An ISO17025 Question About Derivative Measurements from CMC

Started by Hawaii596, 10-09-2013 -- 16:58:40

Previous topic - Next topic
Print
Go Down Pages1 2
User actions

Hawaii596

10-09-2013 -- 16:58:40
Okay, Lets say I am accredited to generate (Fluke 5720A) and measure (Fluke 8508A) (for example) 1000 VDC and to measure current (Keithley 263) to as low as about 1 pA (picoamp). But, for example, I am accredited on my CMC to measure up to 2 Giga Ohms.

But, even though I am accredited to 2 Giga Ohms, I can source 1000 ACCREDITED DC Volts across an unknown resistance, and measure the ACCREDITED DC current to as low as 1 picoamp. This combination theoretically results in 10E15 Ohms (1 Peta Ohm = 1,000 Tera Ohms = 1 Million Giga Ohms). There are of course all of the uncertainty contributors of repeatability, drift, resolution, uncertainties of calibrations, etc.  I know how to deal with them.

So although the cumulative uncertainty of a given established resistance would have lots of contributions, would it nevertheless be accreditable? This would be as a derivative measurement to my CMC? Or even if I am able to make the measurement and subsequent derived (possibly) accredited resistance value, would I be limited to CMC, even though my CMC allows me to directly accredit the input parameters (i.e.: voltage source and current measurement).

I could then (if this is accreditable) accredit these values, and make accredited measurements with them which would not exceed my CMC. Or is this the sort of question I should ask my accrediting body?

This seems logically that if I am accredited for the volts source and current measurement to make a resistance measurement that way, that I should be able to RSS together the factors along with the usual suspect contributors to uncertainty, and derive an accredited resistance value (in this case, a Penn Airborne).  And by the way, does anyone know the max voltage specs on the Penn Airborne resistors?  I have had poor luck trying to find them.
"I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind."
Lord Kelvin (1824-1907)
from lecture to the Institute of Civil Engineers, 3 May 1883

Bryan

#1
10-09-2013 -- 17:05:41
I am curious about that myself, my thoughts are where international agencies are involved creativity & innovation are not welcomed.

RFCAL

#2
10-09-2013 -- 17:24:21
To stay out of trouble, you should need to ask your accrediting body for guidance. Remember, you are accredited to 1000V with your 5720A and 1pA with the Keithley 263, but that probably does not equate to being accredited with the Math. That's why you play it safe and get your accreditating body involved.

USMC kalibrater

#3
10-10-2013 -- 05:23:14
You can absolutely do this.  You should state how the value was derived(which is redundant being that measuring resistance with a meter is doing basically the same thing), you obviously need to combine uncertainties and there is some additional error to consider.
If you think about Ohms is a derived SI unit anyhow so the math portion is inherent.
Jason
"Be polite, be professional, but have a plan to kill everybody you meet." -General James Mattis

Bryan

#4
10-10-2013 -- 11:32:45
I'd think if the result was tagged as calculated or derived from some recognized formula you'd be in good ethical standing, just my personal opinion

Hawaii596

#5
10-10-2013 -- 12:02:43
I had a meeting with our quality manager this morning and the position for the moment is that we need to ask our accreditation body for their blessing.  I am with you, as I believe it will probably be alright.  But that's the temporary verdict.  If I get a firm answer, I'll post it.  I think that's correct also.  But it would be nice to have an official position on it.
"I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind."
Lord Kelvin (1824-1907)
from lecture to the Institute of Civil Engineers, 3 May 1883

USMC kalibrater

#6
10-10-2013 -- 13:08:17
If an AB told me no on this I would challenge it.  If you were trying to calculate a base SI unit then I would say you are questionable but you are calulating a derived SI unit using a perfectly acceptable method. 
Get approval to save hassles later but no way would I concede to a no response.
Jason
"Be polite, be professional, but have a plan to kill everybody you meet." -General James Mattis

Hawaii596

#7
10-10-2013 -- 15:11:33
The AB responded that Yes, we can make the measurements as outlined.  They reminded us to keep in mind that you cannot report a measurement uncertainty better (lower) on our calibration certificate than the CMC on your scope for Resistance; and, the internal calibrations we perform are subject to witnessing during future assessment.
"I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind."
Lord Kelvin (1824-1907)
from lecture to the Institute of Civil Engineers, 3 May 1883

CalibratorJ

#8
10-11-2013 -- 07:14:23
Quote from: Hawaii596 on 10-10-2013 -- 15:11:33
The AB responded that Yes, we can make the measurements as outlined.  They reminded us to keep in mind that you cannot report a measurement uncertainty better (lower) on our calibration certificate than the CMC on your scope for Resistance; and, the internal calibrations we perform are subject to witnessing during future assessment.

So you can measure and accredit higher resistance values than you are scoped for? That's what I am interpreting this to mean. Or am I off?

Hawaii596

#9
10-11-2013 -- 09:37:19
This is why I brought it up with the AB.  The ambiguity is that if we are scoped for sourcing voltage and measuring current that mathematically equates to a resistance, so long as we properly document and account, that it is within our scope.  Our scope is for direct measurement of resistance up to 2 Giga ohms on the Fluke 8508A.  We are (according to the almost verbatim language I posted from the AB in my previous posting) accredited to source the voltage, measure the current and mathematically compute the resistance thereby measured.  So it is higher than what we are scoped to directly measure on our Fluke 8508A, but it is not higher than we are scoped for when measuring resistance by sourcing an accredited voltage and measuring the current (the exact same way the multimeter does it for you).  All the resistance function is on a multimeter is a machine automatically doing the math for you by sourcing a current and measuring a voltage.  The only difference theoretically in many high resistance meters is that they use a capacitive time constant method versus a voltage/current method.  So it is not higher than we are scoped for.  Again, because it is a little ambiguous, and requires understanding the theory of how resistance measurements are made, that also means that I felt the need to make sure the AB was in agreement before proceeding.

This is only strictly under circumstances where we are scoped for both the voltage and current involved in the measurement, that we account for all uncertainties properly, that it does not exceed our scoped CMC for either the voltage or current.
"I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind."
Lord Kelvin (1824-1907)
from lecture to the Institute of Civil Engineers, 3 May 1883

CalibratorJ

#10
10-13-2013 -- 08:01:01
Cool. I understand the theory behind the measurements. Just thought it was interesting that any AB would let you and actually tell you that you could accredit values that aren't specifically listed on your scope, derived or not. Make sure you have it in writing from your AB for audit time.

(I would go on a rant about ABs, but I will refrain, don't want to derail the train!)

RFCAL

#11
10-13-2013 -- 12:00:11
Ok, I'll start! Different AB's will let you do anything to collect your$$. That is a fact!! As in cars, there are different models and makes. Depends if you want to drive a Ford or a Caddilac. Same for AB's. Some will tell you you can do whatever it is that you want to do, others will say no,not correct. Remember, you only have to satisfy the AB that accredits you.

Hawaii596

#12
10-14-2013 -- 08:33:55
I'll stay out of the fray on this one.  I'm just trying to make sure and understand what is right.  Sounds like this is not a definitive already settled issue.  Guess I will have to read some more as well.  And good tip - my quality manager said about the same thing - to keep their response on file.
"I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind."
Lord Kelvin (1824-1907)
from lecture to the Institute of Civil Engineers, 3 May 1883

USMC kalibrater

#13
10-14-2013 -- 10:08:40
Quote from: CalibratorJ on 10-11-2013 -- 07:14:23
Quote from: Hawaii596 on 10-10-2013 -- 15:11:33
The AB responded that Yes, we can make the measurements as outlined.  They reminded us to keep in mind that you cannot report a measurement uncertainty better (lower) on our calibration certificate than the CMC on your scope for Resistance; and, the internal calibrations we perform are subject to witnessing during future assessment.

So you can measure and accredit higher resistance values than you are scoped for? That's what I am interpreting this to mean. Or am I off?

No, thats incorrect.  They are making, voltage and current measurements that they are accredited to do.  Then calculating the resistance.  Same way every other resistance measurement is made except instead of a ohmmeter doing the math, they are

Jason
"Be polite, be professional, but have a plan to kill everybody you meet." -General James Mattis

scottbp

#14
10-15-2013 -- 16:28:36
Okay, I got hammered on an audit for this: We did an accredited calibration for a customer on a power meter that measures watts at 60 Hz (not a watt-hour meter, just a direct reading watt meter). We used a Fluke 5500A to source voltage to the voltage input terminals, and current to the current input terminals. Our AB auditor said we were not supposed to give out an accredited calibration to the customer because we did not have power listed on our scope. I tried to explain that we have voltage and current listed on our scope, and we were mathematically calculating the power and the uncertainty of the voltage and current from the calibrator was well within our scope, and the combined uncertainty was well below the tolerance of the watt meter, but the auditor would not take that as an answer. So now I have to compile an uncertainty budget for power and submit it to the AB to get power put on our scope of accreditation.
Kirk: "Scotty you're confined to quarters." Scotty: "Thank you, Captain! Now I have a chance to catch up on my technical journals!"
Print
Go Up Pages1 2
User actions