Multimeter Calibration

Status
Not open for further replies.
C

cornbread

Senior Member
A bit off topic for safety but perhaps not. We are having a debate concerning multimeter calibrations. We have specific calibrated meters we use to comply with ISO. The question is: should multimeters used for daily troubleshooting be calibrated and if so how often.

The argument against calibration...cost.. and you basically test your meter each time you use it via "verify..test..verify".

The argument for claibration is a false reading that gets someone injured.

As always... your replies are greatly appreciated.
 
Well, what's accuracy worth to you?

Does it really matter to you if the readings are off a bit, or vary by a few volts? Or, do you need the KNOW to the nearest joule?

What are the tasks that require particular precision, anyway?

Well, if you're using a megger to PI or saturation tests, it matters a lot. Your results are interpreted by reference to previous tests.

If you're using a TDR to find a fault, you really need to be sure you're tearing up the correct part of the wall.

By comparison, if you're checking for power, it doesn't really matter if you read 118v when the value is actually 121v. If you're just checking for continuity, a few ohms usually won't matter.

Every now and then, the local Fluke rep has a 'calibration day,' where he'll check your stuff. No certificate, no charge .... but you get the peace of mind of knowing whether it's time to replace. From one of these days I learned I had a meter that was both one of the oldest Flukes (their second model) as well as one of the most accurate he'd seen for awhile.
 
Based on my past ISO experience I would think that having meters that were of a "calibration not required type" that were the functional equivalents of those managed under a tracable cal system would be opening yourself up for some management and certification/surveilence audit nightmares. You'd be opening your system up to some close scrutiny, procedures, cal, etc...
What I have done in the past is establish calibration schedules that matched or exceeded the manufactures recommendations. I have done this both with tracable in-house systems and contracted cal services. I've been out of it for a while but I suspect that it would be hard to compete with a contract service. Any piece of test equipment that could be calibrated was calibrated, any device that could not be calibrated was marked as "calibration not required" and appropriate verification procedures put in place.

I realize that I come form a different world than the electrical but suspect that because you are ISO certified my experience is relevant.

Were I a business owner in the electrical world I would want my test equipment subjected to a calibration or verification system, it may not be tracable, but I woulkd want it in place from both a safety and productivity standpoint. I have seen some real garbage in use in my few short years in the industrial world, voltmeters way off, amp clamps off by >50%, amp clamps that were horribly inaccurate when used in the close proximity of VFD's, you name it.
 
We have all our test equipment calibrated every year, they come to our shop and spend 2-3 days here calibrating. Every test report we do has the equipment, serial #, and calibration date on it.

I have a hard time thinking of a reason not to calibrate your meters
 
zog said:
We have all our test equipment calibrated every year, they come to our shop and spend 2-3 days here calibrating. Every test report we do has the equipment, serial #, and calibration date on it.

I have a hard time thinking of a reason not to calibrate your meters
Click to expand...

We send ours off to get calibrated periodically. Most of the time we are not making a measurement so much as a go/no go type of test where the actual value does not matter all that much.

It is not free to get things calibrated.

Many of our customers have a lot of instruments and readouts that have labels on or near them that indicate they are uncalibrated because it is just too expensive to calibrate them all.

I once had a long and convoluted discussion with one end user over just how they planned to implement their calibration labeling system on an HMI based control system. Their standard called for any display that had a value that came from an uncalibrated source to be labeled as uncalibrated. We never really resolved how they would ever make that work on an HMI that has many dozens of displayed values.
 
Obviously, electromechanical type of meters have a much shorter calibration period than electronic meters and the latter may not really need calibration according to some text book. But still, after any fault event, any meter requires calibration to ensure safety.
 
Last edited:
Just to clarify, calibration of DMMs or clamp meters will NOT include adjusting anything in the meter. A meter may read three volts low its entire working life and still be the most accurate meter in the company.

Calibration measures drift in readings at points in time, allowing you to identify if a meter has an issue. You would get a list of offsets for various measurement points on the scales, and if any are out of spec, the meter doesn't pass calibration and you take it out of service. This also lets you know that ANY measurements you made between the last (failed) calibration and next to last calibration are suspect. As soon as you adjust the meter, none of your measurements are trustworthy.

How often you get it done depends on what kind of quality control you need. I suspect for most cases yearly would be fine, unless you are using your DMM in the aeronautics industry or some similar field.
 
I did work at a Hydrogen Peroxide plant and whether we used our meters or theirs, they were calibrated every Monday. They would send their standard out yearly for calibration
 
We do a fair bit of work for the petrochem sector - mostly drives.
Each panel has a unique tag number and its own factory acceptance test (FAT) document which gets completed usually on the customer witness test.
The FAT document has to have the serial numbers of all the instruments used along with the calibration certificate number and date.
The project documentation includes copies of the calibration certificates.

We have a few bits of kit that are not worth calibrating but are still useful. They get a sticker "Uncalibrated - use for indication only".
 
Status
Not open for further replies.
Top