Easy way to test the calibration of a DMM (Fluke 45)?

[Fried Chicken]

Hah.

So now I have the slightest stupidest insight into what makes pushing the boundaries of metrology difficult.
I have three or four multimeters or multimeter-like devices capable of reading voltages, but I'm not sure the accuracy of any of them.

So what around the house options to I have to verify/test what's accurate and what's not?  Fresh alkaline batteries?  Stick the probes in an outlet and compare with what the power company is reporting?  USB power from computers?

I have two relatively nice DMMs (Fluke 45 and a Hioki handheld), along with other DMM-like devices (o-scope, chinese LCR meter with DMM capabilities, a chinese function generator that let's me dial in a voltage, the mtester thing)...  Resistance notwithstanding.

Are there any reliable around-the-house ways I can test these multimeters?  Reliable resistors?  Pencil lead?  The nicest resistor I have laying around?

[bdunham7]

Do the Hioki and the Fluke 45 disagree on anything you've measured?

[Fungus]

I imagine the Fluke/Hioki won't drift much from spec.

Simple check: Do they agree with each other on things?

eg. A battery? USB power? A resistor?

[adinsen]

Been there... I texted a friend who had gotten himsels an almost new bench meter, we had coffee, discussed life, work and other stuff, and I got my meters checked against his. Recently I got myself one of those small TL431 based voltage references with hand written values on the pcb which are available on eBay, but it drifts a little with temperature and it doesn't say what temperature it was calibrated on, so I can't trust it.

[J-R]

My first thought was this: https://dmmcheckplus.com/shop/ols/products/dmm-check-plus-fully-loaded-with-all-options-lc-board-enclosure-and-dual-frequency

But given the cost and the equipment you have, I am not sure if it makes a ton of sense to spend that kind of money.

If not, maybe an option could be to go the route of sanity cross-checking combined with multiple DMMs and other test equipment.
Different ranges and functions use different paths/configurations.  Different resistor dividers, offset and gain calibrations for each individual range/function, etc.

So apply 3V to the Fluke 45 and Hioki, then manually range up to 30V, 300V, 1000V, and see if everything is consistent.  In some devices it's worth reversing the leads.  The Fluke 45 seems to have multiple negative DC V performance tests but only a single -90mV calibration point.  If you look at the theory of operation of some DMMs, this method isn't always overly useful if the same passives are used in multiple ranges and they've drifted.  But if things align, then that provides some level of confidence I'd think.

You could get some 0.1% resistors from DigiKey/Mouser and use them with Ohm's law.  I picked up a pile some time ago and perhaps since they are designed to hold that spec through various temperature ranges, I found that at room temperature they were typically far better than spec.  So a 1K 0.1% resistor is probably going to easily be between 999.5K and 1000.5K.  Cross check the voltage drop with the current flow and resistor value and see if it all lines up.

If it DOESN'T line up, then you have more work to do, but I suspect things will come out pretty good given your test equipment's pedigree.

[Fried Chicken]

Do the Hioki and the Fluke 45 disagree on anything you've measured?

I imagine the Fluke/Hioki won't drift much from spec.

Simple check: Do they agree with each other on things?

eg. A battery? USB power? A resistor?

Yes they do unfortunately.  The Fluke reads lower than the Hioki when measuring a battery, power supply... not sure how resistors are but they don't completely agree either.

[bdunham7]

Yes they do unfortunately.  The Fluke reads lower than the Hioki when measuring a battery, power supply... not sure how resistors are but they don't completely agree either.

So the first thing you do is quantify that disagreement.  You just need a reasonably stable source for voltage that you measure with the meters in parallel.  Resistors you have to measure sequentially but that shouldn't be a problem at your precision level.  Do this several times with each meter so you get an idea of how repeatable your measurements are.  Then look at the specified tolerances of the meter and make sure your repeatability error (difference between measurements of the same source with the same meter) is much less than the meter's spec.  Once you have measurements that are stable and repeatable, you can see how far off one meter is relative to the other.  Compare that with the specified tolerances to see whether their disagreement is expected and within specs or excessive and out of specs.

You can use a 9V battery, mains power from a wall socket and any random resistor of 1k to 10k as sources to start out.  Calibrators come later.

[CatalinaWOW]

There is another step you should take.  To control the time and money spent on this you should take this step first.

How much accuracy and precision do you need.  Note that those are two different questions.  And in a hobby context want is effectively a need.

For the majority of applications one part in a hundred is good enough.  Only a rare few require one part in ten million which pushes the limits of what can be achieved hard.

If your meters agree to the level of your needs you are done.  Except for training yourself to ignore those other digits on the right of the display.  As you add digits you will have to think about more things.  Leads, contacts, and a bunch of others.

[shapirus]

Yes they do unfortunately.  The Fluke reads lower than the Hioki when measuring a battery, power supply...

How much lower?

[Fried Chicken]

Consistently 0.02v ish

[bdunham7]

You didn't state the Hioki model, but assuming you have a DT4252 (my best guess) the specified tolerance is 0.3% of the reading plus 5 digits.  That works out to +/- 0.020V.  The specified 1-year tolerance of the Fluke 45 is 0.025%/rdg plus 2 digits in the 30V range that you are using which comes to about 0.003V.  So either the Hioki as at the extreme upper end of its tolerance range or the Fluke 45 is out of spec.

You could try using the slow reading 10V range on the Fluke 45 and see if anything changes.  Beyond that, it is going to be difficult to judge which meter(s) are within spec with just what you have on hand.

[shapirus]

Consistently 0.02v ish

Well, even if that may still be within spec (I don't know what is the accuracy spec for these, that's on you to check), a 20 mV difference sounds a bit too much.

There are some good relatively inexpensive voltage references like DMMCheck plus, which are adjusted and calibrated using well-known equipment with valid calibration certificates. One option is to use these to check which of your meters is (or are) off.

Another option is much cheaper, it's to use one of these LM399 boards from Ali. The equipment with which the as-measured values written on them were obtained, and the calibration status of that equipment, are questionable, but reportedly these boards are quite decent, and in most cases good to at least 1 mV. I have one of these, and it's been staying consistently in agreement with my Brymen BM869s within 500uV in the 5V range and 1.5 mV in the 50V range, which is well within the accuracy spec of the BM869s. The question is whether the numbers written on the board are trustworthy remains, and that may be a problem in your case: ideally you would need to check this board with some trusted recently calibrated voltmeter before using it to find out which of your meters is off :).

[Fried Chicken]

You didn't state the Hioki model, but assuming you have a DT4252 (my best guess) the specified tolerance is 0.3% of the reading plus 5 digits.  That works out to +/- 0.020V.  The specified 1-year tolerance of the Fluke 45 is 0.025%/rdg plus 2 digits in the 30V range that you are using which comes to about 0.003V.  So either the Hioki as at the extreme upper end of its tolerance range or the Fluke 45 is out of spec.

You could try using the slow reading 10V range on the Fluke 45 and see if anything changes.  Beyond that, it is going to be difficult to judge which meter(s) are within spec with just what you have on hand.

Oh shoot, I thought the model would appear on the picture.  It's the DT4251 (same as DT4252 afaik, except no current functionality).

So this is where the challenge comes in, what reliable voltage source might I have laying around...  Else as shapirus suggested, one of those aliexpress/ebay voltage sources...  Unless there's some creative way to find a reliable voltage source.  Maybe from a PC computer that might have a built in multimeter for its 5v USB rail?

[shapirus]

Maybe from a PC computer that might have a built in multimeter for its 5v USB rail?

Nope, you wouldn't expect anything like that to have the required accuracy.

You need either a voltage source that is sufficiently stable and has been measured with a sufficiently accurate (preferably better than your DMMs) calibrated (aka verified) meter, or access to such a meter to compare its readings with those of your DMMs on an arbitrary voltage source like a battery.

Realistically, DMMCheck Plus sounds like the best option (but yeah, $225, oh well), since if you get that cheap Chinese board, you still don't have a trusted DMM to verify it and use that initial verification as your reference point. Yes it is done by the manufacturer, but to what extent can that manufacturer be trusted? In case of DMMCheck Plus it is as trustworthy as the calibration certificates published on the manufacturer's website, and that's more than sufficient for your use case.

[bdunham7]

So this is where the challenge comes in, what reliable voltage source might I have laying around... 

This is where/how USA Cal Club got started, albeit at a few more digits of precision.  You aren't likely to have anything "laying around" nearly accurate enough to exceed the Fluke 45 specs.  And a one-point voltage source isn't really all that much help when you want to verify the meter entirely.  If you really care, you need an official or trusted unofficial calibration.  The question is whether you care enough to worry about that level of discrepancy. 

[Majorassburn]

So this is where the challenge comes in, what reliable voltage source might I have laying around...  Else as shapirus suggested, one of those aliexpress/ebay voltage sources...  Unless there's some creative way to find a reliable voltage source.  Maybe from a PC computer that might have a built in multimeter for its 5v USB rail?

Shamelessly self-promoting my products.....

https://www.ebay.com/itm/285781907875

Plus, I have precision resistors, etc....

[Fried Chicken]

So this is where the challenge comes in, what reliable voltage source might I have laying around...  Else as shapirus suggested, one of those aliexpress/ebay voltage sources...  Unless there's some creative way to find a reliable voltage source.  Maybe from a PC computer that might have a built in multimeter for its 5v USB rail?

Shamelessly self-promoting my products.....

https://www.ebay.com/itm/285781907875

Plus, I have precision resistors, etc....

I shall keep this in mind, but right now I've spent so much money on stuff I'm waiting before selling anything.

[bdunham7]

Shamelessly self-promoting my products.....

As I stated earlier, a single-point source is of rather limited usefulness here.  Besides that, is your reference here sufficiently accurate to test or calibrate a Fluke 45?   Does the trim pot adjust the 10V or 6V output?

[bdunham7]

I've spent so much money on stuff I'm waiting before selling anything.

Worrying about anything past the second place after the decimal point can get expensive in a hurry, so be careful!

If you like, I can unofficially calibrate your Fluke 45 for the cost of postage both ways.  I'll post the process and results to the forum here.  PM me if interested.  I don't have and can't find calibration info for the Hioki, so I'd be unable to adjust it, but it can be characterized so that you know how far it is off.

[Fried Chicken]

If you like, I can unofficially calibrate your Fluke 45 for the cost of postage both ways.  I'll post the process and results to the forum here.  PM me if interested.  I don't have and can't find calibration info for the Hioki, so I'd be unable to adjust it, but it can be characterized so that you know how far it is off.

That's very generous, and would be fantastic! assuming you have the right equipment.  I toyed with replacing any recommended wear components (electrolytic capacitors, carbon composition resistors, etc.) prior to having some sort of calibration done at some point.

Unfortunately Hioki has terrible support for their instruments, providing neither schematics, repair manuals, or even calibration instructions.  I was quite fond of Hioki, bordering on fanboyism, until I found this out.  It's a damn shame.

[Majorassburn]

Shamelessly self-promoting my products.....

As I stated earlier, a single-point source is of rather limited usefulness here.  Besides that, is your reference here sufficiently accurate to test or calibrate a Fluke 45?   Does the trim pot adjust the 10V or 6V output?

You just love to pick apart a reply and mis-state the context in which it was offered!
READ WHAT FriedChicken ASKED FOR: a low cost reference like the cheap Chinese devices on AliExpress, right?

So, you want to turn my reply into an argument and compare my $30 device with your favorite DMM Check for $250? Is the DMM Check sufficiently accurate to calibrate a Fluke 45? No F**king way and you know it! So why imply that my ad suggests that mine is? AGAIN and AGAIN in your multiple posts about my devices.

And, with all your expertise, if you read my description of my 10V references, it is CLEARLY EXPLAINED THAT THE TRIM POT TRIMS THE OUTPUT OF THE REFERENCE!

