Products > Test Equipment
+5.5digit DMM with simple calibration?
(1/2) > >>
RogerThat:
Hi,

So, I just read about the Advantest 8.5 digit DMM. I was surprised to read that the only thing used to calibrate this thing was a known 10k resistor and 10V reference.

Is there any other 5.5digit, or higher DMM, that uses the same simple calibration procedure?
tautech:
Similar but different. Each measurement range needs be checked/tweaked to ensure accuracy across all ranges.
Have a peep at this:
http://www.thedefpom.com/siglent_sdm_calibration.php
Kleinstein:
The HP3458 (8 digits)  meters uses a similar procedure and the same 10 V and 10 K starting point for an adjustment.
The Datron 1281 DMM also uses a similar method, though also including a factory calibration for a special ratio transformer, that is usually very stable over time.
Some of the Fluke calibrators use a similar method, though sometimes starting with 2 resistors.
The analysis of the Advantest meter showed that there still seem to be extra correction factors per range involved from a more classic calibration of the ranges, maybe as a seprate factory calibration.

For a reliable calibration one may still need to check the ranges individually, so make shure the transfer between the ranges actually worked as planed
It is more that the calibration constants / adjustment are based on the 2 measurements. The check part of the calibration is usually still done for all ranges.

To make the method of measuring in multiple ranges work well, it needs rather good linearity and thus a relatively good (linear) ADC. The question is of the ADC is more accurate than the resistor setting the gains are stable.
Some meters do a check in different ranges (e.g. part of the self test), but still call for individual adjustment of the ranges.
Some (e.g. the sigilent meter from the last post) even separate for the positive and negative side. The individual adjustment may compensate for some errors (e.g. INL error) and this way give a better result, at least for the shorter time frame.
The method of adjusting to just 2 measurements is in principle not limited to high end meters, and in some aspects it would also make sense (for longer cal. intervals) for the lower end range for a meter.
However so far I know of no lower grade meter to work this way.

Using external reference and the DMM to measure the ratio is possible, but of limited accuracy and a bit time consuming. The internal process has some advantages, though also a few limitations. One would still need a good (low noise) external source, like a calibrator.
alm:

--- Quote from: tautech on October 09, 2022, 08:22:35 pm ---Similar but different. Each measurement range needs be checked/tweaked to ensure accuracy across all ranges.
Have a peep at this:
http://www.thedefpom.com/siglent_sdm_calibration.php

--- End quote ---
That's pretty much as far from artifact calibration as possible, needing positive and negative values for the DCV ranges. This is the traditional way of needing cardinal values for every range, which is very simple to implement in the meter, but has the down side of needing sources for all values it needs, which generally means a multi-function calibrator like the Fluke 55xx and 57xx series.

I'm not quite sure why this was never implemented for meters with less than 8.5 digits of resolution. Maybe their ADCs are not good enough relative to the stability of the reference / gain resistors? Or maybe it's easier to find a calibrator that's good enough to adjust the meter, while I imagine in the late eighties when the 8.5 digit DMMs were designed calibrators may not have been easily available, and the calibration method of the time might have been the slow and tedious process of manual transfers using standard cells, Hamon dividers, ratio transformers, thermal voltage convertors, etc.

bdunham7:

--- Quote from: alm on October 09, 2022, 09:38:41 pm ---I'm not quite sure why this was never implemented for meters with less than 8.5 digits of resolution.

--- End quote ---

Probably because it is complicated and expensive to implement all the required systems internally--and if not executed properly will yield poor results.

https://download.flukecal.com/pub/literature/1260355E_Artifact_Calibration_AN_w.pdf
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod