Keithley 2400 always on with ~ -240.0vdc

[Cyclotron]

I picked up this K2400 as parts and repair on the ePay. The firmware was corrupt, which was an easy fix for the no-boot issue.
Now that I've had a chance to start function verification, I've found that it always has about -240 VDC present on the Input/Output jacks.

I can adjust the voltage and current, and it does output voltage and current, but not at the set values. It won't go through calibration as the outputs are too far off.

It doesn't appear that a schematic is available, though the service manual is and has some good information. I also see that others have been through this sort of issue, though I don't see anything that might indicate a repeatable solution for the issue.

Voltage validations on the digital board are done. I need to run the same on the analog board. I did a thermal image, and U660 is smoking hot. A couple of other chips are relatively warm: an OpAmp and a 74HC409.
I figured I would open a thread to document anything I discover and any possible solution I can find.  And, if anyone happens to know a schematic or potential solutions, I'm all ears.

[Cyclotron]

Well, I thought U210, an LT1124CS8, was bad. An obsolete opamp, so I swapped it out with an OPA145.
The constant -240 VDC output was converted to a continuous +98 VDC output. So I put the original back, and the original symptom returned.

Wish there were schematics available. This poking in the dark is time-consuming.

U500 appears to be doing what it's been told to do, so that side of the board is working as expected.
On a guess, given someone else's experience, I swapped out U660 and U661 for known-good parts.  No changes.

I am going back to U210 and trace things back from there I guess.

[Uunoctium]

Digikey lists LT1124CS8#PBF as direct replacement and has 44 pcs in stock

[Cyclotron]

Digikey lists LT1124CS8#PBF as direct replacement and has 44 pcs in stock

Thanks, I originally misidentified that OpAmp. I've now ordered some replacements.

[Uunoctium]

I can't comment on the device at all—it's beyond my price range. 
But did you test, if an interconnection between sense lines and the IN/OUT jacks changes that behavior? The priciple diagram shows, feedback circuit for source mode works thru the sense.
Don't know, if it helps, but I found an interestig thread in which is mentioned, this SMU has always a Voltage at jacks. They give an solution to complete shutoff the source by remote controlling an sweep, wich starts at 0V and ends at 0V.
https://forum.sweep-me.net/t/switching-off-source-mode-in-keithley-2400/158

[Cyclotron]

I can't comment on the device at all—it's beyond my price range. 
But did you test, if an interconnection between sense lines and the IN/OUT jacks changes that behavior? The priciple diagram shows, feedback circuit for source mode works thru the sense.
Don't know, if it helps, but I found an interestig thread in which is mentioned, this SMU has always a Voltage at jacks. They give an solution to complete shutoff the source by remote controlling an sweep, wich starts at 0V and ends at 0V.
https://forum.sweep-me.net/t/switching-off-source-mode-in-keithley-2400/158

Yes, U210 appears to be fed by TP220 and TP222, which are part of the sense feedback.

I'll take a look at that thread and see if anything pops out as applicable.
Thanks for jumping in.

[MadTux]

Already checked those chained together transistors on the output stage?
On Keithley 236/237, it was common fault, that transistors got degraded and then the whole balanced output stage got stuck at some voltage.
K2400 output stage is similar to K236, but much less accessible.

[Cyclotron]

Already checked those chained together transistors on the output stage?
On Keithley 236/237, it was common fault, that transistors got degraded and then the whole balanced output stage got stuck at some voltage.
K2400 output stage is similar to K236, but much less accessible.

I have walked through some. So far, a poke-and-hope as I study the board. TiN has some xray images on Xdevs from some of the boards through his shop. I've been studying those.  I need to download the schematics for the 236/237 and study those while I'm traveling this week. I've heard the output stage is very similar there. I have a thermal image of the board but I didn't really get much from that other than a couple opamps running warm.

I'll be suffering through the beautiful weather in Chicago this week, so there are lots of reasons to stay in my hotel and study a few schematics.

[Cyclotron]

Primary symptom resolved. Q229 a Siliconix VN0605T was bad. Replaced it  with a 2N7002E and the hard -240vdc constant output is gone.
Still a small issue. I have a -5 constant voltage output now and I'll need to sort that out.

But I'm taking the "W" on the major issue being resolved after nearly a year.   

[KungFuJosh]

A W is a W! 

[Cyclotron]

And it is completely clear now. The precision opamp U222 was gone. Likely a victim of the high voltage flying around in the instrument when it was swung against its low rail?  Anyway, replacing the LTC1050 did the trick and the SMU is able to CAL against my K2002.

I have a display for it out of another instrument and it should be back to great shape.  Glad to get this fixed as its been one of my long term projects.