Author Topic: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?  (Read 3103 times)

0 Members and 1 Guest are viewing this topic.

Offline denimdragonTopic starter

  • Regular Contributor
  • *
  • Posts: 227
  • Country: us
  • "Hole charge chaser and wanna-be o-scope fixer"
 Big guns, I need some guidance. I have been on the hunt for a voltage standard or 8.5 digit multimeter to serve the same purpose. The multimeter would be great because I could have more than one NIST traceable parameter, e.g. DC Volts, DC resistance, etc., but I acquired a Fluke 5450A late last year so the resistance traceability is no longer an issue once I have it calibrated.

Other than the normal "voltnuttery'' addiction we all share, I plan to manufacturer hobbyist level standards that are affordable and sell them, so I must have the best in house references that I can afford to do so. My budget is $3K.

I have been hunting for a 3458A, and I have been able to find reasonable ones, but the logistics has been a nightmare trying to get one. I missed a cal'd opt 001 002 this week because I didn't have the time to travel and get it, nor the extra $340 it was going to take to have it shipped from Californa (the meter went for 4K along with accompanying 34401A, Fluke Hydra, and some other cheap meter. That's 4k, before taxes and buyers premium). That sucked...

So now I am considering the Guildline 4410 10V and 1.018V standard or the Advantest/ADCMT 8.5 digits meters which I have found fit in my budget. I'm leary about the R6581/T, 7840 etc., because I have heard the noise floor is rather high, there isn't much info (...I have read) about repairs nor english schematics. Those are major concerns, but if need be, I should be able to work around that.

I don't know a lot about the Guildline 4400 series, but I have heard good things about it. And I have NOS LM129H vrefs that the unit comes with. I know vrefs don't really die, but the fact that I have those makes me feel a bit better, and from what I have read here, they are pretty stable (single digit yearly ppm stability?).

I have a Fluke 5440B I have had for a while that I have yet to be able to repair because of the outgaurd error, and the unreliability of the relays puts me in another possible high upkeep situation. I intended to repair it, have it cal'd, and use it as a reference since its a legendary piece when it come to stabilty. I don't have the luxury of that anymore since I want to have these products on the market before the year is out.

I am open to suggestion to my end goals stated above if someone has another direction I can take.

Thanks in advance.
"Hi there. I'm short in the hair department, do you mind if I pull a few strands from your head? Thanks!"
 

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #1 on: February 24, 2024, 05:21:00 pm »
What uncertainty are you aiming for? It makes a big difference if you are trying to deliver standards with a 0.1% uncertainty or a 50 ppm uncertainty. Given the equipment you're talking about I'm guessing it's more like 50 ppm or better.

I don't have experience with the Guildline / Statronics voltage standards or Advantest meters, but I can say in general that a voltage standard makes for a better voltage standard than a DMM when they are in the same class. Just look at the 1 year stability specifications which might be sub-ppm/yr for a good voltage standard and maybe 3 ppm/yr for the very best bench DMM. Calibration of a voltage standard might also be cheaper and is certainly simpler, since a DMM generally needs more stimuli. So if the goal is to deliver voltage standards you can rely on with the lowest uncertainty, then a voltage standard would be the superior solution.

The only caveat is that you need to be able to transfer the voltage from the standard to the device under test. With a DMM this is trivial: just measure the device under test. With a voltage standard if they are both 10 V, then you could connect the negative terminals together and measure the difference between the positive terminals with a 6.5 digit DMM. Because this DMM is only measuring a very small voltage, even 1% uncertainty in this small voltage will translate to a small uncertainty in the total voltage. If the values are not close, like a 10V standard and a 5V dut, then transferring is less straight forward. If you have a Kelvin-Varley divider or DMM with a high guaranteed linearity, you can use this to transfer 10V to 5V. A 3458A can do this job well due to its very high linearity. Same with the Fluke 720A or ESI RV722 Kelvin-Varley dividers, though they are certainly much more fiddly and sub-optimal for any kind of production. The 3458A and Fluke 720A have the advantage is that the linearity is inherit in the design or in the case of the Fluke 720A can be adjusted using just a stable voltage source and null meter. In this case you don't use the DMM as a voltage standard, just as a ratio device that measures the ratio between two voltage standards.

Calibration is about having a history saying that at point 1 in time the device was at value X, and at point 2 in time it was at value X+delta, so probably between points 1 and 2 the value was between X and X+delta. The more history, the more confidence you have that this is true. So if you want to use a standard (voltage standard or multimeter) down to its specifications, you ideally want to build up some history of calibrations. Also if voltage references have been turned off for a long time (months or years) and may have sat in a cold warehouse, you expect the initial drift to be higher. So ideally when buying a new standard (of any kind) that you want to keep in calibration, my suggestion would be first power it for at least two weeks and monitor its stability to the best of your abilities, and then send it out for calibration. Make sure you get data and that the uncertainty they deliver is low enough for you. Try to monitor stability within the means you have available. Then after a year or less, send it out again for calibration, and compare the data to the previous calibration. Only after this year (or three months or whatever the time is between calibrations) would I really trust it. So have a look where you could have the units calibrated and what it would cost. Don't be surprised to spend a substantial part of your budget on calibration.

