Keithley 2002 repair (Help please)

[SelectLOL]

Nice clean looking unit, including it's PCBs. Looks like you caught it in time with those leaky caps.......it's a whole different ball game if the PSU side of things fails.....trust me!

Ian.

Thank you very much, I guess you are right, I am quite lucky in an unlucky situation    Leaky capacitors are really the worst thing for the electronics and humans since the stink so much (smells like fish for me)

[SelectLOL]

It seems to me that Keithley instruments are much more prone to have leaky electrolytics, compared to other instrument manufacturers. Especially the 2001 and 2002 models.

Why is that?

Compare this to the Agilent 34401A, I never had one bad electrolytic capacitor, not even on the older models.

You are sadly quite right  I can't argue with the age, since the Agilent 34401A is a quite old mesurement equipment too. One difference that might be game changing is the temperature and how close the caps are driven to the limit. I guess Agilent took a way chillier approach to both  But maybe one of the experts here on the forum might be able to tell us exactly why 

[SelectLOL]

Hello dear people of the Forum

Today my parts have arrived and I successfully soldered them all up, pictures in the attachment.

As usually some really really weird things have happenend. Read carefully now, I will write everything down as detailled as I remember 10 minutes ago. I went thrue every emotion there is, in these minutes, so bear with me 

After soldering everything up I took the other parts of the Keithley (Digital Board, Screen etc.) that are needed for testing and hooked them up loosely to the Analog Board. After verifying all the conncetions I powered it up and... somehow nothing on the screen.

I panicked and pulled the AC plug out, checked for hot spots and took a deep breath, checked all the connections again and plugged the AC power in again. I watched the display closely and saw that the boot message (Keithley 2002, GPIB:xxx etc.) was on the screen for a short time every 3 seconds or so.

I took the ADC board out and tried booting again and it worked fine. I ofcourse got the "Overflow" message everywhere but it booted fine.

I inserted the ADC board, after shutting it down again, to measure the voltages. Then something strange happenend.

I grabbed my phone to take a video. I pressed the on button and waited for the screen to flash away, but after flashing away I saw a voltage reading for the FIRST TIME 

I was so happy that I didnt know what to do  I knew that something is still wrong, because that isnt how keithleys normally wake up, right?  Maybe the ADC board was not inserted fully the first time?

I turned it off again and turned it then on, it started without a problem. I measured the voltage rails, which looked like this:

+15V to VGND -> 12.7V
+15V to GND -> 12.7V
+15V to DGND -> 12.7V

-15V to VGND -> -14.2V
-15V to GND -> -14.2V
-15V to DGND -> -14.2V

+5V to VGND -> 4.91V
+5V to GND -> 4.91V
+5V to DGND -> 4.91V

The voltages are off by quite a bit, so there must be still a problem around 

I have checked the DCV and DCI functions and you can see the pictures in the attachment below. The seem to work fine for the range that I can measure and verify against. I ofcourse cant say anything to noise and stability but the readings arent that far off. But strangly when I switch the ranges myself I get a "Overflow" message, but when in Auto it mesures just fine, altought the voltage should have worked fine in my selected range.

ACV seems to be completely off, I have tried to measure the 230V grid voltage, but it only shows about 150VAC. That is definitley not working correctly... 

I will later on post the error codes I am getting. Should I leave the device running or turn it off? I think it is about time to get my hand on a FLIR for this. 

Thank you all for helping out and the coming help that is still needed

[SelectLOL]

Here some pictures of the limited measurements that I can do :

- Four voltage measurements compared to my Fluke 111.

- One Current measurement against whats set on the Lab Supply.

- One resistance measurement, short with wires...it measures negative resistance. 

- And one temp. measurement, this measurement shows nothing but I find the temp measurement option just cool.

As I said the error messages will follow soon 

[alm]

Congratulations on getting it starting up and reading voltages!

I would start looking further at the voltage rails, particularly the +15V is still way low. Look at the rails with a scope. Is there excessive ripple on the input of the linear regulator? Fixing the rails may well fix a bunch of self test errors.

Keep a record about which tests are failing. Replacing a component and going from 20 failures to 15 failures is a nice measure of progress.

[SelectLOL]

Congratulations on getting it starting up and reading voltages!

Thank you, I am quite happy   

I would start looking further at the voltage rails, particularly the +15V is still way low. Look at the rails with a scope. Is there excessive ripple on the input of the linear regulator? Fixing the rails may well fix a bunch of self test errors.

I will try with the method, where I search for a heat source, because the voltages still arent quite in the range, that the Repair manual specifies. If that doesnt work out, I will try to go component, for component and find the defective component hopefully 

Keep a record about which tests are failing. Replacing a component and going from 20 failures to 15 failures is a nice measure of progress.

You are absolutely right, sadly for me we are talking about 39 failures    The arrors are to be seen below 

[SelectLOL]

So these are the erros that are currently popping up. The device is open without its cover and the measurement was done with the Rear inputs.

200.2 A/D operation
201.2 Testcal (Cannot measure 7V at A/D)
202.2,4,6,8,10,12,14,16,18,20,22,24 Integration (Integration bit I15,14,13,12,11,10,9,8,7,6,5,4)
302.4 ADGND MUX (Divider LO to ADGND)
303.1 Input buffer (Input buffer)
304.1 Open ckt ohms (Open circuit ohms)
304.2 Open ckt ohms (2M open circuit ohms)
305.4 20V range MUX (20V range MUX signal)
306.1 Ohms (9.6uA ohms source)
306.2 Ohms (1.92uA ohms source)
306.3 Ohms (High ohms voltage source)
307.1 100:1 divider (/100 input divider)
406.6 Sample/hold (Sample and hold)
410.1 TRMS (TRMS)
411.1 TRMS filter (TRMS filter)
411.2 TRMS filter (Filter)
500.1 Amps/LO ohms (96uA ohms/200uA DCA range)
500.2 Amps/LO ohms (960uA ohms/2mA DCA range)
500.3 Amps/LO ohms (7.2mA ohms/2mA DCA range)
500.4 Amps/LO ohms (20mA DCA range)
500.6 Amps/LO ohms (200mA DCA range)
500.7 Amps/LO ohms (2A DCA range)
501.1 Amps protect (Amps protection)
501.2 Amps protect (Amps bootstrap)
502.1 ACA switch (AC amps switch)

In total 39 errors. My main goal currently is to get the voltage rails to spec and hope that most of these errors vanish The attachments contain pictures of the errors. If anyone has any other ideas, inputs or feedback are very welcome  Thank you all

[Kleinstein]

The supply seems to use the same strange preregulator as the Keithley2001. So it makes sense to look at the transformer output voltage and the voltag after the rectifiers with a scope, to see if the regualtors have enough voltage to start with and if there is excessive ripple.

A bad regulator is also a possible explaination, though not very likely. An oscillating regulator is also a possibility to get a lower votlage.

Looking if something gets hot would be another point. If there is something drawing enough current to let the 15 V supply drop to 12.7 V this part should consume quite some power.

An easy test would be to run the main board without the ADC card and check if the supply goes up/good.

As long as the supply is not within acceptable limits the self test results are not really relevant. Especiall the 200.2 AD error could make most of the other tests fail.

[SelectLOL]

The supply seems to use the same strange preregulator as the Keithley2001. So it makes sense to look at the transformer output voltage and the voltag after the rectifiers with a scope, to see if the regualtors have enough voltage to start with and if there is excessive ripple.

Thank you very much for the tip, I have used the available schematics for the 2001 to do my thinking.  I have done some measurements, which confused me very much. 

A bad regulator is also a possible explaination, though not very likely. An oscillating regulator is also a possibility to get a lower votlage.

Looking if something gets hot would be another point. If there is something drawing enough current to let the 15 V supply drop to 12.7 V this part should consume quite some power.

An easy test would be to run the main board without the ADC card and check if the supply goes up/good.

I really hope that I can find something out with the temperature measurement. The Regulators for +15V and -15V get toasty (havent measured it but very hot to the touch, cant be right) even without the ADC board. 

I will sum up my current measurement and its results which confused me in the next reply Thank you for your help 

As long as the supply is not within acceptable limits the self test results are not really relevant. Especiall the 200.2 AD error could make most of the other tests fail.

Thank makes sense, without a good voltage many things might not work properly, I will just focus on fixing the voltages for now 

[SelectLOL]

Hello everyone, I hope you are all doing well 

I have measured some points and of course it is as weird as it gets 

I started with the ADC board removed. I used the GND pin where the ADC board normally would sit for my ground. I measured:

+15V -> 12.62V
-15V -> -14.62V
5V -> When the device bootlops its about at 4.18V and when it starts normally at 4.88V or 5.5V depending on how it feels. When I measured the 5.5V in the front from the "GND" that is labeled behind the ADC board. The capacitor which filters the 5V shows 4.95V, so I really wont be measuring from that stupid "GND" point behind the ADC board again.

These measurements can be seen in the picture > VoltagesToGND_noADC.png
For these I then switches to AC mode, to measure the noise > with the same time scale > NoiseSameTime_noADC.png and with a different time scale > NoiseFasterTime_noADC.png

The +-15V rails seem really noisy to me, this is where my measurement of the voltage before the regulators come in handy.

Everything can be seen clearly in the schematic attached, I have drawn in the voltages nicely.

The voltage that comes into the rectifier for the +-15V rail shows me about 25.3VAC, which should mean about 35.67V peak to peak, minus about 1.4V means about 34.27V DC ideally, about with smoothing caps and real world application (I think) 32V (which still is not enough). I measured 29.1V.

The testpoint HTP shows 14.26V and the regulators output voltage is at 12.62V 

The testpoint KTP shows -14.75V and the regulators output voltage is at -14.26V 

As far as I have understood the voltage needs to be about 2V higher than the desired output voltage. For both of these, this clearly isnt the case... Should I solder them out and measure the voltage without load? 

The bootstrap voltages look like this:
+BS = 29.1V
-BS = -28.9V
Input AC = 25.3VAC

The repair manual specifies atleast 34V so there must be also something wrong with that. The input AC voltage here is the same one as for the +-15V regulators. 

Last I measured the 5V supply which seemed fine for the moment. This measurements can be seen in the picture > 5V_Direct_noADC.png and > 5V_Direct_noise_noADC.png

The testpoint JTP shows 5.4V and the regulators output voltage is at 4.95V
The incoming AC voltage is about 5.6VAC.

I will also add the file where I wrote down the voltages and little notices for completion. I also want to add that when I hook up the ADC board doesnt change the values much, but I put it aside for now.

I came to the the conclusion, that this will be a longer process than I thought, these little measurements took me hours. One reason being that the Keithley 2001 schematics use different Designators for the components and that makes it a little harder. I also double and triple checking my every move, because I dont want to fuck up things I will wait until I can hopefully lend out a thermal imager from someone, meanwhile if anyone has any idea what I should measure next, tips and feedback are very welcome. Thank you all 

[picburner]

Since all the voltages are below the limits, as Kleinstein has already told you, it most likely schould depends on a fault in the pre-regulator circuit (I hope for you that it is this and not the electrolyte leaking from the capacitors that does electrolysis frying under some IC  ).
Check that at the transformer primary (black and brown wires) is 230V ac.

[Kleinstein]

With the preregulator the waveform is no longer a clean sine. So the fator of 1.4 for peak to RMS voltage no longer applies. The voltage after the rectifiers still looks a little low, but there is ripple and if the transformer is heavily loaded the peaks go down first.
The question here is if this due to a failure in the preregulator or overloading the transformer in some way (this could be even on the display board).

It would help to measure the AC voltage of a transformer secondary with the scope. Here one could see the waveform and likely also if the MOSFET in the preregulator actually is operating.
With the voltage rather low the control signal to the MOSFET should be constantly active, trying to provide the maximum voltage.