Does my ad anywhere say or imply or suggest that the pot adjusts the 6V tap?  Look closely and you'll see the 0.1% resistor divider that is FIXED at 6V as a "CONVENIENCE" and not a separate reference!

I'm sick of your arrogant sideswipes about most of my posts as I try to be a good participant in this forum. To hell with guys that want to flex their arrogance and inundate us inferior individuals with their self-proclaimed expertise. That's the kind of attitudinal presence that chase away forum participants and narrows the focus of an otherwise limitless exchange of ideas and experience.

So, I may as well get banned forever for saying this but I offer it with a cool head and considerate forethought...

bdunham7.....FUCK OFF! 

[Fried Chicken]

Can both of y'all chill.

Major, your little voltage reference is exactly what I was looking for.  I didn't want to get one of the chinese aliexrpress ones because I don't trust something mass produced out of China.  The only reason I was holding off is because I wouldn't just want the voltage reference, but every reference you have available in one fell order... and I just spent money frivolously on other stuff, so I'm waiting here slightly.

bdunham, your offer to calibrate my multimeter is very generous and something I will take up on, assuming shipping is within the United States and available reasonably, which it should be through whatever freight forwarder service I use.  I can include the return label in the box....  I am curious about what equipment you have in your lab that you might test it with, so I can gloat about know to what standard it's been calibrated to.

[bdunham7]

READ WHAT FriedChicken ASKED FOR: a low cost reference like the cheap Chinese devices on AliExpress, right?

No, he wants to know which of his meters are correct and presumably in spec and which are not.  Neither your device or the cheap AliExpress AD584-based references are likely to be very helpful when it comes to the Fluke 45.  I'm not the one that advocated the DMM Check and it has the same issue regarding only having a single point for DCV.  Its stated uncertainties are marginal for the Fluke 45, but at least would be helpful in determining whether it was in the ballpark.

bdunham7.....FUCK OFF! 

Yikes!  This is a simple discussion involving math and facts, no need to be offended if I don't agree with you.

[J-R]

It might be a bummer to see that your two DMMs disagree that much, but let's analyze a bit.  The Hioki DT4251 is a 6,000 count DMM and you are near the top of the range at 5V.  The Fluke 45 is 99,999 counts but only in slow mode.  In medium mode you are limited to 30,000 counts which means you are in the 30V range.  Try setting it to slow mode in the 10V range.  You could also check the 100V slow range to gather more intel.

Also, was the Fluke 45 allowed to stay powered on for some hours before taking that measurement?

I disagree, the DMMCheckPlus is absolutely good enough to check those DMMs, and in another thread I mentioned leveraging the current source and resistors to generate additional values.  Sure, it's not going to allow you to check everything and the cost for this purpose is iffy as I also mentioned.  Also, the specs are worse case between calibrations, so right after receipt/calibration it is going to be better and that is a common metrology tactic.

Paying some money to get the Fluke 45 calibrated/adjusted seems like a reasonable path at the moment.  Although I wonder if it would be better to snag a calibrated 6.5 digit DMM instead if you're already worried about such things.

[Majorassburn]

To Bdunham7:
OK. I'm never offended when you disagree with me. But, I don't like your seemingly pickiness and taking things out of context regarding my device claims.  I really strive to UNDER-state their capabilities and I would never try to inflate or scam anyone who trusts in me and buys one.

Look, I know that you are a valuable and expert contributor in this forum. I have admired many of your posts and contributions. But, you seem to have a bug up your ass about some of my devices that I offer on eBay. I wish I could resolve that with you because I know that I could learn a lot from you.

I will do my best to eliminate any sensitivity that i have developed about your "disagreeing" with me if you can try to frame your "concerns" about my posts in a more helpful and constructive manner.

I really do appreciate folks questioning my claims and raising questions as to performance, etc. so don't hold back.

TRUCE on my part. I'll probably get banned anyway for engaging in the nasty manner as I have in this post so, if or when I do, Best Regards to all and I'll see you on eBay (SQWARREL) in case I have anything you might be interested in.

Thank You.