Author Topic: Ultra Precision Reference LTZ1000  (Read 1007882 times)

0 Members and 1 Guest are viewing this topic.

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 8266
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1550 on: October 10, 2016, 06:45:29 pm »
Due to the extra about 300 K/W extra internal thermal resistance the A version can operate with a lower heating power if the temperature is choose for the same maximum ambient 
temperature. However the difference is not that big - comparing long an short leads can be a similar difference.
 

Offline d-smes

  • Regular Contributor
  • *
  • Posts: 92
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1551 on: October 10, 2016, 11:22:49 pm »
Due to the extra about 300 K/W extra internal thermal resistance the A version can operate with a lower heating power if the temperature is choose for the same maximum ambient 
temperature. However the difference is not that big - comparing long an short leads can be a similar difference.
I totally agree.   But now that we know it's 150 K/W (instead of 80) for non-A and 400 K/W for A part, the difference is smaller.  Couple that with everyone putting insulation around the parts, and long leads (best for equal lead temperature at the board per this thread) and now majority of heat flow is out the leads to the PCB.  So, not much difference between A and non-A parts.  Makes me wonder if non-A designs run short of thermal margin of heater setting...

I forgot to mention, I looked at the "Die Temperature Rise vs Heater Power" graph from the data sheet again, and it does extrapolate to zero temperature rise at zero heater power.  So the rest of the die elements are assumed de-energized for this graph.  But I agree with what DiligentMinds is saying- when doing design, die (zener et.el.) dissipation will subtract from heater power necessary to reach target temperature.  If that quantity goes zero/negative, you've lost thermal regulation and your reference turns to crap.
 

Online TiN

  • Super Contributor
  • ***
  • Posts: 4236
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Ultra Precision Reference LTZ1000
« Reply #1552 on: October 12, 2016, 04:16:56 pm »
Well, hope nobody mind if I jump in with few practical items into this nice talk.
I was talking to one of volt-nuts (not known member here), and we decided to build and test another KX LTZ-reference for him.
Complete assembly will be tested for tempco and aged for 168 hours and later shipped cold to the final destination via regular mail.

To keep things bit more exiting for me, I decided to try few new things this time.

A. Use VPG VHD200 divider, 1K/13K hermetic oil-filled resistor. My previous LTZ references were using separate resistors.
In theory this should provide better thermal stability, as temperature setpoint divider is now tightly coupled thermally in same oil-filled package.

Rest of resistors are custom 2 x VPG 75K 1%, VPG 120R 0.01%. These have custom PO VFR part-numbers, but I'd expect they are VHP101-based.
Whole lot was bought for 190$USD from eBay, including unrelated resistors on photo.

Chip will be LTZ1000ACH.



B. I want to try ADA4522-1ARZ opamp instead of my usual LTC2057's to see if any difference would be visible on output noise/short-term stability (I expect none, but who knows).



And initial check, using 3458A.



Usual datalog:



Not a bad start, eh? -1ppm from 120 Ω. :)
Ignore first hour when ACAL was ran. 8:15 to 10:30 is measurement with DELAY 5 on 100ohm range. Few samples were overflow, as 3458A's max overrange is just 120 ohms, so data from 11:15 is with 1KΩ range and DELAY 1. All samples NPLC100, AZER ON, with resistor right at therminals as on photo.

P.S. Bonus quiz, what are rest resistors are for?  ;D
« Last Edit: October 12, 2016, 04:43:38 pm by TiN »
YouTube | Metrology IRC Chat room | Live-cam | Share T&M documentation? Upload! No MB limit, firmwares, photos.
 

Offline Nuno_pt

  • Frequent Contributor
  • **
  • Posts: 435
  • Country: pt
Re: Ultra Precision Reference LTZ1000
« Reply #1553 on: October 12, 2016, 05:54:50 pm »
Well, looks like I've to start gathering some parts, to build one too.

 
Nuno
CT2IRY
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 8266
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1554 on: October 12, 2016, 06:02:39 pm »
The extra resistors looks like current shunts, maybe for the SMU project.

