Author Topic: LM399 based 10 V reference  (Read 252228 times)

0 Members and 1 Guest are viewing this topic.

Offline TiN

  • Super Contributor
  • ***
  • Posts: 3612
  • Country: tw
  • xDevs.com
    • xDevs.com
Re: LM399 based 10 V reference
« Reply #250 on: November 06, 2014, 03:21:34 pm »
Revive the dead....



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

Resistors - PTF56 25ppm/C
xDevs.com YouTube | Have test gear documentation to share? Upload here! No size limits, firmware dumps and teardown photos welcome.
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: de
Re: LM399 based 10 V reference
« Reply #251 on: November 07, 2014, 06:04:27 am »
 

Offline MK

  • Regular Contributor
  • *
  • Posts: 184
  • Country: gb
Re: LM399 based 10 V reference
« Reply #252 on: November 07, 2014, 08:28:52 pm »
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.

 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: de
Re: LM399 based 10 V reference
« Reply #253 on: November 25, 2014, 09:20:26 am »
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:

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

With best regards

Andreas
 

Offline metacollin

  • Contributor
  • Posts: 35
  • Country: us
  • Ocelloscopes. You know, for ocelots.
    • Electropimp
Re: LM399 based 10 V reference
« Reply #254 on: December 11, 2014, 04:34:53 pm »
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!
"Violence is the last refuge of the incompetent." - Isaac Asimov
 

Online Dr. Frank

  • Super Contributor
  • ***
  • Posts: 1519
  • Country: de
Re: LM399 based 10 V reference
« Reply #255 on: December 11, 2014, 11:41:46 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
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: de
Re: LM399 based 10 V reference
« Reply #256 on: December 12, 2014, 09:54:35 am »
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




 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 3612
  • Country: tw
  • xDevs.com
    • xDevs.com
Re: LM399 based 10 V reference
« Reply #257 on: December 12, 2014, 11:03:58 am »
Andreas,

I am also interested how is switching done between different channels?
xDevs.com YouTube | Have test gear documentation to share? Upload here! No size limits, firmware dumps and teardown photos welcome.
 

Online Dr. Frank

  • Super Contributor
  • ***
  • Posts: 1519
  • Country: de
Re: LM399 based 10 V reference
« Reply #258 on: December 12, 2014, 07:54:45 pm »

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
 

Offline babysitter

  • Frequent Contributor
  • **
  • Posts: 751
  • Country: de
  • pushing silicon at work
Re: LM399 based 10 V reference
« Reply #259 on: December 12, 2014, 08:27:04 pm »
Offering my travelling LTZ for this kind of project !
I'm not a feature, I'm a bug! ARC DG3HDA
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: de
Re: LM399 based 10 V reference
« Reply #260 on: December 13, 2014, 07:28:11 am »

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


 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: LM399 based 10 V reference
« Reply #261 on: December 13, 2014, 10:10:22 pm »

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..
 

Online Dr. Frank

  • Super Contributor
  • ***
  • Posts: 1519
  • Country: de
Re: LM399 based 10 V reference
« Reply #262 on: December 14, 2014, 01:38:58 am »
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
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 3612
  • Country: tw
  • xDevs.com
    • xDevs.com
Re: LM399 based 10 V reference
« Reply #263 on: December 14, 2014, 01:56:03 am »
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.
xDevs.com YouTube | Have test gear documentation to share? Upload here! No size limits, firmware dumps and teardown photos welcome.
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: de
Re: LM399 based 10 V reference
« Reply #264 on: January 05, 2015, 07:24:39 am »
Day 365 of 2 LM399 ageing (see above)

With best regards

Andreas
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 1130
  • Country: de
Re: LM399 based 10 V reference
« Reply #265 on: January 05, 2015, 08:02:30 am »
Well done, well recorded  :-+
Prema 5000 | 2x Prema 5017 SC | GenRad 1434-G | Tek 2465A | VNWA2.x with TCXO upgrade and access to: Keysight 3458A, Keithley 2002, Prema 5017 SC, 34401A, 34410A, Keithley 2182A, HDO6054, Keysight 53230A and other goodies at work
 

Offline DLWarr

  • Newbie
  • Posts: 3
Re: LM399 based 10 V reference
« Reply #266 on: January 06, 2015, 12:56:13 am »
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..
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: sk
Re: LM399 based 10 V reference
« Reply #267 on: January 06, 2015, 01:03:51 am »
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...)
 

Offline MK

  • Regular Contributor
  • *
  • Posts: 184
  • Country: gb
Re: LM399 based 10 V reference
« Reply #268 on: January 06, 2015, 05:48:24 am »
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.
 

Offline DLWarr

  • Newbie
  • Posts: 3
Re: LM399 based 10 V reference
« Reply #269 on: January 07, 2015, 02:16:21 am »
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...
 

Offline rf-design

  • Regular Contributor
  • *
  • Posts: 57
  • Country: de
Re: LM399 based 10 V reference
« Reply #270 on: January 07, 2015, 03:36:54 am »
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?
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: de
Re: LM399 based 10 V reference
« Reply #271 on: January 07, 2015, 10:15:34 am »
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
« Last Edit: January 07, 2015, 03:33:05 pm by Andreas »
 

Offline babysitter

  • Frequent Contributor
  • **
  • Posts: 751
  • Country: de
  • pushing silicon at work
Re: LM399 based 10 V reference
« Reply #272 on: January 07, 2015, 05:13:43 pm »
Congratulations for 1 year of logging, Andreas!
I'm not a feature, I'm a bug! ARC DG3HDA
 

Offline rf-design

  • Regular Contributor
  • *
  • Posts: 57
  • Country: de
Re: LM399 based 10 V reference
« Reply #273 on: January 07, 2015, 07:30:11 pm »
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
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: LM399 based 10 V reference
« Reply #274 on: January 07, 2015, 07:52:57 pm »
My home-lab spent some time over Christmas measuring delta-voltages, every 16 minutes 4 measurements are taken between 3 references:

http://www.eevblog.com/forum/projects/yet-another-%28lm399%29-volt-reference/15/
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf