Author Topic: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?  (Read 1541 times)

0 Members and 1 Guest are viewing this topic.

Offline rodppTopic starter

  • Frequent Contributor
  • **
  • Posts: 325
Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« on: January 19, 2025, 12:12:42 pm »
Hi,

When sending a multimeter to calibration we have the option to only check if its measurements are according with its specs, OR we can ask the Cal Lab to make adjustments too, reporting the measurements before and after adjustments.

I read some posts here saying that when the equipment to be calibrated has a history of various calibrations, adjusting it will ruin that history.

For equipments that have to adjust physical trimpots or similar, I understand it. 

But for multimeters calibrated without opening the case, only adjusting software parameters, this statment against adjustments to 24h specs remains valid? If yes, could you explain the reasons considering that every calibration + adjustments will provide a report with the measurements before and after ajustments?

Regards,
Rodrigo. 
 

Offline aronake

  • Frequent Contributor
  • **
  • Posts: 287
  • Country: hk
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #1 on: January 19, 2025, 01:10:15 pm »
If you have old calibration history and want to be able to track drift, the possibility to some extent predict drift and adjust for drift out of later measurements you do not want to have it adjusted.

If you have no calibration history, or do not care about being able to track drift, you want also to have it adjusted.

 

Online Overspeed

  • Frequent Contributor
  • **
  • Posts: 545
  • Country: fr
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #2 on: January 19, 2025, 03:02:25 pm »
Hi,

When sending a multimeter to calibration we have the option to only check if its measurements are according with its specs, OR we can ask the Cal Lab to make adjustments too, reporting the measurements before and after adjustments.

I read some posts here saying that when the equipment to be calibrated has a history of various calibrations, adjusting it will ruin that history.

For equipments that have to adjust physical trimpots or similar, I understand it. 

But for multimeters calibrated without opening the case, only adjusting software parameters, this statment against adjustments to 24h specs remains valid? If yes, could you explain the reasons considering that every calibration + adjustments will provide a report with the measurements before and after ajustments?

Regards,
Rodrigo.

Hello

Not all calibration lab run adjustments , depend of their calibration procedures / equipment.
In some cases ( most) that request hardware tests / checks even visual checks as torque / corrosion / ... if there is drift and why .
Making adjustment on a single value is gambling , as a 10 V range tested at 10V don t make any proof  for linearity .

Regards
OS

 
The following users thanked this post: rodpp

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8294
  • Country: us
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #3 on: January 19, 2025, 03:41:29 pm »
But for multimeters calibrated without opening the case, only adjusting software parameters, this statment against adjustments to 24h specs remains valid? If yes, could you explain the reasons considering that every calibration + adjustments will provide a report with the measurements before and after ajustments?

It depends on which calibration service you purchase and the policies of the calibration lab.  You have to understand that those policies are generallly based on the needs of their customers, not a few random voltnuts.  Many basic services don't provide data unless there is an OOT (Out Of Tolerance) measurement.  When you get into services that do offer data, then adjustment policies are often based at least in part on the capabilities of the lab.  A lab might have the abilility (accuracy) to determine that your meter is within its one-year spec but not to adjust it down to the last digit.  In this case, they'll only perform an adjustment when the meter is off from nominal by more than their uncertainty, IOW they'll only make adjustments where they are sure they are improving things.  It is not uncommon for calibration labs to have uncertainties of 5-8ppm @ 10V and still take in 6.5-digit meters and provide calibration certificates for them.

As far as 24h specs, if you look at the entire set of ranges on many meters some will have 24h specs that aren't much better than the 1-year.  A better way of thinking about what the cal lab will do is that they will adjust all the ranges to nominal values and then use the 24h specs as their go/no-go during the performance check after the calibration adjustment.  Adjustment this way, along with as low as possible uncertainties, is what most people would hope for as a best-case when they send their DMM in.  Just getting data and avoiding adjustment is a thing for references and single-range instruments--DMMs have too many ranges to keep track of.  As you said, if you did want to track its drift or other long-term performance characteristic, you have the data and closed-case calibration doesn't affect drift like a potentiometer tweak might.

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.
 

Offline rodppTopic starter

  • Frequent Contributor
  • **
  • Posts: 325
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #4 on: January 19, 2025, 03:42:02 pm »
If you have old calibration history and want to be able to track drift, the possibility to some extent predict drift and adjust for drift out of later measurements you do not want to have it adjusted.

If you have no calibration history, or do not care about being able to track drift, you want also to have it adjusted.

But if no physical adjustments will be done, one can track drift with or without adjustments because only the equipment software will be adjusted.

For example, we are tracking the 10V readings considering no adjustments:

1st cal: reference
2nd cal: reference + 5ppm
3th cal: reference + 7ppm
4th cal: reference + 8ppm
5th cal: reference + 8ppm

Now if that same multimeter had got adjustments:

1st cal: reference
2nd cal: ajusted -5ppm => now reading same as reference again
3th cal: adjusted -2ppm => now reading same as reference again
4th cal: adjusted -1ppm => now reading same as reference again
5th cal: no adjustments needed - already reading same as reference

In both cases it is trivial to track the drift.

