Hi,
I investigated on the obvious threshold of the 50 Ohm and 500 Ohm ranges.
It was strange, that in both ranges 0.000 Ohm were displayed, even when the input was shorted by a cable, which should give a few 10 mOhms.
So I measured a resistor box with short cables on my 3458A, and then applied the same configuration to the 121GW:
Setting 3458A 121GW
0.0 0.02743 0.000
0.1 0.12948 0.079
0.2 0.22764 0.178
...
50.0 50.03912 50.005
Especially at 0.1 Ohm setting, and also with rising settings of 0.2, 0.3, 0.4, you can directly see, that my 121GW had a constant threshold of 0.050 Ohm. (gain error does not apply at low values)
At 50.039 Ohm, the gain error of the 121GW additionally chimes in, virtually reducing that threshold value.
This threshold originates from the zero calibration process. UEI probably does not use a direct, low Ohm shunt, but instead uses a cable and a calibrator, set to zero Ohm, which in sum have these 50mOhm, which are always subtracted from the raw Ohm reading.
If a lower resistance short is applied, this would give a negative reading, so the instruments sets all negative readings to zero (threshold).
What confuses much more, that this threshold also shows up using REL.
If you try to zero the value at 0.1 Ohm setting, you will get a 0.050 reading when setting the decade to 0.2.
That's obviously an error in the firmware:
I suppose, that the REL value is taken from the raw reading, then subtracted subsequently from the raw reading first, and then the calibrated zero value is subtracted additionally, which should not happen.
So I made a re-calibration of the 50 Ohm and 500 Ohm ranges only, using a short of < 1.5mOhm for zero calibration, and reference resistors, including cables with 49.971 Ohm and 499.990 Ohm, as measured again with my 3458A, and good enough for calibration.
Now, this threshold is not visible anymore, and you can also make reasonable REL measurements on the mOhm scale, because these 1.5mOhm do no longer play a role.
Frank
PS: Before this investigation, I stored the original calibration values on the SD card, and during that process recalled these from the SD card several time, that's been very practical.
Anyhow,the cal.bin file, also contains some SETUP parameters like APO, BUZZER, LCD, as these were also restored to their original values.