ISO/TS 16949 Compliant Calibration Method - Laboratory Calibration Levels

[Icy Mountain]

I have a quandary that I have never really considered until now. I am requesting quotes for suppling calibration services for my quality system. Only calibration labs that can demonstrate a scope that is accredited by A2LA to ISO/IEC 17025 for all my calibration requirements are being considered. One of the laboratories has quoted three levels of calibration:

Standard Calibration - is traceable to national standards, reports whether the unit is in or out of tolerance, and reports data only on out of tolerance points. This is accepted throughout industry and meets the needs of many customers.

Calibration with before and after data - This calibration is performed with full data regardless if the instrument was out of tolerance or not. This is a more expensive calibration as the recording of data takes longer. Most FDA and regulated type industries require this level of calibration where product safety is a high concern.

A2LA accredited calibration with full data and measurement uncertainty - This calibration is performed with full data and will also include the uncertainty of the measurements. This calibration is accredited to ISO/IEC 17025. This is a more expensive calibration as the recording of data and the calculations of uncertainty takes longer. The electronic components industry is often required to have this level of calibration.

I believe that the Standard Calibration will allow me to remain compliant with TS16949 as Section 7.6.2 requires ISO/TS 16949 requires calibration records that contain:
"any out-of-specification readings as received for calibration/verification,
statements of conformity to specification after calibration/verification,
an assessment of the impact of out-of-specification condition."
Section 7.6.3.2 requires "the laboratory shall be accredited to ISO/IEC 17025 or national equivalent."

I am I correct in my belief or have I missed a memo, guidance, or implied language somewhere?

 

[vandenbar]

Re: ISO/TS 16949 Compliant Calibration Method

Icy,

If you are using an external lab for your calibration (which from your post is the intention) then you will need option 3 form Joe's Calibration shack. The external lab needs to be accredited and the test/ calibration that they are doing for you needs to be on their scope of accreditation as well. The only exceptions are if the customer accepts a non accredited lab or if you can not find an accredited lab you can use teh equipment manufacturer.

That's my interpretation anyway.

 

[Stijloor]

Re: ISO/TS 16949 Compliant Calibration Method

I have a quandary that I have never really considered until now. I am requesting quotes for suppling calibration services for my quality system. Only calibration labs that can demonstrate a scope that is accredited by A2LA to ISO/IEC 17025 for all my calibration requirements are being considered. One of the laboratories has quoted three levels of calibration:I believe that the Standard Calibration will allow me to remain compliant with TS16949 as Section 7.6.2 requires ISO/TS 16949 requires calibration records that contain:
"any out-of-specification readings as received for calibration/verification,
statements of conformity to specification after calibration/verification,
an assessment of the impact of out-of-specification condition."
Section 7.6.3.2 requires "the laboratory shall be accredited to ISO/IEC 17025 or national equivalent."

I am I correct in my belief or have I missed a memo, guidance, or implied language somewhere?

Hello Icy,

Quote: "an assessment of the impact of out-of-specification condition."

I don't think that Joe's Calibration Shack can provide you with that information.....

Stijloor.

 

[AndyN]

Re: ISO/TS 16949 Compliant Calibration Method

Icy:
To run an effective calibration system, you will need before and after data. That will help you decide when the equipment will need to be calibrated again. Furthermore, for you to be able to conduct any internal calibration or to perform an MSA study you will need the uncertainty statements from the cal. lab.

It's always tempting to go for the least expensive option, but you get what you pay for and in this case it's not much with the standard option........

 

[Icy Mountain]

Re: ISO/TS 16949 Compliant Calibration Method

The external lab needs to be accredited and the test/ calibration that they are doing for you needs to be on their scope of accreditation as well.

The external lab is accredited and the test/calibrations they are doing are within their scope. Do I really need all that data? The calibration method, equipment, technician, methods, etc. don't change.

 

[Icy Mountain]

Re: ISO/TS 16949 Compliant Calibration Method

Quote: "an assessment of the impact of out-of-specification condition."I don't think that Joe's Calibration Shack can provide you with that information.....

Yes, that's true. However, I can make that assessment from the data provided by their Standard Calibration, yes?

 

[Icy Mountain]

Re: ISO/TS 16949 Compliant Calibration Method

Icy:
To run an effective calibration system, you will need before and after data. That will help you decide when the equipment will need to be calibrated again. Furthermore, for you to be able to conduct any internal calibration or to perform an MSA study you will need the uncertainty statements from the cal. lab.

We perform testing but do not provide calibration or calibration certificates so I don't need uncertainties for that. As for MSA studies, I'm in a rather unusual situation in that I have had the requirement waived by my customer for every product that falls under the TS scope. Our volumes with these products are too low to generate any meaningful data. In addition, 120 VAC outputs are highly dependent, in both voltage and current, on how long that the unit has been providing current, amount of load, etc....they're sloppy. You know, 220, 221, whatever it takes.

 

[AndyN]

Re: ISO/TS 16949 Compliant Calibration Method

We perform testing but do not provide calibration or calibration certificates so I don't need uncertainties for that. As for MSA studies, I'm in a rather unusual situation in that I have had the requirement waived by my customer for every product that falls under the TS scope. Our volumes with these products are too low to generate any meaningful data. In addition, 120 VAC outputs are highly dependent, in both voltage and current, on how long that the unit has been providing current, amount of load, etc....they're sloppy. You know, 220, 221, whatever it takes.

Well, that helps - but not much! Does your customer really know what they're doing by waiving the requirements. It might be nice of them to do that, but I've also seen customers who sign off on all kinds of FMEA's, CP's etc that were wrong too.......

I'd also suggest that since you're testing you do need the uncertainties to know what contribution they make to the testing uncertainties.......if you want to run effectively. IMHO this is the area that customers often start to get lost in their understanding, but when things go wrong they'll be all over your case.......

And I stand by the fact that to run an effective calibration system you must use the cal. data to drive the recall date. Just saying do it once or twice a year isn't effective.....

 

[Icy Mountain]

Re: ISO/TS 16949 Compliant Calibration Method

Andy,
You are probably right about customer knowledge. However, we are working with state-of-the-art meters with published accuracies of 0.5% to test AC outputs with published specifications of 5%. The lab gives me an uncertainty ratio of 4:1 or better on every certificate; I just don't get the exact number.

The biggest problem here is that I am not a Calibration Expert, thus the reason for employing an external laboratory, among others. I am just now getting enough history with our calibration database to even consider changing calibration intervals. Prior to my taking over the calibration records, the only data we had were a bunch of paper certs. Try efficiently arriving at any conclusions from that.

 

[Stijloor]

Re: ISO/TS 16949 Compliant Calibration Method

Yes, that's true. However, I can make that assessment from the data provided by their Standard Calibration, yes?

Icy,

Please keep in mind that this assessment is affected by your gage calibration frequency. A lot can happen between two calibration intervals. Unless you can verify the gage in-between..... (I don't want to get into the definition of "verification and calibration" here. We've had some very interesting conversations about this on The Cove recently.....)

Stijloor.