How to calibration if i do not have volt generator

[hongthiet1994]

I buy 121gw EEVBlog. I want to calibrate my DMM but i do not have voltage generator.
I download some file bin have some calibration information

i don't know      can upload file bin to my DMM

thank you so much

[beanflying]

To Calibrate the 121GW given the functionality requires far more than a simple voltage source. The best you can hope for without a full calibration setup is what I would call a sanity check.

An AD584-M (in the Acrylic Case) shipped with voltage measurements. Not all of them now do but they are cheap and reasonably accurate.

The Best Resistors you can afford or better yet a decade box with better accuracy than the meter. The same goes for Capacitors.

You also need an AC source of some sort. These outside of a quality function generator are out of the range of most.

There is all in one options like the DMM Checkplus http://www.voltagestandard.com/DMMCheck_Plus.html this is still only a sanity check but a good one.

This is a list below of the Calibration points needed for the 121GW. Clearly a proper Calibrator or several Calibrated sources are needed.

[hongthiet1994]

i search in this forum and i have some file which have information calibration

How to use it?   

[TERRA Operative]

You can't use a calibration file from another meter for your meter as the values will be incorrect for your meter.
You need to run the tests on your meter to correct for the variations in your particular meter alone.
If you use someone else's calibration file, you will make your meter inaccurate in unpredictable ways.

Why do you think you need to calibrate your meter, is there something wrong with the measurements it is displaying?

[beanflying]

The short answer is DO NOT use the information and DO NOT play with the file. If you don't have the sources you can't calibrate this or any meter. As Terra Operative stated you can't use this file from other meters or manipulate the file in a meaningful way.

If you have broken the file or Calibration in some way then you need to get it to someone with the skills and the equipment to recalibrate the meter.

If you just want to play or try and improve it then some of the gear below is what I have available and apart from the Temperature on mine which is borderline to being out of specification the rest are much better than specification and I wouldn't be trying to improve them or see the point to trying.

Fluke 335D (DC Volts)
Valhalla 2703 (AC Volts)
GPSDO fed Frequency Generator (Frequency)
Sika 18600E (Temperature)
0.02% Resistance Decade Box
0.25% Capacitance Box
A couple of Calibrated Shunts (Current)

If I were to try, taking that collection above and then cross reference the Calibration points for the 121GW to my pair of Calibrated (due back next week from Keysight) 6 1/2 digit Agilents (34401A & 34461A) and improvement would be minor if any at all. I will however be 'testing' the 121GW when my Agilents come back to make sure it still meets Spec.

[NoisyBoy]

That’s always the challenge with calibration, you have to acquire all the sources first, which are not cheap unless you need to do cal regularly.  Even if you have these sources, you still need to verify that they are in spec, and that the value are really what they claim to be, or what the true value are (you will need a calibrated DMM) for that.  Otherwise you are calibrating to a wrong value, which is a waste of time, a half-effort can be worse off than no cal.  It is far cheaper to pay a cal lab to do the work rather than doing it yourself if you want a true cal unless you have access to all these sources that are calibrated.  For the 6.5 digit Keysight meters I have, you will see fluctuation in the last 1-2 digits for the cheaper cal sources like those Chinese voltage calibrator, they are simply not stable enough for high precision cal.  But they are good enough for handheld DMM.

For the accuracy of a handheld meter, I would only calibrate if they are noticeably off.  As others have stated, cal file is hardware component specific, you can't copy them across hardware.  If it is common, they would have just used that one file and be done with it.  It would definitely take you off spec if you copy cal file.

If I want a well calibrated handheld DMM, I would buy the professional Fluke (not the low end Chinese made models).  I don't pay for the traceable calibrated models from Fluke, but even for just the regular models, the reading are consistent with my NIST cal Keysight meters out of the box.  If I buy a cheaper DMM, by the time I add the cost of a cal to it, I might as well buy the Fluke.  Besides, the Fluke has lifetime warranty. 

I don’t have experience with the quality of your meter so I cannot comment on it, my suggestion is to see if you can verify the reading against a calibrated meter, and only touch that if you see a big difference.

[hongthiet1994]

I do not need calibrate correction. DMM's eeprom calib has been earse. The device cannot measure anything

 I just need to calibrate relatively

[forrestc]

I do not need calibrate correction. DMM's eeprom calib has been earse. The device cannot measure anything

 I just need to calibrate relatively

Did it come this way?  If so, it's defective and should be returned.   

Otherwise, it will likely be cheaper just to buy another one than to acquire all of the test equipment to calibrate it.  Or as others have said, find someone with the right gear to calibrate it. 

I understand that you're probably just asking for instructions on how to make it work, even if it's out of calibration.  I'd submit that without proper calibration it's not worth much and one could buy a very cheap meter and get a much more useful (accuracy-wise) meter than a poorly-calibrated 121GW.

[NoisyBoy]

Sorry to hear that, given the cost of the DMM, if I were you, I would have someone calibrate it.  Handheld DMM can be calibrated for less than $100 in the U.S., which is far less than a new meter.  I definitely will not buy any equipment for a one time calc.  Or you can sell it to someone who has the equipment to make it work again to recoup some of your lost. 

I agree with forrestc that if you cannot do a proper calibration because of a lack of equipment, the reading could be a few percents off.  I would rather buy a cheap meter than using a nice meter with suspect calibration. 

In the future, when you try to do any calibration, always back up the existing calc data if there's a way to do it with your test gear.  It is like an operating system for a computer, without it, the hardware is just a dead corpse.

[joeqsmith]

Sorry to hear that, given the cost of the DMM, if I were you, I would have someone calibrate it.  Handheld DMM can be calibrated for less than $100 in the U.S., which is far less than a new meter.  I definitely will not buy any equipment for a one time calc.  Or you can sell it to someone who has the equipment to make it work again to recoup some of your lost. 

I agree with forrestc that if you cannot do a proper calibration because of a lack of equipment, the reading could be a few percents off.  I would rather buy a cheap meter than using a nice meter with suspect calibration. 

In the future, when you try to do any calibration, always back up the existing calc data if there's a way to do it with your test gear.  It is like an operating system for a computer, without it, the hardware is just a dead corpse.

My guess is OP is a victim of not reading and understanding the manual.  A high price to pay but a good  lesson to learn.     

The last time I had my bench meters calibrated, it cost about $100 with a report (not including shipping).   OP's meter will need to be realigned, then calibrated.   I would expect the cost to be much higher.   It would be interesting to come up with a list of cal houses throughout the world that will realign one, along with the costs and turn around times.   

[NoisyBoy]

I am familiar with calibration, but not realignment, can you share what it is about on a DMM? 

For cal labs, I have been a big fan of Keysight for everything (product and services) since my college days, it is my one stop shop for all my Keysight/Agilent gears.  For bench meters with 0.0015% accuracy, that is my preferred cal service.  I have also used Tektronix cal services for the large Agilent system power supplies which I am also happy of.  They both service other brands, it is more expensive but trust is far more important than cost when it comes to calibration for high precision equipment.  I have also heard good things about Tru Cal but have not found a need to use them.

But these are not cost effective for a handheld DMM, a local cal lab may be a better solution.  It will still be less than a new meter with unknown cal accuracy, assuming that the hardware is otherwise in perfect condition.

Edit: Just noticed the OP is in Vietnam, I have no idea of what local cal lab, or their capabilities over there.

[joeqsmith]

I am familiar with calibration, but not realignment, can you share what it is about on a DMM? 

For cal labs, I have been a big fan of Keysight for everything (product and services) since my college days, it is my one stop shop for all my Keysight/Agilent gears.  For bench meters with 0.0015% accuracy, that is my preferred cal service.  I have also used Tektronix cal services for the large Agilent system power supplies which I am also happy of.  They both service other brands, it is more expensive but trust is far more important than cost when it comes to calibration for high precision equipment.  I have also heard good things about Tru Cal but have not found a need to use them.

But these are not cost effective for a handheld DMM, a local cal lab may be a better solution.  It will still be less than a new meter with unknown cal accuracy, assuming that the hardware is otherwise in perfect condition.

Edit: Just noticed the OP is in Vietnam, I have no idea of what local cal lab, or their capabilities over there.

Calibration is checking the meter against a standard to prove that it's still within the manufacture's specs.  No adjustments are made.  If it fails calibration, the device will need to be repaired or adjusted (aligned) to bring it back into calibration.  After alignment, it would be calibrated. 

You can imagine for some equipment, this could be very time consuming.  Some times, the they will automate parts of the alignment.  For the 121GW, Dave would need to look into if there is a way to do it.   Guessing it would be all manual and could very well cost more than the meter is worth when it's all said and done.  The reports are nice if you plan to track it or are trying to shave that last bit out of your measurement.  It adds some cost.   

My advice is ALWAYS make sure you understand what services the calibration house is providing.  In the case of the OP, they should make it VERY clear to the cal house that the meter will need to be realigned.  They could provide the cal house with the manual.  At least see if they can get a quote.   

I posted about updating the firmware on prototype 121GW that Dave provided, which seemed to require that meter be realigned.  I only use the meter for experimental use and didn't require it to be in calibration.  I ended up using my other equipment as a cross check to bring it back in, somewhat...    I wouldn't be surprised at all if some members here couldn't offer this as a service if the OP didn't need the meter to be NIST traceable. 

https://www.duncanaviation.aero/intelligence/2015/May/calibration-v-alignment

[beanflying]

Any 'reputable' Calibration House is going to write a procedure for any new piece of equipment they haven't done before so the costs will be more than the 121GW is worth. Keysight even more so and Dave's interview with them spells that out that they create a worldwide searchable one for their internal use and review which will be part of their QA/Cal system.

Time for a smaller company to look at and even do a manual Calibration on an unknown meter would be high as they will most likely factor that 'time' into their costs of sorting out how it needs to be done. $100 for a known simple DMM in the USA seems reasonable but double for an unknown one is also likely if not more as the 121GW is not that simple and a quote should be obtained.

Depending on UEI (the manufacturer of the 121GW) or their parent company their costs may well be lower as they should have a written test procedure if not an automated one https://www.ueitest.com/service but freight costs each way are going to be fairly high.

[NoisyBoy]

Thank you so much for explaining this to me, I definitely mix these two terms together and call it calibration.

That's a good point that with well established equipment with a remote control interface, both process can be automated with the calibrator using a pre-built script.  But this could be significantly longer for the handheld DMM if it is not a common model or if it cannot be automated.

I am familiar with calibration, but not realignment, can you share what it is about on a DMM? 

For cal labs, I have been a big fan of Keysight for everything (product and services) since my college days, it is my one stop shop for all my Keysight/Agilent gears.  For bench meters with 0.0015% accuracy, that is my preferred cal service.  I have also used Tektronix cal services for the large Agilent system power supplies which I am also happy of.  They both service other brands, it is more expensive but trust is far more important than cost when it comes to calibration for high precision equipment.  I have also heard good things about Tru Cal but have not found a need to use them.

But these are not cost effective for a handheld DMM, a local cal lab may be a better solution.  It will still be less than a new meter with unknown cal accuracy, assuming that the hardware is otherwise in perfect condition.

Edit: Just noticed the OP is in Vietnam, I have no idea of what local cal lab, or their capabilities over there.

Calibration is checking the meter against a standard to prove that it's still within the manufacture's specs.  No adjustments are made.  If it fails calibration, the device will need to be repaired or adjusted (aligned) to bring it back into calibration.  After alignment, it would be calibrated. 

You can imagine for some equipment, this could be very time consuming.  Some times, the they will automate parts of the alignment.  For the 121GW, Dave would need to look into if there is a way to do it.   Guessing it would be all manual and could very well cost more than the meter is worth when it's all said and done.  The reports are nice if you plan to track it or are trying to shave that last bit out of your measurement.  It adds some cost.   

My advice is ALWAYS make sure you understand what services the calibration house is providing.  In the case of the OP, they should make it VERY clear to the cal house that the meter will need to be realigned.  They could provide the cal house with the manual.  At least see if they can get a quote.   

I posted about updating the firmware on prototype 121GW that Dave provided, which seemed to require that meter be realigned.  I only use the meter for experimental use and didn't require it to be in calibration.  I ended up using my other equipment as a cross check to bring it back in, somewhat...    I wouldn't be surprised at all if some members here couldn't offer this as a service if the OP didn't need the meter to be NIST traceable. 

https://www.duncanaviation.aero/intelligence/2015/May/calibration-v-alignment

[joeqsmith]

Thank you so much for explaining this to me, I definitely mix these two terms together and call it calibration.

That's a good point that with well established equipment with a remote control interface, both process can be automated with the calibrator using a pre-built script.  But this could be significantly longer for the handheld DMM if it is not a common model or if it cannot be automated.

Glad to help.   

I would not be surprised if they load special firmware into the meter as part of the functional test and alignment.  Then load the application firmware assuming it checks out.  It would be interesting to see how it's done as I can't see people doing this manually in production.  Then again, cheap enough labor, so maybe.   

