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

[bdunham7]

What do meters normally use?  It's my understanding these DMMs are stable in their calibration once calibrated

Typically either a buried zener for better models or a bandgap reference at ~1.2V for handhelds and lower end bench meters.  Both are supposed to have lower drift than typical zeners.  The zener used in the 45, Fluke p/n 857201, is listed as 6.3V +/- 3% @ 2mA bias current.  They don't even actually specify that it is temperature compensated--I'm just assuming that it is.  I cannot cross-reference or match it to any common part.  This appears to not be an issue as far as temperature stability but I don't know what the long term drift implications are.  In any case, I doubt reference drift has anything to do with whatever is going on with your meter.

[G0HZU]

I've got an old Fluke 45 meter here and I use it daily. I think I've had it for nearly 15 years. In all this time I've not seen any noticeable drift in the meter although I'm not a 'volt nut'. I've never sent it for calibration and I think it has not been calibrated since it was first sold.

I think it still reads just over 0.5mV low with a 5V input and this is quite impressive across this time span. I have never bothered to really put it through its paces other than to check its linearity is OK for my needs. With this in mind, the Fluke 45 owned by Fried Chicken does seem to have some significant issues.

I can post up some tests of mine if it helps? I've got some 6.5 digit DMMs here that are out of formal cal but two of them were UKAS calibrated when last calibrated a few years ago by the previous owners. I've also got a programmable PSU here that has 1mV resolution and good accuracy and this is usually good enough to do a basic health check on a meter across a huge range of test voltages and currents. It can also test current to 0.1mA up to several amps. It's not in the same class as a VI calibrator obviously but it is probably still better than most budget DMMs (and dodgy ebay references) in terms of accuracy. It's not quite accurate enough to actually calibrate a Fluke 45, but it is going to be close enough to prove it's still in pretty good shape...

[bdunham7]

Alright, progress!  I had almost no time to look at it earlier so I set it up to log a 10V reference for a while (overnight) in the 10V range and then I set it up with a 20V (19.99997-ish actually) source all day in the 30V range.  This tests the stability of the 100mV and 300mV ADC ranges as well as the x100 divider circuit.  Screenshots attached, the meter is perfectly stable to 1 count or less.  I did the same thing for a shorter period in the 1V and 3V ranges, no charts but the same result--perfectly stable. 

Looking at the calibration constants it appeared to me that all of the DCV-related constants were way off while all the others seemed to reflect a reasonable amount of correction.  Proceeding on the theory that this and the punctured calibration seal might indicate that someone has tried to calibrate it and failed badly, I ran it through just the DCV calibration steps.  After that I scanned through and checked for linearity and offset on all the ranges and they were good.  As I suspected from the design, the 30,000 count ranges are better than the 100,000 count ranges--the former were perfect while the latter were still good enough but not quite as perfect.  There was a 3-count discrepancy at the negative limit of those ranges. 

However, it appears that all of the other functions--ACV, DCA, ACA and Ohms are now in spec or at least very close to it.  That's all I've got for now, more later this weekend.

[CatalinaWOW]

Lots of good work there. Hope the owner appreciates the value he is getting.

[Fried Chicken]

Lots of good work there. Hope the owner appreciates the value he is getting.

I will be thrilled once it's safely back on my bench xD  But I am excited yes

[bdunham7]

OK, mission complete!  The only calibration constants changed were in the DCV sequence.  The results are pretty good, IMO.  I gave the meter the recommended 1-hour warmup, but it really seems to make very little difference.  that means you don't need to leave it on for extended periods which will extend display life.  There are three different options for testing the resistance ranges plus some extended high-resistance test points that don't have corresponding adjustments.  I did two sets--the decades of 3.0 and 1.0--and the extended tests in the 1.0 since I don't have reasonable 30M or 300M standards.

[KungFuJosh]

Great work!

[Fried Chicken]

Yoooooo the mighty Fluke 45 is back from it's California adventure!

This is some borderline professional work, papers included, packaging excellent, quick rundown of tests as follows (had to go crazy on the compression to get under the 8MB limit):

[J-R]

I think a table with the values would be easier for us to read.

Also, there are a lot of missing digits on the DT4251.  It's a 6,000 count DMM, so 5.3V can have two more digits of resolution, for example.

[bdunham7]

Yoooooo the mighty Fluke 45 is back from it's California adventure!

Also, there are a lot of missing digits on the DT4251.  It's a 6,000 count DMM, so 5.3V can have two more digits of resolution, for example.

I'm glad our friend made it home.  Just a quick pointer--don't use the Auto-V mode on the Hioki if you want any precision at all.  That's not auto-ranging, it is a low-impedance single-range check for voltage that has a specified 2% or worse basic accuracy.  It's appropriate for probing a wall socket or an SMPS primary.

[Fried Chicken]

Yes yes I noticed that.  I excitedly just got it mounted in my station and just took measurements right after having done a brake fluid flush on a car and had some beer.  I'm still running around like a chicken with its head cut off (lol bok bok).  I'll do some more testing stuff.

So professional, so nice.  Love the calibration sticker

[Fried Chicken]

And all these pictures came in a dogshit order.  I apologize, these pictures are probably meaningless.  I might format them properly, or host them myself.

[KungFuJosh]

I'm still running around like a chicken with its head cut off (lol bok bok).

Fried, I assume.

