HP 34401A offset compensation method for resistance measurement

[trobbins]

I'm doing some comparison measurements on the eight 0.1ohm to 1Mohm 0.005% decade reference resistors in an ESI 300 PVB using my uncal 34401A.  Previous testing of my 34401A measurement ranges appear to be within or close to the 1yr accuracy levels (using cross-comparisons that include via a calibrated meter), so this latest batch of resistance tests is adding to the comparisons made so far.   As the 1 and 10 ohm tests go well below the 100 ohm range of the 34401A, I was starting to look at the impact of thermal emf connections and other errors that fit well within the wide 34401A confidence limits, but would none the less be nice to suppress.

Dr Frank identified the following offset compensation method:

For the low value resistors, 1 Ohm to 100 Ohm, you have an additional problem, which the 34401A cannot deal with directly, these are thermal voltages, on the order of µV to several tens of µV, depending especially on your test cables. These e.m.f. have to be removed by a method called 'Offset Compensation'.
.......
There is a trick for the 34401A to manually apply this Offset Compensation method:
With 4W , 6dig slow, manual 100 Ohm range, first connect the Input+ cable to the Input- jack of the 34401A, instead to the 5450A, and take a NULL measurement.
This measures and eliminates the e.m.f. voltage at the resistor w/o the test current.
Then apply the cable back to the 5450A, and you will measure the resistance w/o e.m.f.

I'd like to appreciate this offset compensation method for the 34401A and a 4W connection to a low value resistor, given that two connection schemes shown below seem practical.  One concern is that the 34401A input terminal is at a significantly higher temperature than the DUT terminal.

I started by measuring the 34401A's shorted input voltage for various copper and banana lead shorts and they showed some variation from -40nV to -1300nV using the 100mV range and 100NPLC and sufficient time for temp differences to settle and get consistent averaging.  These levels are within the 100mVdc range spec limits, but obviously nice to suppress for particular measurements.

I then quickly did a comparison of the two shown 4W offset schemes on a known 20 milliohm class 0.2 shunt using basic banana plug leads, and there was negligible difference in the measured 'resistance' of circa -1.1 milliohm.  A 4W measurement of the shunt gave 19.0 milliohm, which would have then measured as 20.0 milliohm after compensation (albeit at 15C ambient).  The shunt had been measured on a Fluke 8588A at 23C as 20.0 milliohm, so I have some confidence in that value.

I hadn't seen any other description or discussion of this offset compensation method for the 34401A, and it seems quite appropriate and convenient to include in certain test scenarios, so I'd welcome any feedback/advice.

Ciao, Tim

[Kleinstein]

I would avoid moving the cables at the dut side, as this can introduce thermal gradients at that end. Idelly one would have a separate switch inbetween the meter and dut for the test current lead (e.g. input +). Handling the cables may also introduce temperature shifts at the meter.

For the very low resistors (especially good ones) it would be better to use a larger test current or bridge circuit.

P.s. I don't know if the 34401 likes it, but one could also do a reversal of the test current, by swapping the input+ and input- connection and than also use the negative reading.

[trobbins]

Thanks Kleinstein.  I'll try the swapped input polarity scheme to see how that is managed.  I'll also try the scheme using Kelvin spring clips and a shorting bar, and a 1 mohm shunt (ie. the same nominal measurement level as the offset).

The aim is just to see whether the 34401A standard measurement can be adjusted for offset, and how that relates to spot measurement comparisons - especially where the DUT value accuracy may be significantly better than the 34401A accuracy spec (like the 300-PVB).  I may get a chance to do the same set of resistor comparisons on two other un-cal'd 300-PVB's, and hence get some confidence that those three ESI instruments have undamaged resistors and that the 34401A is likely within spec.