EEVblog #374 - DIY Multimeter Calibration

[EEVblog]

Dave.

[LEECH666]

"This video is private"

...


NOOOOOOOOOOOOOOOOOOOOO!




(Probably just a temporary thing.)

[GabYoung92]

I'm getting "This video is private"  All good!.  Makes me wonder how accurate my ds1502e is...

[Chris Wilson]

Getting "This video is private". Me or thee?

[samgab]

I watched this video, then went back to my youtube subscriptions home page, and the video was gone, and now there's no trace of it. Maybe Dave pulled it down again to edit something...?
Nice buys on eBay, btw. I wish I could find some great buys like that voltage standard going for that sort of price.
EDIT: The video is back up again now...

[LEECH666]

Works for me now, but my lunchbreak is over now.

[ablacon64]

It works here!

[EEVblog]

I watched this video, then went back to my youtube subscriptions home page, and the video was gone, and now there's no trace of it. Maybe Dave pulled it down again to edit something...?

I didn't edit anything, it just got set private accidentally for an hour or something.
BTW, you can't edit youtube videos after upload, you can only delete and them reupload them, and in that case you lose the views, the unique url, the embedding, the comments etc. It's a one-shot deal.

Dave.

[tom66]

I'm getting "This video is private"  All good!.  Makes me wonder how accurate my ds1502e is...

It's only about +/-3% or +/-4% (from memory) on the vertical axis.

Timebase though is pretty accurate (for a low end instrument), I'd imagine around +/-20-50ppm as it uses a standard crystal oscillator.

[Salas]

What ref gear it takes to cal AC amplitude range in a DMM? For current range AC or DC I see no pots in a couple of low cost DMMs I got at least. Is it just down to shunt resistors and ADC precision? Tinning or grinding the big one in steps can be a means of calibrating the Amps range maybe?

[ecat]

Good question Salas, how do you calibrate a meter without twiddle pots? Comms cable and specialist hush hush propitiatory software?

Another question.
If you get hold of a 5, 6 etc digit, 0.0xy% multimeter or voltage source or one of those little DMM calibrators, what is the best way to maintain their accuracy? Keep them switched on? Switched off? Keep them in the freezer? Within the confines of a lead lined pyramid or under your pillow?

Timely video Dave, thank you.

[ftransform]

Could you make a new case for that reference with the pots on top and fill it with oil for greater temperature stability and accuracy?

[ModernRonin]

I've always wondered how much it would cost to build a low-voltage, high-precision reference. I mean, with stuff like this kicking around:

http://www.maximintegrated.com/datasheet/index.mvp/id/3623

5.000V reference, +/-0.02% accuracy and 3 ppm/C drift for the "A" grade part. Which is $6.27 in qty 1 at Digikey.

Add some 0.01% resistive dividers and you have a pretty nice low-voltage precision reference for probably less than $10 parts cost.

I guess the only real question is how do you extend this to higher voltages? I'm sure there's a way. I want to say, some more precision resistive dividers and a high-precision comparator? Divide your 10V by 2, and then compare with your 5V absolute reference. Use the output of the comparator to correct the resulting voltage. But I'm worried that the input and output offsets of even a high-precision comparator might become significant.

What are the specs again on that chopper-stabilized opamp Dave used for the uCurrent, again? Offset of 2.5 uV worst case for -40 to +85C? Pretty sure the errors inherent in 0.01% resistors are worse than that, even when you figure in the mandatory 10 times gain on that thing. Nice.

[EEVblog]

Good question Salas, how do you calibrate a meter without twiddle pots? Comms cable and specialist hush hush propitiatory software?

Yep, pretty much.
The Fluke 87 for example has a built-in firmware mode for calibration.
And my Gossen meters for example use serial commands.

Another question.
If you get hold of a 5, 6 etc digit, 0.0xy% multimeter or voltage source or one of those little DMM calibrators, what is the best way to maintain their accuracy? Keep them switched on? Switched off? Keep them in the freezer? Within the confines of a lead lined pyramid or under your pillow?

Usually, switched on in a stable temp environment. Reference zeners usually are more stable with operating age.

Dave.

