Calibration Recall

[Wilber100]

This is a question that comes up periodically and the resulting answer (from QA)never seems to be consistent.

If customer equipment calibrated at a laboratory (and now in use by the customer) is recalled due to an out of tolerance lab standard which was used to calibrate that customer equipment, is a full re-verification of the customer equipment required or is a re-verification of only the affected parameter/test point sufficient?

An oversimplified example is this: 34401A is calibrated by 5720A. At its next calibration, the 5720A is found to be OOT enough at 100 mVdc to deem the 34401A calibration suspect and consequently be recalled.
A few questions. 1-Does a FULL re-verification need to be performed on the DMM or is a single test (or at most only the affected range, or perhaps the dcV function only) at the suspect point sufficient? 2-In this example, the DMM has been in use for only one month of its calibration cycle. If it is sufficient that if a full re-verification is NOT required, then the unit would be returned with the original cal label and due date (I think).  How does this sound?

Thanks for any comments.

[Hawaii596]

My understanding is that you only need to cover the impacted measurements.  Even further, if you take data on the 34401A, you can reverse engineer the out of tolerance value and how far off it made the measurand on the 34401A.

[RFCAL]

And how do you know when the 5720A went OOT? What was the relation to the time left on the calibration of the 5720A when the 34410A was calibrated. You will have a greater chance the 34410A was impacted if there was only 1 month left on the 5720A cal rather than 8 months left.

[USMCPMEL]

Don't you just have to assume it has been out for that time period and recheck the instruments anyway? There really is no way to tell when the instrument went out of tolerance.

[RFCAL]

That is why you can calculate the OOT probability between the cal cycle left and the time the UUT was calibrated.

[ck454ss]

When a standard was found to be OOT to the point a recall was needed we perform a full calibration.  Its not that it is really needed its that it gives our customer a "Warm Fuzzy" that their equipment is working properly.  And personally its the least I can do for making them recal their equipment and have possible unscheduled downtime on their end. 

[sdmetrol]

^
l
l
What he said.