Keithley 2000 errors 400.2 and 401.2

[Ferrix97]

Hi!, I've just saved this multimeter (along with a couple more toys, more on that later) from a recycling center, it was in their pile of things to destroy and was stored outside in the rain.. thankfully there were some printers and other junk on top of it so it din't get all that wet.

When I run the self test from cold I get the following errors: 303.1, 400.2 and 401.2
If I let it warm up 303.1 disappears, and 402.2 appeared once.

I did a few basic measurements with it and it seem the higher ACV ranges are off, I'm getting about 1.7V with the two probes shorted.

Inside the unit is clean and there's no capacitor gunk. The cal seal was broken and somebody possibly attempted a repair on it (there are marks on the back nuts on the GPIB connector). The multimeter says it has been calibrated 3 times, and last one in 2014, But the cal sticker reports a date of 2016...

I had a look online and I saw that some resistor networks usually fail in them, In case I have to replace them do I have to re-run the calibration? I don't have instruments capable of calibrating it.

[Ferrix97]

I took the board out of the case. According to what I've read online this is the "later" board revision.

Indeed the unit has been repaired before, one of the relays has been replaced, there's still some flux residue next to the solder joints.

I've been comparing the measurements with a (presumably working) Fluke 8840A that I also scored in the same lot and sometimes in the 100mV and 1V rages the two meters are close (but that might be a coincidence).

EDIT
100mV and 1V ranges are good, when I switch to the 10V and higher ranges the reading is 100x what it should be: Fluke on the 20V range: 0.035V, Keithley on the 10V range: 03.5293V. When Fluke displays 0.35V, the Keithley overflows and displays 35V on the higher range.

According to the service manual, K102 changes state when going from the 1V range to the 10V range, maybe the relay failed again? I can't hear it click when switching between the ranges.

[Samogon]

Try to  wiggle front/rear switch and run self test again. Do it few times and note what errors you get.
Sometimes this switch gets dirty and many errors comes from its unreliable contacts.

[Ferrix97]

Thanks!, I flipped the switch on and off a few times. I ran the test a couple times and in one case the only error I got was 401.2
Sadly I don't have any contact cleaner with me, so I can only keep pressing and depressing it...

Do you you hear a relay click when switching from 1VAC to 10VAC range? or when running the 4xx tests

[Samogon]

Yes relay should switch on 1v <=>10v range change on ACV.

[picburner]

The K102 relay has been replaced with an incorrect type: it should be 2 coils latching
type like K101 and NAIS TQ2-5V are not.

[Ferrix97]

Thanks!, I checked the datasheet and indeed the relay type is incorrect!, TQ2-L2-5V should be the correct part number.
I'll add it to the cart and order it soon, will post updates once it arrives!

datasheet: https://docs-emea.rs-online.com/webdocs/0027/0900766b80027b9b.pdf

[picburner]

Check also that the minimum switchable voltage is equivalent to that of the NEC EA2-5TNJ relay (the original).
In this application it is more important than the maximum values.

[Ferrix97]

They both have a set and reset voltage of 3.75V. I also took a look at the service manual, K102/K101 are listed as TQ2E-L2-5V.
I could not find the TQ2E online, but only the TQ2-L2-5V and TQ2H-L2-5V, according to another datasheet looks like they just have different terminals.

http://www.uni-kl.de/elektronik-lager/447126

[picburner]

I meant the minimum commutable current/voltage on the contacts, not on the coil: it should be 10uA/10mV DC or less.
However, it seems to me that the TQ2-2L-5V goes well and it is pin to pin compatible also.

[Ferrix97]

Sorry, the datasheet calls this "Min. switching capacity", and the TQ2-L2-5V is rated for 10uA/10mV.

[Ferrix97]

Updates!
The replacement relay finally arrived and it solved the problem!

When I replaced the relay I also sprayed some contact cleaner into the front/rear switch. Initially the unit would throw all sorts of errors, I then used some compressed air to dry off all the cleaner and it passed all tests!

[Samogon]

Good job!!!
It is nice meter