So very roughly these are my suggestions from highest uncertainty and lowest cost to lowest uncertainty and highest cost:
  • A good 6.5-8.5 digit bench DMM
  • A voltage standard with decent 6.5 digit bench DMM (great uncertainties for voltages close to the voltage standard)
  • A voltage standard, Fluke 720A/ESI RV722 Kelvin-Varley Divider and null meter (if you have more time than money, though still not cheap)
  • A voltage standard plus 8.5 digit bench DMM, ideally HPAK 3458A (better uncertainty than KVD and much more convenient)
 
The following users thanked this post: denimdragon

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1945
  • Country: us
    • The Messy Basement
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #2 on: February 24, 2024, 05:56:11 pm »
I'd love a 3458 but it's just not in the budget. IMO, anything you can find at the $3k level is going to be risky. Demand is still pretty high. Voltage standards are a better bet, but a single one won't do what you need. You need a minimum of three to intercompare, so you can have one calibrated and still know where everything is. What you want to do can be done on the cheap, but "cheap" is going to cost you well more than $3k, or you have to be very lucky with eBay finds and such. ESI, Fluke and Julie all made good KVDs and IMHO you really should have one.
 
The following users thanked this post: denimdragon

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #3 on: February 24, 2024, 06:33:25 pm »
Voltage standards are a better bet, but a single one won't do what you need. You need a minimum of three to intercompare, so you can have one calibrated and still know where everything is.
To be clear, this applies to any kind of standards. So voltage standard, resistance standard, frequency standard, DMM, etc. Having multiple cells in one, like the Guildline 4410, sort of satisfies this. Sort of, because when sending out for calibration, you are sending them all out, so you have no way to check how much it drifted during transport. But it does depend on how close you want to get to the limits of performance. If your requirements are well below the manufacturer's specifications, then I would be less concerned about this than if you are trying to exceed those specifications (which is definitely feasible for a good standard).
 
The following users thanked this post: denimdragon

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14368
  • Country: de
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #4 on: February 24, 2024, 09:47:07 pm »
For the R6581 (T) there is a thread here in the forum about repairs, that also includes a reverse engeniered schematics. AFAIK noise is not so much the problem. The more common issues are a dim display with no easy fix and quite some input switching spikes. The 6581T version often has a not so good factory calibration and this way has not so good linearity. So the lower price also comes with lower performance. It is still one of the cheaper 8 digit meters with a LTZ1000 based reference.
 
The following users thanked this post: splin, denimdragon

Offline denimdragonTopic starter

  • Regular Contributor
  • *
  • Posts: 227
  • Country: us
  • "Hole charge chaser and wanna-be o-scope fixer"
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #5 on: February 24, 2024, 09:50:14 pm »
Thanks for the rapid repsonses! I was leaning heavily towards the Guildline initially because I have an HP3456A, HP3457A, 2 x HP34401A. So the measure of the terminal differences would be perfect. I plan to offer 10V standards with 50ppm specs primarly, and see how that works out.

I will keep my standards cal'd every year and I already have funds that where set aside to have my 6.5 DMMs (save the 3457A) and the 5440B calibrated. I got a quote last year for those so I would know what to expect, but when the 5440B turned out to be more time consuming than anticipated, I decide to just keep those funds for the new standard/meter that I get and maybe my 3456A (I'm partial to "Big Shirley" and I need screw terminals for measurement monitoring).

I overlooked the 720A and I had a chance to get one for a song early last year, but ruled it out because I had a 3458A in the crosshairs  :palm:

Now, since thinking about the benefits of the reference standards stabililty, it makes the decision a little easier. I'll need a KVD and a Null meter. I might have one already but I need to take a look in storage. I'm about to go check. Back in a snap.
"Hi there. I'm short in the hair department, do you mind if I pull a few strands from your head? Thanks!"
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7950
  • Country: us
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #6 on: February 24, 2024, 10:48:34 pm »
I plan to offer 10V standards with 50ppm specs primarly, and see how that works out.

I would think a 10V standard and a null/microvolt meter would be where to start.  A Fluke 732A/B and a Keithley 155 would be good examples.  Once your 10V standards are out and you want to expand your offering, then a KVD and 8.5-digit DMM perhaps. 

Are you sure there's a market for 10V 50ppm "standard" references?  They'd have to be pretty cheap.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 
The following users thanked this post: denimdragon

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #7 on: February 24, 2024, 11:20:56 pm »
I don't think you need a KVD. I'm sure you'll find one on the shelves of many cal labs, but I doubt they'll see much use. A 3458A has better linearity and is much quicker to use. It's just that KVDs are generally a lot cheaper and can substitute for some of the functionality (DCV:DCV transfers in ratios not too far from 1) with uncertainties that are not that much worse.

I don't think you need to go crazy to get the standards within 50 ppm. A good voltage standard that gets calibrated, or even a good 7.5-8.5 digit DMM, should be all that's necessary. Something like a Datron 1081/1082 would be plenty for this. Also look for Solatron/Schlumberger DMMs (7071/7081). They don't come up as often as Advantest and have an annoying input connector, but they can be quite a lot cheaper.

I would think a 10V standard and a null/microvolt meter would be where to start.  A Fluke 732A/B and a Keithley 155 would be good examples.  Once your 10V standards are out and you want to expand your offering, then a KVD and 8.5-digit DMM perhaps. 
A null meter like Keithley 155 only makes sense if you are adjusting your device (or a KVD) to equal the standard. They are optimized for measuring if a voltage is null, not how far a voltage is away from null. Generally in metrology standards have a value that is closeish to the nominal value but they are not adjusted to have exactly that value, so the voltage standard might get assigned a value of 9.9938V. So if you want to adjust your device to be as close to 10V as possible, you want it to be about 600uV higher. A Keithley 155 is not very useful for this, since it will just be an analog meter with mediocre accuracy. I'd rather use a good bench DMM or if you want to go crazy a nanovoltmeter for this.
 
The following users thanked this post: denimdragon

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 958
  • Country: nl
  • The sky is the limit!
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #8 on: February 24, 2024, 11:32:11 pm »
Thanks for the rapid repsonses! I was leaning heavily towards the Guildline initially because I have an HP3456A, HP3457A, 2 x HP34401A. So the measure of the terminal differences would be perfect. I plan to offer 10V standards with 50ppm specs primarly, and see how that works out.

I will keep my standards cal'd every year and I already have funds that where set aside to have my 6.5 DMMs (save the 3457A) and the 5440B calibrated. I got a quote last year for those so I would know what to expect, but when the 5440B turned out to be more time consuming than anticipated, I decide to just keep those funds for the new standard/meter that I get and maybe my 3456A (I'm partial to "Big Shirley" and I need screw terminals for measurement monitoring).

I overlooked the 720A and I had a chance to get one for a song early last year, but ruled it out because I had a 3458A in the crosshairs  :palm:

Now, since thinking about the benefits of the reference standards stabililty, it makes the decision a little easier. I'll need a KVD and a Null meter. I might have one already but I need to take a look in storage. I'm about to go check. Back in a snap.

50ppm and 10 Volts is already within the annual spec of the 34401 (35 ppm) if you calibrate both annually and cross-check between them for drifts/malfunctions lets say weekly you should already be in pretty good shape.

If you need/want to spent 3K to improve things I'd get a DMM6500 (25ppm/year for 10V DC) with scan card option to monitor (temperature co-efficient test?) and calibrate multiple of your products at the same time.
A upgrade from that would be to use the DMM6500 (or 34401) as a null meter against a voltage standard such as Fluke 732(a/b/c/) (2ppm/year) or a more budget friendly AB-Precision DVR01 (8ppm/year for 10V DC).
https://www.ab-precision.de/products/electrical-standards/

From experience I can tell you that older used test equipment is great to tinker with as a project but if you need a tool that gets the job done without to much hassle it is better to get something more modern that comes with a warranty and simply works straight out of the box.
 
The following users thanked this post: denimdragon

Offline denimdragonTopic starter

  • Regular Contributor
  • *
  • Posts: 227
  • Country: us
  • "Hole charge chaser and wanna-be o-scope fixer"
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #9 on: February 25, 2024, 12:29:25 am »
Ok, so that's good news about the null meter. I had a Fluke 845(A/AB?), but I went to my warehouse and it said "find me". I like the idea of using the 6.5s as it seems this will be the most efficient timewise and the cal quote I have for 34401As won't break the bank if I use those 2 and 3456A. I will still need to get another 10V standard to compare to the Guildline.

So far:
Guildline 4410 10V Standard
2nd "reasonable priced" 10V Standard
2x Cal'd 34401A
1x Cal'd 3456A

Does this look like a decent starting point? Guildline Pdf below.

Thanks again everyone.


"Hi there. I'm short in the hair department, do you mind if I pull a few strands from your head? Thanks!"
 

Offline denimdragonTopic starter

  • Regular Contributor
  • *
  • Posts: 227
  • Country: us
  • "Hole charge chaser and wanna-be o-scope fixer"
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #10 on: February 25, 2024, 12:37:35 am »
Quote

50ppm and 10 Volts is already within the annual spec of the 34401 (35 ppm) if you calibrate both annually and cross-check between them for drifts/malfunctions lets say weekly you should already be in pretty good shape.

If you need/want to spent 3K to improve things I'd get a DMM6500 (25ppm/year for 10V DC) with scan card option to monitor (temperature co-efficient test?) and calibrate multiple of your products at the same time.
A upgrade from that would be to use the DMM6500 (or 34401) as a null meter against a voltage standard such as Fluke 732(a/b/c/) (2ppm/year) or a more budget friendly AB-Precision DVR01 (8ppm/year for 10V DC).
https://www.ab-precision.de/products/electrical-standards/

From experience I can tell you that older used test equipment is great to tinker with as a project but if you need a tool that gets the job done without to much hassle it is better to get something more modern that comes with a warranty and simply works straight out of the box.

I was thinking about 34970A for this task. I'm very familiar with them and my GBIP set-up is mostly HPAK stuff. I'm not opposed, but that's another time sponge learning the Tek/Keithley idiosyncracies. I've had enough problems fighting NI, HPAK, visa conflict issues, that i don't want to endour again if at all possible.
"Hi there. I'm short in the hair department, do you mind if I pull a few strands from your head? Thanks!"
 

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 958
  • Country: nl
  • The sky is the limit!
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #11 on: February 25, 2024, 12:57:33 am »
Quote

50ppm and 10 Volts is already within the annual spec of the 34401 (35 ppm) if you calibrate both annually and cross-check between them for drifts/malfunctions lets say weekly you should already be in pretty good shape.

If you need/want to spent 3K to improve things I'd get a DMM6500 (25ppm/year for 10V DC) with scan card option to monitor (temperature co-efficient test?) and calibrate multiple of your products at the same time.
A upgrade from that would be to use the DMM6500 (or 34401) as a null meter against a voltage standard such as Fluke 732(a/b/c/) (2ppm/year) or a more budget friendly AB-Precision DVR01 (8ppm/year for 10V DC).
https://www.ab-precision.de/products/electrical-standards/

From experience I can tell you that older used test equipment is great to tinker with as a project but if you need a tool that gets the job done without to much hassle it is better to get something more modern that comes with a warranty and simply works straight out of the box.

I was thinking about 34970A for this task. I'm very familiar with them and my GBIP set-up is mostly HPAK stuff. I'm not opposed, but that's another time sponge learning the Tek/Keithley idiosyncracies. I've had enough problems fighting NI, HPAK, visa conflict issues, that i don't want to endour again if at all possible.

Sure, just wanted to point out that there are different options that fit your needs and price range other than ancient gear in unknown condition from a well known auction site.  :-+
 
The following users thanked this post: denimdragon

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #12 on: February 25, 2024, 01:07:12 am »
50ppm and 10 Volts is already within the annual spec of the 34401 (35 ppm) if you calibrate both annually and cross-check between them for drifts/malfunctions lets say weekly you should already be in pretty good shape.
denimdragon will need to calculate their uncertainty budget carefully to make sure they meet the 50 ppm spec, because the 35 ppm figure is only one of the components. You might also need to figure temperature coefficient of both your meter and the device under test, noise, setting accuracy, repeatability, etc. Particularly if you want to adjust something to be within 10V +/- 50 ppm, then if the total uncertainty of your measurement is 45 ppm, that means your measurement must be within a few ppm from 10V to ensure the actual value is within 50 ppm. Things are much simpler if your measurement uncertainty is much better than the uncertainty you are targeting. This is sometimes called test uncertainty ratio, and a test uncertainty ratio of 1.4:1 or worse is a very aggressive value that requires being very careful.

From experience I can tell you that older used test equipment is great to tinker with as a project but if you need a tool that gets the job done without to much hassle it is better to get something more modern that comes with a warranty and simply works straight out of the box.
The problem in this space is that performance of modern instruments is much worse than older instruments in the same price class unless you're willing to spend $10k or so. The basic DC uncertainty of digital multi meters has really not improved much since the late eighties. The HP 3456A from 1984 or so is actually a bit better than the DMM6500 on DC (23 ppm/yr basic uncertainty). So if you want the lowest uncertainty per dollar, then something like a HP 3456A or Solartron 7081 are likely much better value than anything you buy new. I agree with the reliability argument, but what's the point of having something reliable that does not have the performance you need?

I was thinking about 34970A for this task. I'm very familiar with them and my GBIP set-up is mostly HPAK stuff. I'm not opposed, but that's another time sponge learning the Tek/Keithley idiosyncracies. I've had enough problems fighting NI, HPAK, visa conflict issues, that i don't want to endour again if at all possible.
I have not had problems with this. Particularly for later instruments (starting in the early nineties) that use SCPI, programming instruments from different brands usually only requires minor changes. In the older R2D2-language instruments like the HP 3456A the differences are bigger, but back then even two different instruments from the same manufacturer would use different commands.
 
The following users thanked this post: denimdragon

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 958
  • Country: nl
  • The sky is the limit!
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #13 on: February 25, 2024, 01:26:27 am »
Ok, so that's good news about the null meter. I had a Fluke 845(A/AB?), but I went to my warehouse and it said "find me". I like the idea of using the 6.5s as it seems this will be the most efficient timewise and the cal quote I have for 34401As won't break the bank if I use those 2 and 3456A. I will still need to get another 10V standard to compare to the Guildline.

So far:
Guildline 4410 10V Standard
2nd "reasonable priced" 10V Standard
2x Cal'd 34401A
1x Cal'd 3456A

Does this look like a decent starting point? Guildline Pdf below.

Thanks again everyone.

Guildline specs look good and as it contains four independent (*) references you may not need another one per-se.