There is rather little sense in using an ADA4522 for the reference circuit. It could even cause trouble from RF interference and maybe phase shift in the 10 Hz range. Also the noise of the reference is way higher than that amplifier or the LT1013. So the LT1013 is perfectly fine for the LTZ1000 circuit. The critical part of the amplification is done by the transistor inside the LTZ1000. So at the output one sees the noise of the reference zener (which is by far the main component) plus the noise of the transistor (rather low, better than most OPs) and only about 1/200 of the noise (and drift) of the OP used. It is only at higher frequencies (e.g > 100 Hz) where the noise of the OP gets visible to 100% and here the AD4522 is not that exceptionally good. Also current noise of the ADA4522 is a little on the high side, as it sees an about 50-70K source.

The point's where one might want an improvement over the DS circuit is more like RF immunity, tolerance to capacitive loading and may be additional temperature monitoring.

The ADA4522 is a good companion for the shunts.
 

Offline splin

  • Frequent Contributor
  • **
  • Posts: 999
  • Country: gb
Re: Ultra Precision Reference LTZ1000
« Reply #1555 on: October 12, 2016, 07:03:52 pm »
The 3458A's TC spec is 4ppm on the 120ohm range so between 8:15 and 10:30 you have a resistor with -4ppm and a 3458A with a +4ppm TC plus 1.6ppm pk-pk noise. Or vice versa. Or any combination inbetween including both being 0ppm.

The resistance was on average unchanged between 8:15 and 10:27 when the temperature dropped from 26.2 to 24.1C, but then shifted by approx .8ppm when the temperature returned to 26.2. So is that hysterisis in the resistor, the 3458A or both? It seems like quite a lot frohm a temperature change of only 2.1C.

[EDIT] Ignore the above - I was misreading your graphs. I do find them hard to work out what's what though. Any chance of annotating one of your charts sometime to show exactly which lines relate to which scales? E.G. what are the orange dotted and blue dashed lines?
« Last Edit: October 12, 2016, 11:13:06 pm by splin »
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 1943
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1556 on: October 12, 2016, 07:47:17 pm »

P.S. Bonus quiz, what are rest resistors are for?  ;D


The 634 Ohm, 9 Ohm and 0.1 Ohm are intended for your 3458A, current shunts, R209, R211, R213.

I think, we both already discussed to improve their T.C.


Frank

Btw.: Where did you get this 0.1Ohm in small quantities? Directly from Vishay?
Ah, really everything from ebay?!
« Last Edit: October 12, 2016, 10:29:51 pm by Dr. Frank »
 

Offline Vgkid

  • Super Contributor
  • ***
  • Posts: 2610
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1557 on: October 13, 2016, 12:07:30 am »
Either ebay, or Taobao. I'm starting to like the latter.
If you own any North Hills Electronics gear, message me. L&N Fan
 

Online TiN

  • Super Contributor
  • ***
  • Posts: 4236
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Ultra Precision Reference LTZ1000
« Reply #1558 on: October 13, 2016, 12:17:21 am »
Dr.Frank is sharp as always. Yes, everything from eBay this time. 1 ohm I already have VPG VPR5 4wire as well.

I'll redo a graph for better clarity and legend, sorry.
YouTube | Metrology IRC Chat room | Live-cam | Share T&M documentation? Upload! No MB limit, firmwares, photos.
 

Offline Nuno_pt

  • Frequent Contributor
  • **
  • Posts: 435
  • Country: pt
Re: Ultra Precision Reference LTZ1000
« Reply #1559 on: October 14, 2016, 09:44:43 am »
Hi guys,
Does any of you guys have some boards for the LTZ, and are willing to let maybe 1 or 2 of them go, I'm looking for maybe 2 boards to start building my first LTZ.

Thanks.
Nuno
CT2IRY
 

Offline plesa

  • Frequent Contributor
  • **
  • Posts: 965
  • Country: se
Re: Ultra Precision Reference LTZ1000
« Reply #1560 on: October 18, 2016, 10:51:33 pm »
After difficulties with PWW resistors I finally finished all 4 LTZ1000 boards from TiN and started collecting data.
Almost all resistors are PWW made by Edwin except 4x 100k resistors which were out of tempco spec.
He has some manufacturing issues with potting.
Missing PWW were replaced by pre-selected Vishay UXB series resistors.
Initial measurement by 34411A/34410A is rock stable, so I used ratio measurement with 34401A and 34420A.

Very cool!  I'm looking forward to your measurement results!  Is that your temperature chamber on the right side of the photo?

