DMM7510 0V Short Offset issue

[opa627bm]

Hi All,
We have 3 units on site for evaluation, 1 is freshly cal , 2 are not calibrated.
both display about -4uV with a low thermal EMF short installed.
We then inserted the short (34172b) to the 2 units that are not calibrated, let it sit for 1 day then we ran through the DC cal step 0 - 3.
However after cal, the 2 units are still at -4uV when short is installed.
this is giving me 0.5% error at 1mV input.
Anyone having ideas if this is normal?
Regards,
Li

[laichh]

The input terminal of the Keithley DMM7510 is not thermal e.m.f. treated, i.e., not pure copper like Keysight 34470A.

My solution for your case would be just simply NULL (or RELative in Keithley terminology) it.

[MegaVolt]

display about -4uV with a low thermal EMF short installed.

You must abandon the 10 MΩ mode. And turn on the input impedance Auto.

Other bugs and features of this miracle of technology are here: https://www.eevblog.com/forum/testgear/all-about-keithley-dmm7510-bugs-and-features-recipes-advice-notes/

[MegaVolt]

The input terminal of the Keithley DMM7510 is not thermal e.m.f. treated, i.e., not pure copper like Keysight 34470A.

My solution for your case would be just simply NULL (or RELative in Keithley terminology) it.

I will say even more. The input terminals are at different temperatures ( But in this case the problem is the 10 MΩ input impedance which needs to be disabled and never used again.

[HighVoltage]

Very interesting, MegaVolt

It definitely makes a big difference, if the input impedance is set to 10 MΩ or AUTO
I tested this on my two instruments:

10 MΩ
Unit 1: 4.5 uV
Unit 2: 6.3 uV

AUTO:
Unit 1: 0.03 uV
Unit 2: 0.3 uV

See pictures

[MegaVolt]

It definitely makes a big difference, if the input impedance is set to 10 MΩ or AUTO

For 100mV range, 3-5uV error is 30-50ppm which is out of specification

I have no way to check the kind of error. Is this an error near zero? Or is it an additive error?

[opa627bm]

That seems to fix the issue, now my 0 holds at 300nV Avg.

[MegaVolt]

That seems to fix the issue, now my 0 holds at 300nV Avg.

300nV is a good value. If you need to reduce it even more, use Rel. The device keeps zero very well with auto-zeroing.