[blacksheeplogic]

I understand that you're probably just asking for instructions on how to make it work, even if it's out of calibration.  I'd submit that without proper calibration it's not worth much and one could buy a very cheap meter and get a much more useful (accuracy-wise) meter than a poorly-calibrated 121GW.

Pretensions, a good meter out of calibration is fine for most purposes if the readings are repeatable. It's also not hard or expensive to estimate the error and compensate for it.

What the OP asking for is to restore his boutique meter to a functional state. At that point they can decide if it's usable for the intended purpose. It's quite likely that the cal constants from another meter will get close but out of spec. That might be good enough.

[Fungus]

What the OP asking for is to restore his boutique meter to a functional state. At that point they can decide if it's usable for the intended purpose. It's quite likely that the cal constants from another meter will get close but out of spec. That might be good enough.

If he's managed to completely trash his setup then it's a starting point.

It should be quite easy to fix voltage and resistance after loading somebody else's calibration. All you need for that is clean power, some trim pots and another meter that's in spec.

(I think... I haven't actually calibrated^Waligned one of those).

[beanflying]

What the OP asking for is to restore his boutique meter to a functional state. At that point they can decide if it's usable for the intended purpose. It's quite likely that the cal constants from another meter will get close but out of spec. That might be good enough.

If he's managed to completely trash his setup then it's a starting point.

It should be quite easy to fix voltage and resistance after loading somebody else's calibration. All you need for that is clean power, some trim pots and another meter that's in spec.

(I think... I haven't actually calibrated^Waligned one of those).

Check the Cal points even just for DC. 5, 50, 500 and 600V. Resistance Cal needs a large spread too. If you upload someone elses Cal data and change 1 or 2 of the lower points you can reach you will screw with the Linearity of all future measurements.

The only real solution is a full and proper Calibration cycle or maybe sell it really cheap to someone with the gear to do it and buy another. My guesstimate on time for a full manual Cal procedure on this meter given what gear I have will be in the order of 3-4 hours at least.

From Keysight where my Agilents currently are. Even they mix their words a little  Calibration tends to be used as the description for the overall procedure including Adjustment.

We include adjustment in the cost of the calibration. The lab manager advised me you will receive pre and post cal adjustment results.

[tautech]

From Keysight where my Agilents currently are. Even they mix their words a little  Calibration tends to be used as the description for the overall procedure including Adjustment.

We include adjustment in the cost of the calibration. The lab manager advised me you will receive pre and post cal adjustment results.

Totally agree.

We use the term Calibration much too freely as when we send equipment to be 'Calibrated' most often only verification that it meets spec is performed prior to issue of a calibration certificate.

Terms like alignment are the actual adjustments to align an instrument to manufacturer spec.

[Fungus]

From Keysight where my Agilents currently are. Even they mix their words a little  Calibration tends to be used as the description for the overall procedure including Adjustment.

We include adjustment in the cost of the calibration. The lab manager advised me you will receive pre and post cal adjustment results.

Totally agree.

We use the term Calibration much too freely as when we send equipment to be 'Calibrated' most often only verification that it meets spec is performed prior to issue of a calibration certificate.

Terms like alignment are the actual adjustments to align an instrument to manufacturer spec.

Language is defined by whatever the common usage is. No way to fight that.

Would a cal lab send a meter back without touching it if it was within manufacturer's spec, even if it could be improved?

(eg. Brymen makes 500000 count meters that are only 0.03%, clearly a meter that's 0.028% off can be improved)

[Brumby]

Would a cal lab send a meter back without touching it if it was within manufacturer's spec, even if it could be improved?

(eg. Brymen makes 500000 count meters that are only 0.03%, clearly a meter that's 0.028% off can be improved)

This is the point about "calibration".

A cal lab will do the calibration and tell you the meter is 0.028% off on a particular range.  If it's 0.3% off, they will tell you that.  They will not adjust anything - and anyone serious that sends their gear for calibration would not want them to.

[beanflying]

Language is defined by whatever the common usage is. No way to fight that.

Would a cal lab send a meter back without touching it if it was within manufacturer's spec, even if it could be improved?

(eg. Brymen makes 500000 count meters that are only 0.03%, clearly a meter that's 0.028% off can be improved)

Depends on what you are looking for from your 'Calibration' Initial Accuracy or Certainty over time between Calibrations?