*To bad the linked document doesn't contain the schematics, these four references likely share the same power-supply and depending on the design (simple resistor from common psu-rail to set zener currrent?) all four outputs may be affected in the same way from a common event (power-supply voltage drifting, ground loops etc.), 734b handles that better imho (but not relevant to this discussion).
A "second-opinion" (reference) will shine (some) light onto that and pull you from darkness into a grey area.  :) 
 
The following users thanked this post: denimdragon

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 958
  • Country: nl
  • The sky is the limit!
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #14 on: February 25, 2024, 02:04:14 am »
50ppm and 10 Volts is already within the annual spec of the 34401 (35 ppm) if you calibrate both annually and cross-check between them for drifts/malfunctions lets say weekly you should already be in pretty good shape.
denimdragon will need to calculate their uncertainty budget carefully to make sure they meet the 50 ppm spec, because the 35 ppm figure is only one of the components. You might also need to figure temperature coefficient of both your meter and the device under test, noise, setting accuracy, repeatability, etc. Particularly if you want to adjust something to be within 10V +/- 50 ppm, then if the total uncertainty of your measurement is 45 ppm, that means your measurement must be within a few ppm from 10V to ensure the actual value is within 50 ppm. Things are much simpler if your measurement uncertainty is much better than the uncertainty you are targeting. This is sometimes called test uncertainty ratio, and a test uncertainty ratio of 1.4:1 or worse is a very aggressive value that requires being very careful.


The factors you mention, are indeed very important and cannot be discarded but also not that difficult to circumvent, keeping a steady temperature (close to what the dmm or reference was calibrated at) and taking a lot of samples (48 hours per dut minimum I assume the OP is at?) will show repeatability within a few ppm's (depending on dut, but not that difficult to achieve).
TUR, do we need to follow nist guidelines?

From experience I can tell you that older used test equipment is great to tinker with as a project but if you need a tool that gets the job done without to much hassle it is better to get something more modern that comes with a warranty and simply works straight out of the box.
The problem in this space is that performance of modern instruments is much worse than older instruments in the same price class unless you're willing to spend $10k or so. The basic DC uncertainty of digital multi meters has really not improved much since the late eighties. The HP 3456A from 1984 or so is actually a bit better than the DMM6500 on DC (23 ppm/yr basic uncertainty). So if you want the lowest uncertainty per dollar, then something like a HP 3456A or Solartron 7081 are likely much better value than anything you buy new. I agree with the reliability argument, but what's the point of having something reliable that does not have the performance you need?

Ok, assume you have purchased a 20 or 30 year old dmm from ebay, with unkown history, and have it calibrated at your local cal-center, even if "as-found" is within the meters annual spec how you can you be confident it stays within that spec. for the entire coming year?

I was thinking about 34970A for this task. I'm very familiar with them and my GBIP set-up is mostly HPAK stuff. I'm not opposed, but that's another time sponge learning the Tek/Keithley idiosyncracies. I've had enough problems fighting NI, HPAK, visa conflict issues, that i don't want to endour again if at all possible.
I have not had problems with this. Particularly for later instruments (starting in the early nineties) that use SCPI, programming instruments from different brands usually only requires minor changes. In the older R2D2-language instruments like the HP 3456A the differences are bigger, but back then even two different instruments from the same manufacturer would use different commands.
 

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #15 on: February 25, 2024, 02:21:57 am »
The factors you mention, are indeed very important and cannot be discarded but also not that difficult to circumvent, keeping a steady temperature (close to what the dmm or reference was calibrated at) and taking a lot of samples (48 hours per dut minimum I assume the OP is at?) will show repeatability within a few ppm's (depending on dut, but not that difficult to achieve).
TUR, do we need to follow nist guidelines?
My point is that if you start off with a DMM that takes up the majority of your uncertainty budget, then you will need to maintain tighter control of the other factors in your budget than if you start with a 2 ppm/yr voltage standard and could tolerate a 20 ppm contribution due to tempco. It's important to realize that just because a meter has an uncertainty of 35 ppm doesn't necessarily mean that your measurement will have an uncertainty of 35 ppm. I don't care if you follow guidelines, but it is based on basic statistics. The point is that if you want to say a device is 10V +/- 50 ppm with 95% confidence, and you are measuring it to be 10V + 40 ppm with 35 ppm uncertainty, then you don't have 95% confidence of the device being within +/- 50 ppm. TUR is one way of expressing how careful you need to be. The section "Guardbanding" of this Fluke white paper describes it pretty well.

Ok, assume you have purchased a 20 or 30 year old dmm from ebay, with unkown history, and have it calibrated at your local cal-center, even if "as-found" is within the meters annual spec how you can you be confident it stays within that spec. for the entire coming year?
If you scroll up, then you see I argued that only after the second calibration cycle would I start to get confidence in it actually meeting spec. So for a 2 ppm/yr voltage reference it would take multiple calibrations (could be spaced closer together like 3 months if the uncertainty is low enough) for me to trust it within 2 ppm/yr. But if my requirements are 20 ppm/yr then I'd be okay with a single calibration and monitoring the stability using a 6.5 digit bench DMM (which won't guarantee 20 ppm/yr either but would give confidence). I see more problems with brand new instruments showing excessive drift than older instruments (I remember reading about some Siglent bench meters). But either way I agree you need to verify. The lower uncertainty, the more can go wrong, so I'd be more worried about a 2 ppm/yr voltage reference drifting outside that 2 ppm than a 35 ppm/yr DMM drifting outside the 35 ppm spec.
 
The following users thanked this post: The Soulman

Offline denimdragonTopic starter

  • Regular Contributor
  • *
  • Posts: 227
  • Country: us
  • "Hole charge chaser and wanna-be o-scope fixer"
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #16 on: February 25, 2024, 02:53:32 am »
The newer meter have nice tech, but I need the stability of the old meters. The references in older gear have better stability in my experience and I don't mind the upkeep. I will get a new meter one day, but it won't be for data logging. Plus there is the issue of hystersis with a new meter. In the event it gets turned off for some reason, I'm not sure if it will bounce back to previous levels like the old HPs will. I think Mr. Geller used 3456As irc. I didn't get a chance to get one of his standards, but I read they were pretty good.

The real challenge will be getting the uncertainty math correct. I'm already building out a corner in my upstairs lab that will house all the measuring devices and equipment for the endevour and I want the area sealed off and full temp/RH controlled. Set it and forget it. I don't plan to put anybody out of business, I just want to make the absolute best I can make. My day job/biz involves electronics, so this is just and extension of that, but a bit more fun. If I can cover the cost of my annual calibrations, yearly gear improvements - 3458A, Datron 1281, etc., buy a cup of coffee a day, and can make some people happy in the hobby world, it's a success.

I have a lot of NOS parts. I want to play with some of the old voltage references and see what I can do with them.

Another very important question. Once everything is up and running, how long would one monitor the reference before clearing it for release? 90 days? 1 Year? Most of my parts will be NOS, so no burn-in time prior to assembly. Knowing this would help be build out the "hibernator" room more effectively. I have a warehouse with a lab, a home lab where the room will be, and a lab at my business partners home.

"Hi there. I'm short in the hair department, do you mind if I pull a few strands from your head? Thanks!"
 

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #17 on: February 25, 2024, 04:03:27 am »
Another very important question. Once everything is up and running, how long would one monitor the reference before clearing it for release? 90 days? 1 Year? Most of my parts will be NOS, so no burn-in time prior to assembly. Knowing this would help be build out the "hibernator" room more effectively. I have a warehouse with a lab, a home lab where the room will be, and a lab at my business partners home.
This is probably the important question and will depend on how exactly your standard is built. If it involves scaling voltages, like having a ~6.9V LM399 and scaling that up to 10V, means that the scaling resistors play an important role in drift. You should generally assume that any substantial change in conditions like soldering or changing the current at which it operates will start a new aging cycle. The various topics on LM399, LTZ1000 etc have some information on drift. I think you might also be able to find some information on AD586/AD587 on this forum from users like Andreas (e.g. the AD587LW topic). To measure aging it helps if you can measure small differences. If your accuracy only allows you to detect 50 ppm differences, then you might need to monitor the reference for one or two years before you can quantify the drift. But if you can measure 1 ppm drift, and after 60 days you measure a drift of 1.5 ppm, then this already gives you some insight that the yearly drift will probably be well under 50 ppm. Of course extrapolation is risky, but in the case of voltage references a linear extrapolation is generally conservative since drift tends to slow down over time.

I think how well you characterize the voltage standard is a large part of the value proposition vs buying an LM399-based reference from China. I don't know what kind of accuracy figure hits the sweet spot on the effort vs price vs demand curve.
 
The following users thanked this post: denimdragon

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7950
  • Country: us
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #18 on: February 25, 2024, 04:56:26 am »
The problem in this space is that performance of modern instruments is much worse than older instruments in the same price class unless you're willing to spend $10k or so. The basic DC uncertainty of digital multi meters has really not improved much since the late eighties. The HP 3456A from 1984 or so is actually a bit better than the DMM6500 on DC (23 ppm/yr basic uncertainty). So if you want the lowest uncertainty per dollar, then something like a HP 3456A or Solartron 7081 are likely much better value than anything you buy new. I agree with the reliability argument, but what's the point of having something reliable that does not have the performance you need?

I have to disagree with that quite emphatically.  The actual performance of DMMs improved dramatically when they switched to all closed case calibration and laser-trimmed resistors. so a bit later than the mid-80s.  I repair and have all sorts of fun with the old bench monsters like the Fluke 850xA series and the HP 345xA and so forth and while a well-aged, well-maintained and well-cared for instrument that sits on one bench and is used occasionally may seem fairly stable, the odds of one or more of those ranges going sour is much higher than something a bit newer.  An HP 34401A just a decade or so newer (avoiding the early serial numbers) is a far better bet.  And if the DMM6500 is anywhere near as good as the now discontinued Fluke 8846A, it will be even better in actual performance terms, especially tempco where the 34401A is a bit weak.

The HP 3456A is 23ppm + 2 counts for 90 days.  There is no one-year spec, although they state "6ppm per month beyond 90 days".  The DMM6500 has a listed 2-year calibration cycle as well as 1 year.  2 year accuracy at 10V is 30ppm +5 counts.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 
The following users thanked this post: denimdragon

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7950
  • Country: us
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #19 on: February 25, 2024, 04:58:55 am »
I have a lot of NOS parts. I want to play with some of the old voltage references and see what I can do with them.

Another very important question. Once everything is up and running, how long would one monitor the reference before clearing it for release? 90 days? 1 Year? Most of my parts will be NOS, so no burn-in time prior to assembly. Knowing this would help be build out the "hibernator" room more effectively. I have a warehouse with a lab, a home lab where the room will be, and a lab at my business partners home.

Which ones do you have?  You should be the one determining the burn-in period by using your own data.  Maybe start with 1000 hours and see what you have.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 
The following users thanked this post: denimdragon

Offline alm

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: 00
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #20 on: February 25, 2024, 05:35:56 am »
The HP 3456A is 23ppm + 2 counts for 90 days.  There is no one-year spec, although they state "6ppm per month beyond 90 days".  The DMM6500 has a listed 2-year calibration cycle as well as 1 year.  2 year accuracy at 10V is 30ppm +5 counts.
On page 19 of the HP Journal issue introducing the 3457A they describe how they have come up with a more accurate drift figure for the LMx99 reference, and how the datasheet for the 3456A has been amended to reflect this. I have not found this datasheet, but on page 192 of the 1987 catalog (warning, large PDF) they give a 23ppm + 2 counts figure for 1y accuracy on 10 VDC and 15 ppm + 2 digits for 90 days for the 3456A. I attached the relevant table.

I believe this is better than the 3457A which did use close case calibration and laser-trimmed resistors, so I guess their engineers apparently disagreed that the old way of constructing was less stable. But obviously they weren't specifying decades-old instruments.
« Last Edit: February 25, 2024, 06:23:04 am by alm »
 

Offline srb1954

  • Super Contributor
  • ***
  • Posts: 1101
  • Country: nz
  • Retired Electronics Design Engineer
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #21 on: February 25, 2024, 07:12:33 am »
Ok, so that's good news about the null meter. I had a Fluke 845(A/AB?), but I went to my warehouse and it said "find me". I like the idea of using the 6.5s as it seems this will be the most efficient timewise and the cal quote I have for 34401As won't break the bank if I use those 2 and 3456A. I will still need to get another 10V standard to compare to the Guildline.

So far:
Guildline 4410 10V Standard
2nd "reasonable priced" 10V Standard
2x Cal'd 34401A
1x Cal'd 3456A

Does this look like a decent starting point? Guildline Pdf below.

Thanks again everyone.

Guildline specs look good and as it contains four independent (*) references you may not need another one per-se.

*To bad the linked document doesn't contain the schematics, these four references likely share the same power-supply and depending on the design (simple resistor from common psu-rail to set zener currrent?) all four outputs may be affected in the same way from a common event (power-supply voltage drifting, ground loops etc.), 734b handles that better imho (but not relevant to this discussion).
A "second-opinion" (reference) will shine (some) light onto that and pull you from darkness into a grey area.  :)
The attached document does contain schematics albeit as snippets of each section rather than one complete schematic.

The current into each Zener is set by a resistor from the 10V reference output, which is trimmed at manufacture to a nominal 10V level. Consequently, there should be minimal variation in the Zener current as temperature and PSU voltage changes.
 
The following users thanked this post: denimdragon

Offline srb1954

  • Super Contributor
  • ***
  • Posts: 1101
  • Country: nz
  • Retired Electronics Design Engineer
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #22 on: February 25, 2024, 07:32:20 am »
Ok, so that's good news about the null meter. I had a Fluke 845(A/AB?), but I went to my warehouse and it said "find me". I like the idea of using the 6.5s as it seems this will be the most efficient timewise and the cal quote I have for 34401As won't break the bank if I use those 2 and 3456A. I will still need to get another 10V standard to compare to the Guildline.

So far:
Guildline 4410 10V Standard
2nd "reasonable priced" 10V Standard
2x Cal'd 34401A
1x Cal'd 3456A

Does this look like a decent starting point? Guildline Pdf below.

Thanks again everyone.
I have a Startronics VS4 (aka Guildline 4410) and generally find it to be very good although I haven't fully characterised it yet due to problems with my 3458A.

A couple of points to note:
1) The rated ambient temperature range is 16℃ to 28℃. They really mean that upper limit as I find that the oven drops out of temperature regulation at 28.5℃. It still seems to regulate OK below the lower 16℃ limit.
2) The battery life doesn't seem to be quite as good stated in the specs. This will be a problem if you are transporting it to a cal lab unless you enclose it in a well insulated carry case to minimise the power that the oven heaters have to provide.
3) The SLA batteries originally installed are a slightly non-standard size being slightly lower than equivalent SLA batteries available today. The only replacement batteries I could find were 5mm higher that the originals and their terminals are extremely close to shorting to the case. I plan to remove the expanded polystyrene insulation from the base of the battery compartment and thin it down to provide a bit more clearance above the battery terminals.
 
The following users thanked this post: denimdragon

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1945
  • Country: us
    • The Messy Basement
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #23 on: February 25, 2024, 03:36:31 pm »
There's the published specs and there's what you can prove for an individual instrument. My antique 3455A will stay within a couple ppm for years at a time. My Analogic 6.5 digit reference uses an LM399 and is probably 10X worse. Still a very handy unit if you can keep the switches clean.
 
The following users thanked this post: denimdragon

Offline denimdragonTopic starter

  • Regular Contributor
  • *
  • Posts: 227
  • Country: us
  • "Hole charge chaser and wanna-be o-scope fixer"
Re: Guildline 4410 or Advantest R6581/T For Lab Cornerstone Standard?
« Reply #24 on: February 25, 2024, 06:44:10 pm »
I have a lot of NOS parts. I want to play with some of the old voltage references and see what I can do with them.

Another very important question. Once everything is up and running, how long would one monitor the reference before clearing it for release? 90 days? 1 Year? Most of my parts will be NOS, so no burn-in time prior to assembly. Knowing this would help be build out the "hibernator" room more effectively. I have a warehouse with a lab, a home lab where the room will be, and a lab at my business partners home.

Which ones do you have?  You should be the one determining the burn-in period by using your own data.  Maybe start with 1000 hours and see what you have.

I have LM129AH, LT1021BMH-5, LT1027BCN8-5, LT1021CIN8-10, AD588AD, and LTZ1000ACH (saving those). I'm sure I have tons more but I grabbed these for convenience as my warehouse is a mess and I've been trying to organize it for a few years now.

Also, the 1000Hours is a given, as that how datasheets state long term stabilty. I mean in addition to the 42ish days of monitoring. I want to feel confident passed the manufacturers' specs. AN extra 90 days would give the standards 4 months under my monitoring set-up and would give me personal confidence that my standards are solid. Once again, I'm not trying to put anyone out of business, I just want to make some very nice references that will make hobbyist happy.

Ok, so that's good news about the null meter. I had a Fluke 845(A/AB?), but I went to my warehouse and it said "find me". I like the idea of using the 6.5s as it seems this will be the most efficient timewise and the cal quote I have for 34401As won't break the bank if I use those 2 and 3456A. I will still need to get another 10V standard to compare to the Guildline.

So far:
Guildline 4410 10V Standard
2nd "reasonable priced" 10V Standard
2x Cal'd 34401A
1x Cal'd 3456A

Does this look like a decent starting point? Guildline Pdf below.

Thanks again everyone.
I have a Startronics VS4 (aka Guildline 4410) and generally find it to be very good although I haven't fully characterised it yet due to problems with my 3458A.

A couple of points to note:
1) The rated ambient temperature range is 16℃ to 28℃. They really mean that upper limit as I find that the oven drops out of temperature regulation at 28.5℃. It still seems to regulate OK below the lower 16℃ limit.
2) The battery life doesn't seem to be quite as good stated in the specs. This will be a problem if you are transporting it to a cal lab unless you enclose it in a well insulated carry case to minimise the power that the oven heaters have to provide.
3) The SLA batteries originally installed are a slightly non-standard size being slightly lower than equivalent SLA batteries available today. The only replacement batteries I could find were 5mm higher that the originals and their terminals are extremely close to shorting to the case. I plan to remove the expanded polystyrene insulation from the base of the battery compartment and thin it down to provide a bit more clearance above the battery terminals.

Thank you for this information. I know you said that your 3458A is not available for testing the standard, but do you have an idea how stable it is with another meter, say a 6.5?

The HP 3456A is 23ppm + 2 counts for 90 days.  There is no one-year spec, although they state "6ppm per month beyond 90 days".  The DMM6500 has a listed 2-year calibration cycle as well as 1 year.  2 year accuracy at 10V is 30ppm +5 counts.
On page 19 of the HP Journal issue introducing the 3457A they describe how they have come up with a more accurate drift figure for the LMx99 reference, and how the datasheet for the 3456A has been amended to reflect this. I have not found this datasheet, but on page 192 of the 1987 catalog (warning, large PDF) they give a 23ppm + 2 counts figure for 1y accuracy on 10 VDC and 15 ppm + 2 digits for 90 days for the 3456A. I attached the relevant table.

I believe this is better than the 3457A which did use close case calibration and laser-trimmed resistors, so I guess their engineers apparently disagreed that the old way of constructing was less stable. But obviously they weren't specifying decades-old instruments.

This was my first 6.5 meter and I have had it for 8 years now. I absolutely love it. I had no idea that they had revised the 1yr stability. That's pretty amazing for such an old meter.
"Hi there. I'm short in the hair department, do you mind if I pull a few strands from your head? Thanks!"
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf