What is "multimeter calibration"?

[giosif]

Hi,

I am trying to clarify in my head what "multimeter calibration" is and what it is not.
My current understanding is that a multimeter calibration includes checking that the meter readings are in spec and, if they are not, making the necessary adjustments to bring the meter (readings) within spec (we're only talking about adjustments, not repairs).
Is that your understanding as well?

Reason I'm asking is I sent out a multimeter to a company for "calibration" (which was quoted at a certain price), and they are now asking for additional money to perform the adjustments to bring the meter in spec.

Thanks for your responses and apologies if this has been discussed before (a quick search didn't reveal anything, but I probably suck at building the right search text)!

[Simon]

The term is often misused. My employer has all sorts of things "calibrated" including thread gauges, if they fail they fail, it's really a test or calibration check not a calibration check. You need to look at exactly what service you bought. Can you even calibrate a multimeter these days ?

[Wim_L]

Many calibration services usually perform verification first: checking if the instrument meets specifications, and perhaps providing a report of the measured values on the meter vs. the values reported by a far more accurate reference.

Adjustment is not necessarily part of such a basic verification. Always check carefully what they are providing.

And yes, modern multimeters can be calibrated. It's usually a closed-case calibration, where a sequence of specific voltages and currents are provided, the meter measures them all, and then internally adjusts itself based on that information. It can take a bit of time, and isn't always fully automated, especially not in models without bidirectional datalink, as is the case for many handheld meters. There might be 20 to 30 steps in the procedure, and the meter can beep to instruct the operator to go to the next step -- it's not all that difficult if you have the right equipment, but it gets expensive because someone has to sit there the whole time.

[retrolefty]

The term is often misused. My employer has all sorts of things "calibrated" including thread gauges, if they fail they fail, it's really a test or calibration check not a calibration check. You need to look at exactly what service you bought. Can you even calibrate a multimeter these days ?

 So that clears that up. 

 It would be more helpful if the 'calibration industry' defined such services in clearer description. Possibly:

1. Validation only........test against shop standards and publish results.
2. Validation, adjustment if need, revalidate, publish results.
3. Repair service if fault cannot be fixed with just adjustments.

The problem is that most shops could safely give estimates for this level 1 service. Level 2 estimate would be a little more difficult unless familiar with specific instrument. Level 3 estimate are nearly impossible without first physically looking over the instrument.
 

[giosif]

The term is often misused. My employer has all sorts of things "calibrated" including thread gauges, if they fail they fail, it's really a test or calibration check not a calibration check. You need to look at exactly what service you bought. Can you even calibrate a multimeter these days ?

 So that clears that up. 

I was actually wondering if I'm missing the point (non-native English speaker here)... 

[giosif]

Many calibration services usually perform verification first: checking if the instrument meets specifications, and perhaps providing a report of the measured values on the meter vs. the values reported by a far more accurate reference.

Adjustment is not necessarily part of such a basic verification. Always check carefully what they are providing.

And yes, modern multimeters can be calibrated. It's usually a closed-case calibration, where a sequence of specific voltages and currents are provided, the meter measures them all, and then internally adjusts itself based on that information. It can take a bit of time, and isn't always fully automated, especially not in models without bidirectional datalink, as is the case for many handheld meters. There might be 20 to 30 steps in the procedure, and the meter can beep to instruct the operator to go to the next step -- it's not all that difficult if you have the right equipment, but it gets expensive because someone has to sit there the whole time.

Thanks for that!

Didn't realise I need to be that careful for, essentially, something that simple (as a concept, not the process).
Strange that the industry hasn't standardised on the meaning of a standards-related concept such as "calibration"...

Checking on this particular company's web site, it is not clear what the "calibration" service consists of and I also don't see a separate "adjustment" service.
Let's see... 

[tszaboo]

Some time ago, we sent out one of the multimeter for calibration. It came back, the report said, it was outside of the specification (some 500%). So no adjustments were made, as if any adjustment was to be made, the DMM would be within it's acceptable range, right?

BTW I sent the report to Keysight, asking for a repair, as it was under warranty. They replaced it, no further question asked.

On the other hand, we do calibration of our own equipment. What we do, we set a variable, to have the measured quantity match the real world value.

On yet another occasion, there is a 3458A, it can do automatic calibration, where it sets the variables based on an internal reference, called "Auto cal"

So go figure, there is calibration, calibration and calibration, meaning completely different things.

[PA4TIM]

I do not (official) calibrate but I repair instruments for some cal-labs.

Calibration is nothing more then checking the performance of an instrument. If the tolerance specs are known they can include a conclusion.
Adjusting is not standard included. In case they do that they not only have to adjust things but calibrate it a second time as a control. You need to know the difference so you know the consequences regarding former done jobs. Besides that, if a serious quality test gear is out of specs it often is a sign it needs some repair. Adjusting that would be very bad

I would not be pleased if a cal lab adjusted my gear.

I repaired a calibrator that, according to the manufacturer's cal-lab , failed in some range and needed repair. The range that calibrated wrong turned out to be OK after I repaired something that made it look like an other range was OK. If they had adjusted the calibrator, the customer would have a serious problem.

Always ask a cal lab what they do.

Besides that, there are also levels of calibration. Calibration is documenting the history of an instrument.

[madires]

calibration != adjustment

Calibration is the process to compare the measurents with a proven reference and to note the results. The purpose is to verify if your DMM is within specs and to document any drift over time. Adjustment is the process to adjust your DMM to be within specs again. Actually you would need another calibration directly after an adjustment to have a proper track of calibration data.

[giosif]

It would be more helpful if the 'calibration industry' defined such services in clearer description. Possibly:

1. Validation only........test against shop standards and publish results.
2. Validation, adjustment if need, revalidate, publish results.
3. Repair service if fault cannot be fixed with just adjustments.

The problem is that most shops could safely give estimates for this level 1 service. Level 2 estimate would be a little more difficult unless familiar with specific instrument. Level 3 estimate are nearly impossible without first physically looking over the instrument.

That makes more sense to me.

[PA4TIM]

The calibration industry has defined those things, it is not their fault customers do not know or ask 

[giosif]

The calibration industry has defined those things, it is not their fault customers do not know or ask 

That is certainly part of the problem. :-)
However, that still doesn't justify the lack of clarification/explanation on what you get for what price: as I said, the web site of the company in question only lists "Calibration service" and nothing else; there is no "adjustment" service.
And I'm curious how many sites do list those separately...

[retrolefty]

The calibration industry has defined those things, it is not their fault customers do not know or ask 

 And are those definitions clearly shown or available to their customer's order process?

 We have posted here at least one person who sent in meter with much different expectations when what was given by the calibration service.

 A business contract, to 'calibrate' a customer's meter at a specific price has to have access to agreed definition of what that service includes and does not include. That's all that should be asked, are they to measure and fix measurement error(s) or just measure?

[deadlylover]

That is certainly part of the problem. :-)
However, that still doesn't justify the lack of clarification/explanation on what you get for what price: as I said, the web site of the company in question only lists "Calibration service" and nothing else; there is no "adjustment" service.
And I'm curious how many sites do list those separately...

Was there any extra info on the actual quote you received? There should be a definition of what a calibration service means for their quotes.

The quotes I get from my local cal lab has a wall of text which basically says "minor adjustments and re-verifications up to 10% of the cost of calibration are included in the quoted price".

[giosif]

That is certainly part of the problem. :-)
However, that still doesn't justify the lack of clarification/explanation on what you get for what price: as I said, the web site of the company in question only lists "Calibration service" and nothing else; there is no "adjustment" service.
And I'm curious how many sites do list those separately...

Was there any extra info on the actual quote you received? There should be a definition of what a calibration service means for their quotes.

The quotes I get from my local cal lab has a wall of text which basically says "minor adjustments and re-verifications up to 10% of the cost of calibration are included in the quoted price".

What I have in my quote is about pre and post results being included in the price and that repairs would be quoted separately.
The way I read it, this should include adjustments (otherwise, what are the pre result preceeding and the post results succeeding?).

[PA4TIM]

I do not understand the problem. Calibration is just what it is. It is not something special for electronic test-gear. If you know the purpose of calibration you know why adjustments are not part of that. A good call lab will never adjust an instrument without asking the owner first.  A serious company would be very upset if a cal-lab adjusted an instrument without telling it first 

Calibration is about the history of an instrument. Not a one time thing to see if your second hand multi-meter turns out to be OK .

The formal definition of calibration by the International Bureau of Weights and Measures is the following: "Operation that, under specified conditions, in a first step, establishes a relation between the quantity values with measurement uncertainties provided by measurement standards and corresponding indications with associated measurement uncertainties (of the calibrated instrument or secondary standard) and, in a second step, uses this information to establish a relation for obtaining a measurement result from an indication."[1]

And because some people thinks it is something else, you can not blame the industry for this. Calibration is an ongoing process, not a one time thing. 99% of the customers know that. It is a part of a bigger  process done by the instrument owner to guard his "quality". Some must do it because of regulations, some do it because they want it.

Do you want companies to state all the things they do not do ? Like user manuals for the USA market; doing stupid things like drying your cat in the magnetron and then blaming/sueing the manufacturer that he not told all the things you can not do with the device if the cat does not survive.

https://en.wikipedia.org/wiki/Calibration

[deadlylover]

What I have in my quote is about pre and post results being included in the price and that repairs would be quoted separately.
The way I read it, this should include adjustments (otherwise, what are the pre result preceeding and the post results succeeding?).

Ahh I see, it's a big trap for young players unfortunately, I can certainly understand how you were misled! It seems the calibration charge you were quoted includes the calibration (re-verification) after any adjustments, but it does not cover the act of adjustment itself. What a bummer. =/

It's very strange you didn't get a wall of text explaining all of the terminology used in your quote, seems like a dodgy place. You'd think they'd let you know *exactly* what you were getting for your money, like what my local cal lab does which should be standard?

Like the other folks in the thread said, adjustment is always separate and has to be explicitly detailed in the quote. (IIRC for example, Keysight says their calibration service charge includes all adjustments to bring the instrument back to spec)

[giosif]

What I have in my quote is about pre and post results being included in the price and that repairs would be quoted separately.
The way I read it, this should include adjustments (otherwise, what are the pre result preceeding and the post results succeeding?).

Ahh I see, it's a big trap for young players unfortunately, I can certainly understand how you were misled! It seems the calibration charge you were quoted includes the calibration (re-verification) after any adjustments, but it does not cover the act of adjustment itself. What a bummer. =/

It's very strange you didn't get a wall of text explaining all of the terminology used in your quote, seems like a dodgy place. You'd think they'd let you know *exactly* what you were getting for your money, like what my local cal lab does which should be standard?

Like the other folks in the thread said, adjustment is always separate and has to be explicitly detailed in the quote. (IIRC for example, Keysight says their calibration service charge includes all adjustments to bring the instrument back to spec)

I see.
So, it's a case of customer lack of knowledge combined with a misdocumented quotation (let's give them the benefit of the doubt and consider this was unintentional - this company was suggested by a well known multimeter manufacturer, so I can't think they'd be dodgy).
I can live with that.

[Simon]

The term is often misused. My employer has all sorts of things "calibrated" including thread gauges, if they fail they fail, it's really a test or calibration check not a calibration check. You need to look at exactly what service you bought. Can you even calibrate a multimeter these days ?

 So that clears that up. 

I was actually wondering if I'm missing the point (non-native English speaker here)... 

Sorry i should have said that the term seems to mean a check on the current callibration and not a callibration in itself.

[giosif]

I do not understand the problem. Calibration is just what it is. It is not something special for electronic test-gear. If you know the purpose of calibration you know why adjustments are not part of that. A good call lab will never adjust an instrument without asking the owner first.  A serious company would be very upset if a cal-lab adjusted an instrument without telling it first 

Calibration is about the history of an instrument. Not a one time thing to see if your second hand multi-meter turns out to be OK .

The formal definition of calibration by the International Bureau of Weights and Measures is the following: "Operation that, under specified conditions, in a first step, establishes a relation between the quantity values with measurement uncertainties provided by measurement standards and corresponding indications with associated measurement uncertainties (of the calibrated instrument or secondary standard) and, in a second step, uses this information to establish a relation for obtaining a measurement result from an indication."[1]

And because some people thinks it is something else, you can not blame the industry for this. Calibration is an ongoing process, not a one time thing. 99% of the customers know that. It is a part of a bigger  process done by the instrument owner to guard his "quality". Some must do it because of regulations, some do it because they want it.

Do you want companies to state all the things they do not do ? Like user manuals for the USA market; doing stupid things like drying your cat in the magnetron and then blaming/sueing the manufacturer that he not told all the things you can not do with the device if the cat does not survive.

https://en.wikipedia.org/wiki/Calibration

I'm not suggesting we should re-design the world just to fit my understanding (or lack of).
If the consensus in this forum is that calibration doesn't include adjustment (like it is the case), then that's what it is.

However, based on the fact that the quotation I got mentions pre and post results and on the fact that the web site does not specify adjustment services as separate (or not even as excluded from the quotation), I'd hope you see I'm not asking for the obvious here.

[Simon]

as far as I'm concerned the term is being misused. You cannot calibrate a thread guage, if it's "screwed" it's "screwed" you through it away and buy another one yet my employer still has thread guages "calibrated". Calibration infers and adjustment. I think the sales people got a hand in here. "calibration check" would be acceptable terminology.

[giosif]

as far as I'm concerned the term is being misused. You cannot calibrate a thread guage, if it's "screwed" it's "screwed" you through it away and buy another one yet my employer still has thread guages "calibrated". Calibration infers and adjustment. I think the sales people got a hand in here. "calibration check" would be acceptable terminology.

I'd agree with you, especially on the meaning of "calibration".
But, first, I'm not a native English speaker and, second, I'm not that versed in electronics to challenge what's right/accurate and what's wrong/inaccurate in this area.

Oh no, I introduced another debatable term now: accuracy...     

[bson]

Calibration of a DMM consists of going through a routine to measure it against standards and storing the results for the firmware to correct the displayed values.  The fact that the calibration procedure doesn't produce an error is indicative that the instrument is within spec and doesn't require service.  There is nothing adjustable in a modern DMM, so if it's out of spec it will need repair and will refuse to self-calibrate.

[timb]

Calibration of a DMM consists of going through a routine to measure it against standards and storing the results for the firmware to correct the displayed values.  The fact that the calibration procedure doesn't produce an error is indicative that the instrument is within spec and doesn't require service.  There is nothing adjustable in a modern DMM, so if it's out of spec it will need repair and will refuse to self-calibrate.

Define modern? Because most bench meters produced in the last 20-30 years feature "soft calibration" where you apply a known reference voltage/current/resistance to the meter and it stores those constants in some form of non-volatile memory. You might not be physically adjusting pots, but the cal data is being adjusted.

Also, the fact that you have to put these meters into "Calibration Mode" implies to me that the act of calibration isn't just verifying accuracy, but adjusting the instrument to meet that accuracy. Otherwise, why would equipment manufacturers call it that?

If I get something calibrated, I expect it to be adjusted if it's out of spec. Simply verifying if something is or isn't in spec is called Performance Verification, not Calibration. Anyway, that's always been my understanding.

[madires]

Define modern? Because most bench meters produced in the last 20-30 years feature "soft calibration" where you apply a known reference voltage/current/resistance to the meter and it stores those constants in some form of non-volatile memory. You might not be physically adjusting pots, but the cal data is being adjusted.

Also, the fact that you have to put these meters into "Calibration Mode" implies to me that the act of calibration isn't just verifying accuracy, but adjusting the instrument to meet that accuracy. Otherwise, why would equipment manufacturers call it that?

A lot of the T&M manufactures get that wrong too The correct term would be auto/self/soft adjustment. Either you calibrate the instrument after the soft adjustment or those offsets need to be disabled for the next calibration to get the unaltered readings from the instrument. Unless you need a proven record of calibration data of your instrument (used as a reference) you would only adjust it when you think it's out of specs.

calibration = compare measurement results with a proven reference
adjustment = make the instrument show proper results (again)

For example: In German shops / super markets scales need to be calibrated regularly. If they are within specific tolerances they'll get a nice sticker. If not, they need to be adjusted and re-calibrated, or may not be used to weight anything sales related any more.

[highlux]

Calibration should include adjustment.   Making it separate....it just a cheesy way to scam people out of more money.  Electronics shops sure do gouge people.
No wonder why people try  the DIY approach huh?     Local amp shop charges $100 to look at an amp.  If it is fixable or not.   I fix 99% of the amps I see for less than $20.
(Hobby)  But I fix 3-5 a week.

[madires]

Calibration should include adjustment.   Making it separate....it just a cheesy way to scam people out of more money.  Electronics shops sure do gouge people.

It's like in the old modem days when people messed up bps with baud. 

[timb]

Calibration should include adjustment.   Making it separate....it just a cheesy way to scam people out of more money.  Electronics shops sure do gouge people.

It's like in the old modem days when people messed up bps with baud. 

Language evolves. When we're at the point of big name manufactures using the word Calibration to mean "Adjustment+Verification" then that's what it means.


Sent from my Smartphone

[mzzj]

Calibration should include adjustment.   Making it separate....it just a cheesy way to scam people out of more money.  Electronics shops sure do gouge people.
No wonder why people try  the DIY approach huh?     Local amp shop charges $100 to look at an amp.  If it is fixable or not.   I fix 99% of the amps I see for less than $20.
(Hobby)  But I fix 3-5 a week.

Calibration vs. adjustment is really offen mixed. Simon goes for good example. 
But ask any reputable calibration lab and they will tell that calibration does not infer or include adjustment.
And how would you make adjustment to mercury thermometer, thermocouples, pt100 sensor, precision reference resistor or gauge block?

(And what electronic shops have to do with calibrations?)

[jimdeane]

It must differ greatly by industry and company-by-company.

With the data collection sensors we use, "calibration" means to measure against known values and set calibration point(s) to be used by the device when collecting/displaying data. That is how the term is used by the manufacturer/developer and by the users.

The calibration instructions I received from Extech for the EX330 includes sections on "Accuracy Test" and "Calibration Adjustments" along with a cover-off diagram of adjustment potentiometer locations. This is all under the number bulleted menu for "EX330 Calibration Procedure".

When I was setting up instrumentation in physics labs in college and grad school, being told to calibrate the instrumentation meant test-and-adjust. If I had just tested the equipment and labeled it with its performance vs. a standard I think I would have been replaced fairly quickly.

Certain instruments were not adjustable, so "calibrating" one of them meant checking to see that it is still within tolerances. So, a standard meter, a measuring block, etc. might be tested to see if they had been damaged or misused, but unless they were essentially destroyed they were considered permanently calibrated.

I notice that when the local calibration lab replied to my request for a quote on calibrating a multimeter, they returned a quote for "NIST Certification" (emphasis mine). Perhaps that is their way to show they do not mean to calibrate w/adjustment, but only check against a standard.

[deadlylover]

I notice that when the local calibration lab replied to my request for a quote on calibrating a multimeter, they returned a quote for "NIST Certification" (emphasis mine). Perhaps that is their way to show they do not mean to calibrate w/adjustment, but only check against a standard.

I think that just means that it's the lowest level of calibration available, it's kind of like just saying 'traceable calibration'.
The next step up is an Accredited Calibration (ISO 17025 or Z540 or other standards) where they usually state the uncertainties of the references used, in addition to the uncertainties of the instrument's measurements at time of calibration.

