Author Topic: AC calibration of Agilent 34970A (Read 9521 times)

Magnum · « **on:** April 23, 2015, 04:19:42 pm »

Hello,

I have a Agilent 34970A which needs calibration. I don't have a calibrator, so I want to do it with my equipment as good as possible.
I have a Keithley 2000 (calibrated last October), a Keithley 2015, a Rigol DP832 power supply, a Siglent SDG 1010 function generator, a Rigol DS1054 and a RK8511 DC electronic load.
For DC calibration I ordered one of the cheap voltage references from Ebay and a DC voltage divider. I want to use the voltage reference as IMHO the ripple of the DP832 is too high for a stable measurement. I will use the Keithleys for the measurement.
DC voltage and current shouldn't be a problem, but how can I make a stable AC measurement? I bought a cheap adjustable AC transformer, adjustable from 0-250V AC.The powerlines here fluctuate approx. 5%. I can use the signal generator for the low AC levels (up to 5V), but I don't have a solution for 10/100/300V. How can I stabilize the AC voltage?
Thnaks for any info.

Vgkid · « **Reply #1 on:** April 23, 2015, 04:26:14 pm »

What voltage divider did you order?

dom0 · « **Reply #2 on:** April 23, 2015, 04:44:06 pm »

For low frequencies you can just use a 50/60 Hz mains transformer as a step up transformer.

macboy · « **Reply #3 on:** April 23, 2015, 05:10:50 pm »

Maybe try an audio amplifier, driven with a sine from your signal generator, connected to the secondary of a AC wall-wart transformer. The primary will then output a higher voltage by the ratio of the transformer. Don't exceed the rated primary voltage. I'd also put a resistive load on the amplifier (in parallel to the transformer) to help ensure that it isn't seeing a highly inductive load.

TimFox · « **Reply #4 on:** April 23, 2015, 05:30:47 pm »

Cheap lightweight transformers will not give the expected voltage ratio when driven in the opposite direction from the manufacturer's specification. To keep weight (and cost) down, they tolerate excess leakage inductance (reduced coupling), compared with the same voltage ratio and enough iron to contain fully the flux.

dom0 · « **Reply #5 on:** April 23, 2015, 06:18:33 pm »

Well yes you need of course a known good meter to measure it's output, but the same is true of the generators output anyway.

Fun fact: there are in fact precision transformers built like a Kelvin-Varley divider around. But they're huuuge and very expensive.

sync · « **Reply #6 on:** April 23, 2015, 07:55:21 pm »

I have calibrated the AC ranges of my 34401A. I think the 34970A is similar. Fortunately I have an AC calibrator.

You need the frequencies stated in the manual. If they are not within the tolerances the meter refuse calibration. It will need 1kHz and 50kHz for all voltages and additional 10Hz for the 10V source.

For the 10V you can use a simple op amp amplifier and your function generator. I think getting 100V will be difficult. A transformer has to work up to 50kHz. Good luck finding one. Maybe you can build an amplifier which can output 100Vrms. Or just ignore the 100V and 300V ranges and don't calibrated them.

dom0 · « **Reply #7 on:** April 23, 2015, 07:58:15 pm »

Maybe with audio transformers...

sync · « **Reply #8 on:** April 23, 2015, 08:12:53 pm »

Yes, maybe a transformer for 100V audio distribution systems will work.

Magnum · « **Reply #9 on:** April 23, 2015, 09:34:21 pm »

Quote from: Vgkid on April 23, 2015, 04:26:14 pm

What voltage divider did you order?

First I wanted to build one myself, then I found this:
http://www.ebay.de/itm/111605837600?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Less than 12 Euros (shipping is more expensive) and accurate up to 0,005%!!!

Magnum · « **Reply #10 on:** April 23, 2015, 09:41:08 pm »

Thanks for all the replies. Great ideas.
I didn't know about the 1kHz and 50kHz. That doesn't make it easier. I think I'll use a hifi amplifier for the 10V. It can go up to approx. 30V before clipping, maybe that is enough to calibrate the 100V range (if the Agilent accepts it, it is better than not calibrating at all), and will forget the 300V range.

Magnum · « **Reply #11 on:** April 23, 2015, 09:45:27 pm »

Checked the manual: 100V range needs 90-110V input, 300V range 95-303V. Maybe I'll forget 100V and 300V....

sync · « **Reply #12 on:** April 23, 2015, 10:04:10 pm »

Quote from: Magnum on April 23, 2015, 09:34:21 pm

First I wanted to build one myself, then I found this:
http://www.ebay.de/itm/111605837600?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Less than 12 Euros (shipping is more expensive) and accurate up to 0,005%!!!

Thanks! I just ordered one.

Magnum · « **Reply #13 on:** April 23, 2015, 10:13:36 pm »

Quote from: sync on April 23, 2015, 10:04:10 pm

Quote from: Magnum on April 23, 2015, 09:34:21 pm
First I wanted to build one myself, then I found this:
http://www.ebay.de/itm/111605837600?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Less than 12 Euros (shipping is more expensive) and accurate up to 0,005%!!!
Thanks! I just ordered one.

I don't have it yet. I am curious why they limit the 1:10 ratio to max 20V input. IMHO it should take more (18V drop with 90kOhm is less than 4mW) , but they limit it to keep the resistors cold and within specs.

Vgkid · « **Reply #14 on:** April 23, 2015, 10:16:17 pm »

That looks neat, tear it down.

acbern · « **Reply #15 on:** April 23, 2015, 11:12:26 pm »

well, it depends on what you understand by calibration

the methods described are far off what one would do ac-calibrating a 6.5 digit dmm within reasonable limits. you need precsion gear to do that, and well tracable to national standards. I do not want to go into all the details here.
I see you live germany, I live in the munich area, I would be able to help cal your item if you want (for free, with traceability and pretty good uncertainty), if interested send me a PM.

Magnum · « **Reply #16 on:** April 24, 2015, 10:13:05 am »

Quote from: acbern on April 23, 2015, 11:12:26 pm

well, it depends on what you understand by calibration
the methods described are far off what one would do ac-calibrating a 6.5 digit dmm within reasonable limits. you need precsion gear to do that, and well tracable to national standards. I do not want to go into all the details here.
I see you live germany, I live in the munich area, I would be able to help cal your item if you want (for free, with traceability and pretty good uncertainty), if interested send me a PM.

That's why I wrote I want to do it as good as possible with my equipment :-)
Thank you very much for your offer, that's the easiest and best way!

acbern · « **Reply #17 on:** April 24, 2015, 01:02:50 pm »

unless your meter needs cal because it was repaired/parts exchanged, and not simply because the time has elapsed to next cal, you better dont do an ac cal with the methods desribed. the meter's drift from its original cal is most likely much less than what you can ever achieve with your means as decribed.

Magnum · « **Reply #18 on:** April 24, 2015, 02:28:05 pm »

I have to change the chip that holds the cal data, so I don't have a calibration at all.

Magnum · « **Reply #19 on:** April 29, 2015, 08:02:54 pm »

The voltage divider arrived today. It seems to be accurate (at least what I can measure with my equipment)
Can someone tell me what the jumpers 1,2 and 3 are for? IMHO to calibrate it, but I don't know how...

sync · « **Reply #20 on:** April 29, 2015, 09:30:23 pm »

Nice! Hopefully my one will arrive soon.

The jumpers are for calibration. These are hamon dividers. The jumper changes the resistance ratio from 9:1 to 1:1 which is easily to adjust.

Look at these links.
http://conradhoffman.com/HamonResistor.html
Fluke 752A manual: http://us.flukecal.com/literature/product-manuals/752a-instruction-manual
Fluke 752A schematic: http://www.ko4bb.com/manuals/index.php?dir=Fluke/Fluke_752A
https://www.eevblog.com/forum/testgear/hp34401-measurement-of-linearity/msg358713/#msg358713

Magnum · « **Reply #21 on:** April 29, 2015, 09:41:00 pm »

Thanks for the info! I measured the resistances with the jumper in place, so I SHOULD have found out myself :-)

Vgkid · « **Reply #22 on:** April 29, 2015, 09:43:27 pm »

Looks like I should have sent my reply sooner, sync beat me to the hammond divider.

HighVoltage · « **Reply #23 on:** April 30, 2015, 11:16:36 am »

This Russian divider looks like real quality inside.
Thanks for sharing those pictures

Mickle T. · « **Reply #24 on:** April 30, 2015, 01:24:00 pm »

R35 voltage divider is for DC only.

Service Manual is here: http://www.mediafire.com/download/53dj1ua1s65c1r3/R35+Voltage+Divider+-+Service+Manual+%28RU%29.djvu


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: AC calibration of Agilent 34970A (Read 9521 times)

Share me