Uncertainties indicated with an asterisk are traceable, but not UKAS Accredited

[AlbertPaglinawan]

Hi again!

I have here a calibration certificate from Fluke that says

The measurements were made using an automated calibration system. The relevant data concerning environmental conditions, function, range and uncertainty for each point are contained in the following pages.
Uncertainties indicated with an asterisk are traceable, but not UKAS Accredited. Calibrations marked not UKAS Accredited on this certificate have been included for completeness.

What does it mean traceable but not UKAS accredited? is my equipment in specifications or what?

again, any help would be greatly appreciated.

thanks in advance.

It also says that:

UKAS is one of the signatories to the Multilateral Agreement of the European co-operation for Accreditation (EA) for the mutual recognition of calibration certificates issued by the accredited laboratories.

I know that most of you already know what UKAS is but I included this to enlighten the few who doesn't know what it is. Honestly, I am one of those few.

Oh, and while I'm at it, let us further dissect the statements in my calibration certificate. It says:

1) The instrument has been calibrated in accordance with the manufacturer's instrument user's handbook using standards that are directly traceable to or derived from the National Standards maintained at the National Physical Laboratory, Teddington.
-ok, simple enough

2)The reported uncertainty is based on a standard uncertainty multiplied by a coverage factor k=2, providing a level of confidence approximately 95%. The uncertainty evaluation has been carried out in accordance with UKAS requirements and is inclusive of th UUT.
-now, I may sound stupid, but for the love of Light, what does this mean? Is the uncertainty included in the cal cert my equipment's uncertainty? or that of UKAS? or Fluke's standards?

cheers!

 

[Ryan Wilde]

Re: UKAS Certified

Lord Ituralde said:

Hi again!

I have here a calibration certificate from Fluke that says

<i>The measurements were made using an automated calibration system. The relevant data concerning environmental conditions, function, range and uncertainty for each point are contained in the following pages.
<b>Uncertainties indicated with an asterisk are traceable, but not UKAS Accredited</b>. Calibrations marked not UKAS Accredited on this certificate have been included for completeness.</i>

What does it mean traceable but not UKAS accredited? is my equipment in specifications or what?

again, any help would be greatly appreciated.

thanks in advance.

What the statement means is that SOME of the measurements are accredited, while others are not. Those marked with an askterisk are functions/ranges NOT covered under Fluke's current <i>Scope of Accreditation</i>. This means that the equipment was calibrated using a method that has not been assessed by UKAS. So you have a piece calibrated, with most parts of the calibration accredited, although one or more parameters aren't covered under the umbrella of accreditation.

In other words, yes, your equipment is in specifications, although not all parts of the calibration have the UKAS seal of approval. Since it was Fluke, I think I'd take them at their word. The only way around this would probably be to send the item to your national lab, and if your government lab is like NIST in the USA, you will not see your equipment for a few months and you have just financed a third world country for another 6 months.

Oh, and while I'm at it, let us further dissect the statements in my calibration certificate. It says:

2)<i>The reported uncertainty is based on a standard uncertainty multiplied by a coverage factor k=2, providing a level of confidence approximately 95%. The uncertainty evaluation has been carried out in accordance with UKAS requirements and is inclusive of th UUT.</i>
-now, I may sound stupid, but for the love of Light, what does this mean? Is the uncertainty included in the cal cert my equipment's uncertainty? or that of UKAS? or Fluke's standards?

The statement means that with a 95% probability, the actual calibration of your equipment met that particular uncertainty. It means that they not only took into account their standard uncertainties, environmental uncertainties, etc., but they also took into account the uncertainties introduced by your particular piece of equipment (or at a minimum, your family of equipment).

The uncertainties reported on your certificate are those that you would use as a component in your own uncertainty calculations. It does not include equipment accuracy, only the uncertainty of each test point. If this is the Datron 1281 that we discussed in another thread, your actual uncertainty (plus environmental and lead effects) would be the stated uncertainty from Fluke/Wavetek/Datron (pick one) on their specification sheet added RSS (Root Sum of Squares) to the uncertainty shown on your certificate.

Hope this clears up a few things as I muddy the waters...

Ryan

 

[Jerry Eldred]

I think Ryan pretty well covered it.

The only further comment I would make is that (probably a little redundant, but I have to create a logical thought process) traceability is in simple terms the agreement of a measurement with a nationally and/or internationally accepted nominal to some defined uncertainty. THAT is where the accuracy of the measurement is derived.

For financial reasons, it is common to be accredited for only those measurands the a particular calibration lab feels would benefit them. Lack of accreditation in a certain parameter does not in any way imply that the measurement is less accurate. For labs with integrity, it only implies that they did not choose to spend the money to have that particular measurement accredited. The unaccredited measurement in probability is every bit as accurate as the accredited measurement.

I work with a particular third party lab who has been accredited by either A2LA or NVLAP. They selected which measurands they wanted accredited. This did not include everything they do. For the sake of illustration, I'll fabricate some example numbers. They do 115 different types of measurements. They are accredited for 35 of them (these are made up numbers). The 35 measurement types represent 90% of their business in dollars. It was not cost effective for them to accredit the remaining 80 measurements.

However, those additional un-accredited measurements are probably no more or less dependable or traceable than the accredited ones.

