Limited Calibrations - An instrument that will not pass a full calibration, OOT

M

mecramur

I have had two requests for limited calibrations in the last month. By a limited calibration I mean an intrument that will not pass a full calibration, OOT.

One was for an x,y,z measurement scope where the z axis lead screw has to much slop in it to be repeatable causing an OOT when calibrated to OEM procedures. The cost to repair is close to buyingf a new unit. The owner wants to use it stating they only use the x and y axis and can lock out the z measure.

The second is for a four channel oscilloscope, channel four does no pass, the other three do, and rs232 does not pass. Again the cost to repair is close to buying a new unit. The owner wants to continue to use the other three channels.

Our SOP does not have any provision for a limited calibration.

Does anyone use a limited calibration????
 

howste

Thaumaturge
Trusted Information Resource
Re: Limited Calibrations - An intrument that will not pass a full calibration, OOT

I've seen this from time to time, and I think most calibration labs I've worked with were more than willing to issue certificates with limited calibration and provide calibration labels showing the limitations.
 

Jerry Eldred

Forum Moderator
Super Moderator
Re: Limited Calibrations - An intrument that will not pass a full calibration, OOT

They are out there and I have used them. For an external/commercial lab, with some exception, it should be the customers decision (see my description below of exceptions). For an internal lab, it should be governed by what is allowable in your SOP or policy. There are, by the way, many companies that specifically never allow limited calibrations (and they have some good rationale for it).

Typical definition essentially is a calibration to anything less than full specs for a variety of reasons - either standards limitations, customer need, malfunction to an unused range or function, etc.).

A reason to never do a limmited calibration is that there is the possibility that someone may still make an incorrect decision based on using the uncalibrated aspect of the instrument. I previously worked in an FDA regulated environment where they specifically disallowed them. There were times when I gritted my teeth feeling it was a little overkill. But when you are making products that mean life or death to users, it is reasonable that you want an air tight test context. So a limited calibration could be dangerous in that circumstance.

The other time I will not do a limited calibration is on an instrument that does not meet full specifications where there is any reasonable likelihood associated with the limited cal of a reliability failure. If there is something wrong and undiagnosed in the instrument that has a reasonable likelihood of causing further issues, I will not issue a limited calibration.

I have an end user that uses a Fluke 5500A multifunction calibrator only for the thermocouple measurement function, nothing else. This would have to go to the OEM for a high dollar calibration; or they send it to us to verify that single function (true story). There is no reliability issue, and the instrument is clearly identified with its limitations.

I'm not a big fan of limited calibrations, but there are times when they meet customer needs.
 

BradM

Leader
Admin
Limited calibration are a slippery slope. First, if a particular instrument is suitable for the specific need, then "yes", mark it accordingly and have it calibrated for that specific need. If all you are going to do is use a multimeter for measuring 120 volts AC , for example, one really doesn't care that the meter exceeded tolerance at .02 millivolts DC. :)

The slippery slope, though, is having equipment around not suitable for it's purpose, and people don't know it. I have a multimeter sitting there with a calibration sticker. Someone else may use it for other purposes, assuming that it is fine.

1. Clearly mark on the meter what it is calibrated for.
2. If it is a limited calibration, I would suggest using a different colored sticker, so that people realize they need to see what it is calibrated for. Then mark very clearly it's calibrated range, or mark what is not calibrated; depending on how much there is to write.
3. Try to segregate limited calibration equipment as much as possible. Even when it is marked, if it's sitting out in general area for use, someone may use it.

But I don't see any reason to dispose of a measuring instrument, if it serves your intended purpose. Just make sure that no one else will use it outside that limited range.:)
 

Jerry Eldred

Forum Moderator
Super Moderator
If I had my "Druthers", I'druther not have Limited Calibrations exist. As I mentioned that FDA regulated facility I worked for a few years did not allow them ever. If an instrument would not pass full calibration, it was disposed of. It was the most hard nosed policy I've worked with, but in retrospect, I can well understand it. The danger in limited calibrations is, "how do you assure the instrument can not be erroneously used to it's full specs when it actually has not been calibrated to those full specs?"

I think there are times when it is permissible, and I have customers who decide that is what they want. But, as Brad detailed very well, if you do use this compromised method to "certify" and instrument, you need to tread carefully. It is as though figuratively (and pretty much literally, as well), you are only certifying part of the instrument, even though you give the whole instrument back to the user; with a tag that says not to use the whole instrument.

It's Saturday night, so time to go do something less intelligent.
 

howste

Thaumaturge
Trusted Information Resource
There have been times that I didn't give the whole instrument back. I remember several times where I ground off the inside jaws and/or depth rod from a pair of calipers because they were damaged. The limited calibration was pretty easy then - they were only capable of outside measurements.
 

Jerry Eldred

Forum Moderator
Super Moderator
Good point. I hadn't heard of that example. Kind of made me chuckle at my response. I suppose if I were writing a quality procedure on the topic, I might include "IF" I were to allow Limited Calibrations, the caviat that where ever possible, the limited aspect of the instrument should be disabled.

As someone with mostly an electrical background, quite often the limitation would be in an aspect of an instrument that could not be isolated, disabled or removed. A fictitious example, perhaps an Oscilloscope with a 500 MHz bandwidth would only work to 400 MHz on channel 4. The user never uses it at that frequency, but it doesn't make sense to replace, and the instrument is not supported by the OEM, so they request a limitation of "CHANNEL 4 BANDWIDTH ONLY TESTED TO 400 MHZ."

As I've said, it is something that has its place in some areas of the measuring world. But it must be treated carefully.
 
M

mecramur

I like the thoughts and input.

I understand a client not wanting to replace an expensive ITME that is capable of measuring some things but not others. Managing expenses is important and not utilizing what is available is wasteful. I really do not like to see a good piece of test equipment go to waste.

Yet on the other hand managing a ITME to ensure adherence of limited usage with a limited calibration is nearly impossible short of removing the OOT function. It is nearly impossible to control how or who uses an ITME is used once it leaves the calibration service.

If the OOT function can be removed or completely disabled then I would feel comfortable with a limited calibration.

A FDA regulated business is critical but so would be aerospace such as avionics, I would not want to use a bad meter to adjust equipment that keeps an aircraft in flight. Nor would I want the possiblity of medical instruments being manufactured with OOT equipment. Risk is high..

Our SOP has no provision for or against limited calibrations. The client has submitted an NCMR for review to accept a limited calibration for the Oscilloscope. Perhaps if we remove the BNC connection for channel 4 so it cannot be used then it is acceptable. If not I am so uncomfortable with the limited cal.

Our SOP needs to have some clear definitions on limited calibrations.

By the way, I am in an FDA regulated business ;)
 
S

step30044

Within our systems items that are limited calibration or assigned a bright yellow sticker with the limitations listed on the sticker. A shoe tag is then attached to the unit that again states the limitation and the end user who authorized the limitation.

Before a limited calibration is approved the end user must fill out and return a deviation sheet to the lab that is placed on file.
 
Top Bottom