What am I not getting?
 

Offline rodppTopic starter

  • Frequent Contributor
  • **
  • Posts: 325
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #5 on: January 19, 2025, 03:55:56 pm »
But for multimeters calibrated without opening the case, only adjusting software parameters, this statment against adjustments to 24h specs remains valid? If yes, could you explain the reasons considering that every calibration + adjustments will provide a report with the measurements before and after ajustments?

It depends on which calibration service you purchase and the policies of the calibration lab.  You have to understand that those policies are generallly based on the needs of their customers, not a few random voltnuts.  Many basic services don't provide data unless there is an OOT (Out Of Tolerance) measurement.  When you get into services that do offer data, then adjustment policies are often based at least in part on the capabilities of the lab.  A lab might have the abilility (accuracy) to determine that your meter is within its one-year spec but not to adjust it down to the last digit.  In this case, they'll only perform an adjustment when the meter is off from nominal by more than their uncertainty, IOW they'll only make adjustments where they are sure they are improving things.  It is not uncommon for calibration labs to have uncertainties of 5-8ppm @ 10V and still take in 6.5-digit meters and provide calibration certificates for them.

As far as 24h specs, if you look at the entire set of ranges on many meters some will have 24h specs that aren't much better than the 1-year.  A better way of thinking about what the cal lab will do is that they will adjust all the ranges to nominal values and then use the 24h specs as their go/no-go during the performance check after the calibration adjustment.  Adjustment this way, along with as low as possible uncertainties, is what most people would hope for as a best-case when they send their DMM in.  Just getting data and avoiding adjustment is a thing for references and single-range instruments--DMMs have too many ranges to keep track of.  As you said, if you did want to track its drift or other long-term performance characteristic, you have the data and closed-case calibration doesn't affect drift like a potentiometer tweak might.

Yes, your point about the Lab capability to adjust to 24h specs is very relevant.

And your last statment confirms what I am thinking about closed-case adjustments.

My understanding is that the best option is calibration -> adjustments to 24h specs -> calibration, providing a cal report with the measurements before and after adjustments so that I can track the drift.

If the Lab provides that option, it must be the most expensive.

The multimeter is a Keithley 2002, so it will be above USD1K probably.
 

Offline mendip_discovery

  • Super Contributor
  • ***
  • Posts: 1026
  • Country: gb
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #6 on: January 19, 2025, 04:33:47 pm »
Depending on the device and the use of it.

I just go for adjust it if it's more than 75% of the specification. I do checks monthly to ensure that if it drifts outside of the spec over the 12 months I set for the calibration.

Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 
The following users thanked this post: Overspeed

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8294
  • Country: us
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #7 on: January 19, 2025, 05:12:13 pm »

My understanding is that the best option is calibration -> adjustments to 24h specs -> calibration, providing a cal report with the measurements before and after adjustments so that I can track the drift.

If the Lab provides that option, it must be the most expensive.

The multimeter is a Keithley 2002, so it will be above USD1K probably.

the best option is ISO 17025 calibration service -> measurement and recording of "as-found" data" -> adjustments to nominal value -> measurement and recording of "as left" data -> comparison of "as-left" data with 24h specs and expanded uncertainty of the calibration equipment.  Adjustments should be done as exactly as possible, not to within a tolerance.

If you are in the US and want good results, you probably should select a primary-standards affiliated lab. 
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: rodpp

Online IanJ

  • Supporter
  • ****
  • Posts: 1846
  • Country: scotland
  • Full time EE & Youtuber/Creator
    • IanJohnston.com
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #8 on: January 19, 2025, 05:59:33 pm »
Last time I asked Keysight in UK to calibrate my 3458A I asked them if they could make adjustments to zero it in as close as possible in VDC mode.
They told me they could try......but they could end up moving it further away from being spot on as it was now and so best to leave it if it's in spec.

Ian.
Ian Johnston - Original designer of the PDVS2mini || Author of WinGPIB
Website: www.ianjohnston.com
YouTube: www.youtube.com/user/IanScottJohnston, Odysee: https://odysee.com/@IanScottJohnston, Twitter(X): https://twitter.com/IanSJohnston, Github: https://github.com/Ian-Johnston?tab=repositories
 
The following users thanked this post: rodpp

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8294
  • Country: us
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #9 on: January 19, 2025, 06:14:33 pm »
Last time I asked Keysight in UK to calibrate my 3458A I asked them if they could make adjustments to zero it in as close as possible in VDC mode.
They told me they could try......but they could end up moving it further away from being spot on as it was now and so best to leave it if it's in spec.

Wokingham or Telford?
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.
 

Online IanJ

  • Supporter
  • ****
  • Posts: 1846
  • Country: scotland
  • Full time EE & Youtuber/Creator
    • IanJohnston.com
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #10 on: January 19, 2025, 06:22:49 pm »
Wokingham.

Ian.
Ian Johnston - Original designer of the PDVS2mini || Author of WinGPIB
Website: www.ianjohnston.com
YouTube: www.youtube.com/user/IanScottJohnston, Odysee: https://odysee.com/@IanScottJohnston, Twitter(X): https://twitter.com/IanSJohnston, Github: https://github.com/Ian-Johnston?tab=repositories
 

Offline rodppTopic starter

  • Frequent Contributor
  • **
  • Posts: 325
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #11 on: January 19, 2025, 06:49:49 pm »
I´m in Brazil.

Some time ago I asked Tektronix about Cal Labs in Brazil and they indicated Metracal.

Since 2022 Metracal is part of Trescal Group:

http://www.inmetro.gov.br/laboratorios/rbc/detalhe_laboratorio.asp?num_certificado=299&situacao=AT&area=ELETRICIDADE%20e%20magnetismo
 
The following users thanked this post: Alex Nikitin

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8294
  • Country: us
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #12 on: January 19, 2025, 07:04:57 pm »
Wokingham.

That appears to be the better of the two by far.

https://www.ukas.com/wp-content/uploads/schedule_uploads/969151/0147Calibration-Multiple.pdf

Since they show sub-ppm uncertainties for DC voltages then your 3458A must be pretty much spot-on to begin with, much better than just "in spec".  Or perhaps results of the artifact calibration procedure are a little uncertain.  What were the source uncertainties listed on your certificate?
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.
 

Offline CalibrationGuy

  • Supporter
  • ****
  • Posts: 105
  • Country: us
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #13 on: January 25, 2025, 02:11:22 am »
In March, I will be sending my reference 3458a meter to the Keysight standards lab in California. Since the meter has been extremely stable, I will tell them to now calibrate it to 24 hour specs because I'm tired of using offsets for critical precision measurements. But that's just me...

TomG.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15203
  • Country: de
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #14 on: January 25, 2025, 10:07:53 am »
Even if the meter is adjusted digitally, there may be no easy digital read back. Ideally one would get that, but AFAIK not many meters offer this in a well documented form.
Using only the measured value before and after adjustment adds some error to this. So one may not know by how much the scale facters were changed exactly. The drift history would thus be more accurate without the adjustment step.

It also depends on the use case, if it is important that the meter directly displays the correct values, or if it sufficient to know the correction factors to apply afterwards.

 
The following users thanked this post: rodpp

Offline mendip_discovery

  • Super Contributor
  • ***
  • Posts: 1026
  • Country: gb
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #15 on: January 26, 2025, 10:41:31 am »
Wokingham.

That appears to be the better of the two by far.

https://www.ukas.com/wp-content/uploads/schedule_uploads/969151/0147Calibration-Multiple.pdf

Since they show sub-ppm uncertainties for DC voltages then your 3458A must be pretty much spot-on to begin with, much better than just "in spec".  Or perhaps results of the artifact calibration procedure are a little uncertain.  What were the source uncertainties listed on your certificate?

Remember this is the lab's best capability, this might just be for comparison to other 10 V refs and not for the calibration of 8.5digit multimeters. Some labs will have a Golden calibrator and the spare, if they use the spare then your uncertainties might not be as good as you expected, I got caught out by this myself once. So now all our equipment goes out with a comment for them to use their best capabilities.

Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 

Offline rodppTopic starter

  • Frequent Contributor
  • **
  • Posts: 325
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #16 on: January 26, 2025, 11:09:18 pm »
Even if the meter is adjusted digitally, there may be no easy digital read back. Ideally one would get that, but AFAIK not many meters offer this in a well documented form.
Using only the measured value before and after adjustment adds some error to this. So one may not know by how much the scale facters were changed exactly. The drift history would thus be more accurate without the adjustment step.

It also depends on the use case, if it is important that the meter directly displays the correct values, or if it sufficient to know the correction factors to apply afterwards.

I haven't had considered the calibrator errors in the before/after report measurements. And at least in my factory cal report the TUR is very low accros the board, ranging from 0.8 for 1.9KOhm up to only 3.9 for 19mV. So the calibrator error were on the same ballpark of the meter itself.

Not good to track the drift...
 

Offline rodppTopic starter

  • Frequent Contributor
  • **
  • Posts: 325
Re: Shoud we ask the Cal Lab to adjust the multimeter to 24h specs?
« Reply #17 on: January 26, 2025, 11:13:49 pm »
In March, I will be sending my reference 3458a meter to the Keysight standards lab in California. Since the meter has been extremely stable, I will tell them to now calibrate it to 24 hour specs because I'm tired of using offsets for critical precision measurements. But that's just me...

TomG.

Nice, in your case you already know your meter and the adjustment will bring the convenience of not having to use offsets.

My meter only had the factory cal report, I'll think more about adjusting it or not. The point Kleinstein bring is relevant.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf