Test Equipment Anonymous (TEA) group therapy thread

[beanflying]

Just for the records:

calibration is not adjusting.

Calibration means: comparing the DUT to a reference and checking, if it is within specs.
If it is within specs, the differences will be documented.

If the DUT is out of specs, then an adjustment or repair of the DUT can be considered.

So if you send a meter away for calibration, it will come back to you with the same settings as you sent it with? I was under the distinct impression that if the meter was in need of an adjustment, it would be made and then returned to you, with a demonstrable calibration. When I read the service manual for my BlackStar 4503, it gives precise instructions on how to check each range.  Then how to enter the calibration mode and make the required corrections to its parameters and saving those corrections to memory, which is what I did after repairing the damage done by the leaking battery.

Both of my 6 1/2 digits were pre tested and 'adjusted' then re tested and Calibration certificates issued. The 'adjustment' part is I understand onlt done if a meter doesn't meet spec.

The 34461A hadn't been sent for Calibration since new and the 34401A was 5 years out of Cal.

[BU508A]

Just for the records:

calibration is not adjusting.

Calibration means: comparing the DUT to a reference and checking, if it is within specs.
If it is within specs, the differences will be documented.

If the DUT is out of specs, then an adjustment or repair of the DUT can be considered.

So if you send a meter away for calibration, it will come back to you with the same settings as you sent it with?

Yes. The most valuable part of such a meter is its documented history, because based on this one can judge the drifting of a meter and its quality. It can be, that if a meter is drifting out of spec it will not be adjusted, because this means breaking the history.

I was under the distinct impression that if the meter was in need of an adjustment, it would be made and then returned to you, with a demonstrable calibration.

I've sent my 34401A to Keysight in Böblingen for calibration and they stated clearly, that they will document only.
If adjustment or repair is needed, they will contact me, asking, how to proceed. Not only because of the additional costs but
mainly because of the history of the instrument.

When I read the service manual for my BlackStar 4503, it gives precise instructions on how to check each range.  Then how to enter the calibration mode and make the required corrections to its parameters and saving those corrections to memory, which is what I did after repairing the damage done by the leaking battery.

I think, there are different rules for different manufactures. I think, Keysight and Fluke are handling it that way I've described above.

[Zucca]

calibration is not adjusting.

I did not know that, and BU is right:

Strictly speaking, the term "calibration" means just the act of comparison and does not include any subsequent adjustment.

In real life I think calibration means the process to adjust an instrument so it will work in the defined specifications... which means adjusting the reading as well if necessary.

Example: when I paid for the calibration of my good old 34401A, they give me a report of before and after the adjustment to make the instrument work in spec again.

[tautech]

Just for the records:

calibration is not adjusting.

Calibration means: comparing the DUT to a reference and checking, if it is within specs.
If it is within specs, the differences will be documented.

If the DUT is out of specs, then an adjustment or repair of the DUT can be considered.

So if you send a meter away for calibration, it will come back to you with the same settings as you sent it with? I was under the distinct impression that if the meter was in need of an adjustment, it would be made and then returned to you, with a demonstrable calibration. When I read the service manual for my BlackStar 4503, it gives precise instructions on how to check each range.  Then how to enter the calibration mode and make the required corrections to its parameters and saving those corrections to memory, which is what I did after repairing the damage done by the leaking battery.

The Cal lab chaps I deal with and have sold gear to initially verify an instrument meets manufacturer accuracy specs and then if necessary adjust it so that it does.
Therefore calibration can be one or two operations, verification of accuracy or verify any inaccuracies then perform adjustments.
A Cal cert should only be issued if an instrument meets manufacturers spec.

calibration is not adjusting.

I did not know that, and BU is right:

Strictly speaking, the term "calibration" means just the act of comparison and does not include any subsequent adjustment.

In real life I think calibration means the process to adjust an instrument so it will work in the defined specifications... which means adjusting the reading as well if necessary.

Example: when I paid for the calibration of my good old 34401A, they give me a report of before and after the adjustment to make the instrument work in spec again.

Cal labs won't perform adjustments unless they have been authorized to.

[Specmaster]

@BU508A, I can understand that any repairs needed that they would come back to you for guidance, but for adjustments NO, I would expect that would be all included in the calibration process. I can see zero use in having a certification that states on a piece of paper that says that on a certain range it reads xxx high or low. No one is going to take figure into account when using the meter, they will just go by the displayed reading, I would everytime.

[tautech]

@BU508A, I can understand that any repairs needed that they would come back to you for guidance, but for adjustments NO, I would expect that would be all included in the calibration process.

But it's not !
Adjustments are an additional cost and normally not performed without the owners authorization.

