Keithley 2400 SMU - Output + FET-temps

[annieIO]

Hello everyone,
this is my first time posting here

I just started populating my little bench of rescued devices, and so far I was quite lucky: most devices were quite easy to fix.

But I am stuck with my newest addition: a beautiful Keithley 2400 that is optically in a perfect condition - see some pictures I added down here.
I think it is quite interesting that there seems to be two botch resistors over Q232 and Q233, can anyone say something about them?

The actual problem I have with this unit is that the output seems to be enabled all the time: right after the unit is done with it's start up self test the voltage on the outputs reaches 92mv and I can measure a constant current of 222mA across the output terminals. Keep in mind that the output is still not turned on.

When I turn on the output, the 2400 displays an error stating that the compliance current exceeded the programmed value.... which is true I guess.
When other measurement or sourcing ranges are selected on the K2400 the output voltage increases up to 92V wich is a bit scary...

I disassembled the unit to the point shown in the pictures to rule out any obvious component failures. I was only able to notice that the two N-Chanel IRF630 FETs Q514 and Q516 are reaching *very* high temperatures (it starts to smell of hot components and IPA immediately boils off of it when dropped on it). 

My assumption at this point is that this FETs are always driven at their fullest, no matter the settings?
Since I am not able to find schematics for the analog board I'm not sure how to continue from here, am I not looking at the right places?
Can someone point me in a promising direction?

My unit runs on firmware 33 btw

[openloop]

Have you checked voltages as the scripture tells us?

It seems that high voltage is enabled when it should not be.

With the output off:

Check that U500 is powered: pins 4 and 7 (relative to SMU's positive terminal TP500 "floating common" - U500 is bootstrapped). Don't remember voltage, but should be reasonable.

Check U500 inputs: 2 should be at zero (relative to TP500), 3 about the same, but probably not - what is it?
What's U500 output voltage?

(U500 is the op amp that drives the output totem pole of transistors.)

If U500's output is not pinned to the rail, check the drain of Q514: it should be at about 30V (relative to the _negative_ terminal). Be careful!

[annieIO]

Hi!

I finally had enough time to look into this again...

I did what you mentioned and I think the first topic to worry about is that the voltage between TP218 and TP219 is reading ˜15V. TP218 is labelled +5V and TP219 -5V.
when I check them against TP501 0COM I get +/- 7.5V. I think this is out of spec but this test points aren't mentioned in the service manual so I am not sure about this.

I checked all the voltages mentioned in the service manual for the digital & power supply board. all of them are inside their tolerances. but the power for the analog board is also generated on the analog board itself, right?

What would be a good way to tackle the power supply issue?

[BennoG]

See if the capacitors have leaked.
That is the most common cause of high voltage.

Benno

[annieIO]

Hi!

Thank you for your response!

I ordered and replaced all electrolytic capacitors after reading your comment.
Now reassembled the device but sadly no change in the behavior. Like non at all, same error message and the test point voltages are also unchanged.

[MadTux]

Likely the usual dead output stage, I'd look at K236 schematics as reference, 2400 should be somewhat similar.
And there might be some leaked/DIY schematics available at xdevs.com

Output stage isn't easy to repair, since all mosfets are in a chain between V+ and V-, some slightly degraded mosfet and output stage gets out of balance, which might mess up the mosfet next to it. Especially on HV models like K237.

[BennoG]

have you looked over here ?

https://xdevs.com/fix/kei2400/

Benno