-Ken

Measurement with freshly calibrated 3458A started, first reference made by Edwin PWW is within day within +/- 1ppm - cold start.
Each reference will be measured one week minimum to give me some level of confidence in short term stability.
Any tips for low EMF relay for scanner?
 

Offline zlymex

  • Frequent Contributor
  • **
  • Posts: 556
  • Country: cn
Re: Ultra Precision Reference LTZ1000
« Reply #1561 on: October 19, 2016, 12:35:08 am »
Measurement with freshly calibrated 3458A started, first reference made by Edwin PWW is within day within +/- 1ppm - cold start.
Each reference will be measured one week minimum to give me some level of confidence in short term stability.
Any tips for low EMF relay for scanner?
Looks quite normal, except some spikes at the beginning. What is the NPLC used? Is AZ on?
As for the low EMF relay for scanner, try latching relay, such as those used by Fluke 5720A, DS2E-ML2-DC5V. I use similar relays in my DIY scanner: https://www.eevblog.com/forum/projects/diy-low-thermal-emf-switchscanner-for-comparisons-of-voltage-and-resistor-stand/msg610755/#msg610755
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 1943
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1562 on: October 19, 2016, 06:20:34 am »

Measurement with freshly calibrated 3458A started, first reference made by Edwin PWW is within day within +/- 1ppm - cold start.
Each reference will be measured one week minimum to give me some level of confidence in short term stability.
Any tips for low EMF relay for scanner?

As a 10 channel scanner, you may get the old HP3488A plus 44470A, which are quite cheap.

These spikes indicate that your setup may be disturbed by RF noise, bad grounding, and lacks all the blocking caps, like in Andreas design.

Short term stability, like noise and fluctuations may be determined within one hour already, i.e. 10 min and 1h stability. That should be <= 0.2ppm
Up to 24h will give mostly the temperature dependency of your LTZ1000 references, plus the 3458A w/o ACAL (especially U180 test).
Constant RT, dT <= 0.2°C should also give dU <= 0.2ppm.
One week will already show the initial burn-in drift of the LTZ1000, if applicable, and depending of your oven temperature.

Depending on the history of your LTZ1000 (e.g. solder heating, preliminary heating / cycling), the initial burn-in drift may take months or years.

Frank


 
The following users thanked this post: MisterDiodes

Offline plesa

  • Frequent Contributor
  • **
  • Posts: 965
  • Country: se
Re: Ultra Precision Reference LTZ1000
« Reply #1563 on: October 19, 2016, 04:03:00 pm »
Measurement with freshly calibrated 3458A started, first reference made by Edwin PWW is within day within +/- 1ppm - cold start.
Each reference will be measured one week minimum to give me some level of confidence in short term stability.
Any tips for low EMF relay for scanner?
Looks quite normal, except some spikes at the beginning. What is the NPLC used? Is AZ on?
As for the low EMF relay for scanner, try latching relay, such as those used by Fluke 5720A, DS2E-ML2-DC5V. I use similar relays in my DIY scanner: https://www.eevblog.com/forum/projects/diy-low-thermal-emf-switchscanner-for-comparisons-of-voltage-and-resistor-stand/msg610755/#msg610755

Hi zlymex,

NPLC is 200 with synchronous autozero. The spikes are still there, I suspect the switching furnace in basement behind wall.
Thanks for relay type, I will order them. Scanner I'm going to use 34970A, it a pity that it does not have low emf connection.


Measurement with freshly calibrated 3458A started, first reference made by Edwin PWW is within day within +/- 1ppm - cold start.
Each reference will be measured one week minimum to give me some level of confidence in short term stability.
Any tips for low EMF relay for scanner?

As a 10 channel scanner, you may get the old HP3488A plus 44470A, which are quite cheap.

These spikes indicate that your setup may be disturbed by RF noise, bad grounding, and lacks all the blocking caps, like in Andreas design.

Short term stability, like noise and fluctuations may be determined within one hour already, i.e. 10 min and 1h stability. That should be <= 0.2ppm
Up to 24h will give mostly the temperature dependency of your LTZ1000 references, plus the 3458A w/o ACAL (especially U180 test).
Constant RT, dT <= 0.2°C should also give dU <= 0.2ppm.
One week will already show the initial burn-in drift of the LTZ1000, if applicable, and depending of your oven temperature.

Depending on the history of your LTZ1000 (e.g. solder heating, preliminary heating / cycling), the initial burn-in drift may take months or years.

Frank


Thanks Frank,

Scanner is 34970A I only needs to use proper low EMF relays. All cards for 34970A which I have do not have low EMF connection/relays (~3uV), but maybe with Panasonic ones it will be the same. Needs to check.
The grounding and RF noise is not ideal, It is placed in polystyrene container without metal shield.
The LTZ1000 has been burned for few weeks @125°C followed by temperature cycling. If the CERN recommendation works the drift should be minimal now.
Remaining LTZ1000 boards I will measure only few days, but two do not have full set of Edwin PWW resistors (tempco outliers) and for R2/R3 resistor were used Vishay UXB.
I used R4 12k (40-45°C) and 100k R2/R3 resistors for minimal drift.
 

Offline MisterDiodes

  • Frequent Contributor
  • **
  • Posts: 457
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1564 on: October 19, 2016, 05:56:32 pm »

Depending on the history of your LTZ1000 (e.g. solder heating, preliminary heating / cycling), the initial burn-in drift may take months or years.

Frank

Thank You Dr. Frank!!

I'll jump in here with some suggestions & thoughts after a few decades of building these for industrial use.  Again:  I build custom LTZ circuits for my customers that need to be used - not for PPM-chasing  entertainment purposes. 

These are just some observations picked up from decades of making various custom test systems:

1.  When we assemble an LTZ circuit, we power it up (and add power cycles at least a few times per day) and let it run a couple years.  That die inside is going to take a while for the crystal lattice to stress-relieve itself, and there's no way of knowing where your die came from out of the initial wafer.  Every die singulated out of the wafer will have a different strain pattern once it becomes a single die, and every LTZ1000x is going to behave differently.  It also depends on if the die was diamond-scribed or cut with a wafer saw, lasered, etc.  Manufacturing processes change slightly over time, even though the basic  die layout hasn't changed in decades.  The random crystal edge damage seen along every die will have an effect on final performance, good and bad.

Usually one of these things will happen:

A) The LTZ circuit will be initially jumpy, but after a few year's operation it will settle down and become quite stable.  Sometimes these 'grow up' to be the best stable references you've ever seen.

B) The LTZ will be initially fairly stable, but stability doesn't improve as much over time.

C) The LTZ will just stay wobbly and jumpy.  If they are still doing this after 2 or 3 years, they probably aren't going to improve.

D) You might find a Golden LTZ now and then.  These are born pretty stable form Day One, and then just get better.  These are not common, and these are the jewels that would get picked for a Super 3458a (002 option) along with other hand-selected components.

E) There is always the LTZ that starts out rock solid, then as it settles in you see more drift.  These aren't common either.

The point is:  You have to be patient, and plan on a long time scale before you start taking any serious measures of your circuit.

What we do is assemble the LTZ circuits, just do an initial check to see if they are running, then burn them in for a few years.  We use a 'hotter" heater resistor combo with extra 1k's in series with the final top resistor (like 15k~16K over 1k) and then maybe jumper out a 1k or two to get to final heater resistor ratio - and then let that settle in for another 6 months to a year.  The final heater resistor value will depend on it's final running application, but minimum is 13k over 1k, and that is only for coolest environments.  You might be able to get away with less, depending on application.  We have plenty of LTZ circuits running in the field with 14k ~ 15k/1k ratio running 24/7 for decades, and they don't drift maybe 1 or 2ppm per year and get more stable - so I definitely do not agree with Pickering's suggestions about lower heater resistor ratios in every case.  Just speaking from experience here after looking at a lot of LTZ's.  SOME LTZ's might benefit from a lower heater ratio, but it really depends on where you'll be running these in their work place - and each LTZ is different.

For instance right now in 2016 we're processing LTZ circuits that will be delivered in 2019 to a wafer fab production line test system.  This is how HP / Agilent / Keysight Vref boards are done also.  It's not like they are soldered together and delivered the next day.

2.  Stick to the LT datasheet circuit especially if this is your first time building an LTZ circuit. Suggest that you DO NOT use any other op-amp besides LT1013 - that amp is tailor made for LTZ circuit, and will get you best noise performance overall.  The small Vos of the LT1013 isn't really going to have much impact since it's used as a current driver.  If you try another auto-zero amps ('2057 et. al.) you have to give serious consideration to noise generated especially in the rest of you circuit - let alone the current noise on the inputs.  I have never seen any other amp work as good at LT1013 in the long haul.

If you call LT applications engineering, they will tell you exactly the same thing, and they've seen a lot fo LTZ circuits also.  This is good advise.

KEEP ALL SURFACE MOUNT ceramic caps AWAY from any LTZ circuit.  They are microphonic and pick up all sorts of mechanical noise.  Use high quality poly film caps or better, and Thru Hole works better over long term, especially if your PCB is going into an un-stable temperature area.

3.  Adding slots to your PCB may hurt you - I've made boards with and without 'thermal' slots, and never noticed any huge difference either way.  We had one design with slots that all it did was pick up vibration from a vacuum pump running about 20 feet away on the production line - and it showed up on the Vref output as noise.  Make the board compact and thermally uniform.  Keep drafts away, and don't run a serious LTZ out in the open.   Pay attention to about every Linear Tech and HP Schematic that shows star grounding techniques.

4.  If your ambient temperature is stable, then something like cheaper PTF56, 5ppm TC  resistors work just as well in the end for your LTZ as $80+ Vishay Magical Voodoo resistors.  Spending a lot of money on the Vishay Magicals will generally reduce back pain since your wallet is a lot lighter, but that's about it.  You will not get 40 X more performance out of an LTZ made with expensive resistors, guaranteed -  I've never seen that happen, not even close, and I've compared several different assembly methods and parts BOM's over time.

5. PWW 3ppm resistors work fine especially if your ambient temperature is changeable.  I have never seen a Vishay Magical Voodoo work any better in an LTZ circuit than PWW.  Remember, the whole point of LTZ is that the resistor Drift RVal is attenuated by over 100 to 1 - and this is conservative.  You won't see much difference in the Vref if it was made with 1ppm or 3ppm resistors if your ambient temp is changing even several degrees.  The heater resistor ratio is most important to keep an eye on to make sure stays stable AS A RATIO.  The other resistor values are not too critical as long as they are stable...so don't waste money on tight tolerance resistors if you don't need them.

6. Don't be afraid of good quality PWW trimmer pots to make adjustments in a VBoost circuit.  In fact we've got several 3456a that go in for calibration, and they are still in cal after decades with the pots never been touched since the 80's.  You can do the same with LTZ boost circuits IF your final circuit is thermally uniform and doesn't see a lot of vibration.

7. I would NOT recommend ever changing out the heater resistor in a perfectly running 3458a.  I have seen people do that, and all that happens is now they have a driftier 3458a, because the problem wasn't in the heater resistor.  Know what you're modifying!  The ambient temp inside a 3458a is not your lab temperature, and these instruments are designed to run on racks alongside other warm equipment!  I wouldn't try putting a 12.5k heater resistor into a 3458a unless you had a very well ventilated area.

8.  STAY AWAY FROM eBAY LTZ's.  That's only going to end in tears.  If you take a used LTZ from an unknown source - and who knows how it's been mis-treated while it was demounted - and you solder it in to a new circuit:  Your still looking at a multi-year burn-in cycle to see how it's really going to perform over time.  Every time you get a soldering iron anywhere close to an LTZ chip - expect to see at least a year or two to return to stability.  Yes, it really takes that long - You can try other thermal cycle techniques to convince yourself you can speed up the process - and then you realize it took 18 to 24 months before you see real stability again.

With that type of time commitment involved, I suggest that it's never worth your time to save a few bucks on unknown, used garbage from eBay.  It will never save you a dime.

We have always purchased direct from LT, and I can't think of a single time over hundreds of circuits that we've gotten a bad, out of spec LTZ.  Maybe once, but that was an obvious dead transistor we saw right on power up - but LT happily replaced it.

Some LTZ's will be better performers of course - but generally we've never been steered wrong buying direct from the source.  Seen enough problems from eBay garbage to last a lifetime, so we will never purchase there.

Final Suggestion:  If you are running profitable business and need a good, solid, low drift 10V Voltage Reference, stay with a Fluke 731 / 732 series and keep it cal'd.  You won't be able to roll your own and save any money if your time is worth anything.

Don't get me wrong - Of course playing and chasing PPM for entertainment purposes is always fun too!  I've just been around this long enough that chasing the final PPM or two drift is basically the point of vanishing returns.  In terms of profit margin - that last few PPM is not going to ever pay back except in the most critical applications.



