Electronics > Repair
Keithley 2002 repair help
<< < (8/43) > >>
nikonoid:

--- Quote from: Kleinstein on June 08, 2017, 07:18:14 pm ---Measuring leakage / input current is relatively easy. There are two options:
1) use a 10 M or similar resistor across the inputs in the volts (e.g. 2 V) range.
2) have a low leakage capacitor (e.g. 1 nF) across the inputs an watch is charging / discharging.

--- End quote ---

1) 10M standard resistor on 2V scale is giving me around 60uV if I stay 3 meters away and stay very still.

2) I did not have a 1nF capacitor handy, so in a pinch I used 200nF GenRad 1409 capacitance standard. It showed about 1.4V and was discharging approximately at .1mV per second.

Is this good?


--- Quote from: Le_Bassiste on June 07, 2017, 08:26:47 pm ---did you try to slightly(!) knock with a popsickle stick on the board? can that provoke a jumpy ohms reading?

--- End quote ---

Mechanical knocking on the board does not change a thing.

I am planning to do some offset voltage plotting after I get GPIB connection working.

The search continues...

Is there a way to test Q211, Q212 and Q213 in circuit? I do not want to desolder them unless I have to.


Kleinstein:
60 µV at 10 M is only 6 pA. 0.1 mV/s for a 200 nF capacitor would be 20 pA - this includes capacitor leakage and maybe dielectric absorbtion.
So leakage seems to be very low - usual specs are something like below 50 pA.

Checking JFETs in circuit could be difficult - the error still seems to be more intermittent or just marginal. So one would not a expect a fully broken part, but more like a parameter shift, excess leakage or unwanted oscillation. This might be even hard to detect on a separate part.
It might be worth checking the drive signals, that control the JFETs - there have been faults in the LM339 chips and even the shift registers used for control. The fluctuating state might also be due to oscillation - so maybe time to use a scope.

It might also help to get at least a crude schematics of the suspected area. From TiN's tear-down the circuit seems to be similar to the K2001 meter for which schematics are available.
nikonoid:
Thank you fro validating leakage spec.

I really like the suggestion to use scope. This will be my next thing to do.

So far I validated R322, CR217, K201, R334 and R342 (parallel to L206). I used every resistor in the chain as if it would be a current sense resistor, to make sure it is about the same wrong current going through all of them, when device is misbehaving.

I doubt it is the return circuit responsible for the issue (Q246, Q245, Q244, and Q242).

So I am left with everything prior to R322 to look for problem.

The repair manual is referring to /OHMD line, OHMREFHI and REFBUF. What are they? Thanks.
Kleinstein:
There is quite some description in the repair manual:
The /OHMD signal is a digital control signal to set the current source to 2 µA. Due to the / in the name, likely active low. OHMREFHI and REFBUF are a 14 V signal and a buffered 7 V reference.

The 9.6 µA and 2 µA current source seem to use many of the same parts. So one might exclude those used for the 9.6 µA current source. The change from 9.6 µA to 2 µA seems to be changing the reference voltages - thus not in the super higher impedance part. My suspect would be more like U232. It would be really worth checking those ref voltages used for the current source. It may help to make a drawing of the current source - there is quite some descriptions to help with this. However it seems to be a little different from the K2001.
nikonoid:

--- Quote from: Kleinstein on June 09, 2017, 04:37:56 pm ---The 9.6 µA and 2 µA current source seem to use many of the same parts. So one might exclude those used for the 9.6 µA current source. The change from 9.6 µA to 2 µA seems to be changing the reference voltages - thus not in the super higher impedance part. My suspect would be more like U232.

--- End quote ---

You are right on the money! When I checked U232 I discovered that when meter is properly producing 1.98uA in 2MOhm range, on U232 I get:
Pin 9: Low
Pin 10: 12.7175
Pin 11: 12.7154
Basically, Pin 10 connected to Pin 11, as they should.

When current is misbehaving, I get:
Pin 9: Low
Pin 10: 12.7175
Pin 11: FLOATING!!!!

To me it is a clear proof of U232 fault. Thank you, thank you, thank you!


Additionally, I checked U221 register that according to manual supped to be: 11101001, but in my case is 11101011 regardless of meter working properly or not (Bit 2 is different).  It is possible that my firmware is too old and manual was based on new firmware.

Also in checking the repair manual in details, I found surprising number of inaccuracies, like U232 is listed as "CENTRIGRADE TEMP SENSOR LN35DM", instead of "IC SWITCH QUAD SPST 16SOIC".
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod