Two multimeters, two different voltages. Now what?

[ComfyChair]

This is the EE equivalent of the old saying "A man with a watch knows what time it is. A man with two watches is never sure."

The version I've heard is:

Someone with one clock knows exactly what time it is.  Someone with two clocks has no idea what time it is.  Someone with three clocks has a slightly better idea than the person with two clocks.

Suspecting an issue with my old Mastech multimeter, I got a BM235. Now I have three multimeters to compare . The Aneng and The Brymen are practically in agreement. The outlier is the Mastech 

Does it make sense to have it calibrated? If so, how?

[coromonadalix]

mmm  my 2 cents,   

your meter have low resolutions   and the dmm ic's  can round values

the aneng is 20k count,  but resolution ... meh

get a better meter for precision stuff,   resolution is not precision .... not to start a huge debate

you have better meter(s) with at least 40k counts or more who doesn't cost an arm / leg  and do have better precision

find ones  in the dmm thread / compare ...

i would not even try to calibrate them, but if you are able ...

[J-R]

DC accuracy specs on the BM235 are 0.3% + 2 counts.  So 1.503V could read between 1.496V and 1.510V and still be just within specification.  So combine that with the 1.491V reading on the Mastech and you can see that without having any calibrated equipment you are having to make assumptions about which is correct, and by how much.

However, it's somewhat reasonable to assume that the Mastech is the one that is incorrect.  But adjusting it will not be reasonable.  You don't have any proper way to DIY it, and paying someone to do it is probably either not possible or will be $100USD or more.  It's probably not a closed case calibration either, so many of the range selections could still be off a bit (but within tolerance).


Sorry to say, but the BM235 was a mistake to purchase, since it doesn't add much of anything to your situation.  The minimum probably would have been a BM785 or BM867s with a calibration certificate, (or better to get the BM789 or BM869s).

But to rehash some points perhaps, you really want a bench DMM that you can ship out for calibrations every so often.  A typical bench DMM is far more stable and accurate than a typical handheld, and is worth the money spent for the calibrations.

[mendip_discovery]

This is the EE equivalent of the old saying "A man with a watch knows what time it is. A man with two watches is never sure."

The version I've heard is:

Someone with one clock knows exactly what time it is.  Someone with two clocks has no idea what time it is.  Someone with three clocks has a slightly better idea than the person with two clocks.

Suspecting an issue with my old Mastech multimeter, I got a BM235. Now I have three multimeters to compare . The Aneng and The Brymen are practically in agreement. The outlier is the Mastech 

Does it make sense to have it calibrated? If so, how?

Open the mastec up, and have a look to see if there are trim pots that can be adjusted. Hopefully there is and you can just adjust it to be similar to the others.

[tooki]

Suspecting an issue with my old Mastech multimeter, I got a BM235. Now I have three multimeters to compare . The Aneng and The Brymen are practically in agreement. The outlier is the Mastech 

I'm not even slightly surprised. At my old work, they had a few Mastechs --  I'm pretty sure they were the MS8268 -- and they weren't that great. I never got around to trying to calibrate the worst one of them before I left.

Does it make sense to have it calibrated? If so, how?

Paying someone to calibrate it would cost far more than that meter was worth when new. The MS8268 is calibrated by internal potentiometers, so I imagine the lower-end MS8264 does, too. See https://www.eevblog.com/forum/testgear/mastech-ms8268-calibration/

If you can access a good bench multimeter as a reference, use that, but otherwise I'd just twiddle the pots until the measurement more or less agree with the BM235 and AN870.

(For what it's worth, the Anengs have pretty decent accuracy for their price. I'd trust their measurement over a Mastech...)

[J-R]

Since we're in the Metrology forum, I'll point out I don't think it's proper to adjust the Mastech without at least a calibrated DMM or reference.  Yes you can make an assumption that it's incorrect, but just making it match the other two is bad form.

Can you find a friend with a bench DMM to at least add some additional confidence to the situation?

[tooki]

Since we're in the Metrology forum, I'll point out I don't think it's proper to adjust the Mastech without at least a calibrated DMM or reference.  Yes you can make an assumption that it's incorrect, but just making it match the other two is bad form.

Why? Mastech gear is mediocre on a good day, so IMHO going beyond "match it to a new decent DMM" is just putting lipstick on a pig. There's no point in applying metrology traceability principles to such a poor tool.

[J-R]

Since we're in the Metrology forum, I'll point out I don't think it's proper to adjust the Mastech without at least a calibrated DMM or reference.  Yes you can make an assumption that it's incorrect, but just making it match the other two is bad form.

Why? Mastech gear is mediocre on a good day, so IMHO going beyond "match it to a new decent DMM" is just putting lipstick on a pig. There's no point in applying metrology traceability principles to such a poor tool.

There's a solid non-zero chance that the Mastech is correct and the other two are incorrect.  I think it's a terrible step to go from "I think/assume this one is wrong" to "I'll adjust this one to match" even if you're not a stickler for the Metrology "principles".  The OP can do what they want, but we shouldn't condone that.

[tooki]

Since we're in the Metrology forum, I'll point out I don't think it's proper to adjust the Mastech without at least a calibrated DMM or reference.  Yes you can make an assumption that it's incorrect, but just making it match the other two is bad form.

Why? Mastech gear is mediocre on a good day, so IMHO going beyond "match it to a new decent DMM" is just putting lipstick on a pig. There's no point in applying metrology traceability principles to such a poor tool.

There's a solid non-zero chance that the Mastech is correct and the other two are incorrect.  I think it's a terrible step to go from "I think/assume this one is wrong" to "I'll adjust this one to match" even if you're not a stickler for the Metrology "principles".  The OP can do what they want, but we shouldn't condone that.

Nonzero? Yes. "Solid"? No. Nonzero but vanishingly small.

If we were talking about a single Fluke disagreeing with the Mastech and Brymen, it'd be one thing. But Mastech is a cheap brand that never performs that great.

[ComfyChair]

Thank you all again for the very insightful comments once again.

For those telling me the BM235 was a poor choice, well I must adimt I already had it but it was stored at another place. Having two labs is never a good idea but that's another story... 

Anyway, I think that for this power supply project I'll "calibrate" the output aganst the Brymen / Aneng and call it a day. The Mastech will be kept for general troubleshooting duties like continuity testing and the likes.

For the future I'll try to get an accurate reference and/or a reliable benchtop multimeter.

[Smokey]

At least you aren't working with IMUs.  Apparently you need to average the readings from 32 of those to get the confidence level up
https://hackaday.com/2024/10/02/easily-build-this-imu-array-sandbox/
https://github.com/will127534/IMU_Array

[unseenninja]

At least you aren't working with IMUs.  Apparently you need to average the readings from 32 of those to get the confidence level up
https://hackaday.com/2024/10/02/easily-build-this-imu-array-sandbox/
https://github.com/will127534/IMU_Array

This project only really improves IMU data when there is no other data to help - like in space applications. I've flown quadcopters for many years which only use a single IMU. To get good longer term accuracy with just IMU data is very difficult. However, when you add other sensors into the mix, like GPS and barometric sensors and use all the sources of information available, you can easily run a completely programmed flight path then return to home and land very close to your initial take off point.

While averaging the results from 32 IMUs is a great approach for off planet applications, there are better ways to do it here on earth.

[Paul T]

Trick question?  Probably the power supply.

[J-R]

Trick question?  Probably the power supply.

OP stated that the power supply was one they were building, so it can't provide anything meaningful.  It would have been calibrated/adjusted against the unknown DMMs, rather than a reference at a PSU factory.

[Mahagam]

At least you aren't working with IMUs.  Apparently you need to average the readings from 32 of those to get the confidence level up

Some specifications say that you have to separate them mechanically, otherwise it will interfere.

[CatalinaWOW]

Thanks

To summarize:
0. I can consider the measurements equal
1. Get a third (better) multimeter
2. Get / build a reference to compare the multimeters against.