Measuring at the mains input side needs extra care.  Except for maybe a careful check with a good (handheld) DMM this should only be done with the DMM power through an extra isolation transformer.

[SelectLOL]

Since all the voltages are below the limits, as Kleinstein has already told you, it most likely schould depends on a fault in the pre-regulator circuit (I hope for you that it is this and not the electrolyte leaking from the capacitors that does electrolysis frying under some IC  ).

Yeah I just realised now what Kleinstein meant by pre-regulator, I should have measured the left side of the schematic  I see now. Ahahahaha that stupid electrolyte shall stay away from me, lets hope for the best 

Check that at the transformer primary (black and brown wires) is 230V ac.

I have quickly checked it with my Fluke Multimeter, it says 180VAC. hmhmhmhmhm not that optimal right?  Thank you for your reply 

[SelectLOL]

With the preregulator the waveform is no longer a clean sine. So the fator of 1.4 for peak to RMS voltage no longer applies. The voltage after the rectifiers still looks a little low, but there is ripple and if the transformer is heavily loaded the peaks go down first.
The question here is if this due to a failure in the preregulator or overloading the transformer in some way (this could be even on the display board).

Oh interesting so my math wasnt right, but there is still something very faulty 

It would help to measure the AC voltage of a transformer secondary with the scope. Here one could see the waveform and likely also if the MOSFET in the preregulator actually is operating.
With the voltage rather low the control signal to the MOSFET should be constantly active, trying to provide the maximum voltage.

Measuring at the mains input side needs extra care.  Except for maybe a careful check with a good (handheld) DMM this should only be done with the DMM power through an extra isolation transformer.

When I read your last reply I hadnt understood what the preregulator really is and thought that it is what I was measuring. Now I clearly see that the preregulator part is on the left before the transformator, I hadnt seen that, that part has an importance, thank you for that 

I will measure the preregulator out really carefully with the scope and see where I might get an isolation transformer, if not I will cut off the Earthing part of the scopes cable (safety second I guess )

Hopefully I can do that measurement today or tomorrow. Thank you very much for your help 

[Kleinstein]

For the start there is not need to use the scope on the primary side. One can quite a lot by looking at the waveform at the secondary side of the transformer: the waveform should be essentially the same.
It is not a good idea to seprate the ground on the scope - better wait for an isolation transformer to do the measurement in the right, safe way, with the DUT powered trough the transformer and the scope still grounded. This still needs care as the transformer defeats the GFI and the scope provides a path to ground.

Getting less than 230 V for the transformer primary is normal, though 180 V may be a little on the low side. The part with the rectifier and MOSFET is there to reduce the voltage so that the linear regulators don't see much more votlage than needed to maintain a stable supply. One possible fault may be the MOSFET never turning on, e.g. with a bad optocoupler or an error on the secondary siden.
With the current low voltage the The CONT signal should be constant low and thus a signal to turn on the MOSFET. When properly working there should be a 100 Hz PWM like signal, providing the full power for something like the forst half or a little more of the main waveform and than turn off.

[SelectLOL]

For the start there is not need to use the scope on the primary side. One can quite a lot by looking at the waveform at the secondary side of the transformer: the waveform should be essentially the same.
It is not a good idea to seprate the ground on the scope - better wait for an isolation transformer to do the measurement in the right, safe way, with the DUT powered trough the transformer and the scope still grounded. This still needs care as the transformer defeats the GFI and the scope provides a path to ground.

Getting less than 230 V for the transformer primary is normal, though 180 V may be a little on the low side. The part with the rectifier and MOSFET is there to reduce the voltage so that the linear regulators don't see much more votlage than needed to maintain a stable supply. One possible fault may be the MOSFET never turning on, e.g. with a bad optocoupler or an error on the secondary siden.
With the current low voltage the The CONT signal should be constant low and thus a signal to turn on the MOSFET. When properly working there should be a 100 Hz PWM like signal, providing the full power for something like the forst half or a little more of the main waveform and than turn off.

Thank you for your reply. I now have watched some videos related to the Keithley 2001, that I know the similarities in the PSU part from you. Marco reps's video had show a blown out 470Ohm Resistor parallel to the MOSFET. Well looks line mine is broken too

Marco reps's video : https://youtu.be/HVDQeJTEHFI?t=330

Amazing video from IanScottJohnston : https://youtu.be/S8okMaAKYiw?t=1503

I have attached a picture which shows the broken resistor. I was currently doing the thermal measurements which didnt show me anything too suspicious. That being said the Flirs E5's resolution isnt that high, even with the overlay of the real image it is really hard to see anything. I guess I will revert to the good old measure and find method  I will post some of the thermal images later on 

The broken resistor is probably related to the bootlooping problem. It probably turned on many times slowy until some capacitor was charged up enough to get the full cycle going. I hope to find a similar resistor to continue my measurements 

I will try to power the unit on again without it, I mean it worked already when it was open circuit right 

[SelectLOL]

Alright everyone its me again, I hope all of you are doing all right.

I have done some temperature measurements with an Flir E5 which, I was able to borrow. The pictures are attached below as a zip, because it has one thermal image together with an actual image from the camera. The file limit is sadly 10 files per post 

With the thermal imager I saw that the resistor (Ke 2001 R100 near MOSFET) didnt radiate much heat itself and becuase I watched a video from marco reps as I told earlier, I thought it might be a good idea to measure it out. It definitely was an good idea to check it out, since it was broken  (see last reply)

Fast forward. After I took the images I sat down an studied the pre-regulator a bit. I wanted to measure it out definitly.

For the measurements I took my Fluke handheld DMM which is powered with a battery.

I thought that might be the safest bet. I measured a little and saw a 4.3V volatage at pin 1, where 5V should be present. So I thought that my GND point wasnt that good and tried to search for a GND on the IC. When I came to pin 6 where the Gate of the MOSFET is connected, the transformator which wasnt screwed down started to make brumming noises (When measuring between Pin 1 and Pin 6). Of course I went away and then did it again to verifiy that it comes from me and it indeed did. I should have studied the datasheet of the IC, I thought that they put the supply voltage on a diferent part of the schematics to make everything look clean, but it seems, that it just doesnt have a supply voltage. Just as seen on the schematic symbol, it has diodes, not more, not less. 

Well the aftermath now is, that the MOSFET isnt getting hot anymore and probably has a short. The voltage between Brown and Black for the primary is about 220VAC, 40VAC higher than before. The output of the voltage regulators for +-15V are bang on though

I cant measure for a long time, since the regulators are getting hot pretty fast (about 80 degrees in 2-3 minutes) and I didnt dare to let them get hotter.

I probably will have to solder out the optocoupler (Ke2001 U100, Ke2002 U101) and see what comes in from the CONT signal and what lies at the MOSFET Gate. When I do that, I can decide on what I need to buy again.

Thanks everybody and I wish you a good day/night.

The first 4 pics are before the short and last 4 are after short, these are eye candy for visitors  The zips contain the images with the timestamps so one can see the heatup over time.

[SelectLOL]

The zips :

[Kleinstein]

The U100/U101 opto-coupler is a photovotaic type: at the output it can produce power like some solar cells, usually some 5-10 µA and 6-8 V maximum. This is just enough to drive the MOSFET gate, but not much more.

The optocoupler could be a source of error, if the current is no longer enough (e.g. the LED inside getting weaker).
Getting enough voltage with the broken MOSFET point towards a problem with the pre-regulator part in some way.

Having only a 4.3 V supply for the optocoupler (output of U619 in the K2001 schematics) would suggest that there is a problem with the diode or capacitors (e.g. no contact for C625 in K2001 plan) for the charge pump, or maybe an oscillating 5 V regulator.

[Per Hansson]

Thank you for your reply. I now have watched some videos related to the Keithley 2001, that I know the similarities in the PSU part from you. Marco reps's video had show a blown out 470Ohm Resistor parallel to the MOSFET. Well looks line mine is broken too
Marco reps's video : https://youtu.be/HVDQeJTEHFI?t=330
Amazing video from IanScottJohnston : https://youtu.be/S8okMaAKYiw?t=1503
I have attached a picture which shows the broken resistor. I was currently doing the thermal measurements which didnt show me anything too suspicious. That being said the Flirs E5's resolution isnt that high, even with the overlay of the real image it is really hard to see anything. I guess I will revert to the good old measure and find method  I will post some of the thermal images later on 
The broken resistor is probably related to the bootlooping problem. It probably turned on many times slowy until some capacitor was charged up enough to get the full cycle going. I hope to find a similar resistor to continue my measurements 
I will try to power the unit on again without it, I mean it worked already when it was open circuit right 

Fast forward. After I took the images I sat down an studied the pre-regulator a bit. I wanted to measure it out definitly.
For the measurements I took my Fluke handheld DMM which is powered with a battery.
I thought that might be the safest bet. I measured a little and saw a 4.3V volatage at pin 1, where 5V should be present. So I thought that my GND point wasnt that good and tried to search for a GND on the IC. When I came to pin 6 where the Gate of the MOSFET is connected, the transformator which wasnt screwed down started to make brumming noises (When measuring between Pin 1 and Pin 6). Of course I went away and then did it again to verifiy that it comes from me and it indeed did. I should have studied the datasheet of the IC, I thought that they put the supply voltage on a diferent part of the schematics to make everything look clean, but it seems, that it just doesnt have a supply voltage. Just as seen on the schematic symbol, it has diodes, not more, not less. 
Well the aftermath now is, that the MOSFET isnt getting hot anymore and probably has a short. The voltage between Brown and Black for the primary is about 220VAC, 40VAC higher than before. The output of the voltage regulators for +-15V are bang on though

Thank you so much for the link to those videos, I was reading the schematic yesterday but could not understand how in the world the circuit works, the explanation in both videos really cleared it up!
That said I think your conclusion about you shorting out the MOSFET is wrong:
Since your R100* was completely open the only logical explanation for why the meter turned on at all was that the MOSFET Q528 was already shorted.
Otherwise it should have behaved like in Marco Repos video and just remained off: Because U100* is powered from the secondary of the transformer via resistor R100...
* = Keithley 2002: R101/U101 instead

[SelectLOL]

The U100/U101 opto-coupler is a photovotaic type: at the output it can produce power like some solar cells, usually some 5-10 µA and 6-8 V maximum. This is just enough to drive the MOSFET gate, but not much more.

The optocoupler could be a source of error, if the current is no longer enough (e.g. the LED inside getting weaker).
Getting enough voltage with the broken MOSFET point towards a problem with the pre-regulator part in some way.

Thank you for your reply, I have now measured out the opto-coupler and am surprised to say, that it still seems to work just fine. I have desoldered it and measured it out by applying DC voltage to the LED and measuring the output. As to be seen in the attachments.

I measured about 1.25V @ 11mA for the LED, which is OK according to the datasheet (also attached, altough B version insted A). The Output was around 9V unloaded (Just Scope). Should I still order a new one?

I got less luck with the MOSFET, it seems to be totally shot. I will have to replace it, which is quite easy to do. The 10W 470Ohm Resistor is a little harder to find, but I still managed to find some similar things. I will post my basket later on 

Having only a 4.3 V supply for the optocoupler (output of U619 in the K2001 schematics) would suggest that there is a problem with the diode or capacitors (e.g. no contact for C625 in K2001 plan) for the charge pump, or maybe an oscillating 5 V regulator.

I had a hard time finding U619 but wow I saw that, that part is on the digital board side, which I can easily measure out when my parts are bought and arrive. 

[SelectLOL]

Thank you so much for the link to those videos, I was reading the schematic yesterday but could not understand how in the world the circuit works, the explanation in both videos really cleared it up!
That said I think your conclusion about you shorting out the MOSFET is wrong:
Since your R100* was completely open the only logical explanation for why the meter turned on at all was that the MOSFET Q528 was already shorted.
Otherwise it should have behaved like in Marco Repos video and just remained off: Because U100* is powered from the secondary of the transformer via resistor R100...
* = Keithley 2002: R101/U101 instead

Thank you for your input  I always try to post as much info as possible, so that when someone else might see the post with his broken Keithley, has an easier time 

I attached 2 zip folders called beforeShort and afterShort (2-3 replies ago), which contain all the thermal images. Both also include closeups of the MOSFET. Before my accident with the FET it showed me about 60°C and we both agree on, that the Resistor didnt radiate any heat for both times.

You say that the FET was always shorted, then why did I measure only 180VAC across the primary before and now 220VAC?

Where does the temperature difference come from? My guess would be, when it was working normally it had a decent voltage drop over it and created plenty of heat. Now that its shorted, it just doesnt drop any voltage and thus any heat. But when it always was shorted, how would that work?

You are right about Marco Rep's case, that might show another fault in my Keithley which we yet have to find 

I will measure out the MOSFET extensively tomorrow, to see if it really is broken and post again 

[Per Hansson]

Thank you for your input  I always try to post as much info as possible, so that when someone else might see the post with his broken Keithley, has an easier time 
I attached 2 zip folders called beforeShort and afterShort (2-3 replies ago), which contain all the thermal images. Both also include closeups of the MOSFET. Before my accident with the FET it showed me about 60°C and we both agree on, that the Resistor didnt radiate any heat for both times.

It's possible it was not completely dead, the MOSFET is rated for operation with 20v on its gate with transients up to 40v. (I've attached its datasheet).
I don't understand how connecting a 11Meg resistor (inside your handheld meter) between +5v (4.3v) and the MOSFET's gate would destroy it?

You say that the FET was always shorted, then why did I measure only 180VAC across the primary before and now 220VAC?

I just had a thought about the transformer primary:
It works exactly like a variac, and they have an interesting behavior:
A "good for bench use" variac will have 5 terminals, these allow you to get higher than your line voltage.

I don't have such a fancy variac, mine only has 3-terminals but if I'm careful I can still simulate this:
I brought the variac to 92% and then I hook the live wire up to the wiper terminal.
Now when I measure between neutral and wiper I obviously have my line voltage: 240VAC.
But the interesting part is if I measure from neutral to the now unconnected terminal of my variac: 260VAC

Therefore when you measure from the BLK to BRN terminals and get a specific voltage reading I guess you will have no idea which way the Hi/Lo relay K100 is connected!
An example is if you would input 180v between the BLK/WHT & BRN/WHY wires you would read 240v between BLK & BRN.
And further if you would input 90v between the BLK/WHT & WHT wires you would still measure the same 240v on BLK & BRN terminals?
Further complicating matters is the current limiter through CR100, that can be anywhere and it will make all readings confusing at best!
P.S: It's completely possible I'm wrong but then I hope someone will correct me!

[TiN]

I think you probably jumped the train soldering new parts. From your previous photo it looks like you are in full treat with leaked electrolyte. So ALL components from area around 5cm in each direction from where cap was must be desoldered, cleaned with IPA, all PCB via's soldered thru, and cleaned again with Acetone, then distilled water, then IPA again, dried, cleaned again. Then put new parts back in. This procedure is mandatory for any old (15+ year old) Keithley 2001/2002, unless you want to continue "repairing" it till cows come home 

I have old video on my kei2002_u2 unit from 2016. This should show you the process. I still have this very meter, and it's happily calibrated and used as my daily bench meter.

[Zucca]

Excellent Job SelectLOL.

Please check this

https://www.eevblog.com/forum/repair/this-keithley-2001-need-to-be-saved!/msg2883864/#msg2883864

before ordering a new R100 which is difficult to find anyway... such a unicorn...

I do not like thermalfuse, and I would avoid them....