Agilent 34970A strange behavior

[Joe99]

I've used an Agilent 34970A with 20 channel multiplexer to read thermocouples for years with no problem.  Then it starts giving temperatures that don't match other instruments, so I check it with a thermocouple in icewater, and it reads 16C.  At 50C it reads about 40C.  The farther from the reference junction, the worse it gets.  Same on all channels.  An E type and K type thermocouple on two different channels consistently match each other, but are both wrong.  When set to output DC voltage, the voltages match the temperature for each type thermocouple, so voltages and temperatures are consistent with each other, and both are wrong.  I have another identical 34970A on a shelf that hasn't been used in a few years, so I take it out and try it.  Exactly the same problems.  Reading 16C in icewater, etc.  I think it can't be the instrument since they wouldn't both go bad in the same way at the same time.  Maybe its something in the building power, so I take the one from the shelf home to try it there.  Same problems.  So then I think to just make my own calibration curve for it.  I know what the voltage should be at each temperature from the thermocouple calibration, so I start at 50C and measure T with a thermometer and voltage with the 34970A.  As I drop T, I get a curve that can be fit with a quadratic.  At 1C it reads 17C.  Then something strange happened.  I went to read the next point, and the 34970A says 3.6C when T=3.8C on the thermometer.  It suddenly went back to reading correctly.  I tried it all the way up to 50C, and the 34970A consistently reads about 0.3C below the thermometer temperature.  Anyone have any idea what happened or what the problem is?  It seems to be working right now, but I don't think I can trust it.

[nfmax]

Check the setting of the reference junction compensation for the affected channels, on the Advanced button menu - you need to use INTERNAL REF if you haven't made provision to measure the junction temperature any other way.

[Joe99]

Thanks, but that isn't the problem.  It is set to INTERNAL REF, and it is the same problem on all channels.  Also checked the DC voltages on the channels and the problem is consistent.  I'm watching it now and the readout is flipping back and forth between 44C and 51C.  The 51 is correct.  It is doing the same thing on both channels I am checking.  The internal reference is at 35C, which is about right.

[nfmax]

It isn't clear from your post, but are you using the same multiplexer module throughout? Maybe one of the bank select relays is going dodgy

[Joe99]

That is one of the weird things.  It is two different multiplexer modules plugged into two separate 34970A's with two different sets of thermocouples wires, and both are did the same thing.

[Kean]

FWIW I also use a 34970A with 34901A 20 ch mux to log data from four K-type thermocouples for days on end.  Datalogging controlled by python code over serial connection.
I have not seen this issue myself.  If it wasn't for the fact you've encountered it on two separate units I'd say it was a problem with the CJC sensor in the 34901A.

Are they both working now?  If not, what happens if instead of internal CJC ref, you set it for a fixed reference of say 25C?

[Kean]

Anything else you have connected or measuring on the same unit?
Also, what firmware do you have?

Check the service notes at https://servicenotes.literature.keysight.com/litapp/SearchSN.do?method=openExternalSNSearch&prodNum=34970A
Notes #7 and #11 could be relevant - #7 especially.

I'd also say 35C internal temperature seems a little high - recommended operating temp is 18-28C for basic accuracy, but that couldn't explain such a huge error.

[Joe99]

Thanks for this. Nothing else connected on the same unit, just the thermocouples.  Firmware is 13-02-02 on the unit and 2.3 on the 34901A.  Those appear to be more recent than the ones in the service notes.  The 35C was unusual yesterday.  It was a hot day and the room was not air conditioned.  More typical is 20C, but that doesn't change the behavior.  Right now neither unit is working.  One started working for a while yesterday, but then reverted back.  I tried setting the fixed reference, but it didn't help.

I did the self test on one unit after reading the service notes you recommended, and it is giving six errors, 612, 613, 615, 617, 618, and 619.  I need to try it on the other unit, then maybe take one apart and try to find the problem.

[nfmax]

Those error codes are all related to the ohms measurement function - it is possible there is an intermittent fault which is sourcing current into the thermocouple circuits and giving an incorrect voltage. Why both units should fail in the same way at the same time is a mystery, though.

It my help clear things up if you try a couple of simple voltage and resistance measurements, to eliminate the complications of reference junction compensation & voltage-to-temperature conversion.

[Kean]

I did the self test on one unit after reading the service notes you recommended, and it is giving six errors, 612, 613, 615, 617, 618, and 619.  I need to try it on the other unit, then maybe take one apart and try to find the problem.

Your error codes mostly relate to DMM ohms source issues, not really needed for TC measurement but clearly something problematic there.  Maybe a bad capacitor or solder joint.
Check the 7th Ed service guide at https://literature.cdn.keysight.com/litweb/pdf/34970-90012.pdf?id=258467
Self test error codes on page 168-171.  Block diagram, schematics, and parts lists are later in the document if you want to troubleshoot further.

It has been a long time since I ran a full self test on mine, and thankfully it still passes.  I have 34901A, 34903A, and 34907A modules installed.  Slightly older firmware though.

[Kean]

Why both units should fail in the same way at the same time is a mystery, though.

Knowing more about the history of the two units may help solve that mystery.

[Joe99]

Checked voltage and resistance measurements, and they were both off on the unit with the failed tests.  The error became larger with higher voltages, which was consistent with what the temperatures were doing.   It probably does have something loose causing the intermittent problem.

The other unit passed the self test.  After resetting everything, it is now working properly.  Maybe the self test or taking the mux out several times cleared the problem.  Don't know what happened, and will continue to watch it, but it looks like the problem is fixed for this unit, at least for now.  It could be the two units had the same symptoms but for different reasons.  One just needed a reset and the other has hardware problems.