[BU508A]

@BU508A, I can understand that any repairs needed that they would come back to you for guidance, but for adjustments NO, I would expect that would be all included in the calibration process. I can see zero use in having a certification that states on a piece of paper that says that on a certain range it reads xxx high or low. No one is going to take figure into account when using the meter, they will just go by the displayed reading, I would everytime.

It may not be that important for a meter sitting on the bench of an EE but if you are using it in a certified production enviroment where even the cough of a fly has to be documented, then the history of this instrument becomes very important. The history is also important, if such an instrument is used as a transfer standard. Breaking the history of such an instrument is considered as a serious issue and has to be avoided if possible.

[Specmaster]

@BU508A, I can understand that any repairs needed that they would come back to you for guidance, but for adjustments NO, I would expect that would be all included in the calibration process.

But it's not !
Adjustments are an additional cost and normally not performed without the owners authorization.

When I sent my HP3478A away to RS labs for calibration because I lost the stored cal data when I replaced the backup battery, they returned it FOC because they were unable to get the meter to retain any corrections they giving it to correct readings. I then sat down and read and re-read the manual section on how to calibrate it did it all myself. With most modern bench meters, this adjustment can be carried out the front panel push buttons, there is not particularly onerous or time-consuming. After all you have give the meter the correct signals to determine if it meets specs or not, and if it doesn't, it is easy to enable cal mode for that failing range, punching in the correct value that the meter should be displaying and then saving it memory at the push of another button, disabling CAL mode by pressing a button or removing a key plug and checking the reading.

If the error is so great that its beyond the capabilities but the above method and requires component level changes, that is when I would expect them to come and say that it does not meet specs and requires extra work of X cost to enable the certificate to be issued stating that the meter meets its design brief and is within the spec.

[med6753]

To conclude my volt nut activities for now. The 3456A is my new "bench standard" as it has shown greater stability than the 8505A. I've written up a BOM to replace the PSU capacitors as well as ordered the adapter for the obsolete 5 pin C5. The 8505A has been taken off line for now but is my backup. I'm certainly not going to do something stupid like part it out or use it as a parts mule. Tautech sent me the procedure to perform a "manual cal" on the 3055. Into the project cue for now.

Volt nut hat off.  385 Larry awaits more work so back on that today. Besides, that Tek hat is a lot safer. I don't feel like I'm venturing into a minefield.       

[beanflying]

@BU508A, I can understand that any repairs needed that they would come back to you for guidance, but for adjustments NO, I would expect that would be all included in the calibration process. I can see zero use in having a certification that states on a piece of paper that says that on a certain range it reads xxx high or low. No one is going to take figure into account when using the meter, they will just go by the displayed reading, I would everytime.

It may not be that important for a meter sitting on the bench of an EE but if you are using it in a certified production enviroment where even the cough of a fly has to be documented, then the history of this instrument becomes very important. The history is also important, if such an instrument is used as a transfer standard. Breaking the history of such an instrument is considered as a serious issue and has to be avoided if possible.

Both of mine are due to head back for Recalibration in March. Based on their current behavior I doubt they will need a tweak as they stack up against each other well. This is more a confirmation Calibration now they have some history and consistent known use and I have a few newer toys to tweak when they get back and will go over the other toys again.

[tggzzz]

@BU508A, I can understand that any repairs needed that they would come back to you for guidance, but for adjustments NO, I would expect that would be all included in the calibration process. I can see zero use in having a certification that states on a piece of paper that says that on a certain range it reads xxx high or low. No one is going to take figure into account when using the meter, they will just go by the displayed reading, I would everytime.

Sorry, but in many cases you are wrong.

As a simple example, consider where you are checking if your local voltage standard has changed in the past year. If it reads 1.0030V, the meter is adjusted, and later it reads 1.0035V, has the voltage standard changed? Similarly, when I had my voltage standard measured by the PTB's hp3458, I left it at 9.999571V rather than tweaking it to 10.000000V.

OTOH for quick bench measurements that aren't used for formal certification of a UUT, adjusting a meter would be perfectly acceptable and probably preferable.

Summary: BU085 is right.

[BU508A]

Summary: BU085 is right.

      

[McBryce]

A nice fixer-upper for those who go weak at the sight of HP Bubble-7-Seg displays: https://www.ebay.de/itm/265046868409

McBryce.

[Robert763]

I concur with BU. Calibration is confirming that the device meets it's published specification. that is all th certificate says. Many labs charge extra if you want a report on the actual deviations. If it does not meet specification(s) then any adjustment or repair is addtional work. Basic adjustments may be included and considered pre-authorised for simple devices. For higher performance devices it is normally extra and would require authorisation.
For traditional adjustment where a physical item is moved or replaced adjustment to "spot-on" of an item that meets it's specification is considered bad practice as it may affect stability. For digitally stored and appled corrections this concern does not apply but "spot-on" correction is not really needed.
Having accurate history of a devices stability and any adjustments can be used to provide evidence to extend calibration intervals. This can result is significant cost savings.
For devices where adjustment is not possible, but are stable, the use of correction factors by the operator is common practice. For example electrical standards like resistors are not adjusted during their life at all, the error is recorded at calibration. A standard that is 1% off nominal value but stable to fractions of a PPM is much better than one that was at 0.01% error when "calibrated" but is not stable.
 

[tggzzz]

For traditional adjustment where a physical item is moved or replaced adjustment to "spot-on" of an item that meets it's specification is considered bad practice as it may affect stability.

That is precisely the reason I did not tweak my voltage reference from 9.999571V.

A standard that is 1% off nominal value but stable to fractions of a PPM is much better than one that was at 0.01% error when "calibrated" but is not stable.

I know of an instrument that was accurate to 0.1dB, but stable to 0.001dB. Customers purchased it for the stability, and did not care about accuracy. Why? Because it was used to measure impedance changes as a function of temperature - and each test took a week to complete.

[Specmaster]

So it seems then that the word "calibration" is being misused. What you describe I would call a verification that the device is still within the design parameters. To me and loads of other people, calibration is accepted to mean checking and if necessary, adjusting the equipment in order to meet the design specification and then a calibration sticker certificate can be issued to say that the equipment is performing within its design parameters.

Repairs do not in anybodies mind constitute a calibration, merely an effective repair has been carried out to a level that then requires calibration, (or adjustments / fine-tuning / tweaks) to be made against known standards, to bring the equipment back to a position within the original design parameters.

For example https://en.wikipedia.org/wiki/Calibration makes use of both concepts of "Calibration" that have been expressed so far by myself and others in this thread and there is one particular section of the wikipedia that states :-
"In general use, calibration is often regarded as including the process of adjusting the output or indication on a measurement instrument to agree with value of the applied standard, within a specified accuracy. For example, a thermometer could be calibrated so the error of indication or the correction is determined, and adjusted (e.g. via calibration constants) so that it shows the true temperature in Celsius at specific points on the scale. This is the perception of the instrument's end-user.  "
It is that concept that I conjure up in my mind when talking about getting something to calibrated, and it is that I believe most would accept.

[bd139]

Accurate definitions:

Calibration is measuring characteristics and uncertainty against the specification and recording it. 

Adjustment is making sure that the calibration meets the specification.

Getting something calibrated is an informal term for a combination of both.

Always be explicit in what you ask for when you get a calibration or face disappointment and a larger bill than expected 

[Cerebus]

So it seems then that the word "calibration" is being misused. What you describe I would call a verification that the device is still within the design parameters. To me and loads of other people, calibration is accepted to mean checking and if necessary, adjusting the equipment in order to meet the design specification and then a calibration sticker certificate can be issued to say that the equipment is performing within its design parameters.

Repairs do not in anybodies mind constitute a calibration, merely an effective repair has been carried out to a level that then requires calibration, (or adjustments / fine-tuning / tweaks) to be made against known standards, to bring the equipment back to a position within the original design parameters.

For example https://en.wikipedia.org/wiki/Calibration makes use of both concepts of "Calibration" that have been expressed so far by myself and others in this thread and there is one particular section of the wikipedia that states :-
"In general use, calibration is often regarded as including the process of adjusting the output or indication on a measurement instrument to agree with value of the applied standard, within a specified accuracy. For example, a thermometer could be calibrated so the error of indication or the correction is determined, and adjusted (e.g. via calibration constants) so that it shows the true temperature in Celsius at specific points on the scale. This is the perception of the instrument's end-user.  "
It is that concept that I conjure up in my mind when talking about getting something to calibrated, and it is that I believe most would accept.

This is one of those occasions where Wikipedia is not a good reference to use. Look at how many weasel words are used in that quote: in general, often regarded, etc. Compare to a statement from a more reliable source, NASA's "Metrology — Calibration and Measurement Processes Guidelines"

Calibration compares the declared value of an attribute or parameter of a calibrating artifact, such as a reference standard, against the declared value4 of an attribute of a unit under test (UUT).

or Fluke:

What is Calibration?
Calibration is the act of comparing a device under test (DUT) of an unknown value with a reference standard of a known value.

A person typically performs a calibration to determine the error or verify the accuracy of the DUT’s unknown value.

As a basic example, you could perform a calibration by measuring the temperature of a DUT thermometer in water at the known boiling point (212 degrees Fahrenheit) to learn the error of the thermometer.  Because visually determining the exact moment that boiling point is achieved can be imprecise, you could achieve a more accurate result by placing a calibrated reference thermometer, of a precise known value, into the water to verify the DUT thermometer.

A logical next step that can occur in a calibration process may be to make a corrective adjustment or to true-up the instrument to reduce measurement error. Technically, corrective adjustment is a separate step from calibration.

The next part—and commonly confused with calibration—is called adjustment. If a device is found to be performing outside of its published specifications, adjustment is the process of using information obtained during the calibration eliminating the errors, so that the device will measure much closer to its nominal value.

[McBryce]

My brand new Keithley 6500 just arrived. As soon as it has got up to room temperature, I'll turn it on and let it further warm up, then it's time to see how far off all my other equipment is It looks so good you just want to measure things, lots of things.

McBryce.

[tggzzz]

So it seems then that the word "calibration" is being misused. What you describe I would call a verification that the device is still within the design parameters. To me and loads of other people, calibration is accepted to mean checking and if necessary, adjusting the equipment in order to meet the design specification and then a calibration sticker certificate can be issued to say that the equipment is performing within its design parameters.

Repairs do not in anybodies mind constitute a calibration, merely an effective repair has been carried out to a level that then requires calibration, (or adjustments / fine-tuning / tweaks) to be made against known standards, to bring the equipment back to a position within the original design parameters.

For example https://en.wikipedia.org/wiki/Calibration makes use of both concepts of "Calibration" that have been expressed so far by myself and others in this thread and there is one particular section of the wikipedia that states :-
"In general use, calibration is often regarded as including the process of adjusting the output or indication on a measurement instrument to agree with value of the applied standard, within a specified accuracy. For example, a thermometer could be calibrated so the error of indication or the correction is determined, and adjusted (e.g. via calibration constants) so that it shows the true temperature in Celsius at specific points on the scale. This is the perception of the instrument's end-user.  "
It is that concept that I conjure up in my mind when talking about getting something to calibrated, and it is that I believe most would accept.

Metrologists != most people. Metrology != common.

Many words are misused, e.g. troll, accuracy, precision, stability, resolution, hack, etc etc

[Zucca]

My brand new Keithley 6500 just arrived.

Congrats for the new baby, it is a fantastic device!

[med6753]

This is Curly and today's objective is to figure out how to remove those attenuator assemblies. Why? Two reasons. First, the 50ohm/1MEG switch in Larry has a burnt out indicator light. Second, the channel selector switch in Larry is dirty and needs cleaning. So I'm going to salvage the 50ohm/1MEG switch out of Curly and at the same time model how to get all this apart. The service manual is useless in this respect. All it says with respect to the attenuator boards is that they are "delicate". No instructions on how to remove them. The 485 service manual is very un-Tek like. It quite frankly stinks with regard to servicing components and circuit boards. But it does crow about how most of the boards just plug-in. Well big F'ning deal but what about the exceptions?

And while I was reading that section of the manual I did stumble upon a quick blurb on how to remove the Timing Switch board. And I had to laugh. This was the board where I had to drill out the welds on the top plate. I manual states to remove the board out the back after disassembling the entire rear of the scope. Now I've had my head inside these 485's for 3 days straight and there's absolutely no way that would work. Anyone who tried that would wind up drilling out the welds and removing the plate just like I did. 

 

[Specmaster]

Well it doesn't really help much when the manufacturers often refer to what Fluke calls the adjustment as calibration, see the attached sheets from the Black Star 4503 service manual where it describes the method of adjustment via the front panel controls, as calibration, and they are not the only ones doing so either.

So as you can see, I was not being deliberately confrontational by referring to adjustment as calibration .

So in other words then, getting a meter calibrated is not going to help med get a good meter in order to use as a check for all other meters he owns, unless he has a meter calibrated and adjusted so many of us here on the forum are thinking that calibration is making a meter read correctly so 1.0000 V is displayed as 1.0000 +/- the makers design tolerance etc.

[bd139]

LMAO road karma. Cunt overtook me in a Range Rover. Was doing 30 in a 30 limit and that wasn’t fast enough so he blasted round me doing 50ish on the wrong side of an island and nearly hit an Audi coming the other way. Car coming other way? Plain police car which swung round behind me, passed carefully and pulled the cunt over 

Hopefully it was a video car. If so he’s fucked.

[med6753]

LMAO road karma. Cunt overtook me in a Range Rover. Was doing 30 in a 30 limit and that wasn’t fast enough so he blasted round me doing 50ish on the wrong side of an island and nearly hit an Audi coming the other way. Car coming other way? Plain police car which swung round behind me, passed carefully and pulled the cunt over 

Hopefully it was a video car. If so he’s fucked.

No, You're the cunt for doing the speed limit.  What the fuck is wrong with you?  I would have blown your doors off too. 

J/K.  I love seeing Karma when is comes to assholes who road rage.

Shit...I'm talking about myself.