 




« Last Edit: October 19, 2016, 06:14:38 pm by MisterDiodes »
 

Offline Echo88

  • Frequent Contributor
  • **
  • Posts: 628
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1565 on: October 19, 2016, 06:30:40 pm »
Thank you for that informative post! Your post and the infos from Mr. Pettis encourage me to build one or two LTZ-references for myself, since i now see that i dont need to spend that much money on vishay-resistors.

 
The following users thanked this post: MisterDiodes

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 2650
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1566 on: October 19, 2016, 07:11:06 pm »
Hello,

thanks from my side too for the comprehensive summary.

to point 2.
what about using LTC2057 for long term stability?
The hermetically LT1013A are very expensive compared to plastic LTC2057.
And in my cirquit every input of the OP-Amp carries a filter capacitor anyway to reduce EMI.

to point 5.

As a volt-nut you may want to have the tolerance of the temperature setting resistors less or equal to 0.1% especially if using a lower temperature setpoint. ( e.g. 1K/12K resistor).
Of course you can use 1% resistors if using a higher temperature setpoint.

With best regards

Andreas
 
The following users thanked this post: MisterDiodes

Offline Edwin G. Pettis

  • Frequent Contributor
  • **
  • Posts: 435
  • Country: us
  • The plural of anecdote is not data.
Re: Ultra Precision Reference LTZ1000
« Reply #1567 on: October 19, 2016, 10:03:51 pm »
As noted, the LT1013 is intended as the best match for the LTZ, its drift is inconsequential as it is operated as a current device, it is very stable long term.  I do not think that a hermetic LT1013 is important, the circuits referred to are not hermetic in practice and the LTZ apps and notes do not specify a hermetic.

For external inference (EMI) is it better to use a shield (if necessary) for the enclosure rather than bypassing inputs all over the place, a shield made out of copper PCB built around the LTZ module and grounded can be much more effective, ceramics can also be sensitive to some RF signals and couple them into the circuit themselves.  Bypass the power supply input at the LTZ module, possibly using an RCL filter if needed.  LT specifically rules out using ceramic caps, no exceptions, I've talked to them too.

The exact ratio of the temperature setting resistors is not so important as reasonably close tracking of the two resistors, ±0.1% tolerance is quite acceptable as long as the two resistors have very similar TCRs.  The exact temperature is not important as long as the resistors are reasonably long term stable, we're not talking PPM in this case.  15K:1K is about right for just about all circumstances, it won't hurt the chip in the least.

Another advantage of the copper PCB shield is that you can completely shield against external air drafts, the PCB doesn't pass outside air temperature changes very well, helping to keep the internal temperature stable.

Linear Tech has a vast amount of long term data on their LTZ chips, HP (Keysight) also keeps such data and shares with LT, the LT engineers know more about how these LTZ chips work and perform than anybody else, they know what works for breaking in these references and will tell anyone who asks, it isn't a secret (well, maybe Keysight might not tell all).
« Last Edit: October 20, 2016, 04:39:34 pm by Edwin G. Pettis »
 
The following users thanked this post: lowimpedance, TiN, MisterDiodes

Offline MisterDiodes

  • Frequent Contributor
  • **
  • Posts: 457
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1568 on: October 19, 2016, 11:25:10 pm »
Hello,

thanks from my side too for the comprehensive summary.

to point 2.
what about using LTC2057 for long term stability?
The hermetically LT1013A are very expensive compared to plastic LTC2057.
And in my cirquit every input of the OP-Amp carries a filter capacitor anyway to reduce EMI.

to point 5.

As a volt-nut you may want to have the tolerance of the temperature setting resistors less or equal to 0.1% especially if using a lower temperature setpoint. ( e.g. 1K/12K resistor).
Of course you can use 1% resistors if using a higher temperature setpoint.

With best regards

Andreas

#2. What long term stability issue?  The LT1013 has a slight Vos drift which doesn't matter a lot as a in an LTZ current driver circuit. The Zener + Transistor base stability in the LTZ will overwhelm that.

