Products > Test Equipment
Random 3458A question
(1/1)
mendip_discovery:
I have a good condition 3458A and its reading fine except when I do resistance measurments.

I'm still learning how best to operate this unit so don't shout at me for daftness.

With a short on the front terminals I get a low error on voltage but 2wire resistance is -0.46798. After NULL it measures spot on and it's been out for calibration and it's ok.

What I would I like to know is should I do a CAL on the front and rear terminals. If I was to do this, can I only do those two as I don't want to disturb the 10V and 10Kohm settings as that would void the calibration certificate.

Dr. Frank:
Hello,
I guess you mean if you could do a ZERO CAL w/o affecting the GAIN CAL?
In the calibration manual, p.23,  that's called 'Front Terminal Offset Adjustment' and 'Rear Terminal Offset Adjustment', either performed by command CAL 0.

I think, that's indeed possible, and will not affect the gain calibration of DCV and OHM (CAL 10 and CAL 10000).

I have written several articles about optimizing OHM readings, i.e. by using 4W, APER 1 instead of NPLC X command, OCOMP with proper delay, and how to improve noise by using shielded PTFE cable, a 'silent' and stable environment,  statistics and holding your breath during this measurement.
You can either use internal statistics function on e.g. 16 samples, or read out via GPIB, to achieve (stable) 8 digits resolution.

This way, I am able to achieve 0.2 ppm transfer accuracy and about 0.15 .. 0.2 ppm rms noise (StD), better than specified in the FLUKE 3458A/HFL data sheet (but not at all in the HP3458A manual).

The DUT resistor has to be designed for that purpose, i.e. it should have a proper shield around the resistive element, no leakage and small polarization at it jacks. Otherwise, you would inherit additional errors or noise from there.

Frank 
mendip_discovery:
Thanks for that Dr.Frank.

I spent yesterday reading the manual in between building a uncertainty budget for all ranges so brain was in excel overload. So my brain was a little mushy on remembering the CAL modes.

I have ordered a short but I might see if i have a nice bit of copper to make my own zero link.
Dr. Frank:
I'm already impressed by your "small" lab.

I also use that simple copper wire as described in the manual, but first remove oxide before every use.
As I know best what to search for, here are the links to threads where we discussed Ohm mode problems:

https://www.eevblog.com/forum/metrology/volt-nut-meeting-2019-in-stuttgartgermany/msg4427035/#msg4427035
https://www.eevblog.com/forum/metrology/mm2022-delay-and-ocomp-on-the-3458a-with-different-resistors-and-cables/msg4434751/#msg4434751

In the 2nd link, first thread, you'll find further references to two basic articles.
Now here's a long term stability measurement, which I haven't shown before, I think.
I compare these 5 VHP202Z resistors, each 10kOhm, against each other, by making consecutive Transfer measurements within about 10minutes.
Noise / stability values can be found in the first thread.

All resistors have a constant -1ppm/yr. drift (determined over the years), so if you look closer at the latest measurements, you'll encounter that their predicted individual values are uncertain, or "smooth" to less than 0.2ppm, i.e. that's an indicator for the Transfer Uncertainty.
I'm using these shielded 4W PTFE cables since about 2019, before that I used 4 low e.m.f. cables only, being visible in more recent measurements, where you see bigger "roughness" in the measurements.
I also compare this group to my 5450A 10k resistor, which is more long term stable, but as this big box catches more noise, these measurements show a higher degree of "roughness".
Hope, that illustrates the possibility of greater performance of the 3458A Ohm mode.
Frank

   
Navigation
Message Index
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod