a bit below the nominal voltage would be better in that case.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on June 09, 2022, 06:09:53 am
Directly measuring the difference also helps when looking at the noise. With a modern meter to read the difference and the still relatively noisy LM399 there is no big problem if the difference is in the 1 V range of the meter. The full 7 or 10 V meter can be a bit demanding for the DMM and the reference to compare to is than fixed to the DMM internal one.

There are a few other divider ratios that can be obtained with a reasonable small number of equal resistors. With 6 equal resistors one can get a 10/7 gain (1 :2.333 resistor ratio).
It would anyway be a good idea to still have access to the raw 7 V to at least check the 7 to 10 V stage stability seprately. Long term stability is hard to judge upfront and in the ppm range there are sometimes unexpected small effects even with good parts. With enough time modern DMMs are not that bad in measuring the 7 to 10 V ratio and a resistor ratio has a hard time (expecially not guarantied) to be that stable over a long time.
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on June 09, 2022, 09:24:11 am
what i think you are trying to say :
every component you add to the output of your lm399 or other ref
you dont add accurasy or stability, but you add more problems.
Title: Re: LM399 based 10 V reference
Post by: mawyatt on June 09, 2022, 02:01:00 pm
TL;DR . Bob pease or jim williams had a circuit where they used like 10 of the lm199's. anyone know where to find that schematic ?

Think that was Bob Pease in response to the newly introduced LTZ1000 or whatever it was originally called way back, but can't recall the exact details nor schematic.

Best,
Title: Re: LM399 based 10 V reference
Post by: dietert1 on June 09, 2022, 04:15:49 pm
In this thread (above) i reported a build of a 5x LM399 array, with image and schematic including a PWM gain stage 7 => 10 V. There were some measurements on initial aging and a six months log.
Meanwhile i built a 10x LM399 reference with 14 V output (2 by 5) with a PWM divider 14 V => 10 V and programmable for use as calibrator. Hope to report measurements later.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: free_electron on June 09, 2022, 05:45:01 pm
In this thread (above) i reported a build of a 5x LM399 array, with image and schematic including a PWM gain stage 7 => 10 V. There were some measurements on initial aging and a six months log.
Meanwhile i built a 10x LM399 reference with 14 V output (2 by 5) with a PWM divider 14 V => 10 V and programmable for use as calibrator. Hope to report measurements later.

Regards, Dieter
why pwm ? use a multiplying DAC ... send output of the dac into summing amplifier.
take 10V ref output , divide by 10 send into dac 1. , divide ref by 100 , send into dac 2 , take ref , dac1 output and dac2 output into summing amplifier.... if you use a 10 bit dac ( 1024 steps)

you don't care what the absolute value of your 10v is. it can be trimmed using the dac.
you take the reference and create a band around it. let's say 9.9 and 10.1 volt using two opamps. use that as the DAC ref+ and ref-. the output of the dac is post-divided 1/10. sum the referenc ewith the post-divided output and you can trim the output very precisely. store trimdata in nvram.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on June 09, 2022, 05:55:49 pm
DACs are not that stable on the long run.  Many internaly use resistors for the R2R chain and are thus not more stable than the resistors.
Trimming with resistors is possible and this may be combined with a DAC for the very fine part. Combining multiple DACs is often not helping very much, as often (the PWM DAC is one of the few exceptions) the stabilty of the DAC for the coarse part is not much more stable than it's resolution.

The PWM DAC can be very long term stable. There is some drift (e.g. from charge injection that might change over time), but usually not much.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on June 09, 2022, 06:29:14 pm
Those projects require some patience to try various revisions. For example i first used an integrated MOS switch to implement the PWM. Then i found that the delays of those switches (50 to 100 nsec typ.) have temperature dependence, a detail largely unspecified in any datasheet, yet important to get the best precision. So the calibrator (second revision) got a discrete implementation resembling those Fluke calibrators, with switching delays of less than 5 nsec. Then temperature dependence and aging effects mentioned by Kleinstein shall also be 10x less than with the integrated switchess.
I think a good test of a PWM DAC is its native precision, i mean without any calibration. If one can get the output within some ppm of the predicted value, it can be made sub-ppm by calibration. If using 10 or 20 MHz as clock, a one second integrated DC with sub ppm resolution is possible. My calibrator test setup uses a MSP430F2616 running at 16 MHz. So one may even use a temperature measuement to numerically compensate residual TC.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: David Hess on June 09, 2022, 11:25:20 pm
DACs are not that stable on the long run.  Many internaly use resistors for the R2R chain and are thus not more stable than the resistors.

Analog Devices has a 20-bit DAC which is suppose to be stable to 22 bits.
Title: Re: LM399 based 10 V reference
Post by: oz2cpu on June 10, 2022, 11:16:32 am
i dont understand what is the idea about pwm a voltage ratio,
since the voltage is critical in the sub ppm range, the timing demands and later filtering, to make it DC again,
is super critical..

the fundamental : pwm ratio, of voltages, sounds like a good idea, since the ratio is a digital value,
but we want to have a stable, noise free DC
Title: Re: LM399 based 10 V reference
Post by: dietert1 on June 10, 2022, 12:08:16 pm
The PWM circuit can be built very stable. This was the result of Fluke research many years ago. I remember a list of download links in another thread in this forum with Fluke papers describing how and why this is a good method.
Here the discussion may be a bit off-topic.
The basic idea is like this: ppm precision in a 1 second DC integral requires timing correct to 1 usec or lets say 100 nsec, which is easy to do. Many parts we are using nowadays work at bandwidths of 500 MHz, so that would be less than 1 nsec. When you start with circuit design, it will be a similar adventure as building a high resolution DVM. You can read one of the threads about building DVMs, e.g. Kleinsteins.
You will find that you shouldn't use the built-in clock generator of the MCU, but a separate clock generator.
You will find that you shouldn't use a 74HC74 or a FPGA for synchronization, but you should rather select a 74lcv74 or the like.
Then the MOS switches. The discrete solution i found uses a dual n-channel mosfet with pair specs. The Rdson difference that shifts output voltage can be as little a 0.1 Ohm. When the filter impedance is 50 KOhm or so, this is an error of 2 ppm at most.
The AD DAC 5791 mentioned above is a good part and it will beat many of those calibrators available at ebay. Yet it won't replace a Fluke 5700 calibrator. I'd guess to build a PWM DAC of that quality one needs a 1 GHz digital scope.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: guenthert on June 10, 2022, 10:07:51 pm
The PWM circuit can be built very stable. This was the result of Fluke research many years ago.

Minor correction, Valhalla beat Fluke to it and their 2720GS was preceded by research and patent of a French fellow in the 60s: https://adret-electronique.fr/appareils_adret_metrologie/104a/104a.html
Title: Re: LM399 based 10 V reference
Post by: iMo on June 13, 2022, 07:17:06 pm
At the end of the day you have to create the DC out of the PWM train. Do we have data on how the low pass filter components stability affects the created DC?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on June 13, 2022, 08:31:41 pm
The PWM to DC filter can indeed effect the stability. If usually uses a relatively large resistor in the 50-200 K range and this makes it somewhat sensitive to changes in the bias / leakage current.
For the level of accuracy of the LM399 this is usually OK with a reasonable moder AZ OP like LTC2057. The version with gain adds some additional complications. The possibly more tricky part than the filter / buffer after the filter can be the driver to provide a low impedance reference signal to the PWM switches. Here the settling performance to the current spikes can matter.

Using PWM as a long term stable divider is not necessary the easiest way to go, but definitely an option worth considering.  The brute force way with very stable reistors. Another, a bit more exotic way is to use a precision transformer, somewhat similar to the ACAL part in the Datron 1281. This can also be quite stable on the long term, though the resolution is a bit limited to integer turn numbers.

Resistors tend to drift over time. The PWM ratio an a transformer turn ration can be very stable over time, but it takes extra effort and includes some extra interference even for a short time scale. For a short time scale the resistors are the obvious winner but the longer the time, the more attractive the alternatives get.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on June 13, 2022, 10:53:36 pm
Making a precision PWM source is much easier than making a precision DVM. Once time division has been implemented with good quality, the filter is very easy in comparison to the integrator. The caps are selected for low leakage, dielectric absorption is no issue as they run at constant voltage. I found caps rated at 160V would perform very well at 10 V or so. In fact i even implemented kind of a guard scheme to run the caps at near zero voltage. 1 pA leakage on 50 KOhm gives 50 nV.
Also i found on the web a ripple cancelation circuit that simplifies the filter and that reduces the ripple of the PWM DAC input as well. It's in the schematic i posted above.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on June 14, 2022, 09:43:33 am
@dietert1: do you plan to sell the final version of your PCB (inside the EU, for example).. ?
Title: Re: LM399 based 10 V reference
Post by: dietert1 on June 14, 2022, 10:52:45 am
To be honest i looked into this after reading the 2019 reports of Andreas and branadic on their experiments with PWM gain stages (see above). Meanwhile they seem to have better results, too. As far as i remember they initiated kind of a calibration circle in May, using a PWM.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: dietert1 on June 15, 2022, 10:00:04 am
If an analog filter for the PWM divider cannot be used, it can be implemented in the digital domain as well. This would be a hybrid design with a resistor divider including a high resolution resistor DAC for adjustment. This gain stage would by slaved to a PWM divider that is wired as precision ratiometer, with a primitive analog filter to make it a sigma-delta ADC (similar to integrator in a precision DVM). With some digital control loop to continuously measure the division ratio and readjust the DAC one should get a ripple free output with the precision of the PWM.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on June 20, 2022, 06:28:39 pm
While playing with LTSpice and simulating the "classic" schematics (with an npn buffer transistor after the opamp) I can see the transistor significantly lowers the PSRR of the entire circuit. The typical PSRR values I see are 30-100x only (the higher Vcc the higher PSRR). Is that caused by the transistor amplifying somehow the Vcc noise? I always thought the diodes (after the opamp) and the transistor are within the op-amp loop..
For example - below - the 10mVpp square noise pulses (50kHz) injected into the Vcc translate into 175uVpp at the 10V output..

PS:
With C1=100p and 10mVpp 50kHz square the 10V ripple is 144uVpp.
With C1=100p and 10mVpp 5kHz square the 10V ripple is 133uVpp.

With C1=100n and 10mVpp sine 100Hz the 10V ripple is 62uVpp.
With C1=10n and 10mVpp sine 100Hz the 10V ripple is 6uVpp.
With C1=1n and 10mVpp sine 100Hz the 10V ripple is 0.95uVpp.
With C1=100p and 10mVpp sine 100Hz the 10V ripple is not measurable (nil).

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on June 20, 2022, 06:43:06 pm
The capacitor C1 slows down the OP loop quite a bit. So it can not compensate for the fast transients. If the relatively large capacitor is needed, one should also filter the supply sufficently to reduce the fast part in the supply ripple.
Title: Re: LM399 based 10 V reference
Post by: branadic on June 20, 2022, 07:19:38 pm
Quote
To be honest i looked into this after reading the 2019 reports of Andreas and branadic on their experiments with PWM gain stages (see above). Meanwhile they seem to have better results, too. As far as i remember they initiated kind of a calibration circle in May, using a PWM.

The reference only uses the modification suggested by Kleinstein (see attachement). It was then monitored with a K2182A @ 1NPLC without anything obvious and afterwards trimmed to a certain value that I can not tell at the moment as the (blind) ring comparison is still ongoing.

-branadic-
Title: Re: LM399 based 10 V reference
Post by: n_haku on September 25, 2022, 09:40:23 am
Sorry of sudden question (and hope it's right place):
Though I start interest in voltnutting pretty recently (maybe approx a year) I see a lot of zener's buffer here, in datasheets (like on 1st img) and in DVM's schematics, and almost in all cases zener startup is like on 1st img,  but why not use jfet startup, like in img2? It provide reliable startup and almost no current after, so high PSRR. I see no drawbacks in this setup, but as almost no one incl big companies use it, seems I miss something? Well, I could guess a HF noize feedthru via parasistic capacitance, but resistor behaves same, and this could be reduced with bead.
PS: I ended assembling alpha version of pure DAC reference scaler, it's intended to use w/lm399, but as it's hard to buy it for me now, I use jfet (as in JVR thread). It's ok to post here, or create dedicated thread?
Title: Re: LM399 based 10 V reference
Post by: MiDi on September 25, 2022, 11:00:10 am
Nothing wrong in using JFET for startup, just additional cost.
With non RRO op amp and op amp V- >= Vz- the startup circuit could even be omitted.
PSRR for startup resistor is dRz * dVsup / (Rs *  Vz), dRz: dynamic impedance of zener, dVsup: change in supply voltage, Rs: startup resistor, Vz: zener voltage.
For Rs = 200k, dRz = 1.5 it is around 1ppm/V (ADR1399 around 0.1ppm/V).
So if the supply is stable to (or noise) better than 100mV, a startup resistor has negligible effect on zener voltage.
Title: Re: LM399 based 10 V reference
Post by: Andreas on September 25, 2022, 12:21:51 pm
Well, I could guess a HF noize feedthru via parasistic capacitance, but resistor behaves same, and this could be reduced with bead.
Hello,

a JFET is much more sensible to RF than a resistor.
I used single JFETs as 3-5V voltage regulators.
But without 100 nF at in and output the JFET likely oscillates.
So your cirquit is missing at least 1 capacitor.

See here:

https://www.eevblog.com/forum/metrology/ad587lw-10v-precision-travel-standard/msg1449488/#msg1449488 (https://www.eevblog.com/forum/metrology/ad587lw-10v-precision-travel-standard/msg1449488/#msg1449488)
https://www.eevblog.com/forum/metrology/ad587lw-10v-precision-travel-standard/?action=dlattach;attach=402886 (https://www.eevblog.com/forum/metrology/ad587lw-10v-precision-travel-standard/?action=dlattach;attach=402886)

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: n_haku on September 25, 2022, 02:17:43 pm
Nothing wrong in using JFET for startup, just additional cost.
Well, I had feeling that answer is likely as you described: it definetly makes sense with jfet based reference, as it have pretty high impedance (typ few kOhms), and just useless additional cost for ordinary zener.

I used single JFETs as 3-5V voltage regulators.
What a coincidence, my scaler use jfet as zero-current regulator for pwm microcontroller. Of course here and there capacitors on power lines.  Just checked noise with ACV mode in DVM, seems all ok.
Also, I think this irrelevant to startup, as there will be almost no current. Anyway, thanks for information.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on September 26, 2022, 02:19:56 pm
Another cheap and perfect solution is with a second zener, connected from Vcc. Of course for this to work, a stable supply is required - which is a good idea anyway.
Or a voltage divider from Vcc with a diode as gate (2 resistors + diode).

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: DavidAzulay5000 on March 22, 2023, 10:55:07 am
my 10 volt ref
base on LM399A /LT3042 /LTC2057

https://www.youtube.com/watch?v=6UVrzmMod7Y (https://www.youtube.com/watch?v=6UVrzmMod7Y)
Title: Re: LM399 based 10 V reference
Post by: DavidAzulay5000 on March 25, 2023, 08:28:33 am
Can you be interested in precision cookies?

48 hours in a row
with a voltage of 10.000.00
the LM399A zener output voltage 7.086.15V
the lt3042 output volt 15.000.2V
Total current consumption of the circuit 27 milliamps.

During the transitional season in Israel
that at night it is relatively cold 12 degrees and in the day 23 degrees
The change ranges from
10.000.00
to
10.000.03
This LM399A has gone through about 500 working hours
https://www.youtube.com/watch?v=NoLZ1ftXLwc (https://www.youtube.com/watch?v=NoLZ1ftXLwc)
Title: Re: LM399 based 10 V reference
Post by: jorgemef on January 16, 2024, 10:16:38 am
Hello,

I am planning to build two of these boards. Since is not just pennies invested I would like to ask your advice.

1 - For the Opamp I have 3 options. LT1001, LT1012 and LT1037 (and OP07). Which one is best for this application?
2 - Should VOffset compensation be used or not used for the Opamps?
3 - Is it better for dual layer or single layer? With or without ground planes? Remove copper area around LM399? Remove FRP as well for betther thermal isolation?
4 - Power filtering, on board or off board? I am planning to add the layout for it if in future I want to add some permanent low noise DC source but at the moment was planning to feed it with some X20 parkside battery. :)



Cheers,
Jorge
Title: Re: LM399 based 10 V reference
Post by: iMo on January 16, 2024, 11:02:06 am
All those opamps are usable, imho, and "I would.."
1. not use the voffs compensation,
2. .. btw ADI recommends as less copper as possible beneath the 399 (to limit the thermal flow)
3. add 1u ser 5ohm at the 399 zener (as the option for the 1399)
4. use no ground planes (use tracks in the star config)
5. not put the voltage regulators on the board (they will change their temperature based on the input voltage change)
6. use a current limiter made of a npn transistor (Q2 with 1k in the base) and a shunt resistor, ie 27ohm for 20-25mA limit (none R2)
7. wire the C2 to the opamp's output (and something like 4n7-10n)
8. not use R12 (or optional on the pcb)
9. ..btw the R11 shall allow some 1mA for 399 and 3mA for 1399
10. add 1-10u ser 5ohm at the output
11. add a 12V TVS diode at the output.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 16, 2024, 11:59:16 am
For the OP amps suitable candidates are the LT1001, OP07 and newer  ADA4077, OPA202 or OPA205 or OPA207.
The LT1037 is not at all working as it is not stable at such a low gain.
The LT1012 is a super low bias type and a bit high in voltage noise - not a total fail, but still not a good choice.
The precision OP-amps are usually good enough with the offset and not need to trim - the trimmers are an additional possible point of failure and drift source.
Offset from the OP-amps would be seen as part of the reference voltage.

The capacitor C2 in the circuit is at the wrong place. There should be a capacitor in direct feedback at the OP instead (between pins 7 and 2).

For the circuit a single layer would be enough, but the PCBs from China a 2 layers anyway. So one can use 2 layers. A ground plane is good for fast circuits that have possible EMI isssues. A star ground is the more suitable way for a precision more DC circuit.  One can still fill areas with ground, but a start ground should have priority.

The area around the LM399 should have not too much copper and may be relatively thin traces to the LM399 to reduce the heater power a little. Extra PCB cut outs are more overkill.

The supply parts as shown in the PDF is bad - the LM399 does not like when the negative side of the heater is not connected and thus gets lifted higher than the ref. ground.
The logical and simpler solution is to use only a positive supply and have the heater neg. side at GND, like the reference part, just with separate links to the center ground.
For the voltae regulator one might consider using more than just 12 V for the heater - ideally as much as avialable from the battery (e.g. 15 V with a low drop regulator). That should be about the lower end for a 5 cell Li battery pack.  There is usually little need for extra supply filtering. The battery is essentially noise free. The main other alternative would be based on a classical transformer - so one could include a rectifier and buffer cap if one considers this.

I would consider to also have a 2nd OP-amp to buffer the 7 V voltage, so to have a 2nd output that is not effected by resistor drift.
Another small addition would be space for a RC series element (e.g. 1 µF + 5 ohm) parallel to the zener so that one has the option to use an ADR1399 as a upgraded version for lower noise.


The use of 8 equal resistors for the 7 to 10 V step makes sense especially if the resistors are from a resistor array. Chances are one would have space to allow both options: a resistor array like NOMCT (SO16) or TDP (DIP16) and the seprate resistors as currently planed. Depending on the resistors used one may not need the copper for TC compensation.
Title: Re: LM399 based 10 V reference
Post by: Eraldo on January 17, 2024, 09:23:21 pm
I have been thinking, what about using a boost converter at the input with a 7815 in series. Would the noise be too high to be usable for the reference?

The biggest problem would be high frequencu noise but a simple filter would do the job. Also some of these converters offer quite good noise values in the single digit mV rms range. You could also put a metal can on top of it

Anyone tried it before or any reason this would be a bad idea
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 17, 2024, 09:30:12 pm
With enough filtering one could use a boost converter, but it still need effort and care with the layout. It can be done, but the return is not that great. One can as well start with a higher voltage from the transformer / battery. Ideally this would be in a way (e.g. with diodes) to allow seamless change from one to the other. The power consumption of the reference is not very low ( e.g. 200 mW range) and one would thus anyway want a relatively large battery.
Title: Re: LM399 based 10 V reference
Post by: Wolfgang on January 17, 2024, 10:20:28 pm
With enough filtering one could use a boost converter, but it still need effort and care with the layout. It can be done, but the return is not that great. One can as well start with a higher voltage from the transformer / battery. Ideally this would be in a way (e.g. with diodes) to allow seamless change from one to the other. The power consumption of the reference is not very low ( e.g. 200 mW range) and one would thus anyway want a relatively large battery.

What can be tried are converters that support limiting switching speeds in order to reduce RFI (they more or less create a rectangle with slow ramps). I think TI has a chip for that.
Title: Re: LM399 based 10 V reference
Post by: iMo on January 18, 2024, 09:49:07 am
..and perhaps get rid of the negative supply.. Single 15V would be ok, the heater current will be higher.

What Kleinstein suggests above with two power inputs isolated with the diodes is important - I've been using that in my ADR reference - I have the internal linear source (trafo+723) and 2 posts for an external battery. Below a simplified wiring. I even put a resistor in parallel with the isolation diode such I can peek the voltage inside (the input of the 723) and also trickle charge the battery a little bit (not using trickle charging much). 99% of time it runs off the trafo.

While on the shelf it runs off the trafo, when moving it to LABs or measuring it is powered off the batteries (5x 18650). You may wire the batteries anytime on/off..

At the output I've been using a C-L-C filter as well (RF suppression) - like 1n ceramics and 3-4 turns bifilar on a small FT43 toroid.
Title: Re: LM399 based 10 V reference
Post by: jorgemef on January 31, 2024, 11:57:00 am
I have redone my circuit and implemented it.
I am seing some fluctuation with temperature which I find strange. I traced it to the Zenner but I cannot tell if is leading or lead as is bootstraped to the 10v. The gain resistors are all from same batch so TCR influence from them should be minimal or none.

Is it coming from the zenner, and is it an acceptable level?

Adding the charts on the zenner, buffering opamp (gain1) and 10V output.

R3 and D2 are not populated neither R11/R15.

Cheers,
Jorge
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 31, 2024, 12:11:42 pm
The voltage still looks a bit noisy and the temperature effect is also not normal. Some of the noise can be from the meter, but I would not expect that much (depends on the settings).

The layout is not perfect in seprating the ground for the heater and reference. So some of the heater current can couple to the reference.
Title: Re: LM399 based 10 V reference
Post by: Gyro on January 31, 2024, 01:05:49 pm
I notice that you have uses SRBP for the PCB rather than FR4. SRBP has poor dimensional stability with temperature. I don't know if this is going to introduce additional temperature related stresses on the SMDs and opamp legs?
Title: Re: LM399 based 10 V reference
Post by: iMo on January 31, 2024, 01:12:18 pm
You should see something like 3-5uVpp noise directly at the zener in quiet periods, imho (always use the same NPLC).
The "ratio TC" of the divider causes the 10V follows the temperature as well.
Mind the zener itself is usually equal or better than +/- 3.5uV/C.
Your "ratio TC" of your divider has to be similar or lower.
The same batch guaranties nothing.. You may calculate the TC of your divider when you plot the 10V voltage against temperature in the excel and you make the linear trend, you get the number (do the "scatter graph"). Do the same for the 7V output.. Subtract TC_7V from TC_10V then..
Also try to compute the standard dev of your 7V output, it helps with the noise comparison against your future improvements..

PS: below an example of TC (+1.28uV/C)

PPS: you may also post your data in a .csv or excel file, and we may have look at it..
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 31, 2024, 07:24:54 pm
If a temperature variation of 29 - 27.9 = 1.1 K gives 57 - 37 = 20 uv this is about 3 ppm/K of 7V after the amplifier, more than expected for a LM399. For the 10 V output the TC is more like 6 ppm/K. Maybe the meter used to determine these graphs exhibits some TC to start with.
The difference between 3 and 6 ppm/K is certainly caused by the voltage divider. What is the TC spec of those resistors?

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on January 31, 2024, 08:21:10 pm
..and what is your voltage regulator there? The resistors values around it are quite large..
Title: Re: LM399 based 10 V reference
Post by: jorgemef on January 31, 2024, 08:28:27 pm
If a temperature variation of 29 - 27.9 = 1.1 K gives 57 - 37 = 20 uv this is about 3 ppm/K of 7V after the amplifier, more than expected for a LM399. For the 10 V output the TC is more like 6 ppm/K. Maybe the meter used to determine these graphs exhibits some TC to start with.
The difference between 3 and 6 ppm/K is certainly caused by the voltage divider. What is the TC spec of those resistors?

Regards, Dieter

1/2W 207. As per seller specs is about +-10ppm/C. https://pt.aliexpress.com/item/1005006320806775.html

So you mean the circuit is performing as per component specs, and I need to look for strategy to reduce the circuit TC like adding copper resistance on the lower side of the feedback network (and other resistor on the upper side to maintain the 10v)?
Title: Re: LM399 based 10 V reference
Post by: jorgemef on January 31, 2024, 08:32:06 pm
..and what is your voltage regulator there? The resistors values around it are quite large..

It is a HT7550. I keept the resistors high to keep low bias current. The circuit is slow due to the time to charge the capacitors. Takes around 1 minute to reach final voltage of 14.5V. I am feeding it with a 5S battery.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on January 31, 2024, 09:44:48 pm
With parts from a somewhat dubious source I would not really count on the TC. They only say TC down to 10 ppm/K. This means this series of resistor may offer down to 10 PPM/K grades, but the parts actually sold may still be a lesser grade.
Using resistors from the same batch can result in good TC matching, but it is not guaranteed.

The first point to improve would be to improve the ground routing: take out the shart GND wire near the LT1001 and give the heater current a separate return path to the regulator or power connector.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on January 31, 2024, 09:45:48 pm
If a temperature variation of 29 - 27.9 = 1.1 K gives 57 - 37 = 20 uv this is about 3 ppm/K of 7V after the amplifier, more than expected for a LM399. For the 10 V output the TC is more like 6 ppm/K. Maybe the meter used to determine these graphs exhibits some TC to start with.
The difference between 3 and 6 ppm/K is certainly caused by the voltage divider. What is the TC spec of those resistors?

Regards, Dieter

1/2W 207. As per seller specs is about +-10ppm/C. https://pt.aliexpress.com/item/1005006320806775.html

So you mean the circuit is performing as per component specs, and I need to look for strategy to reduce the circuit TC like adding copper resistance on the lower side of the feedback network (and other resistor on the upper side to maintain the 10v)?
In general 0.1 % resistors are specified with TC of 10 or 15 ppm/K, e.g. PTF56 resistors. So those 1 % resistors you got may be more like 50 or 100 ppm/K and even if they are from the same lot and you use kind of averaging, 3 ppm/K as a residual TC is better than expected. Yes, one can use copper wire to tune the TC. In order to do that one needs some kind of temperature chamber in order to vary the temperature of the circuit while the meter remains at near constant ambient temperature.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on January 31, 2024, 10:08:41 pm
..and what is your voltage regulator there? The resistors values around it are quite large..

It is a HT7550. I keept the resistors high to keep low bias current. The circuit is slow due to the time to charge the capacitors. Takes around 1 minute to reach final voltage of 14.5V. I am feeding it with a 5S battery.

It is a CMOS 100mA low drop regulator with low 5uA quiescent current (at no output current), but why you want to have those resistors so big? I would use something 50-100x smaller.. Bias current is not important here. With such large resistors it may induce some instability, noise etc..
Title: Re: LM399 based 10 V reference
Post by: Andreas on January 31, 2024, 10:17:03 pm
Hello,

the PCB looks not clean, I would wash the PCB before testing.
The "hook" may also introduce errors with many contacts of not equal metals.

1.5 / 2 degrees span to measure T.C. is also not enough.
You may have several degrees temperature differences across a board depending on where you put your temperature sensor.
(especially when you have a heated reference).

That is also the reason why I would not try to compensate several ppm/K with copper resistors. Some 0.1 ppm/K would be ok.
(you never know which temperature should be sensed by the copper).

Did you check wether the heater is working. (how is the current consumption of the PCB?)
The last time when I had a LM399 with large T.C. I recognized that the heater was not working (had a bad contact).

with best regards

Andreas

Title: Re: LM399 based 10 V reference
Post by: iMo on January 31, 2024, 10:34:15 pm
3ppm/C of the divider's ratio TC is would be a pretty good result, but as Andreas wrote the 1C span might be not enough, moreover, his temperature resolution is rather low. I saw that in my measurements too - I had large diffs in TC between smaller and larger temp spans.
Title: Re: LM399 based 10 V reference
Post by: iMo on January 31, 2024, 11:03:50 pm
The Vreg variant.. The two 10u input/output capacitors C1 and C2 around the Vreg should be wired to ground. In your wiring the two important decoupling capacitors C1 and C2 are "missing", because they are wired to ground via 300k..
Title: Re: LM399 based 10 V reference
Post by: jorgemef on January 31, 2024, 11:34:26 pm
The first point to improve would be to improve the ground routing: take out the shart GND wire near the LT1001 and give the heater current a separate return path to the regulator or power connector.

The heater GND and Power follow different path. They meet almoust at power main decoupling cap (470nF).
The ground to the zenner is almoust common to the ground from the LT1001 which is buffering it, and the VZ is very close from IN+ from same buffer opamp.
I increased now the size of the ground and power lines up to the heater lines derivation to try to reduce eventual effects by adding a layer of solder of 1mm.
I also noticed that the 10nF may not be enough to the amplifier stage bypass capacitor. I think I will increase that to 100nF. I see some examples on LT1001 datasheet with it and I have a second board where the noise was close to 300uV. After much scratching my head and many tests I decided to test with outer 22nf cap in the bypass position and the noise immediatly went down, so I guess will go with 100nF.

I think I will follow iMo suggestion for the reduction of the biasing resistors. Makes litle sense to spare 1mA of biasing current when the LM399 are consuming 17-20mA to the heater. :)

As for the controlled temperature room to test this stuff still need to think how I can do it. The air flow of the air conditionair is not the best way to test. The best I can do is turn it off during the night and goes to 18ºC and turn on during day and goes to 24ºC. :) But this also impacts the Keythly2000. :)
Title: Re: LM399 based 10 V reference
Post by: jorgemef on January 31, 2024, 11:39:23 pm
The Vreg variant.. The two 10u input/output capacitors C1 and C2 around the Vreg should be wired to ground. In your wiring the two important decoupling capacitors C1 and C2 are "missing", because they are wired to ground via 300k..

I took this from the datasheet. I found this strange but I guessed it may be to keep the noise low between the output, input and gnd.

Title: Re: LM399 based 10 V reference
Post by: iMo on January 31, 2024, 11:47:49 pm
..and you have to recalculate the R1 and R2 - see below, the same principle as with the LM317. I would wire the C1 and C2 to ground..
For Vout=15V   
R2=2*R1
Title: Re: LM399 based 10 V reference
Post by: jorgemef on February 01, 2024, 12:06:37 am
Went back to #1251 to iMo simulations. :/ I guess will use only 22nF on the one misbehaving or even test a different SMD part of 10nF not to make the circuit more susceptible to the supply noise.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 01, 2024, 12:14:32 am
Went back to #1251 to iMo simulations. :/ I guess will use only 22nF on the one misbehaving or even test a different SMD part of 10nF not to make the circuit more susceptible to the supply noise.

The #1251 is a simulation only.. The opamp models are mostly behavioral (not sure about OP07 there), thus my simulation might not be perfect (or it could be imprecise). Anyhow, the larger the capacitor in the feedback the slower the opamp's response and therefore it may respond to the power rail noise/ripple slower (experts here may clarify).. Also mind your schematics is a little bit different.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 01, 2024, 06:45:02 am
For the heater ground I somewhat mixed up the heater and zener GND. So the layout in this respect is OK, though not great with the OP amp supply shared with the ref. votlage.

For the capacitor in the LT1001 feedback (C7)  10 nF should be large enough. It may be an issue if the output emitter follower sees very little current. Currently this is only R16. At least for the PCB this is 1 K and not 5 K as in the plan. 5 K would be borderline low current for the reference.  A larger capacitor for C7 would be needed with really large capacitive load..


The position of the capacitor (C1 in the HT75xx datasheet) at the regulator input is defintely wrong - it looks like a mistake in the regulator data sheet.  As shown the input side capacitor would couple supply noise directly to the regulator feedback, forming a divider with the capacitor on the output side. So one would get a PSRR of only 6dB, once the capacitors take over. The capacitor at the output side may be OK.

edit:
The capacitor at the output side is also bad, limiting the speed how fast the regulator can work. 
The
Title: Re: LM399 based 10 V reference
Post by: jorgemef on February 01, 2024, 07:59:59 am
I replaced the 10nF SMD with another one of same value on the second board and noise came down, so probably stress from soldering or bad part.
So now noise on the second board at 10V output is of 10uV over 30 samples/seconds like in the other board.

I have used 3K for the current refference. Need to update the schematic later. :)

Will update the C1 position and reduce the voltage divider resistances there to make the response faster
Title: Re: LM399 based 10 V reference
Post by: iMo on February 01, 2024, 08:59:36 am
..Will update the C1 position and reduce the voltage divider resistances there to make the response faster
Your board will take perhaps 80-100mA when cold, after couple of seconds when already hot the current drops down to perhaps 30-40mA. If I were you I would go with R1=1k and R2=2k (in my Vreg schematics) or something like that.
C1..100uF/35V and in parallel 100n/50V ceramics.

.. and the power loss (==heat) of your voltage regulator: with 21V input (with your 5S) and 15V output the loss when zener is cold will be aprox 0.55W (!!), when zener is hot 0.17W (!)..
Your HT7550 will be hot (like 40-45C at T=23C ambient, add say 5-10C coming off the 399 and others). Do not use more than 16-17V input. It could be your 5S battery is too much (21V max), try with 4S instead (16.8V max).

.. and re #1262 point 5. - when the battery voltage goes down by 1V, for example, the heat dissipation at the 7550 goes down by aprox 30-40mW.
That creates a temperature drop around the 7550 by aprox 3-4degC (doublecheck that).. Multiply it by your TC..
Title: Re: LM399 based 10 V reference
Post by: jorgemef on February 01, 2024, 11:33:26 am
Do not use more than 16-17V input. It could be your 5S battery is too much (21V max), try with 4S instead (16.8V max).


I considered using the 4S but then the lower range of the battery pack would be 12V. That is why I added the jumper to be able to set the config of the LDO to 12V.
But I see the 10V output changes when switching from 15V to 12V by 40 uV. Not sure if the LM399 behaves well with 12V or something else going on with the opamp headrooms.
The datasheet specifies 2V dropout for the HT7550 so that would be additional challange for the regulation with a 4S.

I will do some measurements. If I see problematic on keeping the regulator onboard will probably build a doughter board to keep the regulator temperature out and just shunt VIN/VOUT on the board layout for the HT7550.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 01, 2024, 03:05:20 pm
I can see "100mV" dropout in my DS (at 1mA out), with 30-40mA it could be more..
Would be better to have the Vreg off the reference board and use a higher input voltage, sure..
Title: Re: LM399 based 10 V reference
Post by: jorgemef on February 01, 2024, 03:15:28 pm
I remembered Dave video that opamp offset changes with supply voltage. I have two in series. Now I realize why changing from 15V towards 12V the 10V output changes. :) I imagine that same happens in a smaller scale as the battery empties or you have changing load  but the LDO keeps this in check for some degree.
Title: Re: LM399 based 10 V reference
Post by: David Hess on February 01, 2024, 04:27:29 pm
I remembered Dave video that opamp offset changes with supply voltage. I have two in series. Now I realize why changing from 15V towards 12V the 10V output changes. :) I imagine that same happens in a smaller scale as the battery empties or you have changing load  but the LDO keeps this in check for some degree.

PSRR (power supply rejection ratio) is the change in offset voltage for a change in the magnitude of the supply voltage.  CMRR (common mode rejection ratio) is the change in offset voltage for a change in the voltage at the non-inverting input.  If the voltage at only one of the supply pins changes, then both apply.

For precision operational amplifiers, PSRR can usually be ignored, and CMRR can be ignored if the non-inverting input does not change much.  Changes in offset of better than 1 microvolt per volt of power supply or common mode change is typical for precision parts.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 01, 2024, 04:52:31 pm
With battery operation it would make sense to have a low drop out regulator. This way one could use more of the battery voltage. This is more improtant than the ground current of the voltage regulator.

The PSRR of the OP-amps is usually quite good. A second path is via the reference heater. The heater voltage has a slight effect on the actual temperature.
The voltage regulator should keep the voltage variations small enough to not have a problem from the supply side.

With a 12 V supply the output swing for the LT1001 may not be sufficient. The specs give some 1.5 to 2.5 V that are lost at the upper end and there is also the emitter follower that looses some extra 0.6 V. So one could just hit the upper output limit and to drive the output that hard it may need a little extra voltage at the input. From this I would consider some 13 V as the lower limit for the LT1001.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 01, 2024, 08:24:43 pm
Hello,

there are better output cirquits for low headroom.
See e.g. DATRON 4912 cirquit for 10V output from a 12V supply.

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: iMo on February 01, 2024, 10:43:21 pm
I can see "100mV" dropout in my DS (at 1mA out), with 30-40mA it could be more..

In Farnell's datasheet for the HT7550 I've found the definition for the "Dropout Voltage"  ::)

Quote
Dropout voltage is defined as the input voltage minus the output voltage that produces a 2% change in the output voltage from the value at VIN= VOUT+2V with a fixed load.

For the 7550 it is 25mV typ., max 55mV (at 1mA out). Thus it seems it cannot be powered from 4S (for the 15V out), indeed..

Holtek's datasheet calls it the "Voltage Drop", for 7550 it is 100mV typ. (at 1mA out), with no further explanation.

PS: you may have a look at the TI's slup239a.pdf talking the dropout voltage confusion..
Title: Re: LM399 based 10 V reference
Post by: jorgemef on February 06, 2024, 08:41:36 am
I had to change the feedback network on one of the sets to have headroom to compensate with coper wire and have adjustment floor to the 10v.
The thing went from +10uv/C to -50uV/C so the resistors are not that great. :)
Anyway with two coper wire resistors I made the thing come down to 3uV/C. I think I will call it done now. :)
I added the coper wire resistors in the other side of the board where the feedback resistor network are with thermal glue to have proper thermal coupling.

Also some picture of the heat distribuition on the board with Vin=20v.

Cheers,
Jorge
Title: Re: LM399 based 10 V reference
Post by: ivo on February 14, 2024, 10:24:50 am
1/2W 207. As per seller specs is about +-10ppm/C. https://pt.aliexpress.com/item/1005006320806775.html (https://pt.aliexpress.com/item/1005006320806775.html)

So you mean the circuit is performing as per component specs, and I need to look for strategy to reduce the circuit TC like adding copper resistance on the lower side of the feedback network (and other resistor on the upper side to maintain the 10v)?

You can't expect cheap good resistors from aliexpress! They are probably gathering the lowest-spec rated TC parts, not the best (yes the best is 10ppm rated, but which rated parts did they decide to sell? If they don't tell you...). But it is not impossible to get okay parts.

https://www.lcsc.com/product-detail/Chip-Resistor-Surface-Mount_Milliohm-HoAR0805-1-10W-10KR-0-1-TCR5_C2912582.html (https://www.lcsc.com/product-detail/Chip-Resistor-Surface-Mount_Milliohm-HoAR0805-1-10W-10KR-0-1-TCR5_C2912582.html) single 10k resistors rated at 5ppm.

Even better is a Vishay TOMC16031002BUF which is 8 resistors for ~$6 all thermally matched in a single part, divide and combine it how you like for a divider. 10k & 22k is a good start for 6.9V->10V noninverting opamp. Include one or two if you have an order from a western supplier you need to make.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 14, 2024, 11:00:06 am
Poor quality resistors not only have a problem with the TC but also with long term drift and noise.  For a reference source the long term drift is the more serious issue. Even with high grade parts it is hard to get specs on the long term drift and already the possibilty for drift is an issue, as one would not easily know if they drift.
The TC part is only the point that is easy to measure and compare.  A resistor array would really be the more sensible choice.
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 19, 2024, 08:23:16 pm
I decided to build second PCB version of my LM399/ADR1399 reference board.
One improvement point was - buffering the opamp output /as proposed many times in this thread with two npn transistors/
But I sow major drawback of proposed circuit - It is excelen voltage source and almost zero current sink. Zoutput source ~ 0 ohm    while Sink capability is limited to pulldown resistor ~22..100KOhms. see attachment.  I come to conclusion that this solution introduce more problems that it solves. Am I correct? Do I miss something?
I tend to use - combined Opamps Outer loop opa2182 for DC accuracy  inner loop//AD706.707 have many in my drawer / - for lower noise, filter/remove chopper artifact and reduce thermal stress on OPA2182

 
Title: Re: LM399 based 10 V reference
Post by: iMo on February 19, 2024, 09:19:20 pm
People sometimes use the opamp+LT1010 as well..
Title: Re: LM399 based 10 V reference
Post by: David Hess on February 19, 2024, 09:53:29 pm
People sometimes use the opamp+LT1010 as well..

I am not sure what that was a response to.

Buffering a precision operational amplifier preserves precision and there are lots of ways to do it.  A simple class-a emitter follower may be enough if there is a significant load.  A diamond buffer or current feedback buffer also works well and is likely less expensive than an LT1010.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 19, 2024, 10:02:05 pm
If one wants a seprate current buffer inside the loop, that extra buffer should be fast (faster than the outer loop, if there is no extra AC feedback), but it does not have to be accurate. So the AD706/7ß7 are not a good choice. It would be more like TL071 or NE5532.

The version with just the emitter follower is asymmetric in the source / sink capability, but one could add more resistive load (e.g. 5 K instead of the 22 K) if needed. The output impedance would still be just 1 value, unless the output needs to sink a much of the current from the resistor.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 19, 2024, 10:09:04 pm
People sometimes use the opamp+LT1010 as well..
I am not sure what that was a response to.
What is wrong w/ it? LT1010 is 20MHz BW, 75V/us, +/-150mA.
Within the loop with the opamp ("opamp+LT1010") it could provide source or sink capability easily.
Title: Re: LM399 based 10 V reference
Post by: David Hess on February 20, 2024, 12:39:20 am
What is wrong w/ it? LT1010 is 20MHz BW, 75V/us, +/-150mA.
Within the loop with the opamp ("opamp+LT1010") it could provide source or sink capability easily.

There is nothing wrong with using the LT1010; it is just more expensive than the alternatives.
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 20, 2024, 07:21:04 am
Thanks, Kleinstein, David and Imo. I've got your point.
My point was that emitter follower promoted many times in this thread makes more harm than good.  Better choice is without it or alternative solution with composed amplifier. I agree that current feedback is the most suitable. Unfortunately I don't have experience with single ended current feedback amplifiers. Can somebody propose an suitable component. Requirement are low power, at least +15V Vcc, low noise and THD.

PS: I use such combo regularly to drive my ADCs input and Vref- OPA189/182 + Opa2625/AD8065. I find A8065 over the top for this application

Title: Re: LM399 based 10 V reference
Post by: iMo on February 20, 2024, 07:54:05 am
..
My point was that emitter follower promoted many times in this thread makes more harm than good..

Could you elaborate a little bit what kind of harm it may cause?
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 20, 2024, 08:23:05 am
..
My point was that emitter follower promoted many times in this thread makes more harm than good..

Could you elaborate a little bit what kind of harm it may cause?
Output impedance of 10..100K is no go for.
questions are
- how this circuit handle the chopper input of any DMM.
- Can I connect this output to any switched capacitor input? Sayying in simple enginering  jartgon - any CMOS  DAC ADC input or Vref
- Is there an use case where this circuit works?

My conclusion - this circuit is is only sutable to heat 10MOhm resistor. The switching behavour of DMM frontend and modern CMOS circuit requires low impedance sink and source output
Title: Re: LM399 based 10 V reference
Post by: iMo on February 20, 2024, 08:59:21 am
That circuits people use with their 10Mohm (and higher) loads because it de-loads the 7->10V opamp and it has a current limiting function. For coping with xx nanoseconds wide fluctuations at the inputs of modern choppers or modern adcs you would need something better, indeed.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 20, 2024, 10:30:10 am
The emitter follower has a relatively low output impedance. Typical a few ohms from the parasitic emitter resistance and than some 26 Ohm divided by the current in mA. So with a 20 K load (0.5 mA of current for the transistor) this would be a little more than 50 ohm - about comparable to the open loop output resistance of OP-amps.  This is the resistance without the OP-amp for the control loop - with the feedback one would have an output restance similar to the case without the emitter follower, just without the thermal error and no output stage cross over as the output stage is class A.  If there is no problem with oscillation due to excessive capacitive load the output should be able to also drive larger loads.

When driving directly a switched capacitor type ADC (most SD and SAR ADC chips) it would need extra care / filtering close to the ADC and ideally a ADC driver at the ADC. These ADCs just need a defined drive characteristic and generally don't work well with a variable impedance - the cable length alone could make a difference. So directly to the ADC is not a good input design.
Also the ADC / DAC ref. inputs may need a dedicated driver close to the chip and some want quite some capacitance there.

Chopped inputs at a DMM should include the filtering needed, so that the source impedance is not as critical. This is one difficulty in using AZ OP amps at a DMM input. If the input is sensitive to the source impedance this is more a problem with the DMM not the signal source.  The HP like AZ switching produces fewer (e.g. 5-50 Hz compared to some 50 kHz) current spikes than AZ amplifiers, but the spikes tend to be larger and longer. Some settling time is usually alowed, but this may not be enough for a slow driver.   The current spike from AZ switching could be an issue with the output, if build to tolerate a large capacitance, as this slow down the settling.

There is nothing bad with the LT1010. Similar to the AD8065 it may be a bit over the top. The in loop driver does not need to be low drift, low noise and low THD. One would like good speed and good tolerance to capacitive loading, whicht often comes with a bit higher power consumption. Quite some of the faster OP-amps are made with audio in mind and are thus still low noise and low THD, but this is not really needed. So the NE5532 was suggested for it's speed and availabilty, not the noise or THD. It is however quite power hungry. For a 10 V output is usually OK to power the amplifier from one the +15 ond GND, with no need to also have -15 V.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 20, 2024, 10:47:12 am
The output impedance in Ohm..
Title: Re: LM399 based 10 V reference
Post by: r6502 on February 20, 2024, 08:18:56 pm
@MIRO123:
Does your schematic contain an error?
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=2026559;image)

It helps, when you generate a part with multiple units. This units you can than pace like single parts on the schematic, see sample below, just a sample, but much clearer schematic and helps to prevent errors.
(https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/?action=dlattach;attach=2026517;image)

Guido
Title: Re: LM399 based 10 V reference
Post by: iMo on February 20, 2024, 09:53:47 pm
Despite the fact the output impedance of the emitter follower buffer is low, it handles the pulses coming from DMM's side poorly. I've made sims where I can see the output voltage changes based on pulses parameters.

If somebody knows how the DMM pulses look like we may make the sim (I think there was a post in past on the DMM's output transients comparison, but I cannot find the post).

The buffer with ie AD8065 works much better in the sim, of course.


Edit: Nope, the buffer with the emitter follower (with the current limit) behaves the same as with the AD8605. My error in the simulation..
I deleted the next post on the droop as well..
Title: Re: LM399 based 10 V reference
Post by: Wolfgang on February 20, 2024, 10:53:01 pm
The regulating loop is slow, and the output cap is a bit meager. So pulse loads will not be regulated out in time.
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 21, 2024, 09:49:26 am
If somebody knows how the DMM pulses look like we may make the sim
See attached screenshot from Keysight 3446x data-shit. "Noise and injected current: Keysight Truevolt DMMs contribute less than 30% of the injected current than alternatives. Compared to some lower cost alternatives, Truevolt DMMs offer almost 100% less noise"

My trial and error experience is that my PICO M3500 is the worse of all my own DMMs - If I have time I can measure it with my Keithley DMM6500 digitizer. I have bunch of HP3456, HP34970 and Pico/Array M3500.
Back on topic
  - My goal is to make reliable voltage reference. Reliable translates to lower uncertainty
  - It must work in forward and reverse polarity on battery as well as transformer powered.
  - I want to use the voltage reference in real modern world. I dont blame for fout the DMMs auto-zero implementation or the modern zero-drift OpAmps. I assuem that they exist.
  - cope with EMI challenges from modern world - from solar Inverters, to EV chargers. from LED bulbs to BLDC air-conditioning /refrigerator.
Title: Re: LM399 based 10 V reference
Post by: David Hess on February 21, 2024, 10:09:47 am
In the same sim schematics and the same pulses: with the transistors the voltage droop is say 12ppm, when transistors are replaced by a faster opamp like the AD8065 there are just 0.1ppm transients left.

An class-ab stage doubles the transconductance so should have half of the voltage droop.

I suspect this explains why Sziklai pairs are sometimes used at low current, especially when the first transistor is an FET.
Title: Re: LM399 based 10 V reference
Post by: guenthert on February 21, 2024, 10:42:05 am
The motivation for the transistor behind the OpAmp were
      i) boost output current
      ii) cheap device to be replaced (there was a report regarding failed OpAmp potentially due to ESD) and
      iii) avoid uneven warming of the OpAmp due to excessive load, which might change input offset voltage

The pulses of DVMs are intense (in the order of a few nA -- observable using an oscilloscope), but brief.  Unless one uses very short integration times, the effect ought to be small.  It would be interesting to compare implementations with and w/o output transistor.  I recall having read the advice to observe the deviation when connecting a DVM using a passive null detector (I suppose a galvanometer was meant, but I don't think the passive ones were sensitive and certainly not fast enough)

To put things into perspective, the Fluke 515A portable standard has an output resistance (DC 1V, 10V ranges) of 300Ohm, the HP735A transfer standard even 1kOhm (comparable to Weston standard cells), but yeah, they can hardly be called 'modern'.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 21, 2024, 11:26:10 am
In the same sim schematics and the same pulses: with the transistors the voltage droop is say 12ppm, when transistors are replaced by a faster opamp like the AD8065 there are just 0.1ppm transients left.

An class-ab stage doubles the transconductance so should have half of the voltage droop.

I suspect this explains why Sziklai pairs are sometimes used at low current, especially when the first transistor is an FET.

See above I corrected the post on the sim -> no 12ppm droop, my mistake. There is a transient visible, like 1uV with 100nA/50ns pulse injected into the output.

Added 500ns long 1uA pulse, some 6uV response.

Added 500ns long -1uA pulse, some -6uV response.

Added 500ns long -1uA pulse with 2u2 capacitor, some -4uV response.

The 4.7ohm is not "esr" but "ser".
Title: Re: LM399 based 10 V reference
Post by: iMo on February 21, 2024, 12:50:01 pm
And the stability plots (FRA) with 47p/470p/4n7/47n FB capacitor and 2u2 ser 4R7 load.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 21, 2024, 02:02:49 pm
It absolutely makes sense to want an output that is no effected by a modern DMM. So good tolerance to current spikes and fast settling is really desirable. For this it help, if there is no large capacitance (e.g. 100 nF range) at the output. Getting fast settling with a large capacitor gets really hard. It could also help to have some actual resistance (e.g. 50 ohm range) in series, to reduce RF oddities from cable reflections.


The current spikes from AZ switching should be larger and with most meters also longer. The precharge part can compensate for this somewhat, but there is still the nonlinearity in the capacitance of the FETs.
Especially with a non zero voltage expect something like some 1-10 pQ of charge pulse, that would be some 1-10 µA for 1 µs or maybe a bit more stretched out.
The pulse from an AZ amplifier are shorter and smaller charge, but a lot more of them. There should be at least some filtering meter internal to smear out the pulse and make the peak current smaller.


I don't know what meters KS used in the comparison. A candidate for the really bad one is the ADT R6581.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 21, 2024, 02:43:50 pm
.. a voltnut here made the measurements years back, for several meters afaik, the pulses repetition rates and amplitudes..
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 21, 2024, 07:05:29 pm
Hello,

the question is: are these peak values, DC-values or RMS values? (with a large peak to RMS ratio?)

  - It must work in forward and reverse polarity on battery as well as transformer powered.
  - I want to use the voltage reference in real modern world. I dont blame for fout the DMMs auto-zero implementation or the modern zero-drift OpAmps. I assuem that they exist.
  - cope with EMI challenges from modern world - from solar Inverters, to EV chargers. from LED bulbs to BLDC air-conditioning /refrigerator.

So you will have a lot of work and additional components to spend like
- shielded transformer
- EMI filters on output (which increase the output impedance again)
- a good (double shielded) enclosure
...
And how do you check wether the requirements are really met?

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: jorgemef on February 22, 2024, 11:38:08 am
Poor quality resistors not only have a problem with the TC but also with long term drift and noise.  For a reference source the long term drift is the more serious issue. Even with high grade parts it is hard to get specs on the long term drift and already the possibilty for drift is an issue, as one would not easily know if they drift.
The TC part is only the point that is easy to measure and compare.  A resistor array would really be the more sensible choice.

Finally found a local provider that can provide the NOMCT1603 5K at a reasonable price so will give it a try. :) Shipping costs from Mouser or Digikey are around 20 euros so as much as what the local provider will charge for four pieces plus shipping cost inside the country.

At least the tracking TCR should be more stable. I think will attach the part on top of my boards with thermal glue as dead bug configuration with the required connections between pins and just 3 connections to the board.

PS: would it be better to thermocouple to the board the resistor network on on the contrary to try to thermally isolate it like with some plastic spacer? :)
Title: Re: LM399 based 10 V reference
Post by: iMo on February 22, 2024, 01:23:00 pm
PS: would it be better to thermocouple to the board the resistor network on on the contrary to try to thermally isolate it like with some plastic spacer? :)

The temperature differences matter as they may create the EMF (in case of material differences), thus they mess with the voltages somewhere on the board. Also the temperature differences may create thermal waves bouncing in your box, creating "oscillations". So it is a rather difficult area - the modeling of thermal processes on the board and in the box.
The best solution, imho, would be to keep the board and the components in a thermal equilibrium, with constant thermal gradients (whatever they are).

Not an easy task, but the 399 is a cheap noisy reference, so one would not invest too much into it.. As Kleinstein wrote above the long term stability of the resistors is the key you have to focus on. The temperature could be measured (and you can make the TC compensation in math), but long term stability is something you cannot easily predict or adjust.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 24, 2024, 10:45:36 am
While playing with the new LTSpice24 - here are couple of sims with the traditional 7V to 10V buffer made of a single n-jfet and OP07.
The current limit value depends on the Idss, in the sim it is about 48mA.
Mind in case of a short the current flows from the OP07's output via n-jfet's gate, therefore at least 3k3 in the gate would be nice to have.
The low value resistors in series with capacitors are their esrs.

Title: Re: LM399 based 10 V reference
Post by: ivo on February 24, 2024, 11:49:11 am
If one wants a simple buffer in-front of the opamp for the slight bit extra power/safety over its bare output stage, isn't this a classic push pull design? the 1k resistor allows the opamp to fine control the level when it's settled on the exact value, but probably only has to contribute single digit mA if the output is yanked on by a bigger load. I remember reading it from some app note.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 24, 2024, 01:08:40 pm
The transistors in the simple push pull would only work one at a time and not at all a light loads. So this is more like a class B output stage.  One would like more like a class AB output stage, so that both transistors would be active at the same time.  A simple for may work with an NJFET and PNP transistor. To limit the current one would likely need a resistor at the source side and somewhat selected threshold and / or extra diode in series.
If one takes into account current limiting an OP-amps as driver does not look that complicated.

For the response wanted, this may be different dependent on the meter / sensing amplifier in mind. With an AZ amplifier at the meter input one would wont a well behaved response with not much ringing or resonance. Quite some capacitance at the output could be a brute fore way to get there, maybe combined with an RC combination.

For an AZ switching DMM (like more HP meters) this may not be enough, as a large capacitance also slows down the pulse response. So the ideal would be more fast settling after a pulse. There should be at least some 10 µs before the meater actually reads the singal, and for the first few µs one does not need to be settled to sub ppm, just that one average the error is not to large in error. Setting in some 100 µs may still be fast enough if not starting to large in excursion.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 24, 2024, 02:48:17 pm
The settling time sim - after the same NEG pulse as above in my ADR1001#1 buffer.
With the POS pulse it is 46us. The CLC filter adds the damped resonance, of course. Not counting capacitance and inductance of the leads.
Without knowing how the typical pulses look like it would be just wild guessing, imho (after 11 years of this thread)  :D
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 24, 2024, 03:14:37 pm
The settling time should not depend much on the pulse. The pulse can effect how much the settling part is excited, but the time constant should not change.
Title: Re: LM399 based 10 V reference
Post by: iMo on February 24, 2024, 04:29:09 pm
Here is a doc with chopper's pulses at their inputs.
10-15ns wide and 4-15uApp of amplitude..
https://www.eevblog.com/forum/metrology/zero-drift-amplifier-input-bias-current/msg4991131/#msg4991131 (https://www.eevblog.com/forum/metrology/zero-drift-amplifier-input-bias-current/msg4991131/#msg4991131)
The pulses from DMMs may differ significanlty, of course..

PS: Not related to the DMMs inputs directly - a 15ns pulse (2ns edges) -10uA amplitude will create aprox 2uVpp response with 26us settling in above sim with OP07 and jfet transistor. Also mind the modern opamps have the pulses X usecs apart..
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 24, 2024, 10:46:05 pm
The settling time sim - after the same NEG pulse as above in my ADR1001#1 buffer.
With the POS pulse it is 46us. The CLC filter adds the damped resonance, of course. Not counting capacitance and inductance of the leads.
Without knowing how the typical pulses look like it would be just wild guessing, imho (after 11 years of this thread)  :D
Hi Imo,
Few remarks on your simulation
Common mode chokes are not ideal inductance. their core is optimazed for max loss instead of efficiency. It is nice to look at IR images of such chokes.
Common mode chokes also behaves as diferential.
Common mode has parasitic capacitance. Capacitance depend on inductance - number of turns.
Common mode chokes are in many flavors. Depends on application that you want to use it.
More info at manuafascture sites datasheets and ANs. Top manufacture Wurth, TDK - few links https://www.we-online.com/components/products/datasheet/744282100.pdf (https://www.we-online.com/components/products/datasheet/744282100.pdf)
I tend to use them sparely and only If I know which frequency range I want to attenuate.  If you dont follow this rule you end up with worse solution. Having an spectrum analyzer with all accesories arround it help a lot.
The spectrum also depends on type/coax or twisted pair or random wires / and length of used cables
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 24, 2024, 11:18:55 pm
Hello here is the draft schematics revision of 4,66 , 7V and 10V
Few  requirements
Long term stable 6,9...7.1 Vout
4,6V short term stability and lower noise
10V - the most lose specified - I dont know do I really need. It was just available by the  bootstrap circuit.
Opamps are still not specified. I will build at least two  boards + one for test. I will try different opamps.
The board wil be in separated case - Bigger case will contain   Vref + R reference cases. Resistors are almost ready. I still waiting for final cure of the sealant. I wil post some pictures later on. Four selected Vishay foil resistors + NTC+PT100 are submerged in silicone oil  bath.

Comments and remarks are welcome.  I'm sure there are many mistake at this stage. Keep in mind that I'm beginner in CAD systems. Last time I did it ~30 years ago when I was junior engineer :)
BR,
Miro

Title: Re: LM399 based 10 V reference
Post by: iMo on February 25, 2024, 07:24:38 am
..Common mode chokes are not ideal inductance. their core is optimazed for max loss instead of efficiency. It is nice to look at IR images of such chokes.
Common mode chokes also behaves as diferential.
Common mode has parasitic capacitance. Capacitance depend on inductance - number of turns.
Common mode chokes are in many flavors. Depends on application that you want to use it..

My common mode choke is made of a small ferrite toroid core with two or three (I forgot) turns of twisted enameled copper wire coming from the Vref board and going to the terminal posts. Inductance around 10-20uH I guess. The CM choke is intended to suppress the higher frequency bands (like XX MHz up)..

Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 25, 2024, 08:58:35 am
The OPA2209 is a rather expensive choice for the output buffers inside the loop. A NE5532 should be about as good though maybe with a little  more power consumption.
The filters before the output buffers may effect the loop stability. One may want an extra capacitor in direct feedback to get the fast feedback part for U3B and U1B, at least as an option in case it oscillates.

The capacitor C5 at the 10 V output should not come from the ouput, but directly the OP-amp output or maybe from the the other side of R6. The idea is to get the higher frequency feedback this way and this is needed to get the tolerace for large capacitance at the output.

The filtering for the reference voltage looks rather modest. If there is space large capacitors (possibly a few µF) for C1,C6 and C12 can make sense

As a simplification the 4.6 V part could get away be using the 7 V output as a buffer to start with. So one could get away without U3A.

The 10 V output would be about 10 V, not an accurate 10 V, with no trim.

The LM399 heater voltage should ideally also be regulated. There is a slight effect of the voltage on the temperature and this way the ouput voltage.
Title: Re: LM399 based 10 V reference
Post by: Andreas on February 25, 2024, 09:25:39 am
Just some remarks:

U4 seems to have no power supply.

I would use the same protection zener for all outputs.
(All Op-Amps are supplied from the same 15V. So the zener only needs to limit to ~15V with (current limited) voltage at the output).
I prefer the PTVS11VP1UP,115 Diode for my references up to 10V.

Is there any reason to  put R13, R16, R17 outside the feedback loop (R14, R22).
With 10 Meg input resistance of the DMM 20 Ohms give already 2 ppm error.

How do you dimension the output capacitors? (for which frequency coupled into the output)
C4 with 100 pF is rather low compared to ~1 nF of the transient zener.

I would wire C8, C10 not to GND but to the negative Input (PIN 6) of U1B / U3B to get stability against capacitive loads at the output (Coax cables etc.)
(See also AN86 of LT for the combo outputs).

with best regards

Andreas
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 25, 2024, 09:26:50 am
The OPA2209 is a rather expensive choice for the output buffers inside the loop. A NE5532 should be about as good though maybe with a little  more power consumption.
The filters before the output buffers may effect the loop stability. One may want an extra capacitor in direct feedback to get the fast feedback part for U3B and U1B, at least as an option in case it oscillates.
Many thanks for quick and relevant review Kleinstein.
I will evaluate bunch of opamps in first one or two prototype boards. OPA2209 is know component for me but expensive. NE5532 is in my order list for evaluation. I have doubt about 22186 too. It is new component for me. I don't know how bad is compare to my well known OPA2198 opa2182. According to datasheet is more noisy an slower I have no idea about switching frequency.
All other remarks are relevant too.
Thanks
Miro
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 25, 2024, 10:12:26 am
The OPA186 seems to be a lower power version of the OPA187. The noise curves give a crude hint where the chopper frequency should be.
Similar the OPA182 is a slightly lower power / lower speed variant of the OPA189. Here the lower speed can be a good thing as the OPA189 is quite fast, when used with low gain.

The LM399 reference is still relatively noisy (e.g. 100 nV/sqrt(Hz) of white noise part and quite some 1/f noise on top) . So the op-amps for the buffer should not be that critical. The question is more if one needs AZ op-amps with there possible spikes or could use a classic precision OP-amp like OP07, OP177 or newer OPA207, OPA205,  ADA4077.
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 25, 2024, 10:29:26 am
U4 seems to have no power supply.
Thanks you save two patches in PCB
I prefer the PTVS11VP1UP,115 Diode for my references up to 10V.
I will order a bunch of TVS and select the one with lower leakage across T range
Is there any reason to  put R13, R16, R17 outside the feedback loop (R14, R22).
I have doubt too. I still did not make a choise between OpAmp stability vs Performance/LowZ
How do you dimension the output capacitors? (for which frequency coupled into the output)
C4 with 100 pF is rather low compared to ~1 nF of the transient zener.
Still work ToDo - I put some random numbers :-) Thanks for the tip I will on LT AN86

I would wire C8, C10 not to GND but to the negative Input (PIN 6) of U1B / U3B
I dont like placing this capacitors to GND too due to worseing of stability  but I was also aware of choper injection back input. I will place them both and decide which to use later on
Thanks
Miro
Title: Re: LM399 based 10 V reference
Post by: miro123 on February 25, 2024, 10:48:01 am
The LM399 reference is still relatively noisy (e.g. 100 nV/sqrt(Hz) of white noise part and quite some 1/f noise on top) . So the op-amps for the buffer should not be that critical. The question is more if one needs AZ op-amps with there possible spikes or could use a classic precision OP-amp like OP07, OP177 or newer OPA207, OPA205,  ADA4077.
OPA2186 took my atention due to extremely low price. I'm reluctant to use it blindly since it seems different from other TI modern Az opamps. The max voltage is 24V. All other TI AZ OpAmp members have 36V. Is it different architecture or different Fab process? - I will order few and evaluate them. I like OPA207, OPA205,  ADA4077 and especially OPA2205 but I don't like their price.
OPA2140 is another chiose - since one opamp can serve as both - High accuracy, low Ib, and high BW - no need of dual stage opamps. So one expensive instead of two cheap
Title: Re: LM399 based 10 V reference
Post by: iMo on February 25, 2024, 12:33:20 pm
The NE5532 needs a small resistor at the output (within the main loop) when wired as the unity gain final stage in the buffer, otherwise it oscillates with capacitive loads (at least in my sims).

PS: below with 1uF ceramic multilayer at the output. The output impedance is max 1.9ohm, peaks at some 100KHz.
Note: sim only..

PS1: with 180pF FB there is still 47deg PM at 82kHz, but it starts ring..

PS2: replaced the 1uA/500ns pulses with 10uA/1us pulses in order to see the responses better. Output settles in 11/15usecs..
Title: Re: LM399 based 10 V reference
Post by: iMo on February 26, 2024, 08:22:51 am
.. and yes, the LTspice can draw nice smooth graphs in sub uV range too..  :D
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on February 26, 2024, 08:59:41 am
The LM399 reference is still relatively noisy (e.g. 100 nV/sqrt(Hz) of white noise part and quite some 1/f noise on top) . So the op-amps for the buffer should not be that critical. The question is more if one needs AZ op-amps with there possible spikes or could use a classic precision OP-amp like OP07, OP177 or newer OPA207, OPA205,  ADA4077.
OPA2186 took my atention due to extremely low price. I'm reluctant to use it blindly since it seems different from other TI modern Az opamps. The max voltage is 24V. All other TI AZ OpAmp members have 36V. Is it different architecture or different Fab process? - I will order few and evaluate them. I like OPA207, OPA205,  ADA4077 and especially OPA2205 but I don't like their price.
OPA2140 is another chiose - since one opamp can serve as both - High accuracy, low Ib, and high BW - no need of dual stage opamps. So one expensive instead of two cheap

If you look for a cheap precision OP-amp, the OPA202 / OPA2202 are good candidates. They can kind of replace the OP07 with less power consumption and slightly better performance in most aspects.

The extra buffer is not just for high BW, but also to remove the thermal effect from the precision OP-amp. With the thermal effect the OPA2140 may be equally or even more sensitive than other precision amplifiers.

The simulated pulse response looks good enough. It should be well fast enough to be used with a DMM, likely even the R6581 with it's rather strong current pulse (but at least quite some waiting time).
A real world NE5532 and many other OP-amps may want more than 1 ohm in series to avoid oscillation from capacitive loading. This would slow down the response a little.
Title: Re: LM399 based 10 V reference
Post by: David Hess on February 27, 2024, 02:23:05 am
If you look for a cheap precision OP-amp, the OPA202 / OPA2202 are good candidates. They can kind of replace the OP07 with less power consumption and slightly better performance in most aspects.

The price and availability are so good that I wonder what the catch is with the OPA202.  The drift is only average and there is no offset null to correct it.

This is the first time I have noticed a high voltage complementary bipolar process with super beta transistors.  Manufacturers used to publish articles describing their semiconductor processes, but that data all seems to be secret now.

I am not sure why, but it does not show up in the selection guide at Mouser, at least when I search for it.
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 05, 2024, 03:25:46 am
Hi all! The time has come to finish a reference project that has been on the back burner for a long time. The project is very simple, based on an older design by Mickle T.. I don't need anything fancy, just to be able to check some 4 1/2 digit meters for drift. I have a few MAB399, OP07CZ (ceramic) and a 5k NOMCA network. The power supply is based on a transformer and uA723 set for 15V output.

I have a few questions about the general design:

- Does the heater need bypass caps? I was thinking to add a 10uF tantalum + 100nF. The thing is, won't the tantalum cap die from the heat if it's mounted near the MAB399 pins? Would a 100nF suffice?

- Does the "zener" need a bypass cap (CZByp)?

- The bias compensation resistor R38 makes a convenient RC filter with CZFilt (as seen in AD587 datasheet). What kind of capacitor would be needed here? I was thinking to use 1uF 250V polypropylene (or polyester). C16 and C17 would then be of the same type.

- It would be nice to also have access to the raw voltage of the reference. Should it be buffered with a unity gain op amp?

My apologies if some of the questions have been answered. I have read this thread and many others ~1 year ago but I tend to forget.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 05, 2024, 08:24:53 am
The LM399 gets hot inside but has already some insulation. So the power consuption in at about 200 mW (at bit more with short leads) and it does not heat up the PCB and surrounding parts very much.
The larger capacitor for the decoupling can be a bit away from the reference. 100 nF or maybe 1 µF at the reference heater would be enough. Some larger capacitor  for the 15 V supply in general could be good. C18 could already do that job.

The zener sides does not need a parallel capacitors. Just a capacitor could even cause more problems than good. The ADR1399 wants some 5 ohm and 1 µF as series element. A series RC would probably also better for the LM399.  So it makes sense to include the series RC in a PCB to have the option to also use an ADR1399. The capacitor could be a SMD ceramic (not too small form factor to no loose to much of the capacitance).

CZfilt should be a low leakage type. A polyester type should be good enough. PP tends to be bulky. There is no need for a low TC resistor for R38.

For the raw voltage it depends on the use if one wants / needs a buffer. A buffer offers some extra protection and may help with more lower resistance meters. A high Z DVM could be OK without.

The capacitor C16 to slow down the buffer is a bit on the large side. A smaller capacitor (1-10 nF range) would give faster settling, though lower maximum tolerated capacitance at the output.
For the trim resistor R52 the value depends on the individual reference voltage. Depending on the unit one may also need to trim up, e.g. with a resistor in parallel to R50.
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 08:38:03 am
I had several MAB399 and MAC199. It seems to me there is a design flaw with the heater, without a series protection resistor many of my chips died (the heater died). Also I would recommend the 1u ser 5ohm on the MAB399 as well for the stability.

I found out the soldering any stuff onto the 399 pins will increase the noise (ie by 20% in 34401A) (thermal flows). So use thin wires or tracks to the pins in some length.

Below my mods.

PS: you may go with higher voltage than the 15V, of course.
Title: Re: LM399 based 10 V reference
Post by: macaba on March 05, 2024, 11:07:27 am
What are the practical advantages (no theory please!) of having these discrete transistor output buffers over using a buffer opamp?

Just wondering if I should change my design...
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 05, 2024, 11:29:50 am
I'd guess the 2N2222 with its 800 mA current spec is more robust than most opamps, so the idea is a decent short-circuit behaviour. The diode between amplifier and output stage together with a bigger R39 help, too.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: bastl_r on March 05, 2024, 12:26:13 pm
I'd guess the 2N2222 with its 800 mA current spec is more robust than most opamps

And you can easily create a very fast over current- and short- protection via Q8 and R40.
I think.
Correct me if i'm wrong.
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 12:46:43 pm
A practical advantage of a fast opamp buffer (with built in current limit) would be a faster response to some spikes coming from the DMM. As has been discussed here, people claim there are such pulses coming off the DMM's inputs (I've seen a picture in the "Sampling with 3458A" book, like 15nA and couple of us long coming off the 3458A, but it could be anything).

What are the practical advantages (no theory please!) of having these discrete transistor output buffers over using a buffer opamp?

Just wondering if I should change my design...
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 01:22:51 pm
Some notes to my "MOD" above:
1. the resistor in the heater could be 10-33ohm, I would definitely put there something with TESLA's clones, indeed
2. C27 100pF - what is the role of that cap there? The simulation shows no impact on the stability, afaik
3. I would put a 100-1k resistor into the Q8 base, thus there will be no "direct path" from 15V to 10V output
4. I would go with a higher Vcc (when possible), like 18 to 24V.
Title: Re: LM399 based 10 V reference
Post by: David Hess on March 05, 2024, 02:48:35 pm
What are the practical advantages (no theory please!) of having these discrete transistor output buffers over using a buffer opamp?

A discrete buffer can be less expensive and provide a lower open loop output impedance, although that is not the case with this example.
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 05, 2024, 03:40:18 pm
Thanks for all the answers and suggestions.

I'll add the RC damper, lower C16 a bit and add a buffer for the raw 7 volts as well. It's always nice to be able to push out some milliamps. And then the output can also be protected with a TVS.

It's sad to hear about the MAB399 heater problems. I have 2 pieces, both have been aged for ~1000 hours on a current limited power supply. It was set for about 100mA and that might have protected them on each start. Inrush limiting as suggested by iMo is a must then. Would be interesting to see what failed inside of a reference with dead heater. I don't think this part saw much use because it arrived too late (1990 judging by the date code), when the soviet economy was already down the drain. No use means not much testing, so it could be flawed :(.

It might be wise to buy a LM399 when found at a decent price then, to avoid the nightmare of the MAB stopping working after thousands of hours of aging. What would be better, a vintage NatSemi or a new LT?

Regarding the 100pF C27 cap, it's not included in similar circuits. I think it speeds up the response of Q5 in case of a sudden short.

As for the input voltage, I already bought the 18 VAC transformer some time ago so it can't be much higher, but it could probably do 18V DC out.



Title: Re: LM399 based 10 V reference
Post by: mawyatt on March 05, 2024, 04:08:30 pm
Quote
It's sad to hear about the MAB399 heater problems. I have 2 pieces, both have been aged for ~1000 hours on a current limited power supply. It was set for about 100mA and that might have protected them on each start. Inrush limiting as suggested by iMo is a must then. Would be interesting to see what failed inside of a reference with dead heater. I don't think this part saw much use because it arrived too late (1990 judging by the date code), when the soviet economy was already down the drain. No use means not much testing, so it could be flawed :(.

I'm sure you are aware but others may not, a typical Power Supply usually has significant output capacitance, usually the better quality ones this is minimized. When the active supply is connected to the DUT in CC mode, the supply has achieved an initial voltage (Voltage Limit) across the output capacitors before connection. If this is set much higher than the DUT requires (a dynamic low impedance) significant energy discharge into the DUT can occur, from prior work (EMP sensitivity) the amount of energy required to damage a typical IC is on the order of just 10mJ on a single IC pin.

Seasoned folks usually set the PS Voltage Limit just slightly above the nominal required DUT voltage, and why quality accurate Power Supplies are desirable for development work.

Best,
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 04:19:11 pm
..It was set for about 100mA and that might have protected them on each start.
..I don't think this part saw much use because it arrived too late (1990 judging by the date code), when the soviet economy was already down the drain. No use means not much testing, so it could be flawed :( ..
I had none current limit and lost perhaps 5 pieces out of a dozen.
The part was in the development in late 80ties at Czechoslovakian TESLA, came into production perhaps around 1990, and soon gone with the decline of TESLA corp, where the microelectronics divisions were later acquired by Motorola/OnSemi, afaik.

There were Chineese and Soviets clones as well - https://www.eevblog.com/forum/projects/tesla-mab399-mac199-voltage-references/ (https://www.eevblog.com/forum/projects/tesla-mab399-mac199-voltage-references/)

Noopy did the die shot of the MAC199 I sent him too - https://www.richis-lab.de/REF02a.htm (https://www.richis-lab.de/REF02a.htm)

TiN messed with the MAB399 a lot - https://www.eevblog.com/forum/metrology/tesla-mac199mae299mab399-and-mac01-teardown-and-tests/ (https://www.eevblog.com/forum/metrology/tesla-mac199mae299mab399-and-mac01-teardown-and-tests/)

Title: Re: LM399 based 10 V reference
Post by: 2N3055 on March 05, 2024, 04:28:35 pm
With LM399 PSU current limit does not matter because heater is driven by current limited circuit internally... Sometimes external current limit was used to limit the spike for various reasons, but it is absolutely unnecessary for LM399 survival.

Maybe TESLA 399 clones had something different in that part of circuit. And so these need some current limiting...
Thanks iMo for the info!

P.S. as a kid I had TESLA bass guitar amplifier. It had just average sound, but was built like tank.. Undestructable ... ^-^
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 05:07:59 pm
Btw I compared the noise (0.1-10Hz) of the MAC199s (one of them with broken heater) with the LM399A/Hs and they were the same..
https://www.eevblog.com/forum/metrology/low-cost-lm399-noise-indicator/msg5055250/#msg5055250 (https://www.eevblog.com/forum/metrology/low-cost-lm399-noise-indicator/msg5055250/#msg5055250)

PS: My bet all MAB/MACx99 refs available today at ebay or in various eshops were already carefully "sift through" (mind they are 30+ years out of production), also the markings on their hats might be replaced/forged. All chips/dies inside are "MAC199" afaik, and the "MAB/MAC/MAE" versions were just stamped onto the easily replaceable plastic hat after the binning during the testing..
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 05, 2024, 05:39:23 pm
Interesting point mawyatt. I must admit I didn't think of this before messing with the references. In my setup, there was the K7200 diy PSU (100uF output cap), ~1.5m of wires and a small board with 2 refs that also had a 10uF capacitor and a sliding switch. I tried to recreate the possible surge in LTSpice. Oddly enough the current pulse on the heater pins is bigger with the 10uF cap than without it. The pulse current seems to largely depend on how fast the switch is closed. If the rise time is faster than 1 ms, the chip would be in trouble (transients get up to a few amps).

Thanks for the information about the clones iMo. Those Chinese and Russian ones must be really rare, haven't even found pictures around.

P.S. as a kid I had TESLA bass guitar amplifier. It had just average sound, but was built like tank.. Undestructable ... ^-^

I am very fond of TESLA, had a B100 stereo reel to reel player, it sounded great.

PS: My bet all MAB/MACx99 refs available today at ebay or in various eshops were already carefully "sift through" (mind they are 30+ years out of production), also the markings on their hats might be replaced/forged. All chips are "MAC199" afaik, and the "MAB/MAE" versions were just stamped onto the easily replaceable plastic hat after the binning during the testing..

Who knows... The probability for forgery might be low though, since they are cheap and not that popular. There's still one shop in the Czech Republic selling them for 1,18 EUR / piece, but they sell only locally. The MA* dies are indeed the same. In the factory I think they just sorted for tempco, so there's not a big difference between the MAB and MAC versions regarding noise / other parameters.

[edit]
Found a few meters that used them, the Metra M1T 390 (4.5 digits) and M1T 380 (5.5 / 6.5 digits, quoted as the most advanced meter produced by Metra Blansko). Some pictures are available here: https://forum.elektrolab.eu/resources/metra-m1t-380.629/ (https://forum.elektrolab.eu/resources/metra-m1t-380.629/). I found the schematic for it and in the part list they wrote LM199 but in reality they used MAB399 (picture 8 in the link above). In any case, I was curious to see if they limited the surge current somehow, but nope, the heater is simply connected to 15V (generated from a MA7815) and that's it. The zener is supplied from 20V (generated by a MAA723) via a 12K resistor, so 1mA current.

[attach=1]

[edit2]
On a second look, there is actually a current limiter for the 15V rail (that powers the heater of the reference) in the M1T 380. At first I thought it was a capacitance multiplier, but it actually limits the current to about 375 mA before the input of the MA7815. This may not help the reference too much since I doubt the heater surge is bigger than 150-200 mA on a cold start. I will add a current limit in my circuit as well at the 15V regulator.

[attach=4]
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 08:46:31 pm
Your Metra link says the M1T380 meter was produced a) with "standard stability" - with the TKZD13/D zener (we discussed here the samples), and b) with "enhanced stability" - with the MAB399.
Now, the big question is how many b)s they actually produced (as Metra Blansko stopped its production in 1991 - in the same time as the MAB399 arrived)..
https://www.eevblog.com/forum/reviews/czechoslovakian-measure-equipment-collection/msg85691/#msg85691 (https://www.eevblog.com/forum/reviews/czechoslovakian-measure-equipment-collection/msg85691/#msg85691)
PS: the meter in your link (the teardown pictures) is using the TKZD13 zener. Frankly, I doubt there were any Metra meters with the 399 ever sold..
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 05, 2024, 09:05:29 pm
Interesting read, seems like the parts were notoriously unreliable. I bet not many of those multimeters were made, possibly a few tens at most. The other components like the CPUs were probably in high demand at that time, slowing down production.

In any case, the high failure of your 199/399s, the cracked dies found in MAA723s and MetraCollector's testimony make me think that some eastern components are better suited as a collectible than for actual use. I'll search for some proper LM399s, just to be on the safe side. If this project gets completed, I'll make a new thread and present it there, so this one stays clean.

[Edit]
Regarding the picture, see "D1639.jpg". The MAB is in the middle towards top right.
Title: Re: LM399 based 10 V reference
Post by: iMo on March 05, 2024, 09:18:38 pm
I got three Metra 4.5-6d meters in my hands (long time back) for free. All not working. Spent a week elaborating them. At the end of the day I removed the ww resistors, some transistors and analog chips for my collection, the reed relays, and the TKZD zeners. The rest went into the shredder..
PS: ..and the D1791.jpg is with the zener.. :)
But nice to see MAB399 in an actual application  :-+
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 18, 2024, 11:44:08 am
Still playing with the circuits in this thread. Went back and read all the posts that I missed, but it made things worse because now I can't decide which output buffer to use, opamp (e.g. NE5532) or current limited emitter follower :-DD.

There is another thing that I'd like to do before starting the build, a way to know if the heater is operating correctly. From what I gather, the LM399 gets from ambient temperature to ~90C in a few seconds. I imagine that the die is subjected to a big amount of thermal stress and that the heater could sometimes break, especially with some parts (like iMo said above). If the reference is placed inside a case, under thermal isolation, it would be hard to know if the heater is working or not.

A "passive" way to implement this would be to place a NTC close to the pins of the 399, on the PCB island, and then have a comparator circuit made of a jellybean op-amp light a LED if the heat gets over a threshold. But I don't think that there is enough heat at the legs to work reliably, especially on a hot summer day when the ambient temperature might trigger the comparator. Maybe there's a better way to do this?
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 18, 2024, 11:53:38 am
As the heater gets so much hotter than ambient temperature, heater power is roughly independent of ambient temperature. So inserting a 1 Ohm resistor into the heater supply and looking at the heater current with a dual comparator should be enough to assert correct heater operation.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on March 18, 2024, 11:59:54 am
Also do not put anything with low thermal resistance close to the 399 pins as it will increase the thermal flow (the "something" will cool down the reference). It is a recommendation of ADI afaik (and I saw the effect - an increase of stddev - in-situ here).
Title: Re: LM399 based 10 V reference
Post by: IconicPCB on March 18, 2024, 12:08:38 pm
The "something " could be a small incandescent bulb akin to the wien bridge nonlinear gain element.
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 18, 2024, 12:28:41 pm
Thanks for the ideas. The shunt + comparator might be the best way. And a bigger value resistor might also work as an inrush limiter. I'll follow the advice and not put stuff around the legs then.
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 22, 2024, 09:54:56 am
The heater current monitor proved to be a bit more difficult than expected because one needs to work close to the 15V rail and a simple comparator + shunt won't work. Luckily there is a nice high side current sense IC (ZXCT1009) that makes life easier, so here's a try at a monitoring circuit. It will light up an LED when the current through the heater is smaller than 5mA.
Title: Re: LM399 based 10 V reference
Post by: iMo on March 22, 2024, 11:22:39 am
The ZXCT may indicate to you the temperature inside the box - the heater current changes with the surrounding temperature.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 22, 2024, 12:35:17 pm
MAX4373/MAX4374/MAX4375 supervisors may be interesting, too. They include the high side shunt amplifier, a reference and two comparators. Max input voltage is 28 V.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: David Hess on March 23, 2024, 09:20:23 am
OPA2186 took my atention due to extremely low price. I'm reluctant to use it blindly since it seems different from other TI modern Az opamps. The max voltage is 24V. All other TI AZ OpAmp members have 36V. Is it different architecture or different Fab process? - I will order few and evaluate them.

I am in the process of updating my preferred operational amplifiers list and I have used the low power OPA187 recently.

I suspect that the OPA186 has a lower supply voltage because its input stage is different from the other parts in the series to support rail-to-rail input operation.  My guess based on the offset voltage versus common mode voltage graph and input bias current is that both PMOS and NMOS stages are used in parallel and a lower gate-to-source voltage rating of one limits the supply voltage.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 23, 2024, 10:41:33 am
Thanks, i was looking for a zero drift opamp with input rail to rail and couldn't find one. All i found excluded the upper 1.5 to 3 V of common mode voltage or limited operation voltage to 5 V.
The OPA186 datasheet figure 6.7 seems to indicate some anomaly at the lower 3 V of the input range. Maybe that disappears at lower supply voltage. Anyway now there is a choice..
The f = 0.1 Hz to 10 Hz input noise spec 125 nVRMS for 90 uA supply current is amazing.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: macaba on March 23, 2024, 12:38:46 pm
I am in the process of updating my preferred operational amplifiers list and I have used the low power OPA187 recently.

I would like to hear more about your list when you are ready. I would suggest adding LTC2057 to the evaluation list - I recently did some evaluation on a group of zero drift opamps, measuring the input bias current over CM voltage range, whilst varying the R/C of input/feedback. LTC2057 was the only device to have both: 1. Nearly constant input bias over CM voltage range, 2. Maintain that constant input bias over a variety of R/C conditions. It was quite astonishing to see how badly behaved other parts were under various conditions and LTC2057 would always have a flat clean line on the charts.
Title: Re: LM399 based 10 V reference
Post by: David Hess on March 23, 2024, 06:10:16 pm
Thanks, i was looking for a zero drift opamp with input rail to rail and couldn't find one. All i found excluded the upper 1.5 to 3 V of common mode voltage or limited operation voltage to 5 V.
The OPA186 datasheet figure 6.7 seems to indicate some anomaly at the lower 3 V of the input range. Maybe that disappears at lower supply voltage. Anyway now there is a choice..

Also the input bias current get significantly worse starting at the same low common mode input voltage.  For operating near the negative supply, there are better parts.

Maxim had a chopper stabilized part which used an internal charge pump to get clean rail-to-rail input operation.

Quote
The f = 0.1 Hz to 10 Hz input noise spec 125 nVRMS for 90 uA supply current is amazing.

I am having trouble replicating the 0.1 to 10 Hz noise specifications in the recent Texas Instruments datasheets from their input voltage noise spectral density graphs and specifications, and Texas Instruments has a history of lying on their noise specifications, so I would not trust them blindly.  I would hate to build a dedicated 0.1 to 10 Hz testing jig, but I may have to.  It is not that difficult to do, but I currently lack a suitable sampling multimeter to make the measurement.

That 125 nVrms specification should be about 0.75 uVpp so it is by far the noisiest part in that series of chopper stabilized parts, and I suspect that is why they used nVrms instead of nVpp like the others.

I would like to hear more about your list when you are ready. I would suggest adding LTC2057 to the evaluation list - I recently did some evaluation on a group of zero drift opamps, measuring the input bias current over CM voltage range, whilst varying the R/C of input/feedback. LTC2057 was the only device to have both: 1. Nearly constant input bias over CM voltage range, 2. Maintain that constant input bias over a variety of R/C conditions. It was quite astonishing to see how badly behaved other parts were under various conditions and LTC2057 would always have a flat clean line on the charts.

I will add the LTC2057 at least for comparison purposes.  I am trying to get away from parts supplied by Analog Devices because their prices have increased so much.

I was unpacking my electronics stuff after moving and reorganizing everything, cleaning parts cabinets, etc., and decided to go through my prototyping parts in detail because so many are either no longer produced, have outrageous prices, or now have better alternatives.
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 27, 2024, 04:50:13 pm
Back with a (stupid) question. What is the difference between using a PI bifilar common mode filter on the output vs. using separate cores for each wire like in older voltage standards (e.g. Cropico ESC1 (https://www.eevblog.com/forum/metrology/cropico-esc1-electronic-standard-cell-a-look-inside/msg896395/#msg896395))?
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 27, 2024, 05:00:02 pm
The common mode choke would mainly act on the common mode signal, e.g. to filter a so perfectly isolating DCDC converter and reduce an injected signal from there. The output regulation is much less effected (the CM mode chokes still have some differential mode inductance) and can still be fast (e.g. with somewhat transisent loads).
With separate inductors the filtering would also effect the output drive / regulation speed. Transient loads may lead to more LC ringing and slow regulation. This could be an issue with separate drive an sense terminals.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 27, 2024, 05:06:58 pm
Common mode filters are usually made with high inductance cores that are less useful for making chokes as they easily saturate. So a typical inductance for chokes like those in the image is 10 to 100 uH, while a typical common mode choke is at 5 or 10 mH.
I am sometimes using Würth 744866104 common mode chokes with 100 mH. Those have a stray inductance of about 1 mH.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: iMo on March 27, 2024, 05:48:36 pm
..What is the difference between using a PI bifilar common mode filter on the output vs. using separate cores for each wire like in older voltage standards..

Provided all three coils are perfectly identical (incl the winding direction) there is none difference.
As on your above shot the proper way would be to use single toroidal core with trifilar winding. Also the return path wire (ie gnd) must be included (the number of wires/signals in the single CM choke is unlimited, ie you may pass entire flat cable with say 100 different signals and their returns through a single CM choke).
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 27, 2024, 06:05:50 pm
Ah, that's good to know. So I can put the 10V, 7V and GND signals through a single choke instead of using 2 (i.e. 10V and GND on one core and 7V and GND on another). Helps save a few parts.
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 27, 2024, 09:50:22 pm
..What is the difference between using a PI bifilar common mode filter on the output vs. using separate cores for each wire like in older voltage standards..

Provided all three coils are perfectly identical (incl the winding direction) there is none difference.
3 separate coils are different. They could have the same effect on the common mode signal, but the effect on a differential signal is quite different, with much more inductance seen to the differential signal.
Separate coils (especially if not identical) also tend to have stronger coupling from the CM to the differential signal.
Title: Re: LM399 based 10 V reference
Post by: Birb on March 29, 2024, 05:36:02 am
Hi, here's an attempt at designing a lm399 based Vref. Note that the OP077 will probably be replaced with something else and is currently a placeholder.
Also, the input voltage is 15V, which gets regulated by an LM7812, giving sufficient headroom for a lower noise.
Lastly, the entire reference section should be heated by a small PTC heater to around 50C. It should be good enough to keep the resistors more stable (Using weird BWL resistors - they seem to be ok, but I don't trust the tempco).
Though I worry that the buffer may not be sufficient?
Any feedback is appreciated, thanks a lot.
Title: Re: LM399 based 10 V reference
Post by: dietert1 on March 29, 2024, 07:02:00 am
R2 and R3 are exchanged.
R4 should be about 3K instead of 10K.
U3 needs a heatsink and is mounted outside the oven. A higher supply voltage is preferred, like 15 or 18 V. 15 Vac will give you about 22 V from the rectifier. Depends on the heater power.
U4 should be mounted close to the PTC.
10 V tuning is better with fixed resistors (as far as possible).
Concerning the output buffer it depends on what will be the application of this reference.

Regards, Dieter
Title: Re: LM399 based 10 V reference
Post by: Kleinstein on March 29, 2024, 07:46:46 am
An additional oven around the LM399 is tricky, as the reference starts with a relatively high power and the thermal insulation thus can not be that good. So one would end up with a rather high heater power.
The LM399 itself hardly need the temperature regulation and the gain stage it depends. Same series resistors may show some compensation of the there TC. Precision use normally is also done with not so extreme room temperature, as the other instruments would also be effect from the temperature.
Title: Re: LM399 based 10 V reference
Post by: argintviu on March 29, 2024, 09:35:12 am
@Birb, you could try to ovenize just the resistors / opamp and simply thermally isolate the 399 (top and underside). Kind of how branadic did with the QH40 crystal heater on top of the gain resistors here (https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg277738/#msg277738).