[mzzj]

I notice that when the local calibration lab replied to my request for a quote on calibrating a multimeter, they returned a quote for "NIST Certification" (emphasis mine). Perhaps that is their way to show they do not mean to calibrate w/adjustment, but only check against a standard.

Americanos have the most confusing ways of referring to calibrations 

"NIST Certification": probably/usually means that calibration lab references are calibrated at NIST
"NIST Traceable calibration" references are calibrated at NIST or the reference of reference is calibrated at NIST --> papertrail and traceability leads to NIST.

To make things more confusing there are calibrations and "calibrations"
1. "OKAY sticker" -calibration: someone has somehow determined that calibrated meter is probably within manufacturer specifications.
2. "calibration with  data"   as above but you get calibration certificate with  measurement results.
3.  Z540-1 Accredited calibration: Independent third-party(like http://www.a2la.org/ ) has verified the calibration laboratory to work according to (American) standards.
4. ISO 17025 Accredited calibration: Independent third-party(probably member of ILAC-MRA http://ilac.org/about-ilac/ ) has verified the calibration laboratory to work according to (international) standards.
5. NIM- National Institute of measurements calibration: even more complicated case.   

Option 4. is internationally the most widely accepted option.

[VK5RC]

I thought NIST (? NATA) certified calibration is quite a high level certification as I thought it includes the 95% confidence intervals of the errors of instrument under test AND the reference.
I think Keysights standard calibration includes adjustment but only to the original specifications.
Recent quote in Aus for a 3458a
Standard Calibration $ 1331 +gst
NATA calibration $1730 +gst
Fixed repair $3343 +gst

[mzzj]

I thought NIST (? NATA) certified calibration is quite a high level certification as I thought it includes the 95% confidence intervals of the errors of instrument under test AND the reference.
I think Keysights standard calibration includes adjustment but only to the original specifications.
Recent quote in Aus for a 3458a
Standard Calibration $ 1331 +gst
NATA calibration $1730 +gst
Fixed repair $3343 +gst

"NATA calibration" is  the the 4.th option on my list "NATA =National Association of Testing Authorities" == member of ILAC-MRA
And with 99% probability Agilent is going to ask your permission before any adjustments if you order "NATA calibration"

"NIST certified calibration" on the other hand can be "anything" and there is no guarantee of any sort over the results.
U.S equivalent of "NATA calibration" would be  "A2LA-(accredited) calibration"

 

[retrolefty]

  Are there any home business types or hobbyist here that pays to have some firm place a updated in-cal sticker on their equipment?

  I think the best aspect of the on-running volt-nut threads is that if it's your money, you will find a way to collect/build your own references to perform your own calibrations, be it check only or check and adjust.

 Paying someone with your own money to tell you your stuff is still in calibration seems silly. If required by law or company policy then fine, pay the tax and ship your stuff in and out But for personal labs I would think only repair activity on equipment to be useful.

 Lets ask Dave if he pays for any external cal services for his equipment?


 

[jitter]

Over here it exclusively means the check if an instrument is within spec, not the adjustment. But if I google for the meaning of the word (in English, not translated to my language), some definitions include adjustment, others do not.

Would the origin of "to calibrate" have something to do with weaponry, i.e. measuring the calibre of a gun. If so, that would imply just the measurement, not the adjustment.
Any linguists here with some insight on the semantics of this word?

[madires]

  Are there any home business types or hobbyist here that pays to have some firm place a updated in-cal sticker on their equipment?

I doubt it

  I think the best aspect of the on-running volt-nut threads is that if it's your money, you will find a way to collect/build your own references to perform your own calibrations, be it check only or check and adjust.

There are some inexpensive references from voltagestandard.com and ebay. I've build my own references and when I find someone with a nice bench meter (with a cal sticker ) in my near I'll adjust my references.

[giosif]

Ok.
Thanks to everyone for your contribution to this topic!
It is appreciated.