HP 3458A 8.5 digit calibration questions

[Signal32]

Hello,

I was wondering if the 3458A has an option for storing a calibration string(the date of the last calibration) and how would I go about querying it ?
I tried many of the "CAL? XXX" commands but they seem to store only calibration values, not strings.

Also anyone have any suggestions for competitively priced calibration centers that can calibrate a 3458A with option 002 ?
Out of curiosity, how does option 002 impact calibration, I get that it requires special gear vs a regular 3458A. I was under the impression that it just maintains stability better, not that it would have any special instrument requirements for calibration.

Thanks!

[TiN]

Calibration procedure for 3458A-002 is same as regular one. It's just people who have 002 option usually like to get best accuracy, hence send for standards cal instead of usual.

[Signal32]

Calibration procedure for 3458A-002 is same as regular one. It's just people who have 002 option usually like to get best accuracy, hence send for standards cal instead of usual.

What do you mean "Standards cal", calibrated to direct standards vs to calibrators ?
The reason I am asking is because I've been told by multiple cal labs that while they can calibrate 3458A, they can't calibrate 3458A-002. I was surprised too, I guess I should have asked them directly.

Also do you know if there's a cal string for the 3458A ? I guess if you don't know no one would (Thanks for the very detailed article on the  3458A ! )

[Dr. Frank]

Of course there is a command for storing a calibration string.
RTFM, please use calibration manual, page 74, or user manual, page 165:

CALSTR string[,security_code] stores the string in non-volatile memory, maybe secured by the calibration security code, if set. This command is available over the GPIB only.

CALSTR? will return this string.


A calibration usually verifies (i.e. PASS/FAIL) the 3458As 1year limits against external calibrators, like a Fluke 5730A, plus another ' Gold Standard ' 3458A.
In case of the standard reference, for the 10V range, about 10ppm  test limit has to be checked, for option 02, this is about 5ppm.


Therefore, for the standard case, the 5730A and the reference 3458A may just manage that level of uncertainty with enough T.U.R.  (Test Uncertainty Ratio), but will fail for the latter case, as about half of the uncertainty is needed. That's also valid for verification of the other DCV ranges.

This can only be achieved by using a well characterized Fluke 732B reference, with <1ppm uncertainty, and a Fluke 752A ratio standard, < 0.5ppm uncertainty, or a JJA system, especially, when additionally a precise adjustment (10V, 10kOhm) to 24h specification shall be done. This equipment and uncertainty grade are available only in higher grade laboratories, like Keysight Loveland.

Good information about this subject can be found in the Fluke AN:"Calibrating the Agilent 3458A DMM with the 5730A Multifunction Calibrator" , and on Keysights metrology youtube channel "Keysight Calibration & Repair": https://www.youtube.com/channel/UCG8BVoXwdZqz-PlZxeO56Tw
which can also be found as documents on the Keysight training site: http://www.keysight.com/main/facet.jspx?t=79831.g.1&cc=US&lc=eng&sm=g

Frank

[Signal32]

Of course there is a command for storing a calibration string.
RTFM, please use calibration manual, page 74, or user manual, page 165:

Thanks, this seems to be the only documented remote-only command  .
Your other points also make sense. Thanks again!

[TiN]

Yes, Dr.Frank beat me to it. While calibration procedure itself is same as usual 3458A, high-stability option 002 units need uncertainty better than most of calibration facilities without direct access to JJA's can provide.

Same reason why local Tektronix lab here in Taiwan are unable to calibrate my Keithley 2002's and need them to be sent to USA main calibration facility, but for lesser 7.5-digit 2001 they do all in-house using annually calibrated 5700+5720A and 3458A.

[Signal32]

Yes, Dr.Frank beat me to it. While calibration procedure itself is same as usual 3458A, high-stability option 002 units need uncertainty better than most of calibration facilities without direct access to JJA's can provide.

Same reason why local Tektronix lab here in Taiwan are unable to calibrate my Keithley 2002's and need them to be sent to USA main calibration facility, but for lesser 7.5-digit 2001 they do all in-house using annually calibrated 5700+5720A and 3458A.

I see, thanks.
One calibration-unrelated question. In your 3458A article you mention:

can replace battery-backed NVSRAMs with Ferroelectric RAM (FRAM) such as Cypress FM18W08 and FM16W08 which does not require battery power to retain its data

Has anyone tried this with success ? I'm asking because in the FRAM datasheet I see certain differences that may indicate incompatibility such as "F-RAM memories latch each address on the falling edge of chip enable" and some slower timings. Does the 3458A just happen to do a chip enable before each read/write, making FRAM compatible ?

[TiN]

Hm, I did not try that as has no FRAMs, but I'll stand corrected if those not work. I don't recall seeing anyone used FRAMs in 3458A, but to be precise not many info on web covering these DMMs anyway.