I have a few ideas for the third point. I'll open another post for more info regarding the voltage reference design.

You missed the most important thing.  How accurately do you need to know the voltage?   For some purposes those two meters can be considered identical and the answer is 1 volt.   For others they are impossibly far apart.  In the second case the only solution is to discard (or at least set aside for lesser tasks) the two meters and go purchase a meter that meets your needs.   Since this can get quite spendy as the number of digits increases it is fortunate that a great many applications don't care about more than a couple of digits of accuracy.

[Smokey]

someone needs to get two old Simpson 260s and put them side by side measuring the same voltage and complain that they can't get the needles to line up perfectly on the same value.  Every time they move their head the reading changes!

[CalibrationGuy]

I have 5 3458A meters connected to the same 732 that have different values displayed on them all the time. It drives me crazy. As @CatalinaWoW said, all that matters is how accurate the measurement has to be to satisfy your requirements. There's a guy on eBay that sells a small reference, he's in Germany, here: https://ebay.us/m/BTqc22 I don't know him, and don't have any affiliation with him. Having said that, maybe this can satisfy your needs for a known reference voltage.

TomG.

[J-R]

Follow the 3458A specifications to the letter and I think you'll be questioning the last three digits.  But typically you could expect much better, assuming you have a way to check for the dreaded drift.

For US customers wanting a 10V reference, I think you will want to go with this unit: https://voltagestandard.com/001%25-10v-reference
He has a 3458A that is calibrated yearly and the reference can be ordered with a TC graph.  Shipping within the US will be cheaper/faster than overseas.

[CalibrationGuy]

Right, Doug, forgot about his reference. Yes, that would also work.

TomG.

[J-R]

I suppose I should clarify that the shipping cost is important because you'll typically ship the unit back every 6-12 months for calibration if you're serious!

[ITArchitect]

Every bench should have a PDVS2Mini.  If he's looking for a steady 1.0v reference, it would also be a nice source for those meters to compare.

[J-R]

Every bench should have a PDVS2Mini.  If he's looking for a steady 1.0v reference, it would also be a nice source for those meters to compare.

The PDVS2Mini is nice and all, but surely the OP needs to invest in a decent calibrated bench meter first?  You're looking at around $600 USD for the PDVS2Mini which is quite a distance from the handheld DMMs we've seen mentioned so far.


For cheap ideas, there are the old Ronan Engineering calibrators, the X85, X86 and X88, although you definitely have to look out for a deal because a lot of asking prices are unreasonable.  $50 for the X85, $100 for X86 & $150 for X88 should be possible.

Also, there are the Advantest R6142/R6144 voltage/current generators, for around $350.

In the case of both the Ronan and Advantest units, you would use them with your calibrated bench DMM to generate some very stable outputs for use with your other, lesser, DMMs.  Although once adjusted, they are probably good enough alone for use with handheld DMMs.

[bastl_r]

Hi
For the multimeters shown here, I would say that a universal reference with the AD584, which you can get cheaply on eBay, is perfectly adequate.
If necessary, you can also send it to someone who has better measuring equipment for recalibration.
I myself have an Aneng 870 here, which I bought out of curiosity. Considering the price, it is surprisingly accurate. Only the switch causes problems from time to time because I lost one of the switch's contact elements and had to replace it with another one that fits reasonably well.

Regards

[J-R]

For the multimeters shown here, I would say that a universal reference with the AD584, which you can get cheaply on eBay, is perfectly adequate.
If necessary, you can also send it to someone who has better measuring equipment for recalibration.

In my opinion those AD584 units don't bring much confidence though.  There are a lot of random sellers many of whom have the same name just with extra letters or numbers at the end.  No calibration certificates, and even if they did have one there still is the issue of trust.

So as you say, probably necessary to send it to someone trustworthy who can write down the calibration data for you.  But you could just have them do that with one of your handheld DMMs instead and be further ahead with more ranges/functions.

Handheld DMMs are available with calibration certificates/data, so if you really want to be cheap and don't need a lot of digits, it can be an option.