DMM calibration

[eccerr0r]

What do people do to calibrate meters... or perhaps you just take the reading as gospel?

I don't own any high quality meters and likely all are out of calibration. Without paying for a tech to calibrate your meters, what do you do to get meters calibrated?

[Kiriakos GR]

Re: DMM calibration

You can not get something with out paying. Rule of life No1

[modemhead]

Re: DMM calibration

If the job requires it, then a professional calibration is necessary.

But for hobbyists and weekend warriors there are other solutions. You could find a friend with a new or recently-calibrated meter and compare readings. There are also some reasonably-priced voltage references available:

http://www.voltagestandard.com/
http://www.gellerlabs.com/Voltage%20References.htm

You can make your own with one of the various voltage reference chips available, but the advantage of ordering one from the above vendors, is that they will have trimmed it using their own calibrated DMMs, and will allow you to send it back later for recalibration.

[retiredcaps]

Re: DMM calibration

I don't own any high quality meters and likely all are out of calibration.

If these are your meters

http://www.vanade.com/~blc/html/hardware/meters.html

then I suggest top 6 in your list are within their respective specifications.

Some of them may actually give the same reading if you connect them to the DCV source at the same time. As an example, I connect multiple multimeters to an ATX power supply.

The HF Centech $3 specials can be a hit or miss in terms of accuracy as blogged by modemhead in the name of science. Click on each picture to read the captions.

http://mrmodemhead.com/blog/gallery/cen-tech-92020-dmm/

Reliability, build quality and durability are, of course, a different story.

[retiredcaps]

Re: DMM calibration

There are also some reasonably-priced voltage references available:

http://www.voltagestandard.com/
http://www.gellerlabs.com/Voltage%20References.htm

I can't remember where right now, but at least one person bought multiple $3 HF meters and checked them against Doug's voltage standard every 3 months.

He preferred this approach vs buying "expensive" multimeters.

[Kiriakos GR]

Re: DMM calibration

I will play the role of wikipedia here for the correct use of words.

calibration = comparing the meter against calibration standards (volt Ohm Ampere Resistance)
And feeling up one simple report of how off the meter is in all those parameters.

Adjustment = The use of adjustment pots OR in meter special software mode,
so the merer to return close to their specifications.

In conclusion one meter with low accuracy specifications of 1% or 2% it does not considered as one device which worth any calibration.

One high quality meter with the accuracy of 0,025% it translates that it worth to be inspected regarding calibration, when what you do with it, it does have high importance.

[eccerr0r]

Re: DMM calibration

It's absolutely correct that the cheap low precision meters are bound to deviate quickly even if calibrated against a known source, but if it's checked, it will still be more accurate than one that has never been checked against a known source.

I was most worried about measuring the voltage of Li-Ion cells and being 100 mV off is kind of bad. The range of the meters I have is fairly large, over 100mV, meaning that at least one of them can be quite a bit off.

I tried to test them all within a short period of time with each other to make sure I don't have the test batteries self discharging on me but it doesn't help with absolute accuracy.

I just wonder how many people who read BCN actually calibrate their meters, assuming a lot of us don't have contracts with Fluke or Agilent or something to keep our meters calibrated. The Li-Ion cells is what bothers me the most right now, I wish my meters were within 20mV instead of 100mV, but 2mV accuracy is unneeded... And based on the fact 20mV is about 0.5% of nominal Li-Ion cell voltage, I just hope temperature compensation will hold it through.

[Kiriakos GR]

Re: DMM calibration

Yesterday I discovered this gadget which is a crippled Li-Ion capacity tester, after three minutes of thinking, and by reading 20 customer comments I did press the buy Now button.

http://www.ebay.com/itm/121127895422...84.m1497.l2649

[Sparks88]

Re: DMM calibration

buy various 1% resistors to check the readings you are getting is the cheapest way as for voltage get hold of a lab power supply and do a quick check there, calibration is expensive and most meters will do a good job, even the expensive fluke models have a error which is based on a percentage of the reading so the higher you go the further out the reading could be - notice i said COULD.

[mariushm]

Re: DMM calibration

A laboratory doing official calibrations rarely - if ever - adjusts a multimeter. For professional tools, it's more important to have a HISTORY of how the multimeter or tool drifted or stayed within the specifications so that you can predict how the thing will work in the near future.

These videos are very informative and explain things better than me :

EEVblog #420 - What Is Calibration?
Peter Daly, metrologist at Agilents world leading standards & calibration laboratory in Melbourne explains what calibration is.
http://www.youtube.com/watch?v=bN1y1my4gEE


EEVblog #424 - $3M Agilent Portable Calibration Lab Tour

A tour of Agilent Australia's new VOSCAL (Volume On-Site Calibration) portable NATA accredited calibration lab in a customised expandable shipping container, with Metrologist Peter Daly

http://www.youtube.com/watch?v=G3QK31zotoQ

Now for cheap multimeters like we use, and for just making sure the measurements are closer to reality, you can "calibrate" it yourself by building several voltage references, compare what the multimeter shows with what the voltage reference shows and adjust accordingly.
There are reasonably cheap (6-7$ a piece for +/- 0.02% tolerance and 5ppm/c) voltage references out there ... most multimeters are about 0.05-0.1% +/- 1-3 counts on DC voltage, so even such cheap voltage reference is good enough.
Just let it stabilize for a few hours, make sure the temperature stays constant and compare...

There are also ready made voltage references, some also have AC voltage and current options. This guy for example makes such voltage references :

http://www.voltagestandard.com/

He measures them with a relatively high end multimeter that he keeps calibrated and tested.


Now keep in mind before tweaking your multimeter that some are more sensitive than others. Don't try too hard to get your meter exactly to a reference voltage.. it may affect other ranges or the meter itself may show other value once it warms up or once battery voltage goes down a bit etc etc

For example, my Uni-T UT61E is more sensitive to temperature. I could open it up and use a voltage reference that outputs 1.024v/2.048v to tweak it and make it show 1.0240v/2.0480v on the screen (it's a 22000 count multimeter) but 5-10 minutes after closing the case it may show 1.0241v or 1.0242v because with the case closed the temperature inside increases a bit. It's still valid, the specs for the multimeter say dc voltage +/-0.1% + 2 counts so at 22000 counts, about 2-5 counts drift from 1.0240v is perfectly within specs of this multimeter.

[eccerr0r]

Re: DMM calibration

I have a few REF198G ICs which are 4.096V +/- 10mV... Wonder if this is worthwhile to try to test against these - well, also need to find a few resistors and know their exact values to calculate a voltage divider, and use that voltage, would like to stay below the 1999 count.

Lots of places error can creep in however, including solder stress... ugh. But being accurate to 20mV would be nice.

[Kiriakos GR]

Re: DMM calibration

would like to stay below the 1999 count.

This is a good point as realization.
For electronics even a budget meter it should be at 6000 counts.
20000 counts with speed and good specifications is a must have today.
Fresh meters from good brands are now available at 30.000 to 60.000 counts.

Therefore even three years back, I did got my reference DC chip at 10V at a price of 24$, with out considering counts limits.

For mV calibration tests the "ticket" is even higher at 500$.

http://www.ittsb.eu/DMM%20Test%20tools.html#MICROCAL

The point is that behind the manufacturing of reference sources there is one industry which it does not kidding regarding retail prices.

For AC mV or Volts, this is the field that one good quality function generator will be able to produce volts and frequency, so to test even the bandwidth of your DMM.