[PA4TIM]

I made several Vrefs. The last one a LM399 and 0.01% low tempco precision resistors.
The biggest problem ( I think)  at this moment is drift due to humidity changes. But it is only powered on for about a month so not burned in yet.
I'm monitoring this now every day 3 times, noting temp and humidity. The room temp is between 19 to 24 degrees and Hr is 30-35%. Biggest changes seems to be humidity.

The Vref (LM399 with a chopper LTC 1052 and a opa277) is now floating between 10.000,040 and 10.000,062 depending on temp and humidity. The next step will be ovenizing the whole Vref and adding battery power to suppress common mode problems. The 7.5 digit meter monitoring is also on 24/7 and modified by me to keep it inside at a steady 37 degrees.

Use good shielding, avoid thermal and mechanical stress, clean everything very good with IPA, never, and I mean never, do what Dave did in the video over his Voltstandard. Use your " dirty" fingers to grab the reference or opamps. ( when he pulled out the piggyboard, that was mounted like that for several good reasons. Removing it and replacing introduces mechanical stress, the greas an salt from your fingers degrade isolation resstanse with a factor upto 100. And then you got things like seebeck effect, triboelectrical effects ect. Keithley has a downloadable book on their site about measuring uV's that has a lot of good advise. Jim Williams wrote some good application notes on making Vrefs.

It takes several thousants of hours to get the best stability.
If you buy a 0.02% Vref  you have a voltage somewhere between 4.999  or 5.001. That is not rather accurate. Good for a 2,5 digit meter. Add to that the resistors, if you use two 0.01% 10K and they turn out to be 9999 and 10001 ( murphy rules) you get an even bigger fault.
So to make a good reference you need a good meter to check it, and to check the meter you need a reference. Chicken and egg.

[Salas]

What about ACV calibration though? Having a great spec DMM reading a function generator and the one to cal. in parallel, then tweaking its AC trimmer (if any)?

[tom66]

You can tweak True RMS ACV using a square wave of your reference voltage, the reading should be equal to the reference.

In general ACV is about +/-0.5% even for a good meter, it's hard to get it right.

[Dread]

A well calibrated meter is a must have item but on the flip side I always find it amusing
how we all get bothered If the meter reads 5.0012V yet we build circuits that should output 5.000V and then say "Good enough" when it reads 5.09v!  I think the extreme level of accuracy we all want from our meters is rarely ever needed in real world circuits.

[LoyalServant]

I know someone that got one of these and claims that it's as the aussies say: ducks guts
http://www.voltagestandard.com/DMMCheck.html
I really have to wonder if any such device could be used as a standard for anything.

I suppose it's better than nothing but I would rather have the real deal like Dave does.

[PA4TIM]

He also sells a better one but the downside is the expected 162 uV drift the first year powered on. On the other side, i think 5 dollar for recalibration is real cheap and the first couple a times are free. But it is just one voltage. You need more to calibrate a meter.

AC can be made by a squarewave using a Vref to drive it. For a scope just use the DC. A handy feature of old Tek tube scopes. They have a squarewave from a few mV to 100V or so. I use my Fluke 760, Fluke 510 10V AC standard, Tek 067-0508-00 and 067-0542-99 or my R&S AGFU
For DC the real deal, a Fluke 332B, 3 KV dividers ( Fluke, ESI) , Guildline 9152T/4, Fluke 731, Fluke 750, 760, Philips PM2480. I monitor this with a Solartron 7061 and a Prema 7.5 digit.
For Resistance some ESI and GR stuff. For capacitors GR stuff, current through the 750 and Philips or 332 and a ESI resistor. For Frequency a thrimble thunderbolt and a HP-105.

Get on of these:

[Salas]

A well calibrated meter is a must have item but on the flip side I always find it amusing
how we all get bothered If the meter reads 5.0012V yet we build circuits that should output 5.000V and then say "Good enough" when it reads 5.09v!  I think the extreme level of accuracy we all want from our meters is rarely ever needed in real world circuits.

It even matters we know when and where it does not matter. And that takes we know what it is to our field's adequate accuracy demands. Me I make silly leisure time audio circuits, another is positioning satellites though. Lucky me $50-100 DMMs can do. So lets tweak them.

