EEVblog Electronics Community Forum

EEVblog => EEVblog Specific => Topic started by: EEVblog on February 11, 2013, 01:45:20 pm

Title: EEVblog #422 - How To Calibrate A Calibrator
Post by: EEVblog on February 11, 2013, 01:45:20 pm
Les Thomaidis shows and explains the calibration rack at the Agilent Melbourne Standards Lab that is used to "calibrate the calibrators". This is where other calibration labs send their multimeter and oscilloscope calibrators to get calibrated.

EEVblog #422 - How To Calibrate A Calibrator (https://www.youtube.com/watch?v=JID8bS2-skg#)

Dave.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: free_electron on February 11, 2013, 04:12:56 pm
Fluke has a specific scope calibrator . They acquired it by borging another company. These machines use a special pickup head that screws directly onto the bnc input of the scope. Darn it. Can't remember what it's called.

-edit-  wavetek

http://totalsolutions.fluke.com/calibrators/newsletter/printer_friendly.asp?id=87&type=article (http://totalsolutions.fluke.com/calibrators/newsletter/printer_friendly.asp?id=87&type=article)

That head can create programmable signals with risetimes up to 70 picoseconds... The jim williams pulse generator doesn't even go that fast...
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: Marco on February 11, 2013, 04:28:34 pm
I want to see deeper down this rabbit hole ...

What are the official central standards which his equipment is calibrated against? I know that there are fundamental physics derived voltage references based on superconductors ... is there anything similar for current or resistance? Or do they just have some gold standard set of reference resistors at NIST/whatever?
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: c4757p on February 11, 2013, 04:36:40 pm
Not that I have much experience in this "super-metrology", but I'd guess that it's not too hard to come up with a standard for current, since it's defined as an actual count of electrons flowing per second. Once you have a standard volt, ampere, and second, you can create derived standards for just about anything.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: jucole on February 11, 2013, 04:51:25 pm
nice one!  the bit about the type a/b analysis was a real gold nugget for me! ;-)
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: jahonen on February 11, 2013, 04:56:39 pm
Primary standard for volt is the Josephson junction used as a voltage controlled oscillator, so that the voltage measurement can be done via definition of a second, which in turn can be referenced to an atomic clock:

http://en.wikipedia.org/wiki/Josephson_effect (http://en.wikipedia.org/wiki/Josephson_effect)

Regards,
Janne
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: Neilm on February 11, 2013, 06:31:53 pm
Not that I have much experience in this "super-metrology", but I'd guess that it's not too hard to come up with a standard for current, since it's defined as an actual count of electrons flowing per second. Once you have a standard volt, ampere, and second, you can create derived standards for just about anything.

The definition of an Ampere is:
The ampere is that constant current which, if maintained in two straight parallel conductors of infinite length, of negligible circular cross-section, and placed 1 meter apart in vacuum, would produce between these conductors a force equal to 2 x 10-7 newton per meter of length.

Voltage is not actually a base SI unit. Voltage is also derived unit.

Neil
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: SeanB on February 11, 2013, 06:49:43 pm
I asked at my local Metrology lab how they do calibration. They keep 3 sets of masses, and at any time one is away for certification, one is in use as the primary reference ( never leaves the shop, used to verify the working masspieces only) and the third is in a safe for verifying the primary reference. The working references go out to sites, and get verified once a week. Mine were done by comparing to the one primary one, so they are as accurate as you can get, 0,5g error in 20kg. the 1g unit is accurate to 1 milligram. most importantly is they vary very little between check cycles, mostly only due to dust or dirt being wiped off them.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: Marco on February 11, 2013, 07:23:46 pm
From cursory googling it seems the Ampere is the least "fundamental" unit from a calibration point of view (with both voltage and resistance having practical quantum effect primary standards in use at NIST for calibration).
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: ftransform on February 12, 2013, 10:11:20 am
I wish a eevblog member would do experiments based on pain voltage calibration, i.e. determining the voltage by licking the wires.

Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: Ed.Kloonk on February 12, 2013, 10:43:00 am
Calibrate A Calibrator. It sounds to me like that would cause a rip in space or something.

Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: PA4TIM on February 12, 2013, 11:04:12 am
There are nice pdfs on the nist site. Voltage is a dervivated standard but through modern technic a very known one. In the beginning they useds cgs units. centimeter, gram, seconds ( if i recall this right) all others were derivated by that and the terms like Volt were in the same order as feet, tumb ect. They were workman units. 1 Volt was the value of a standard cell. Lord Kelvin ( thompson) was te first to measure the real volt ( as function of resustance and current) and found out the standard cell was not 1V. The first meters were galvanometers. They meaure current but you can convert that to voltage using resistance. The first accurate real voltmeter ( 4 quadrant electrometer) was an invention of Kelvin.

For resistance they used length coupled to current through the resistance of some material. That is why we still know the name potentiometer. The resisance was measured in meter ( cm/mm) it was a long straight thing with a sliding contact and a scale in cm beside it. Like measuring frequency ( wavelength) with a slotted line.

For current they used a current balance but I still do not know how they work. It looks like some mechanical divise.

Just a sidestep: calibrating pur sang is only testing if it is still within specs. So adjusting withing specs is not allways included. What you want to know is the drift over time. If you calibrate your meter every year and 1V is 0.9V after one year, they adjust it to 1V again ect without telling it was 0.9V you think your meter is perfect. The next year the calibrate it and its 0.8V, they adjust it to 1V again. So while you think its perfect, it is far from ( but it can be dtill within specs, so technical it is also perfect)
That is why specs are so important and things like min 1% or best 0.2% specs from our Chinese friend tell you nothing. ( besides they do also not state referenced to what, it can be 1% to theire 5 % calibrtor ( rather extreme but to get the idea) but it can be even be 0.1 % to the prototype, or in theory because the used parts, those numbers are pure commercial) a good brand gives it per range, over time under notification of temp and humiity. Most times 24H, 3 months and a year.

But if it is new 1V and the report states after a year it is 0.99V, and next year 0.995 and the 3rd year it is still 0.995V ect you know it is now aged and stable over time so if you then adjust it to 1V and next year it is fi 0.997V you know that if you meaure something in the year between calibrations that the value will probbly be somewhere , with extra margin, between 1 and 0.995V.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: EEVblog on February 12, 2013, 11:23:07 am
Voltage is not actually a base SI unit. Voltage is also derived unit.

There are 7 fundamental physical properties in the SI system:
http://en.wikipedia.org/wiki/Physical_quantity (http://en.wikipedia.org/wiki/Physical_quantity)
http://en.wikipedia.org/wiki/SI_base_unit (http://en.wikipedia.org/wiki/SI_base_unit)

Quote
metre for length (US English: meter)
kilogram for mass (note: not the gram)
second for time
ampere for electric current
kelvin for temperature
candela for luminous intensity
mole for the amount of substance.

Only the ampere is fundamental, all the other electrical units are derived.

At high RF frequencies for example (above 40MHz), cables and connector standards are actually calibrated in terms of their physical dimensions.

Dave.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: PA4TIM on February 12, 2013, 11:28:44 am
Hmm, just found out typed a lot for nothing, you allready have a nice video on calibration ;-)
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: EEVblog on February 12, 2013, 11:39:04 am
Hmm, just found out typed a lot for nothing, you allready have a nice video on calibration ;-)

There is another one rendering now, a tour of $3M worth of gear in a portable cal lab. Plus lots of extra talk on calibration for the aficionados, 30min total.
I could have left out all the cal talk and just did the tour, but there is so much good stuff in there.

Plus I have another two on top of that I have to edit. One a further tour of the standards lab, and another a tour of the regular cal lab. But I'll leave those for a while or I'll probably get lynched.

Dave.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: jucole on February 12, 2013, 11:42:50 am
But if it is new 1V and the report states after a year it is 0.99V, and next year 0.995 and the 3rd year it is still 0.995V ect you know it is now aged and stable over time so if you then adjust it to 1V and next year it is fi 0.997V you know that if you meaure something in the year between calibrations that the value will probbly be somewhere , with extra margin, between 1 and 0.995V.

I know I'm still a beginner but surely having a meter that gives it's most accurate as possible reading is far better than a table of offsets for each of my best dmms?   also when I took my Solartron 7150 to be calibrated recently,  the adjustments are all done in software, so the difference between an adjusted one and the not adjusted one is all in the mind! or the offset tables in memory ;-)  but you can still track the aging of the device from the cal lab results.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: EEVblog on February 12, 2013, 11:52:57 am
BTW, for those who are not aware, if you have a bit of Agilent gear you can get the original factory cal certificate online:

https://service.tm.agilent.com/infoline/single-lookup.aspx?form=orig-cert-query

Useful if you have bought a recent Agilent product on ebay for example, or you have lost yours.  :-+

Dave.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: PA4TIM on February 12, 2013, 01:14:13 pm
Dave, you get lynched if you do not show those cal lab tour videos  >:D
Dave is becoming a Voltnut  :-DD
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: robrenz on February 12, 2013, 02:55:26 pm
These are great Dave :-+ show them all and let the whiners whine.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: free_electron on February 12, 2013, 04:16:38 pm
BTW, for those who are not aware, if you have a bit of Agilent gear you can get the original factory cal certificate online:

https://service.tm.agilent.com/infoline/single-lookup.aspx?form=orig-cert-query

Useful if you have bought a recent Agilent product on ebay for example, or you have lost yours.  :-+

Dave.

Thanks for that one. I'm going to DOS that server when i start pulling em in.  ;D
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: jnd on February 12, 2013, 10:49:02 pm
Dave, you get lynched if you do not show those cal lab tour videos  >:D
Dave is becoming a Voltnut  :-DD
Love it! More Volt-nuttery please :-+
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: EEVblog on February 12, 2013, 11:05:46 pm
These are great Dave :-+ show them all and let the whiners whine.

Well, yeah. But it is kinda flooding the channel. Latest 30min vid is already uploaded, but that will be it for a while.

I'm actually not all that happy with these videos, shooting one-take impromptu videos on-site never really works as well as you think it might. But there is plenty of good info in these.

Dave.
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: quarks on February 14, 2013, 11:17:50 am
Hello Dave,

very interesting to see. Please show more if you can.

bye
quarks
Title: Re: EEVblog #422 - How To Calibrate A Calibrator
Post by: jucole on February 14, 2013, 05:20:42 pm
If you're interested to know more about calibration uncertainties, there is some great information in:
NIST Technical Note 1297 - Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results.
http://physics.nist.gov/Pubs/guidelines/TN1297/tn1297s.pdf (http://physics.nist.gov/Pubs/guidelines/TN1297/tn1297s.pdf)