One of my Agilents has never been calibrated since new and the other is a few years out of Cal. This time around is looking at them first to establish a point, tweaking then as necessary then retesting and confirming they are in Cal. Next cycle it will be possible to look at this cal and compare it to where they are now to establish if there is a longer term drift or ongoing aging issues. It will also over time give you a better idea of measurement certainty throughout the Cal cycle if drift turns out to be minimal from Cal to Cal in future.

My 34461A for example is likely worse than my slightly older 34401A due to initial aging based on what testing I could do with it while the 34401A was Agilent calibrated over time prior to me buying it and was most likely more stable in terms of drift/aging.

For a one off Calibration on an unknown 3 or 4 1/2 digit Handheld you have brought then set it as accurate as you can and 'hope it doesn't drift' or better get it rechecked some time in the future even if not regularly you can still get an idea of drift. If it is well used chances are it won't drift as much as a new one.

That is where having the likes of a DMM Checkplus tested every year or so by Doug has a lot of value as a sanity check from a known source it would be cheaper to do than getting a full calibration done even on one meter let alone several like most of us have. Where people go wrong is believing they are a 'calibration' device which they are not.

As I mentioned a few posts back my 121GW is great apart from the Temp range and I need to go over the current ranges properly. When the Agilents get back it will make sense to hit all the Cal points and look at and record them. Use the meter for another 6 months and do it again. This procedure will be done on my entire fleet of meters and calibration gear against the two Agilents. The collection can then be used to check against itself if you get a suspect device throughout the Cal Cycle. Get meters Re Calibrated and repeat on the fleet next year.

Rambling but maybe one to many Beers 

[Brumby]

My understanding is:

It has been mentioned before, but the point of calibration is to ensure the highest possible confidence in the values any bit of kit measures between calibrations.  If a testing station in a production facility has a unit that is know to measure 0.028% low at the last calibration, then that will be taken into consideration when the measurements are taken.

At the next calibration you will again find out how far off it is - and it is not uncommon for this to be different to the previous value.  If you then look back over several calibrations and see a trend that the latest calibration follows, then you can have very high confidence that the meter has been measuring consistently during the period between calibrations.  It is this confidence that calibration - especially the calibration history - provides.

These changes, identified each time calibration is performed, are hopefully the result of aging and will be predictable.  If a calibration shows a drastic change when compared to the previous calibration history, then that's when the T&M department at the factory get anxious.  In situations like this, there may need to be a recall and/or retest of the product that was tested by this unit.

Here it is important to understand that if you adjust, realign or change anything, you will very likely change how the unit will "age" - which means all the previous calibration history loses much of its value.  You will need to go through several more calibrations over a period of time to build up that calibration history.

In a production testing environment, absolute accuracy at any point in time is nowhere near as important as knowing how far off it is - and having confidence in that between calibrations.

[tautech]

Would a cal lab send a meter back without touching it if it was within manufacturer's spec, even if it could be improved?

(eg. Brymen makes 500000 count meters that are only 0.03%, clearly a meter that's 0.028% off can be improved)

This is the point about "calibration".

A cal lab will do the calibration and tell you the meter is 0.028% off on a particular range.  If it's 0.3% off, they will tell you that.  They will not adjust anything - and anyone serious that sends their gear for calibration would not want them to.

Yep.

From a older post by a real Cal lab technician:

<snipped>

The method you use to 'certify' a 'scope (calibration verification) is to wait for the minimum warm-up time – people forget this, and it is important as a cold 'scope is almost always going to be significantly out – then record the ambient temperature (and humidity, usually, for a lab) and run the 'self-cal' routine.  After it passes that routine you then perform the recommended verification procedure (or the customer requested one, or your in-house procedure if the manufacturer doesn't have one).  Usually the temperature can vary by a couple of degrees with no significant impact, so recording the ambient conditions at the end of the procedure is usually going to be sufficient (unless your lab is in a barn or something  ).

This provides reasonable proof that the 'scope meets specifications (or doesn't, or is ambiguous) under the manufacturer's recommended usage conditions – warm-up time and 'self cal', generally.  If it is out of spec, or very close to it, then the customer may request adjustment and a post-adjustment verification.

[joeqsmith]

Language is defined by whatever the common usage is. No way to fight that.

Would a cal lab send a meter back without touching it if it was within manufacturer's spec, even if it could be improved?

Yes they would as they should.  Normally some sort of tamper proof seal is added.  You only know that it's in or out, nothing more, unless you have them supply a report.  Again, this may add additional cost. 

Always ask before sending off the equipment. The internet's use of the language how ever popular it is, may not be what the cal house uses.