[ejeffrey]

Add some 0.01% resistive dividers and you have a pretty nice low-voltage precision reference for probably less than $10 parts cost.

I guess the only real question is how do you extend this to higher voltages? I'm sure there's a way. I want to say, some more precision resistive dividers and a high-precision comparator?

The annals of electronics history are filled with really clever solutions to this type of problem.  Basically, if you are clever and you have enough patience you can use bridge circuits and a null detector to construct whatever voltages / resistances you want from a single reference of each.  Even if you don't have a good resistor reference you can still construct a decade series of resistances that are all in perfect agreement with each other, and from that you can generate any voltage ratio you need from your voltage reference.

[hobbs]

Nice gizmos, very useful.  Of course now that you've slashdotted them, there'll be a run-up in the prices, but oh well.

BTW you could make a nice 100V voltage reference using something like an OPA2118 chop amp, your voltage calibrator, and a programmable power supply (or a MOSFET on the op amp output).  Set the loop up for a gain of -10 using your Vishay resistors as the feedback network.

I'm a big fan of the old HP 611xA precision power supplies, with four decade thumbwheel switches and a pot for the last decade.  Not as good as your voltage calibrator, but then one of mine goes to 3 kV, and another to 1A, Also cheap on eBay, at least until recently.

Cheers

Phil Hobbs

[Matje]

I've always wondered how much it would cost to build a low-voltage, high-precision reference. I mean, with stuff like this kicking around:

http://www.maximintegrated.com/datasheet/index.mvp/id/3623

5.000V reference, +/-0.02% accuracy and 3 ppm/C drift for the "A" grade part. Which is $6.27 in qty 1 at Digikey.

Add some 0.01% resistive dividers and you have a pretty nice low-voltage precision reference for probably less than $10 parts cost.

Well, a bit more probably. You need a case (cheap), some connector stuff (can be quite expensive), preferably a (pre-)-regulator, a diode to protect against input power connected the wrong way around, some random stuff.

I think I paid about 30 to 40 Euros for something using LT1021s with 5 and 10 Volts, 0.05% nominal. This is more than enough for the usual DMMs, and still enough to check that a 34401A is doing reasonably OK.

The reason for having 5 and 10 Volts is that different DMMs have different ranges, checking near maximum range value is a good idea.

I guess the only real question is how do you extend this to higher voltages? I'm sure there's a way. I want to say, some more precision resistive dividers and a high-precision comparator? Divide your 10V by 2, and then compare with your 5V absolute reference. Use the output of the comparator to correct the resulting voltage. But I'm worried that the input and output offsets of even a high-precision comparator might become significant.

What are the specs again on that chopper-stabilized opamp Dave used for the uCurrent, again? Offset of 2.5 uV worst case for -40 to +85C? Pretty sure the errors inherent in 0.01% resistors are worse than that, even when you figure in the mandatory 10 times gain on that thing. Nice.

I don't think that is worth it for hobbyist use. Resistors with 0.01% cost real money. Precision OpAmps cost real money.

As far as I know for affordable DMMs you don't get to adjust every range individually, the ratios are fixed. So why bother?

[ModernRonin]

I made several Vrefs. The last one a LM399 and 0.01% low tempco precision resistors. The biggest problem ( I think)  at this moment is drift due to humidity changes.

Sounds like you might want to put it in a container, then fill the container with dry nitrogen, and then pot the whole thing so it's air-tight. No more humidity. (Yeah, I know -  this kills the heat transfer.)

If you buy a 0.02% Vref  you have a voltage somewhere between 4.999  or 5.001. That is not rather accurate. Good for a 2,5 digit meter.  Add to that the resistors, if you use two 0.01% 10K and they turn out to be 9999 and 10001 ( murphy rules) you get an even bigger fault.

I agree we can do far better. But if the voltage measurement mode of my meter is calibrated to +/- a thousandth of a volt, that's way more accuracy than I will ever need for the kind of stuff I use the voltage mode for. (Current mode is a different story, of course...)

Also, that part I mentioned is guaranteed for currents up to 10mA, so 5V / 0.01 A = 500 ohms. So we'd be well within spec to use 560 ohm, 0.01% resistors and have only 0.056 ohms of error introduced by the resistors.