Help me repair a Keithley 2000

[carpelux]

According to the Service manual it's a MMBZ5247BL 17V sot23 Zener diode.

[ddavidebor]

put a damn 17v zener of the same wattage and live happy

[mimmus78]

I found same part series but 18V at it.rs-online.com (only 6€ shipping), it should be ok just for checking if it solve this noise problem.
I checked opamp and 18V is max rating, so it should be OK also as long term fix, but I need to check specs better.

[mimmus78]

Just some updates.

After changing the "zener voltage regulator" the voltage is now under control but noise was still too much.

I decided to check if opamp was burned by this 24V, so I removed it and:



I cannot believe this puppy after all my abuses is still spot-on :-)

I'm going to spend a lot of money to repair this meter, the opamp is 7€ plus shipping, and I already spent 30€ for changing relay and the zener ...

[Galaxyrise]

Very nice

[mimmus78]

Some updates.

I removed all components relative to current source U123 and now I have only errors in tests where U123 is engaged (starting from 300).

I found also that by connecting the shield of both meter (TP102) I can use the voltage source of the working meter to measure Ohm with the other ... both meter read almost the same value on all the ohm ranges (and I like it).

I ordered a bunch of components including this U123 ... we will see what happens when all the stuff arrives.

[Galaxyrise]

Since this isn't the only bad VR109 thread, I figured I should check the ones in mine.  Of the two K2k on my bench right now, one had a good VR109, the other seemed bad.  So I replaced it with what seemed to be the same thing (and then promptly lost the original, doh!)  But now the meter is "noisy" (fails 100.2), and indeed, compared to the working k2k, the voltage has some AC coming through it.  (In both  cases, there's a strong sine wave component coming into it.) Since it wasn't failing that test before, it seems likely that I used the wrong part, but, what was the correct part, then?

(The repair manual calls it a "DIODE, ZENER 17V, MMBZ5247BL (SOT-23)")

[mimmus78]

Sorry but I really don't know ... when the meter was sitting down to the bench it build up another problem (ADC always overload) so I decided to not continue anymore with it.

[Galaxyrise]

Thanks for the reply!

I've actually made some progress on this; I'm waiting on some parts to arrive before I'll know more.  But what I know is that the zener I put in was running way too hot.  The resistor in series with it (R168 iirc) is only 270 ohms, and that put my zener at its max wattage. Maybe you should check yours? Curiously, my other k2k has a 1k for R168, even though the repair manual specifies 270.

[mimmus78]

On the K2K that I repaired this zener is just worm.

This part of the circuit around R168 is the current source for ohm measurements.

Removing Q119 fixed noise on my meter but it will "disable" your ability to measure ohm (you will get also related errors).

I tried to understand how this current source works but my knowledge and time where not sufficient for it.

[flolic]

On my K2k voltage measured across the VR109 was 18.6V, so I replaced it with regular small glass 16V zener diode. Now I am measuring little less than 17V across it. Diode is getting quite warm, and I will probably replace it with bigger 1W one.

I have some 0.01% resistors for testing, and my meter is spot on on low ranges. But on higher ranges it reads little higher than it should, 100.001k resistor reads as 100.015k. It can be simple calibration issue, or maybe there is fault somewhere... I must do some more tests.

Then there is also noise issue visible if I disable filtering. Last digit or two randomly flickering up or down. I can't find some pattern in that behavior, sometimes it is more and sometimes it is less noticeable... 

There are some more issues, I will document them after more testing. But I think they are related with calibration.
I've dumped original content of calibration eeprom so I can safely play with calibration.

My K2k passes self test without error.

[Galaxyrise]

What's the voltage from R168 to ground? Mine was around 23.2.  If that's high, it could be the reason my zener died in the first place (and 270 Ohms would make a lot more sense.)

[Galaxyrise]

Then there is also noise issue visible if I disable filtering. Last digit or two randomly flickering up or down. I can't find some pattern in that behavior, sometimes it is more and sometimes it is less noticeable... 

Yeah, that's normal.  But if you have the meter sample the same voltage 1000 times at the 10V range, you should get a std deviation of like 3uV and 20-30uV between the min and max reading.    My noisy meter gets a std dev of 12-14 uV and max-min will be 70-100uV.

[flolic]

I will check and report those values later when I get home 

[macboy]

On my K2k voltage measured across the VR109 was 18.6V, so I replaced it with regular small glass 16V zener diode. Now I am measuring little less than 17V across it. Diode is getting quite warm, and I will probably replace it with bigger 1W one.

I have some 0.01% resistors for testing, and my meter is spot on on low ranges. But on higher ranges it reads little higher than it should, 100.001k resistor reads as 100.015k. It can be simple calibration issue, or maybe there is fault somewhere... I must do some more tests.

Then there is also noise issue visible if I disable filtering. Last digit or two randomly flickering up or down. I can't find some pattern in that behavior, sometimes it is more and sometimes it is less noticeable... 

There are some more issues, I will document them after more testing. But I think they are related with calibration.
I've dumped original content of calibration eeprom so I can safely play with calibration.

My K2k passes self test without error.

If you are testing the 100.001k? with the 1 M? range, then it is measuring within spec at 100.015k?. If you are using the 100k? range, then it is slightly out of spec (spec limit would be 100.011k?). This would likely be calibration drift, not a fault. If it was measuring say 130k, then you would suspect a fault. The resistance measurements are dependent on internal resistance references, and these do drift over time as they age. Aging of other components like the FETs, analog switches, and relays in the signal path also affect accuracy over time.

As for noise, this is normal, depending on the settings. You can to use the 10 PLC integration time to reduce noise, and averaging helps too. You need to decide on measurement speed vs. noise. In my experience, at 1 PLC, I see 5.5 digits essentially noise-free (not with Keithley 2000 but with Keithley 2001 and Keithley 199). If there is excessive noise, that may be a fault.

Making any repair can throw out the calibration. Even new fingerprints on the PCB can have an impact (no joke).

[flolic]

Yes, you are right, those values fall within the specs 
In fact, I must correct myself. Actual resistor value is 100.06k and I am measuring 100.174k on a 1M range and same on 100k range. That's even better.
But 10.002k resistor reads 10.0133k on 100k and upper ranges, and 10.00311k on 10k range. The similar happens with 1.0001k resistor. I don't like discrepancy in reading between ranges, and that can probably be fixed with calibration.

Another thing, on DCI mode i have cca. 5uA reading on all ranges, with test leads disconnected.
 

[flolic]

What's the voltage from R168 to ground? Mine was around 23.2. 

I read exact that value, so it looks ok 

[Galaxyrise]

I read exact that value, so it looks ok 

Thanks!

I don't understand how it was supposed to work originally, then... I don't see anything around that voltage bigger than 350mw in a SOT-23 on digikey or mouser.  What kind of zener did you use, mimmus78?

[flolic]

Yes, that poor little diode is pushed beyond maximum allowable disipation (377mW if I calculate correctly).
I think all sot-23 zeners are limited to 350mW dissipation, that housing is simply too small to efficiently sink more heat.
First next more powerful devices are 500mW diodes in sod-123 housing.

[mimmus78]

I remember was not able to find an smd with same specs in Italy so I a bought a tht zener to check if replacing it would fix the problem, but it does not worked.

[Galaxyrise]

I put a 5ppm 1k resistor with a low Iz 16V zener, and that seemed to settle the voltage across VR109 a lot.  Both parts run about 40C with the case open.  But there's still ADC noise.  I replaced U123 just in case it was causing some kind of supply noise, but the ADC noise remains unchanged.

It's probably something I did bodging R271, but I'm conceding defeat; I found someone who would be happy with it as a 5 digit meter.

[kubatyszko]

//A bit of post necromancy but better to keep it relevant.//

Here's a short blog post on repair of my Keithley 2000, it was initially failing with 600.1, 600.2, 601.2 errors.

http://kubatyszko.com/2015/08/10/keithley-2000-bench-multimeter-repair-part-2/

Cheers

[TiN]

Thanks for post.  Seems lot of relay problems people discovering lately