Hermetically sealed?  If humidity is a concern use PTFE or parylene coating, both of which make the circuit immune to humidity effects.  And then seal it again in a good EMI enclosure  Just because you get a plastic IC package doesn't mean you don't seal the system yourself.   If you're building a robust circuit for industry or military, modules will almost always be sealed anyway.  These guys don't want a bare circuit board - you supply them a sealed box...power in >>> stable voltage out.  Bulletproof and reliable. 

Look at the huge current noise on '2057 inputs, and limited bandwidth - both are also concerns if you're driving a fast SAR ADC VRef input.  A different story though.

LT recommends the LT1013 only, and that's what we use, and we seal up the circuit also.  They do NOT want a '2057 in there creating current noise issues.

Never been a problem with LT1013 and I have circuits running in Finland, Georgia, USA and very humid Thailand.  That means we go from about 5% humidity to about 100% -  They are doing fine and stable over long time periods.  Other amps have been a problem either with noise on the output or having to be very careful with supply bypassing, and messing up a circuit someplace else.  Those chopper switches inside any A-Z amp don't come cheap when you're keeping the analog system as quiet as possible, and watch out if you're using them as a buffer.  Those op-amp response poles will come around and bite you. <Grin>

It's not that the AZ amps don't have their place - but you have to consider the high current input noise carefully.

#5.  I suppose.  But I've never found a location for a properly sealed, robust circuit container that would be running that cool in our application on a test rig.  A 1% matched resistor pair - as long as they have a stable ratio and good TC in ratio mode - are usually fine.  You just need a resistor ratio that is rock solid - absolute value is not as critical unless ( as you correctly point out) - you get down closer to 12k over 1k.  Now you have to watch out if you're that close to the edge   That's kind of pushing the envelope on the cool side, but if your environment tolerates it...

Listen, I just got a 4yr old Keysight 3458a back from Cal lab, with Z-540 data - and this has the standard 15k/1k heater pair, LT1013 plastic - and it hasn't moved 2ppm over the year, and it doesn't even have option -02.  Running 24/7 in a warm cabinet.  You set the heater resistor ratio for the environment it's running in, and you want it warm enough to regulate the thermal changes of the surrounding area.  So just because it's 15k over 1k doesn't mean it'll drift more.  Reducing the heater temp will save a little power (good for battery power), but from what I've seen - it doesn't have a HUGE effect on EVERY LTZ drift rate... that's all I'm saying.  The drift rate of an LTZ is what it's going to be..
« Last Edit: October 20, 2016, 05:43:35 pm by MisterDiodes »
 

Offline MisterDiodes

  • Frequent Contributor
  • **
  • Posts: 457
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1569 on: October 19, 2016, 11:50:53 pm »
Thank you for that informative post! Your post and the infos from Mr. Pettis encourage me to build one or two LTZ-references for myself, since i now see that i dont need to spend that much money on vishay-resistors.

You'll be fine.  Now the only other thing to consider before you build an LTZ1000/ LTZ1000a circuit is monitoring.  You see lots of people here presenting data from a 3458a - which is sort of OK "if that's all you have" (not joking - a 3458a by itself is not necessarily good enough) - but you have to remember that in the case of an LTZ circuit:  You are building basically the same reference circuit that the 3458a is using - and so in a sense what the measurement data shows is the drift rate of the 3458a compared to the circuit under test - IF you're just looking at absolute voltage value on the DMM.  Which doesn't tell you much about how your circuit is drifting or not.

Ideally if you have access to a 732b (which is about the only source stable and quiet enough) and a 7-decade voltage divider, you can use the 3458a, 3456a or other capable NULL meter to watch your Vref noise & drift voltage.  You really want to do this occasionally during your Vref burn in period or else you'll never know when you've achieved real stability on your new circuit.  Plan on it taking a while.

If you're making an LTZ circuit with a 10V boost circuit added, you can use a null meter comparison directly at 10V, and watch for drift from there.

The concept here is that you don't use the reference in the 3458a to measure the absolute value of your newly built reference - because you really want to measure it against a better reference if possible.

In the case of the LTZ1000x, that gets very hard to do unless you have a JJ-array laying around - because LTZ's are about as good as it gets!

In fact my lab 3458a jsut came back from Cal lab - and the Fluke calibrators aren't accurate enough to do the complete cal routine on the 3458a.  You need also a 732b (or averaged 734b preferred, which is basically four 732's in a frame) and /or a "golden" 3458a which is freshly cal'd with -02 option.  Those get categorically more expensive.

