My Keithley 6517A electrometer randomly reboots, any ideas? (+photos)

[tty]

I've just acquired a secondhand (from USA) Keithley 6517A. Switched from USA to NZ power and changed the fuse. It is broadly working OK, but with two annoying problems. It's a nice enough piece of equipment I would like to make it work perfectly and reliably (but sending it to Keithley is not attractive due to high shipping costs from the bottom of the world)

1) The unit randomly reboots every hour or so.
2) Current displayed when nothing is connected to the output is ~30 fA and drifty

30 fA is worse than it should be (~3 fA) but I'm hoping it might just need some time to settle down after turning on. I did try cleaning the top side of the input circuit with isopropanol, which made very little difference (less than I thought it would, after reading some posts here suggesting after the isopropanol is removed it might take a long time to reach normal operation).

But because of the random reboots problem it won't just stay running a long time without needing attention. Does anyone have any hints on how to fix? I can't spot any signs of capacitors bulging and leaking, and the power rails do not look particularly ripply. Firmware is C05. I will attach some photos.

[Smith]

The offset problem will probably be related to the input circuit being dirty.  Cleaning it with IPA and letting it dry thouroughly should do the job. Mine had exessive noise and offset, and after a good clean and letting it dry in an dry cabinet all was well below 3 fA. If I remember correctly, there is an calibration option for zeroing the input in the device too. This might help you get the offset down aswell.

I cannot help you with the resetting. But as normal, check voltages first (but beware of the 1200V or 1050V rails, cant remember exactly).

Are you sure it's rev C05? It does show on the eeproms, so it must be true. The C05 firmware I got from Keithley was  non functional,  and the original C04 was fine. I never saw an actual unit with this version. Even Keithley told me they never officially released C05.

[tty]

Thanks Smith, I've cleaned it very carefully with isopropanol, blow dried with R134a air duster, then oven bake at ~60 for 1 hour. About how long did yours take to stabilize / leave it in the cabinet after rinsing? Immediately after this, mine is at -80 fA.

It's definitely C05, says this on boot as well as on the eeprom. Did yours not boot at all?

[TiN]

Does it reboot if you smack it a little?

I had issue with customer's 6517A, ended up dodgy ribbon 26-pin cable between analog and digital boards. After replacement it start to behave nicely.
Of course make sure ALL electrolytic caps are replaced on all boards. 

If you can backup that C05 firmware into our collection, I would appreciate beyond limits. 

[tty]

Sure, I will have a look at reading the firmware after sorting out the other things, I guess I'd need to find that kind of wide old IC socket
I'm plotting the bias current now, it moves agonisingly slowly.. so far not rebooted. But I put it somewhere far away so I am not tempted to take it apart and clean it again, instead plan to wait ~few days to see if the bias current improves.

Do you always replace all the capacitors or only when they have problems? I'm curious how old it is, because it looked pretty new on the inside, but more worn than I thought an electrometer could be on the outside (like, printing worn off buttons, plastic yellowed).

Really like your website BTW! I'm glad I joined this forum.