What really gets calibrated?
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it
gets "adjusted", not "calibrated".
I can see that on the part of the calibration lab, it is easier and faster
to simply check that a device meets spec and not have to spend time
adjusting anything. It should also be less expensive for the customer.
My question is, is just meeting spec good enough? If an instrument is
capable of exceeding spec, shouldn't it be adjusted to the best standard
available? In other words, if spec says 2 ppm, but it can be adjusted to 1
ppm, wouldn't you want to do that?
Joe Gray
W5JG
Joe,
I agree with your analysis of 'calibration' and 'adjustment'. However, if a
piece of equipment was able to meet a higher specification, I would suspect
the manufacturer would have claimed that.
The benefit of the 'as found' and 'as left' measurements is that it tells
you not only whether the unit is 'in spec' but also where it is in the
range. Might be useful.
All of this, of course, is dependent of the 'quality' of the 'standard' that
is used to compare with. That is probably where the price goes up and the
'metrolophilosophy' comes in. In other words, do you trust your 'cal lab'?
Good luck.
Joe
-----Original Message-----
From: volt-nuts-bounces@febo.com [mailto:volt-nuts-bounces@febo.com] On
Behalf Of Joseph Gray
Sent: Sunday, July 28, 2013 5:16 PM
To: volt-nuts@febo.com
Subject: [volt-nuts] What really gets calibrated?
I recently had it explained to me that "calibration" is really just checking
a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it gets
"adjusted", not "calibrated".
I can see that on the part of the calibration lab, it is easier and faster
to simply check that a device meets spec and not have to spend time
adjusting anything. It should also be less expensive for the customer.
My question is, is just meeting spec good enough? If an instrument is
capable of exceeding spec, shouldn't it be adjusted to the best standard
available? In other words, if spec says 2 ppm, but it can be adjusted to 1
ppm, wouldn't you want to do that?
Joe Gray
W5JG
volt-nuts mailing list -- volt-nuts@febo.com To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Many manuals I've seen have Calibration and Adjustment sections.
In general, 'If it ain't broke, don't fix it'.
Many adjustments are done via pots that do not have infinite resolution,
and adjusting them may not really improve matters, and could make matters
worse or less stable.
My inclination has always been not to touch any adjustment that soesn't
need it.
YMMV,
-John
===============
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it
gets "adjusted", not "calibrated".
I can see that on the part of the calibration lab, it is easier and faster
to simply check that a device meets spec and not have to spend time
adjusting anything. It should also be less expensive for the customer.
My question is, is just meeting spec good enough? If an instrument is
capable of exceeding spec, shouldn't it be adjusted to the best standard
available? In other words, if spec says 2 ppm, but it can be adjusted to 1
ppm, wouldn't you want to do that?
Joe Gray
W5JG
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
It is not about cost but about building a history. If you send in a meter and they adjust it, send it back you will not know how much it is off. And also not if some projects, adjustments you made before that are done right. Also you do not know the amount of drift. It could be 10% off or only 0.1 ppm.
But if you get it back and know 10V is 10,000.005 V and after the next session it is 10,000,004 and a year later 10,000.003 you can predict better what it will be half way in between. And you know it is very stable. But if you get it back every year at 10,000.000V and it was 10,000.010 the first year, 9,999.993 the next time and 10,000.040 the last time and do not get those values you think you have a very good meter but no history and no confidence.
And if they adjust it, will the deviation be caused by something like mechanical stress on the trimpot or is the reference changing ?
And if they calibrate and adjust it at 25 degrees , 1000 mbar and 70 % humidity. What will it be at 22 degrees, 1200 mbar and 50 % rH.
And that is the things of Voltnutting that can drivecypu nuts ;-)
Fred PA4TIM
Op 29 jul. 2013 om 00:15 heeft Joseph Gray jgray@zianet.com het volgende geschreven:
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it
gets "adjusted", not "calibrated".
I can see that on the part of the calibration lab, it is easier and faster
to simply check that a device meets spec and not have to spend time
adjusting anything. It should also be less expensive for the customer.
My question is, is just meeting spec good enough? If an instrument is
capable of exceeding spec, shouldn't it be adjusted to the best standard
available? In other words, if spec says 2 ppm, but it can be adjusted to 1
ppm, wouldn't you want to do that?
Joe Gray
W5JG
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Hi guys,
The "as found" measurements are most likely useful to determine if gross
errors were made while the instrument was in use. Say you were using it to
calibrate DMM multimeters in a production environment, you would like to
know that a batch of meters went out with defective calibrations.
If you are getting an instrument calibrated for the first time after
replacing a NVRAM battery, there would not be a need to have that known bad
data recorded.
Many of the newer instruments require entering a known voltage and telling
the instrument to record a calibration constant in its NVRAM.
I think that is mostly correct but if not, please feel free to criticize it.
Regards,
Tom
----- Original Message -----
From: "J. L. Trantham" jltran@att.net
To: "'Discussion of precise voltage measurement'" volt-nuts@febo.com
Sent: Sunday, July 28, 2013 6:42 PM
Subject: Re: [volt-nuts] What really gets calibrated?
Joe,
I agree with your analysis of 'calibration' and 'adjustment'. However, if
a
piece of equipment was able to meet a higher specification, I would
suspect
the manufacturer would have claimed that.
The benefit of the 'as found' and 'as left' measurements is that it tells
you not only whether the unit is 'in spec' but also where it is in the
range. Might be useful.
All of this, of course, is dependent of the 'quality' of the 'standard'
that
is used to compare with. That is probably where the price goes up and the
'metrolophilosophy' comes in. In other words, do you trust your 'cal
lab'?
Good luck.
Joe
-----Original Message-----
From: volt-nuts-bounces@febo.com [mailto:volt-nuts-bounces@febo.com] On
Behalf Of Joseph Gray
Sent: Sunday, July 28, 2013 5:16 PM
To: volt-nuts@febo.com
Subject: [volt-nuts] What really gets calibrated?
I recently had it explained to me that "calibration" is really just
checking
a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it
gets
"adjusted", not "calibrated".
I can see that on the part of the calibration lab, it is easier and faster
to simply check that a device meets spec and not have to spend time
adjusting anything. It should also be less expensive for the customer.
My question is, is just meeting spec good enough? If an instrument is
capable of exceeding spec, shouldn't it be adjusted to the best standard
available? In other words, if spec says 2 ppm, but it can be adjusted to 1
ppm, wouldn't you want to do that?
Joe Gray
W5JG
volt-nuts mailing list -- volt-nuts@febo.com To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
On 28 July 2013 23:15, Joseph Gray jgray@zianet.com wrote:
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it
gets "adjusted", not "calibrated".
I think it depends somewhat on who cals it, and under what conditions
are stated.
I think if something is specified with an uncertainty of 10%, and is
found to be 9.9% off, then I feel it should be adjusted as it is close
to the limits of being acceptable, even though it is within spec.
Certainly with Agilent, you will pay more for the cal if you want both
"before" and "after" data.
Dave
On Mon, Jul 29, 2013 at 12:25 AM, David Kirkby david.kirkby@onetel.netwrote:
On 28 July 2013 23:15, Joseph Gray jgray@zianet.com wrote:
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it meets the
manufacturers specification. If it is outside specified values, then it
gets "adjusted", not "calibrated".
I think it depends somewhat on who cals it, and under what conditions
are stated.
I think if something is specified with an uncertainty of 10%, and is
found to be 9.9% off, then I feel it should be adjusted as it is close
to the limits of being acceptable, even though it is within spec.
Certainly with Agilent, you will pay more for the cal if you want both
"before" and "after" data.
Here is what Agilent say about the new 3446xA DMMS:
"Adjustment is Recommended
Whatever calibration interval you select, Agilent Technologies recommends
that complete re-adjustment should
always be performed at the calibration interval. This ensures that the
instrument will remain within specifications for
the next calibration interval and provides the best long-term stability.
Performance data measured during performance
verification tests does not mean that the instrument will remain within
these limits unless the adjustments are
performed. Use the calibration count to verify that all adjustments have
been performed. "
This is however, a closed box calibration where the calibration process
stores correction constants in non-volatile memory. There are no concerns
about flaky trimpots etc..
Orin.
The big difference is that nowadays, adjustment is done in software without moving parts. There is no wear out involved in doing frequent adjustment and through the software you can record by how much each setting was adjusted.
These were the main reasons for not adjusting an instrument that was in spec when going to the cal lab.
Didier KO4BB
Orin Eman orin.eman@gmail.com wrote:
On Mon, Jul 29, 2013 at 12:25 AM, David Kirkby
david.kirkby@onetel.netwrote:
On 28 July 2013 23:15, Joseph Gray jgray@zianet.com wrote:
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it
meets the
manufacturers specification. If it is outside specified values,
then it
gets "adjusted", not "calibrated".
I think it depends somewhat on who cals it, and under what conditions
are stated.
I think if something is specified with an uncertainty of 10%, and is
found to be 9.9% off, then I feel it should be adjusted as it is
close
to the limits of being acceptable, even though it is within spec.
Certainly with Agilent, you will pay more for the cal if you want
both
"before" and "after" data.
Here is what Agilent say about the new 3446xA DMMS:
"Adjustment is Recommended
Whatever calibration interval you select, Agilent Technologies
recommends
that complete re-adjustment should
always be performed at the calibration interval. This ensures that the
instrument will remain within specifications for
the next calibration interval and provides the best long-term
stability.
Performance data measured during performance
verification tests does not mean that the instrument will remain within
these limits unless the adjustments are
performed. Use the calibration count to verify that all adjustments
have
been performed. "
This is however, a closed box calibration where the calibration process
stores correction constants in non-volatile memory. There are no
concerns
about flaky trimpots etc..
Orin.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
--
Sent from my Motorola Droid Razr 4G LTE wireless tracker while I do other things.
For those who have not heard of it, Fluke's "Calibration: Philosophy in Practice" is an excellent reference and quite readable. I got a copy as a holiday gift last year.
Robert G8RPI.
From: Didier Juges shalimr9@gmail.com
To: Discussion of precise voltage measurement volt-nuts@febo.com
Sent: Monday, 29 July 2013, 18:56
Subject: Re: [volt-nuts] What really gets calibrated?
The big difference is that nowadays, adjustment is done in software without moving parts. There is no wear out involved in doing frequent adjustment and through the software you can record by how much each setting was adjusted.
These were the main reasons for not adjusting an instrument that was in spec when going to the cal lab.
Didier KO4BB
Orin Eman orin.eman@gmail.com wrote:
On Mon, Jul 29, 2013 at 12:25 AM, David Kirkby
david.kirkby@onetel.netwrote:
On 28 July 2013 23:15, Joseph Gray jgray@zianet.com wrote:
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it
meets the
manufacturers specification. If it is outside specified values,
then it
gets "adjusted", not "calibrated".
I think it depends somewhat on who cals it, and under what conditions
are stated.
I think if something is specified with an uncertainty of 10%, and is
found to be 9.9% off, then I feel it should be adjusted as it is
close
to the limits of being acceptable, even though it is within spec.
Certainly with Agilent, you will pay more for the cal if you want
both
"before" and "after" data.
Here is what Agilent say about the new 3446xA DMMS:
"Adjustment is Recommended
Whatever calibration interval you select, Agilent Technologies
recommends
that complete re-adjustment should
always be performed at the calibration interval. This ensures that the
instrument will remain within specifications for
the next calibration interval and provides the best long-term
stability.
Performance data measured during performance
verification tests does not mean that the instrument will remain within
these limits unless the adjustments are
performed. Use the calibration count to verify that all adjustments
have
been performed. "
This is however, a closed box calibration where the calibration process
stores correction constants in non-volatile memory. There are no
concerns
about flaky trimpots etc..
Orin.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
--
Sent from my Motorola Droid Razr 4G LTE wireless tracker while I do other things.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
I too, highly recommend this book. Once I started reading it I couldn't
stop, and read it from cover to cover probably twice.
Regards,
Mitch
----- Original Message -----
From: "Robert Atkinson" robert8rpi@yahoo.co.uk
To: "Discussion of precise voltage measurement" volt-nuts@febo.com
Sent: Monday, July 29, 2013 2:28 PM
Subject: Re: [volt-nuts] What really gets calibrated? A reference
For those who have not heard of it, Fluke's "Calibration: Philosophy in
Practice" is an excellent reference and quite readable. I got a copy as a
holiday gift last year.
Robert G8RPI.
From: Didier Juges shalimr9@gmail.com
To: Discussion of precise voltage measurement volt-nuts@febo.com
Sent: Monday, 29 July 2013, 18:56
Subject: Re: [volt-nuts] What really gets calibrated?
The big difference is that nowadays, adjustment is done in software without
moving parts. There is no wear out involved in doing frequent adjustment and
through the software you can record by how much each setting was adjusted.
These were the main reasons for not adjusting an instrument that was in spec
when going to the cal lab.
Didier KO4BB
Orin Eman orin.eman@gmail.com wrote:
On Mon, Jul 29, 2013 at 12:25 AM, David Kirkby
david.kirkby@onetel.netwrote:
On 28 July 2013 23:15, Joseph Gray jgray@zianet.com wrote:
I recently had it explained to me that "calibration" is really just
checking a piece of gear against a known standard, to see if it
meets the
manufacturers specification. If it is outside specified values,
then it
gets "adjusted", not "calibrated".
I think it depends somewhat on who cals it, and under what conditions
are stated.
I think if something is specified with an uncertainty of 10%, and is
found to be 9.9% off, then I feel it should be adjusted as it is
close
to the limits of being acceptable, even though it is within spec.
Certainly with Agilent, you will pay more for the cal if you want
both
"before" and "after" data.
Here is what Agilent say about the new 3446xA DMMS:
"Adjustment is Recommended
Whatever calibration interval you select, Agilent Technologies
recommends
that complete re-adjustment should
always be performed at the calibration interval. This ensures that the
instrument will remain within specifications for
the next calibration interval and provides the best long-term
stability.
Performance data measured during performance
verification tests does not mean that the instrument will remain within
these limits unless the adjustments are
performed. Use the calibration count to verify that all adjustments
have
been performed. "
This is however, a closed box calibration where the calibration process
stores correction constants in non-volatile memory. There are no
concerns
about flaky trimpots etc..
Orin.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
--
Sent from my Motorola Droid Razr 4G LTE wireless tracker while I do other
things.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.