Good Luck!



« Last Edit: October 20, 2016, 05:42:38 pm by MisterDiodes »
 

Online ManateeMafia

  • Frequent Contributor
  • **
  • Posts: 720
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1570 on: October 19, 2016, 11:56:14 pm »
@MisterDiodes

What percentage of LTZ's make the final cut in your projects? In other words, if someone wanted to build a LTZ based reference, how many would you recommend starting with for the 2-3 year process?
 
The following users thanked this post: MisterDiodes

Online TiN

  • Super Contributor
  • ***
  • Posts: 4236
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Ultra Precision Reference LTZ1000
« Reply #1571 on: October 20, 2016, 04:07:36 am »
No less than 5, I'd expect. You can't really tell if measurement equipment acting up with less than 3 references anyway.  :-DMM

Btw, I'll just leave a link here for LTZ module study I've done last week, in case thread get forgotten.
YouTube | Metrology IRC Chat room | Live-cam | Share T&M documentation? Upload! No MB limit, firmwares, photos.
 
The following users thanked this post: MisterDiodes

Offline MisterDiodes

  • Frequent Contributor
  • **
  • Posts: 457
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1572 on: October 20, 2016, 12:45:49 pm »
@MisterDiodes

What percentage of LTZ's make the final cut in your projects? In other words, if someone wanted to build a LTZ based reference, how many would you recommend starting with for the 2-3 year process?

I would start with at least a minimum of 4~5 units.  All of them will probably work fine in-spec, but if you want to be able to compare, pick and choose the best performers for an extra-critical application - build yourself a few extra to sort thru.  The more the merrier.

For instance if we do a production order for 20 units, we'll start processing say 25~26 units, and do a sort after 18 ~24 months or so depending on application.  Some of them will get in their comfortable place sooner than their siblings.  Even the leftovers will still be running better than most any reference and can be used for other less-critical backups - or save them on the shelf for the next order or in case the customer physically damages a unit, warranty replacement etc.  Stuff happens.

As with any production situation, on occasion a unit won't pass muster for some reason - and that's why you really want to have some other better reference to null-check check against - like a 732, or similar (like a parallel group of known working refs, more 3458a's, etc.).  If you JUST have one 3458a - again no joke here - all you can check is the drift rate of your new circuit vs. drift rate of 3458a, which will tend to be on same sort of scale.  So you won't learn everything - the new reference you're measuring is PROBABLY fine if it looks good on one 3458a, but no real way to tell what the true absolute drift rate is of either the reference OR 3458a - all you know is how they are drifting relative to each other over time.

You can gain more confidence if your circuit is measured against a better reference (732 is at least somewhat better), or perhaps cross-check across multiple 3458a, etc.  At most cal labs the goal is to try to measure against a standard that is 10 times more accurate than the unit under test - but that is harder to do with an LTZ circuit, because that IS about as good as it gets. 

So now you have a good excuse to get more 3458a's or some 732b / 734 - or wherever your PPM adventure takes you!  :-DD

If you get in touch with your Volt-Nut side, just remember: If it's something worth doing, it's worth doing to excess!!

« Last Edit: October 20, 2016, 05:45:55 pm by MisterDiodes »
 
The following users thanked this post: TiN

Offline d-smes

  • Regular Contributor
  • *
  • Posts: 92
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #1573 on: October 31, 2016, 04:01:05 pm »
How important are guard rings around the collectors (pins 5 & 8 ) of the LTZ and why are they used in many LTZ layouts?  If for control of leakage currents, what about the 10's of nA bias current of the LT1013?  If to reduce capacitive coupling, what about 4-layer designs where you already have appreciable capacitance to ground?
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 8266
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #1574 on: October 31, 2016, 05:03:00 pm »
The guard traces are not that important, but the effort may be rather low to have them. The typical current through the transistors is 100 µA, and for every 100 ppm change in current there will be about a 1 ppm change in reference voltage. The input bias of the LT1013 is relatively stable - it is only the change in current (e.g. due to temperature or aging) that really matters. Leakage currents are prone to be changing but hardly reach the nA range unless in a dirty humid environment.

Capacitive coupling is not a problem - this might give slight changes at high frequencies like above 10 kHz, nothing one cares about in an DC reference. The circuit should be tolerant to parasitic capacitance.
 
The following users thanked this post: d-smes


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf