posted 27 June 2000 09:14 AM               

-----Original Message-----
Date: Mon, 05 Jun 2000 15:37:25 -0700
From: Tim Holder
To: "[email protected]"
Subject: Z540, ISO 17025 and Calibration Data

I have been receiving an unusual number of request for calibration data with our calibrations. Some customers are saying that Z540 requires data with each calibration as a verification that a proper calibration has been performed. I have always understood that Z540 does not require data with calibrations. Has anyone else been experiencing an unusual amount of request for data? Also, will anyone verify that Z540 and ISO 17025 does not require data with each calibration?

-------------------------------------

Hi Mike and members,

I have tried to raise this issue several times in this forum but with little or no discussion.

Since it has been raised again by yourself, and in such an explicit and unambiguous way, I am hoping that it will be debated this time and therefore felt I would like to comment.

I spent 17 years managing an accredited calibration laboratory who were responsible for the calibration of electronic test and measuring equipment from the military. The experience I gained there was that the end user, very often a "computer operator" or "button pusher" doesn't have the faintest idea of what points on an instrument need to be calibrated and with what uncertainty.

This same scenario, I believe, applies pretty much in the broader industry as well (we also calibrated for commercial industry). I am therefore critical of those who believe that we will reach a stage where the majority of customers, (persons requesting calibration),will be in a position to enter into a meaningful technical "contract review" with the calibration laboratory.

Furthermore, I believe that in many instances, the calibration laboratory, based on years of experience calibrating instruments of similar design and from a handful of reputable suppliers, is in a better position to guide the customer to having a meaningful calibration performed than just relying on the customer.

As a technical assessor I have seen too many calibrations where the calibration laboratory merely performs a measurement and documents the result without doing any further investigation to determine whether or not the instrument is within any agreed upon limits, namely verification. This leads to large numbers of instruments in industry which are not fit for their intended measurement purpose.

Lastly, I agree with you that the best and safest default calibration action to be performed should be according to the manufacturer's performance verification procedure. This will at least restore the instrument to the accuracy conditions which applied when it was purchased. I believe that at least at purchase, some technical ly competent individual was involved in the selection based on accuracy requirements. Therefore calibration according to the manufacturer's performance verification would restore this situation by default.

Lastly, I feel so strongly about this that in my humble opinion, any calibration performed using any different procedure from the one recommended by the manufacturer should result in the label stating "Partial Calibration refer to certificate".

As for those instruments where no such procedure is described, well that is where technical guidance documents from the accreditors in the world play a role.

I'm afraid that the people designing these pie in the sky ideals of having all calibration requests being backed up by a technically competent customer must spend some time in the real world where millions of instruments are performing critical measurements and being sent for calibration by the company "buyer" who shops around for the cheapest price because his bonus depends on it.

Cheers
Eddie

Eddie Tarnow

dc Metrology Project
National Metrology Laboratory
CSIR
P.O.Box 395
Pretoria
0001
South Africa

Tel: +27 12 841-3138
Fax:+ 27 12 841-2131

E-mail: [email protected]

>>> Greg Gogates 00/06/13 03:36:53 >>>
Date: Mon, 12 Jun 2000 17:08:21 +0100
From: "HUTCHINS,MIKE (A-England,ex2)"
To: 'Greg Gogates'
Subject: Defining adequacy (was ...Calibration Data)

Tim Holder's question (below) produced several replies but none offered the possibility that more customers are now wanting the cal data in order to check, and hopefully confirm, that the extent of testing is adequate for their usage of the equipment. The standards all agree that the onus is on the instrument user. However, a majority of users consider that the scope of calibration and its adequacy to assess spec conformity is the responsibility of the calibrating lab. Moreover, many of those who seek an accredited calibration believe that the accreditation agency defines what constitutes "adequate" cal. In Britain at least, this is incorrect and highlights a need for official clarification of the common, though non-descript, order stipulation "Cal to spec".

UKAS believes the problem should be addressed by educating those placing calibration orders. A recommended solution is that they create a schedule for each item which outlines its usage and corresponding technical cal requirements. This approach may be feasible for scientists in a laboratory environment who know exactly what they want to learn from the calibration, but not for an inventory manager responsible for the maintenance of thousands of items throughout a multi-sited company. It's hardly necessary to say that the latter represent the major part of the cal business.

An alternative argument made by UKAS is that the enhanced requirements in ISO17025 for contract review (consistent with ISO9000 with which many labs have satisfactorily complied for years) will lead the cal lab to enter into a dialogue with the customer to establish exactly what's required and what can be provided. However, for the previously mentioned reasons, this ideal scenario isn't any more likely to happen in the future than now.

Where a manufacturer publishes recommendations regarding the extent of testing necessary for confidence in the product's overall performance, surely that ought to be considered the default (or minimum) calibration level? If the authorities responsible for policing calibration stipulated this, it would help equipment-users and reputable cal labs alike. It wouldn't preclude the possibility of agreement between customer and cal lab to provide a customised cal (i.e. only functions and ranges actually used) but would help in vendor selection if the fundamental aspect of the service was known to be consistent between potential suppliers. Of course, the actual deliverables could be different -- that's competition and part of meeting customers' particular needs -- but there would be comparable confidence in the equipment's overall performance capability.

A challenge to the proposal has been that some manufacturers have not published such advice. This could yet be countered since I understand that the French AFNOR (military) standards group have been developing guidance (which could become mandatory in France as COFRAC are also involved) for several hundred generic product types in all fields of measurement.

It would be useful to get feedback from the list members on experiences of cal adequacy and particularly whether there's support for the proposal that manufacturer's recommendations should be used by the accreditation agencies to establish the default calibration for particular models.