[J-R]

I vote for posting a couple photos because it's always fun to see people's setups, but make them smaller because they are grainy anyway and we don't need much detail.  The smiley sticker is fun, nice.

Then for the measurements, just make a table with values from both DMMs so we can compare, no need for those photos.  And as mentioned don't use the AutoV mode so you get all the digits of resolution on the Hioki.

[Fried Chicken]

Got some more time to dick with the unit.

The inexpensive but seemingly well built power supply I have has fixed outputs for useful voltages.  My guess is it intentionally errs on the conservative side for safety, but who knows.  Maybe now I can calibrate it that I have a calibrated and hopefully reliable multimeter.

DC Voltages:

Here it's measuring at "5 volts":

Hioki: 5.174v
Fluke: 5.163v



And at the "12 volts" output:

Hioki: 11.81v
Fluke: 11.804v



And at the "-12 volts" output:

Hioki: -12.10v
Fluke -12.086v



And I managed to get it in the mV scale using the variable voltage option on the cheap but seemingly well built DC power supply.  The Fluke 45 cuts out of the mV range relatively early, the Hioki goes up to 600mV:

Hioki: 276.1mV
Fluke: 275.58mV




High Voltage AC:

I don't have any good high voltage sources besides the wall, so that's what I used.  While I could have measured them simultaneously, I decided against doing stupid shit.

Hioki: 120.8v @ 59.98Hz
Fluke: 120.63v @ 60.02Hz
Ercot: 60.014Hz

Here is the Hioki:


And the Fluke:


And the power company:




Blah, the pictures didn't come out how I like them.  I wanted expanding thumbnails not links to full images.  I don't think I can do expanding thumbnails inline.

[Fried Chicken]

Resistors:

For the (presumably, I didn't read it) 470Ω resistor:
Hioki: 470.7Ω
Fluke: 469.9Ω

Hioki:



Fluke:



For the (again presumably) 1kΩ resistor:
Hioki: 0.998kΩ
Fluke 0.9973kΩ

Hioki:



Fluke:



And finally a (presumably) 220Ω Resistor (I just randomly grabbed these out of my misc. can of electronic stuff):

Hioki: 220.0Ω
Fluke: 219.61Ω

Hioki:



Fluke:




I have to say in doing all of this, it feels really weird to not trust the Hioki and instead trust this new (to me) Fluke.  The Hioki has been my companion for everything, being my only multimeter after someone stole my cheap-o craftsman (still angry/sad about this one).  The Hioki has this very endearing way of beeping at you when taking measurements.  It's also very fast, and been rock solid reliable whenever I've used it.  It's a weird thought to think that this Fluke 45 is now presumably my gold standard and I should trust it over my Hioki.  That said, after finding out about Hioki's after sales support/information including service manuals (it pretty much doesn't exist), I can't really recommend Hioki anymore.  That is a damn shame but oh well.

[Fried Chicken]

Frequency Counting

So now I gt to dicking with the frequency generator that I bought along with the Fluke 45, and trying out the Fluke's frequency counting capabilities.

It's a BK Precision 4040 (no 4040A, 4040B or anything, just the 4040).  This must be a pretty old unit, b/c I had a very hard time finding any information on it online.

The frequency generator's display and the Fluke are almost perfectly bang on all the way up to about 2.3MHz-ish on the Fluke that then got pissy and stopped reporting accurately.  This might have been b/c the signal became too weak, but when I hooked it up to the scope, I saw an appropriate p2p voltage, so I'm guessing it's just out of range for the Fluke?

Here is 10.004kHz:



And then at 1.6603MHz.  The scope is set to 5 volts per division, but the Fluke is only reading 45.58mV RMS.  It must be at its limit.  Also bonus Tektronix 2230 display.





I also just noticed that for many of the above voltage and resistance numbers I had the Fluke in auto ranging mode.  Forgive me, it got a bit finnicky taking these measurements.

[J-R]

I think all of those numbers look good.  You could further explore the minor disagreements by doing some additional cross-checking of different values and ranges (and the slow mode of the Fluke 45).  Use Ohm's law for further calculated verifications.  Although odds are that the Fluke 45 is right and the Hioki is slightly off.  But as mentioned before, if such things are bothering you then you're headed for some kind of higher-end bench DMM with frequent calibrations...

The Fluke 45 ACV accuracy is not specified above 100kHz, and frequency measurements are not defined above 1MHz.  So any displayed values above that are for entertainment only!

[bdunham7]

Here it's measuring at "5 volts":

Hioki: 5.174v
Fluke: 5.163v

So calibrating the Fluke 45 has only reduced your ~20mV discrepancy to 11mV--but at least you know which is which.  The tolerances for the Hioki work out to 19 counts, so 11 is OK.  Welcome to the "within specs but not as good as you hoped" zone!

[Fried Chicken]

Here it's measuring at "5 volts":

Hioki: 5.174v
Fluke: 5.163v

So calibrating the Fluke 45 has only reduced your ~20mV discrepancy to 11mV--but at least you know which is which.  The tolerances for the Hioki work out to 19 counts, so 11 is OK.  Welcome to the "within specs but not as good as you hoped" zone!

This is more than adequate for my needs.  The Hioki has similarly always been more than adequate for my needs.  This is by no means a way to verify the calibration of the Fluke.... that question... and the original inspiration for this whole thread I have no way to test your calibration   I gotta trust your work, but I do