Accreditation does not make any measurement better. Measurement accuracy/uncertainty is a factor of all those variables that go into performing the calibration (accuracy of measurement standards, environmental controls, training, good procedures and following them, and so forth).

As I need to use only accredited third party labs, I may not be able to use them for this particular measurement type. But not because I don't trust them, because of paperwork requirements.

I'll be going to 2PCP training sometime soon, so I may audit them for those unaccredited parameters later this year.

My bottom line point is:

Traceability is important to "good" calibration.
Accreditation is important to satisfying quality system requirements (such as QS9000, etc.).

 

[AlbertPaglinawan]

<i>Darn Scope of Accreditation!</i>. This just blows the bananas off my top shelf!

I requested a copy of Fluke's Scope of Accreditation so i can at least have a geist of what it looks like.

That's the problem with calibration. It seems like the metrologist speaks different language. I mean, what in the name of Charlie's Rotten Gums is Scope of Accreditation?, ok, I know some, but i guess i'll never know it really until I get an example. Maybe even then I would still be confused.

That's why I can't reply to the earlier thread, I promised to hold my peace, or my questions really, until I have an actual scope of accreditation.

Thanks for the info though. And mostly, thanks for the patience.

I hope, for the love of Chip Chippewanggo, that I would have a clearer view of metrology. I thought I know enough but I was mistaken. (but that's just me rambling )

Thanks again! I bow before all of you!

cheers!

oh, btw, the equipment is Wavetek 9100 multifunction calibrator. thanks again!

 

[Ryan Wilde]

Lord Ituralde said:

<i>Darn Scope of Accreditation!</i>. This just blows the bananas off my top shelf!

I requested a copy of Fluke's Scope of Accreditation so i can at least have a geist of what it looks like.

That's the problem with calibration. It seems like the metrologist speaks different language. I mean, what in the name of Charlie's Rotten Gums is Scope of Accreditation?, ok, I know some, but i guess i'll never know it really until I get an example. Maybe even then I would still be confused.

Here is the Fluke (Feltham, Middlesex) Scope of Accreditation http://www.ukas.org/calibration/schedules/actual/0183C0_19.pdf . Eighteen pages of pure metrology joy.

I will add that I looked through it quickly, and if you have any questions, post them and one of us will happily offer a bit more clarity in non-metrology lingo.

Ryan

P.S. It should also be noted that in the UK it seems that "Scope of Accreditation" is known as "Schedule of Accreditation". I just learned this today, so I apologize for the language barrier.

 

[AlbertPaglinawan]

Thanks!

From my understanding, the <i>Scope of accreditaion</i> is a compilation of all of your standards' tespoints and accuracy.

Say if my standard is calibrated in 10V and the next testpoint was 20V, I cannot calibrate a UUT using a 15V-testpoint and still say that it was under our scope of accreditaion (if we ever make one).

The question is: <i>did i get it right?</i>

<font face="comic sans ms" size="+2" color="red">Thanks again.</font>

cheers!

 

[Ryan Wilde]

Well, no, actually the Scope of Accreditation (Schedule of Accreditation in the UK) shows the "Best Measurement Uncertainty" that a lab can attain. The uncertainty is not generally for specific test points, but encompasses ranges. It also not just the uncertainty of your standards, it is the uncertainty of your process on a "nearly perfect device", and quite often in electrical, no device at all. I'll give you an example using your Datron 1281 from 1 to 10 VDC (I like the 1281 because with the selfcal I don't have to factor in temperature, which makes me happy).

Parameter:
VDC Measurement
Range: 1 to 10 VDC

Uncertainty components (completely contrived - do not use these as your measurements):

Error of Standard (1 Year - with Selfcal performed within ± 1°C): 6 PPM of reading + 0.5 PPM FS = (Conservatively) 65 µV as the worst case in the range. This would be a rectangular distribution, therefore in my budget I would use 46 µV


Uncertainty of calibration on 10 VDC scale on your standard(Again, contrived) with k=2: 3 PPM = 30 µV. Reducing to one standard deviation would give me 15 µV.

Uncertainty of accessories (Such as errors due to leads): Okay, this requires some experimentation to nail down, but let's say that you did and found that you have an 12 µV lead error uncertainty with your best condition, shortest, shielded leads, with one standard deviation.

Now we crunch them together using an RSS of the components:
SQRT(46 µV² + 15 µV² + 12 µV²) = 49.85 µVDC at one standard deviation. We always give Scope (Schedule) of Accreditation uncertainties at 2 standard deviations (k=2), therefore we would report (very simplified - I wouldn't combine the 6 PPMR with the 0.5 PPMFS normally):

Uncertainty of DCV Measurement in the range of 1 to 10 VDC: 99.7 µVDC.

That is what would be reported on the Scope (Schedule) of Accreditation - The uncertainty of your calibration system as a whole, generally broken down by ranges determined by the ranges of your standards.

In other words yes, you can calibrate the UUT at 15 V, because that RANGE is calibrated, and every point within that range is considered calibrated as well. More than likely, the test point at 20 V is in the same range, and the uncertainty of that range would apply.

Hope this helps,

Ryan

 

[AlbertPaglinawan]

Thanks.

It did help. I'm trying to make one just for the fun of it (if you can call it fun).

cheers!