The calibration of HP 3458A "Gold" version

[dl1640]

Recently our lab equipment 3458A was calibrated in a local Keysight affiliate.
The calibration standards used were 5730A for source, 3458A for reference and a function generator 33250
They said the 3458A is a "Gold" version and is calibrated in Keysight US affiliate
I'm curious what differences are between "Gold" and normal vesion or option-002 version,
and how it is calibrated, method? equipments?

[chriswebb]

I am not sure if they have any hardware difference to the meters, but from what I've gathered from Dave's video series basically they are the instruments that are calibrated against the primary standards in the US:

https://youtu.be/G3QK31zotoQ?t=19m18s

There are like 4 or 5 great videos in that series that I highly recommend.

[Kleinstein]

The HW is supposedly the same, maybe with a little more of selection (e.g. only use those meters that have a proven good drift history - thus not brand new one).

The main difference would be the calibration (e.g. more often, to primary reference) and the cal data are handled. So they might include predicted drift.

[ap]

The gold standard in this case usually is a meter that has tighter specs due to frequent calibration and lower uncertainties used for it and a better tolerance 3458A (HFL or 002). They probably dont take the effort to characterize it additionally, just too much effort, the HFL will do. The standard 3458A calibration by Keysight, as per manual, is based on standards with 2ppm (10V) and 3pm (10k) uncertainty, which needs to be added to the spec uncertainties of the 3458A. This is not very tight. Uncertainties under 1ppm are possible without too much effort. Not sure how they applied the 3458A in your specific case, but in most cases (DCV...) the 5730 is the more relevant standard. Also, theer are different calibrations performed by Kesight, there is e.g. info here on eevblog. Also, there is a document by Fluke talking about how to calibrate the 3458A with a 5730 and what is critical and how to combine uncertainties re. a pass/fail. When you go through it you will see the limitations of this procedure. Your cal report will also give you some insights into what uncertainties they apply and what they did specifically. Might be interesting to see it here (if you have permision, and keep in mind the cal report also has certain limitations, e.g. reproduction only in its complete form). As per standard calibration procedure of the 3458A cal manual, the gold standard 3458A is used e.g. for AC comparision (SYNC versus ANA mode).

[TiN]

One should be aware that metrology people have observed shifts up to 5ppm after inter-continental shipments - especially during (cold) winter periods.

Can you elaborate a bit more with data points? I'd appreciate that, since I did not see that big shift on my unit (with my trust levels around 2ppm).

[dl1640]

Later, I will post a copy of the certificate 

[alm]

Not sure how they applied the 3458A in your specific case, but in most cases (DCV...) the 5730 is the more relevant standard.

These posts from the volts-nuts mailing list may still be relevant. They talk about using the multi-function calibrator merely as a stable source, and purely relying on the well-characterized 3458A for traceability to whatever standards were used to calibrate it. Has this HP/Agilent/Keysight STE9000 procedure, which I believe I saw referenced in recent 3458A calibration certificates, changed significantly with the introduction of new multi-function calibrators?

[CalMachine]

A gold standard meter is one they have tons of history on and is proven to be extremely stable over time.  This allows them to predict its drift very precisely.

At a regularly spaced interval, one can start to accumulate history of their own 3458A and "gold"ize their own meter.  Ofcourse, this is relies solely on the performance of your meter.

[Kleinstein]

A gold standard meter is one they have tons of history on and is proven to be extremely stable over time.  This allows them to predict its drift very precisely.

At a regularly spaced interval, one can start to accumulate history of their own 3458A and "gold"ize their own meter.  Of course, this is relies solely on the performance of your meter.

It is not only the quality of the meter that is important, also the quality of the calibration data make a difference. For the 3458 one would need a series of good standards Lab calibrations - not just the normal cal. Having that calibration every 90 days is part of the added value and costs.

[CalMachine]

A gold standard meter is one they have tons of history on and is proven to be extremely stable over time.  This allows them to predict its drift very precisely.

At a regularly spaced interval, one can start to accumulate history of their own 3458A and "gold"ize their own meter.  Of course, this is relies solely on the performance of your meter.

It is not only the quality of the meter that is important, also the quality of the calibration data make a difference. For the 3458 one would need a series of good standards Lab calibrations - not just the normal cal. Having that calibration every 90 days is part of the added value and costs.

Ofcourse, the quality of a meter and the quality of its documented history/calibration go hand in hand!

How is one to determine how "good" their meter is if the reported uncertainties are as wide as the barn door

[VintageNut]

One should be aware that metrology people have observed shifts up to 5ppm after inter-continental shipments - especially during (cold) winter periods - for people outside USA this might be a bummer considering KS Loveland USA standards Lab.

For european customers 3458A standard calibrations are only performed in KS standards Lab in UK - all accredited calibrations - apart from standards Lab - are also executed in Böblingen (Germany).

One should also wonder if your really need accreditation and official Keysight paperwork - wonder if high end AC performance is needed, otherwise one can use local labs that have their set of Fluke 732A/732B batteries calibrated against a local JJA - it might be cheaper and give same or even better results (on certain ranges) if operator is trained to do so.

A standards Lab will not use directly a JJA to calibrate the customer's unit but a golden reference unit which is calibrated at regular intervals and a (large) set of Fluke 732A/732B DC sources with known history.

Flinstone

This is a very insightful post. If you look at the 5730A calibration info, you will see that the 732B is used for calibration of the 5730A. So, why not bypass the transfer from the 732B to the 5730A to the 3458A? Go directly from the 732B to the 3458A.

I know from discussions with the standards lab that I visit, that on any given day, they can transfer the volt to the instrument of your choice with about 1/3 ppm uncertainty. They inter-compare all of their standards once per week and they have been doing this for many years. 

[Moon Winx]

This is a very insightful post. If you look at the 5730A calibration info, you will see that the 732B is used for calibration of the 5730A. So, why not bypass the transfer from the 732B to the 5730A to the 3458A? Go directly from the 732B to the 3458A.

I know from discussions with the standards lab that I visit, that on any given day, they can transfer the volt to the instrument of your choice with about 1/3 ppm uncertainty. They inter-compare all of their standards once per week and they have been doing this for many years.

So a 732B can be used in the artifact calibration of the 3458, but to perform a verification it is convenient to automate the procedure as much as possible. You *could* use a 732B w/ divider to check the 3458 DC volts, but using a calibrator (characterized by a 'gold' meter) that can be computer-controlled is much quicker and good enough.

[dl1640]

Not sure how they applied the 3458A in your specific case, but in most cases (DCV...) the 5730 is the more relevant standard.

These posts from the volts-nuts mailing list may still be relevant. They talk about using the multi-function calibrator merely as a stable source, and purely relying on the well-characterized 3458A for traceability to whatever standards were used to calibrate it. Has this HP/Agilent/Keysight STE9000 procedure, which I believe I saw referenced in recent 3458A calibration certificates, changed significantly with the introduction of new multi-function calibrators?

Hi,

The procedure is STE-50109421-E.05.03

Let's see a picture of dcv items calibrated.
The reference value is not an integer, like 10.000000V, it is the so-called "expected" value which is displayed on that "Gold" 3458A.
Or it is a compensated value against the calibrated value/factor of that "Gold". (I guess)
Reported calibration uncertainties are not so "good", I think it is just a standard procedure as we sent in the UUT 3458A and the next day they shipped it out. (just hook it up after receive and run the program then print out papers)

Any idea of such calibration is welcomed because we calibrate 3458A by KS for the very first time and before that we calibrate it by NMI.
NMI calibrate it by using 8508A + 5720A if I recall correctly, but they don't calibrate it to 8.5 digits...while KS does.

I think the "Gold" is calibrated to primary standards very often then it can get very tight spec or low uncertainty.