Is calibrating <1ohm in 2 wire mode make sense ?

[opa627bm]

I have 4 34470DMMS, 3 are calibrated from Transcat , but when I apply the 0 ohm shot provided by keysight, they all read either negative resistance or 0.01XX ohm range.
I want to use it to calibrate a HP 5.5 digit meter, but from the calibration instruction for the 2 wire, I need to apply short to the terminal in 2 wire mode, but different kind of short yield different "0". I dont have a good short at hand so I was using a Keysight 1mohm shunt module but after calibration, with a ponoma banana plug box with silver wire short, I was reading 0.11ohm. (4 wire calibration I can do with my Fluke resistance standard and all ranges are spot on..)
I understand no one will use 2 wire mode to measure low resistance but that inconstancy in 0 ohm short reading is bugging me...
Also I noticed that 34410A and 34401A reads far better and stable with my own short (they all out of cal for at least 6 years)

[mendip_discovery]

To annoy my 1705 assessor I did a 2 wire measurment at 10 ohm with a 34401A. Only the last digit was different from the calibrated value. But that is a rare meter. This was after a basic null. I don't rely on it below 1 k but I do find it handy to check.

[Conrad Hoffman]

The spec for the meter is something like +/- 0.0030 ohm, plus 0.003% of the range (If I understand it right). Is the meter within that? Is your short low thermal emf?

I have several cables for my old 6.5 digit meter and they give different "zero" readings because of different materials. A good short never does, but you'll see things with that meter I can't.

[Berni]

It does not matter. How would you actually practically use the last digits on it?

By the time you hook up probes and poke something you already have a fair few miliohms on the probes themselves. So in pretty much any low resistance use case you would short the probe tips together and hit the NULL button on the meter before doing the measurement. So this removes not only offset caused by probe cables but also the meter itself.

Sure 4 wire is the way to do it at these resistances, yet the humble 2 wire mode can still work surprisingly well, but only once you manually null the offset away. Very useful when using just regular DMM probes to poke around a board looking for a short.

[strawberry]

probemaster probes ~50mohm
4W should not matter for 6.5digit and 10k range or above
except thermal voltages  (amplifier input sensitivity is mostly constant ~1V) therefore zeroing some uV

100ohm range is sensitive uV and nV  ( 100.0000mV ) any mechanical connection must be clean

[Kleinstein]

Many meters have a relay and or switch (e.g. front / back terminals) in the 2 wire ohms path. From these contacts there is some natural uncertainty in the 2 wire mode.
The other weak point are the cables.
To really get good low resistance readings in 2 wire mode one essentially needs a zero adjust / relative measurement as part of the measurment or just before use.
If low ohms accuracy matters one would use the 4 wire mode. So the zero in the 2 wire mode is not that import. With many meter the 2 wire ohms zero is user adjustable (or use a relative mode) anyway.

[opa627bm]

Hi All,
Thanks for the response.
The problem is during the calibration procedure (E.g from Agilent 3458A) want me to short both input then connect to a resistance standard.
but depends on what kind of short you have, your offset will be different.

for the 34470 case, we have transcat calibration and now with agilent factory DMM short block, in 2 wire mode, we have reading from -0.2ohms to -0.09 ohms and I am not sure if that suppose to happen...

[HighVoltage]

The 34470A is not a meter to be used for low resistance measurements in not even in 4W mode but definitely not in the 2W mode. The older 34410A is much more stable in this regard.

Your observation of  "reading from -0.2ohms to -0.09 ohms" with a short is very typical for the 34470A. Again, the 34410A will give you more reliable and repeatable values.

[opa627bm]

The 34470A is not a meter to be used for low resistance measurements in not even in 4W mode but definitely not in the 2W mode. The older 34410A is much more stable in this regard.

Your observation of  "reading from -0.2ohms to -0.09 ohms" with a short is very typical for the 34470A. Again, the 34410A will give you more reliable and repeatable values.

Hi Sir,
Thanks for the reply.
Do you know why 34470A is behaving like this? I constantly getting more stable reading from 34401 and 34410. We are budgeting to get new meter this FY year and I am researching what should we get to displace 34470A...

[HighVoltage]

Hi Sir,
Thanks for the reply.
Do you know why 34470A is behaving like this? I constantly getting more stable reading from 34401 and 34410. We are budgeting to get new meter this FY year and I am researching what should we get to displace 34470A...

I had several 34470A and had the same observation in regards to the 34410A and even the 34401A.

We discussed some issues here:
https://www.eevblog.com/forum/metrology/keysight-34470a-calibration-and-acal-problems/

The 34470A is a good meter for many applications but you need to know its limits.
It is essentially a 6 1/2 digit meter, showing one extra digit.

If you want a 7 1/2 digit replacement, look for the Keithley DMM7510, they are far more stable than the 34470A. Even the smaller 6 1/2 digit Keithley DMM6500 is more stable than the 34470A, especially for low resistance measurements.