Keithley 2400 SMU - shows OVP but works

[TheSteve]

I bought a pair of Keithley 2400's the other day. One of them has some cooked parts and a broken VFD. The second one works fine in all respects yet always shows OVP (over voltage protection) when the output is switched on.

-The OVP protection is of course switched off, and when enabled actually works fine on all ranges.
-The unit has been fully reset.
-Firmware has been upgraded/downgraded.
-I have tried the analog board with a second digital board to confirm the problem is on the analog board.

The OVP protection seems to be fully handled on the analog board, I assume when it detects an OVP condition it sets a flag of some sort and sends it to the digital board to display. Whatever sets that flag is stuck.

If the schematic was published this might be a pretty quick/easy fix, but without it there is more challenge.
Assuming I can get the second 2400 fully working I may consider probing various points while causing a true OVP to attempt to find a line that toggles and work backwards from there.

Until then anyone care to take a guess as to where the problem lies?

btw, both the analog and digital board are reasonably early, they are rev E.

[kawal]

Did you get this resolved?  Was the meter on 4 terminal  measurement  ?

[goaty]

+1 for releasing the schematic.
I'm fighting with a 2420 where I replaced some cooked parts too.
But then it's still not working and from there on finding the problems is hard...

[TheSteve]

The problem was resolved. The 2400 worked, the error indication was false and was a bad transistor sending a flag to the digital board through an opto.

[kawal]

+1 for releasing the schematic.
I'm fighting with a 2420 where I replaced some cooked parts too.
But then it's still not working and from there on finding the problems is hard...

The closest to schematic is the XDevs 236 manual with schematic.


Also read pages 90 to 101 very helpful to understand  how these work. I attached  the principle of operation below.


And this attacchmet

[kawal]

+1 for releasing the schematic.
I'm fighting with a 2420 where I replaced some cooked parts too.
But then it's still not working and from there on finding the problems is hard...

What problem do you see on your unit?
Mine seems to have broken current measurement loop and broken NIDAC and overheating NVdac  - ordered DAC but want to fix the loop issue before replacing.

https://www.eevblog.com/forum/testgear/keithley-2400-sourcemeter-review-and-teardown/msg2940898/#msg2940898

[goaty]

Mine had a dozen or so parts blown (totaly destroyed) from the 12.1Ohm shunt resistors, the mosfets, several of JFets, the analog switches ics (so14).
Which I replaced all. I think someone connected a high power source to the input which sank a high current.
Now all user interface is ok, power is ok (85V+-, etc.) but I cannot hear relais clicking when enabling output and foremost there is no output voltage whatsoever and display reads random values in Volts and Amps.
For now I have just been probing around for defective analog parts (diode, transistor) using multimeter and thermal camera.
But problem is understanding the circuit is very hard (I am software engineer, not hardware), and there is no schematic.
I will continue again with trying to measure the DAC outputs and maby understand which ground is for what (FCOM, etc).
Usually I repair simple stuff, but I though I can do this, but this is way over my current knowledge.

I tried to get this "LeBassiste" schematic, but alas no response.

If you need thermal images, I can make some for comparison, but it´ll take a while since the device is currently packed away (due to frustration)

I put picture of mine here in this german forum
https://www.mikrocontroller.net/topic/478884

[kawal]

Mine had a dozen or so parts blown (totaly destroyed) from the 12.1Ohm shunt resistors, the mosfets, several of JFets, the analog switches ics (so14).
Which I replaced all. I think someone connected a high power source to the input which sank a high current.
Now all user interface is ok, power is ok (85V+-, etc.) but I cannot hear relais clicking when enabling output and foremost there is no output voltage whatsoever and display reads random values in Volts and Amps.
For now I have just been probing around for defective analog parts (diode, transistor) using multimeter and thermal camera.
But problem is understanding the circuit is very hard (I am software engineer, not hardware), and there is no schematic.
I will continue again with trying to measure the DAC outputs and maby understand which ground is for what (FCOM, etc).
Usually I repair simple stuff, but I though I can do this, but this is way over my current knowledge.

I tried to get this "LeBassiste" schematic, but alas no response.

If you need thermal images, I can make some for comparison, but it´ll take a while since the device is currently packed away (due to frustration)

I put picture of mine here in this german forum
https://www.mikrocontroller.net/topic/478884

The Partial Schematic of the power stage from "LeBassiste" did not match the actual unit that I have so watch out. This is not the fault of the user "LeBassiste" as the PCB is 4 layer and very hard to figure out because of that.
The Best you can do is use the 236  schematic are reference  and  they are mostly the same in principle but used different component types.  Best advice is to take the block diagrams and schematic from 236 and apply to the 2400. Attached are just these pages.   

The analog section has to be measured to FCOM with the exception of the output stage that measures to OCOM. So basically all probing for voltages will need to be in reference to FCOM .  Do not connect any earth grounded equipment as it will damage parts (  Thats what I think I did and damaged DACs's) which are now on order.

The complication with this unit is the number of switches in the device. All these DG444, optoMOs,  Jfet,  Mos  and relays  are needed for range and quadrant switching.  these switches is the main difference between the 236 and 2400's analog section.

I think It is quite complicated but the more I look at it the simpler it becomes.  There are  Test points (TP) on the board and a lot of them have names which helps.  Some do not have names,  but user "Openloop" has experience with these and knows some of the Test Points without descriptions.
Dont give up.

In the German forum they asked you if you checked the Dacs.  These should be marked with NIDAC and NVDAC  close to the blue resistors in the middle of the board.  These will output negative voltages based on your setting. Anything  from 0V min to -10V max  but only works when output is enabled.

Check the feedback voltages  there is one for Voltage VFB and One for current IFB ( Middle , lower  side of the board  closer to relay). Mine has error here on the IFB  always showing the same number and leads be to hunt this area. Maybe you can narrow down the issue to concentrate on the part that is actually left broken.  One by one you will fix it.

[Le_Bassiste]

The Partial Schematic of the power stage from "LeBassiste" did not match the actual unit that I have so watch out. This is not the fault of the user "LeBassiste" as the PCB is 4 layer and very hard to figure out because of that.

 the original schematic i did upload was actually labelled Keithley 2410, not 2400 see here:https://www.eevblog.com/forum/repair/keithley-2400-in-reset-loop/msg1630664/#msg1630664.
so i'm actually not understanding why one would try to use it to repair a Keithley 2400.

[kawal]

 
[/quote]

 the original schematic i did upload was actually labelled Keithley 2410, not 2400 see here:https://www.eevblog.com/forum/repair/keithley-2400-in-reset-loop/msg1630664/#msg1630664.
so i'm actually not understanding why one would try to use it to repair a Keithley 2400.
[/quote]
Le_Bassiste
I was so glad I found that little piece as it got me started. Thank you .
I know it was for a different model but that way better than starting from zero.  The whole family is based on the same concept as the 236/237/238
Now I verified your 2410 schematic against the early 2400 i have an we have 2 schematics. Better for all.

[Le_Bassiste]

I was so glad I found that little piece as it got me started. Thank you .

you are welcome.

Now I verified your 2410 schematic against the early 2400 i have an we have 2 schematics. Better for all.

"better for all" that you stopped faking the caption of the original schematic so people would think that i had done the 2400 version wrongly.
that level of ignorance is absolutely beyond me. 
to get this straight: whatever i'm posting here i basically don't care about what you are using it for. so, fingers crossed, you get your unit up and running again!
however, once and for all: don't ever f*ing again change anything in any of the contents i'm posting and then repost it without further clarification of what exact changes you did to it. 
clearly state what you have changed in it, and that you are the author of such changes (which you, albeit silently, did now.  )
there is a reason for doing it that way, it's called your responsibility.

[kawal]

I was so glad I found that little piece as it got me started. Thank you .

you are welcome.

Now I verified your 2410 schematic against the early 2400 i have an we have 2 schematics. Better for all.

"better for all" that you stopped faking the caption of the original schematic so people would think that i had done the 2400 version wrongly.
that level of ignorance is absolutely beyond me. 
to get this straight: whatever i'm posting here i basically don't care about what you are using it for. so, fingers crossed, you get your unit up and running again!
however, once and for all: don't ever f*ing again change anything in any of the contents i'm posting and then repost it without further clarification of what exact changes you did to it. 
clearly state what you have changed in it, and that you are the author of such changes (which you, albeit silently, did now.  )
there is a reason for doing it that way, it's called your responsibility.

WOW  - That is  what  I call culture.  Did you notice the topic is 2400 ? If you dont want to help out just leave the forum.  No need to stay.   Are you claiming to own the schematic to Keithley 2410  that you  reverse engineered ?  All I am claiming is that a 2400 is different from your schematic  and  i posted  the modified schematic that matches my 2400.  Whats the big deal  I wrote that its for 2400 unit .  Thanks for the lesson.

[goaty]

You should learn German and join the "Microcontroller" forum here, it´s - let´s say - interresting.

Thanks for giving me hope. I was starting out hopefully, but then frustration came ;-)

The schematic 2410 from LeBassiste helped me around the blown 12Ohm resistors and FETs. Also I made "xray" photos of the board, as it has no GND plane, it´s somewhat possible to look through.

I will try and start again measuring test points.

[kawal]

Nice "Xray" picture . I tried to figure out the tracks that way too but it became too hard to follow.  I ended up using a combination visual inspection and resistance testing.  I would say we both need to stay calm about getting these fixed and it is definitely possible.  In my case I clueless unitll i started going through modes and ranges. I was doing measurements on the unit and noticed that in Current mode the unit controls voltage well (But the actual current is portion is not working) . That narrowed my search to first current compliance and now to current measurement section. 
You should definitely  check to see what is working and that could make it easier to find what is not working, Check the Reference  voltage (always on when unit on).  Then check if the NVDAC and NIDAC outputs are working ( only works when output on ). To do this change the output settings and compliance settings and measure these at a few different settings. These should be 0..-10V  range depending on the  range and settings selected.   
Then check  VFB , IFB ,  this is the feedback section.  These points feed back to the compliance section and to the A/D section so basically this is  your feedback loop coming back from the output terminals.  Check the EAMP   this feeds directly to the driver amp input. Should be 0V when  at target  voltage/current so basically always zero. But is its any other number it will drive the output toward the direction of the  EAMP voltage . If this is a negative number the output will go negative if its positive it should go  all the way positive.  Basically think of it as a direction  of change.  So when the unit is settled into the correct setting it should be 0V .
I started to put together a table of testpoints and hope to finish one day.

BTW  - I used to know a little German  - learned 30 years ago but haven't been using it so cant really communicate in German any more.

Updated based on Udok  - thanks
Updated again tanks to PMs from users.

[udok]

Some more testpoints:

TP200 - NAVMDAC: Negative Aproximate Voltage Feedback Mode DAC
TP215 - AFB: Approximate Feedback Output (provides Feedback during switching ranges)

TP218, TP219 - VL+ and VL-: 5 Volt Local Supply for Lowside Buffer derived from +-30V referenced to Power GND

TP233 - ADC Input after multiplexer between VFB and IFB

TP224 - IGUARD: Current Guard Buffer output used to reduce leakage in Cmos switches

[goaty]

Thanks kawal, I will try and collect some energy to start again and fight frustration. Also my problem is with this analog stuff my head spins very quickly as I loose focus and cannot remember what I just measured and what it means. I am not hardware engineer, so the routine is missing.

Also I understand why LeBassiste was upset: You should not have edited the image (especially the 2410 number!) without asking him. Maby annotate for your purpose (2400), but not overwrite his text with 2410.

I copied some text over to a list I just made in the 2420 thread-
https://www.eevblog.com/forum/repair/keithley-2420-sourcemeter-repair-help/msg2948016/#msg2948016

There are some differences is the test points and their designation on silkscreen on the version of 2420 I have. (Which seems to be some preliminary development model as it has a number "2420-100H2 Dev# 40897" which sounds like a development model.

[kawal]

For sure there will be differences between 2400 +/-200V @ 1A,   2410 _+/-1100V@1A,   2420 +/-60V @ 3A
they all must have differences  due to range switching and output capability.  But fundamentally the same building blocks are incorporated into theses.

Good luck and keep us posted on what you find wrong and what fixed the unit.

[Ordinaryman1971]

I have one of those not working as well. Mine is reporting "no comm link".. and don't have a clue where to start the troubleshooting.
I've reprogrammed the firmware thinking that it may solve the problem but no change...
What would be your next step guys.

[TiN]

Check the digital power? No comm link on Keithley's means it's braindead, front panel controller cannot receive data from main processor, so it outputs default message as result.

[goaty]

Also, no comm link could possibly be fixed with swapping all optocouplers, no ?

[Ordinaryman1971]

Optocouplers... that was what I thought to do too.... lets see, will report as soon as changed.

[kawal]

Check the digital power? No comm link on Keithley's means it's braindead, front panel controller cannot receive data from main processor, so it outputs default message as result.

There is no optocoupler between front and digital board that I know of.   The mainboard should be using ground referenced voltages so I assume you already checked the voltages.  Maybe something to do with ram ?  Pull the front out and inspect the front display board for damage.

[goaty]

I think it's referring to the link between digital and analog board.
There's several optocouplers to the Altera.