Keithley 480 Picoammeter repair

[Shock]

Fault:
Have a Keithley 480 Picoammeter I pulled off the pile for repair, the fault is when powered on, no display.

Troubleshooting:
Opened it up and checked the instruction/service manual and performed the power supply checks on Table 6-2. All voltages present except step 6, the +5V supply.

On this version PCB (will get the board version later) it has tantalum caps namely C115 on the input of the +5V regulator VR103 and C116 on the output of that regulator, so was immediately suspicious (due to the devices age).

I confirmed (as it came with no battery option) that the internal switch is on the LINE position. Saw Eevblog #406 which works when set to line voltage as well, also noticed the +5V supply is using electrolytics on that revision, Daves is a black case, this one is a beige case.

I confirmed voltage measured approx 10-11V at TP2 and was present at the leg of voltage regulator input, it needs at least 7V to account for the drop.

I disconnected everything, discharged the axial caps then tested the resistance between ground and the +5V  supply (output of VR103). Easy to identify as on the underside of the PCB there is a noticeable star ground. Approx 500ohms which I thought could be low, I didn't notice any capacitor charging either.

Isolating the fault:
I removed the +5V supplies output tantalum C116, measured again between ground and the +5V and got 1.4k ohms (getting better?). The tantalum measured about 20% low on capacitance out of circuit, predictable but I didn't get around to testing the ESR.

Checked again and still no +5V present on the output of the VR103 regulator so I removed that as well. Again 1.4k ohms present on the output side to ground (both ways) with it removed. Then I tested VR103 out of circuit at 9V on a current limited supply, it shows good so far and measured +5V on the output. But I did not get a chance to load test it.

I ran a diode test in both directions and discovered 0.5V drop positive to ground and approx 1V drop the other way. So this along with the resistance has got me curious.

Since no battery is present I think I can ignore U301 as that is on pin 1 of the battery connector and I confirmed this with continuity test.

Looking around on the +5 supply rails...
+5V to Q102 Q104 driving the U104 opamp for the front end zero adjustment
+5V to pin 13? and 14 on NAND IC of the clock circuit?
+5V to pin 1 and 4 of U101 the processor
+5V to Q101 on the D5 line to the display module
+5V to pin 3 and 16 of U202 on the display module

I have removed the display module when testing but I didn't confirm what exactly was isolated.

What should I do next:

A. There is a low ohms short still present. Remove the ....... next.
B. Apply a 5V current limited to ..... mA onto the PCB where the output of VR103 was.
C. Test the C116 tantalum caps ESR and try to reform it if possible.
D. I believe the C116 tantalum cap was the problem just replace it.
E. Test the +5V regulator VR103 with a load.
F. Just replace that regulator it's dead.
G. Do this....

More info needed:
If you have a Keithley 480 can you please take a peak and tell me what voltage and capacitance electrolytics you have installed for C115, C116 and what VR103 regulator you have installed? Let me know if it's been repaired.

[2N3055]

Voltage regulators are vanilla 7805 for 5V and 7815/7915 for +-15V rails.
C115 should be same voltage as C112, 25V something.. C116 is on 5V bus, so 10-16 volts is more than enough. But you can put 63V electrolytics for both if there is space in there..

[Shock]

Further troubleshooting:
I went with B. and applied 5V current limited to 30mA to the output pad of the 5v regulator. With the substitute 5V now present I connected mains and powered up and switched on the Keithley 480. Dim display shown, so turned up the current limiting slowly and it stopped at 97mA with a nice bright display. ITS ALIVE!

So now I'm thinking the voltage regulator is no good. Yet it was able to output 5V before when I tested?

Not fixed yet:
I decided on a hunch to wire it out of the circuit back to the board and put 9V current limited to 100mA on the input instead of using the 10-11V from the Keithleys own supply. It worked again! So the regulator is not faulty at all.

Disconnected everything and discharged the caps and then traced out the tracks through the power switch and back to the bridge rectifier and tested continuity for the voltage rail and ground. No problems found.

I powered it on again and double checked I had AC voltage still, powered it down and did a diode test on the bridge rectifier, it sort of looked good (no shorts anyway). So I removed the bridge rectifier from circuit and performed another diode test. No problems at all.

Now I'm wondering if I'm missing anything I applied the 9V at the recifiters DC positive output and powered the Keithley 480 up again. Of course it had to work.

What should I do next:

A. Do I put the seemingly working rectifier I have on 12VAC or something and try loading it down?
B. Perhaps I have a faulty secondary on the mains transformer should I try manually loading that down?
C. Has C115 the input tantalum come now to haunt me? With it isolated in circuit it's up in the megaohms do I need to remove that as well?
E. Is the 97mA just a little too much for the transformer or bridge rectifier?
F. Why didn't I remember to check the current draw to the input of the regulator that might have helped me determine a sagging supply.
G. Do this....

[The Soulman]

A. No need if it measures ok with the multimeter (all 4 diodes).
B. This is starting to look like the case, instead of powering up and loading down measure the resistance of that output winding, should read fairly low, if it reads high it's probably broken, measure the resistance of the other winding's to get a ballpark figure.
C. probably not.
E. 97mA sounds reasonable.
F. Same reason you've skip't  D.? 
G. Do B.
 

[Shock]

No need if it measures ok with the multimeter (all 4 diodes).
This is starting to look like the case, instead of powering up and loading down measure the resistance of that output winding, should read fairly low, if it reads high it's probably broken, measure the resistance of the other winding's to get a ballpark figure.

D is for diode. Yeah I measured all 4 diodes in the rectifier with a 0.5V drop and also confirmed I got a 1V drop from negative to positive. I also measured all backwards for any shenanigans.

I really don't want it to be the winding, that would make me sad. I recall I checked it without consciously doing so and it was low resistance, I'll take another measurement and get it's inductance/resistance and measure against another winding.

I was seeing voltage present on the regulators input right from the start before I removed anything which is the confusing thing. You would think it would cause the input voltage to drop down when the output was loaded, but I only got the output dropped down.

I'm ordering replacement caps and also a bridge rectifier today.

At least the large electrolytics are getting a chance to reform while I'm doing this. Now that has me wondering if anything else causing the transformer to saturate.

The other thing I'm wondering about is if I have a load dependent high resistance pathway from the rectifier to the regulator. There is only a couple of switches I see in the circuit path. I thought I ruled this out though powering the circuit from the vacant bridge rectifier.

I might have to check what drop I'm getting from the rectifier to the regulator as a sanity check in that case. I've been working mostly from the top of the board and couldn't probe the secondary and rectifier easily till now.

[Shock]

Picked up some replacement electrolytic input and output caps for the voltage regulator, a bridge rectifier and a smoothing cap.

Adding new parts:
I only checked the input tantalum for a short previously and not out of circuit so I decided to remove that first and install the 2 new small electrolytics. Bodged in the smoothing cap as the leads were short and I didn't have an axial as a replacement.

So now I had ruled out the unknowns or at least some preventative maintenance.

Fired it up and still not working, ok time to test the new rectifier, it's the wrong size but I was only going to bodge it in anyway to lift a leg on the positive output so I can isolate the transformer secondary if I need to and power it from my bench supply.

This proved a good idea as I still could get no 5V out. So now I'm suspecting the transformer secondary winding next. I had a look at the schematic again and saw that if you switch to battery it enables the extra winding to allow for the battery charging.

More power required schottky:
I decided to move the internal switch to battery and jumper pins 4 and 5 to bypass the open circuit (no battery option installed). Tested again an still no 5V output. Measured something like 19V, way too high. I measured the AC and DC at the rectifier and it wasn't low and looked healthy, I couldn't even get a measurement of ripple it was smooth DC I was getting at the regulator. I undid the hack and double checked that I could still power up the logic and display from my bench supply.

Please let it not be the transformer:
I was considering if it was time to remove the rectifier and isolate and measure the windings to see if I could get the ratios to make sense. What a pain, the secondary felt healthy all the numbers made sense and it wanted to work.

Before I did that I wanted to confirm I had no high resistance path under load between the rectifier and the regulator input.

So first ohmed it out which was fine, it was at this point I decided I might as well just load the rectifier positive output as I had more voltage than I needed anyway. So took a voltage measurement and bodged in the old rectifier as a couple of diode voltage drops. It measured 1V less. There is nothing wrong with this transformer!

I know whats broken:
At this stage I knew it, connected the regulator back into circuit and powered the Keithley 480 on and saw a nice display.

Seeing the finish line immediately I switched it off and disconnected the rectifier bodge and put my bench supply back in circuit at 9V... 10V... 10.5V... 11V... (display goes out).

Hahaha, oh well.

I could literally just move half a volt and turn it on and off over and over. Nice power saving feature I guess. So looks like a new T220 7805 regulator will get it back up and running. I'll work out power dissipation and probably test with a heat sink before committing.

[The Soulman]

 

[Shock]

I ended up using a preloved Motorola 7805 regulator and attached a new heatsink. Input voltage sat at about 10.2V. Ran through the service manual and checked all test points were in tolerance as per the troubleshooting proceedures.

Cause of the fault:
Measured the ESR on C116 the regulator output cap (I was originally suspicious of this and it did test 20% low). Turns out one of my ESR testers measures it as 80 ohms, so looks I was on the right track after all before finally discovering the 7805 regulator was sensitive to input voltage.

Time to play:
Here is a photo of a used electrolytic cap under test off my bench supply. Showing 17.6nA leakage current prior to some reforming/self healing (which increases then decreases the current before settling out). Other photos include what came out, what went in and the picoammeters new home.

[zrq]

Does anyone have experience with fixing the old mechanical range switches on Keithley 480 or similar meters. I'm trying to repair one but it seems the range switches got very unreliable. It's behavior is the reading randomly jumps around unless I exert the exactly the right amount of force to set the switches. Do I have to desolder the entire switch, disassemble it and polish the contacts? That would be a shit job and I'd rather just dispose the entire meter. Can Deoxit (or an equivalent that is sold in Europe) be a magical solution?

[Shock]

Watch this video and read it's comments. First thing I would try is switching them several times as that can suddenly bring pots and switches back to life.

Apparently they are designed to be cleaned from the rear drained out the front. If you decided to clean them this way with contact cleaner I would disconnect everything in the front of the switches (display, panel and buttons etc) cover the rest of the pcb with something esd safe to prevent splashes.

My main concern would be that I don't want cleaner draining onto the pcb especially if it has a lubricant. If contact cleaner can soak in the top it may leak out the bottom. So perhaps vertically angled slightly upside down is best?

What contact cleaner would I use? Well that depends if it was lubricated and gold which is not meant to corrode. I'd be inclined in that case to just squirt some pure IPA cleaner with a fine/straw nozzle if so. It might be enough to dislodge the contaminant and leave any contact coating mostly there.

Otherwise you may need an electronics contact cleaner that both deoxidizes and lubricates. If the shaft has a lubricant you have to consider this as well. I've not seen inside one to best determine how to clean them properly. Which product to use is very contentious question. If you have problems with a suitable response start a new thread in this repair forum. Something like "How do I clean old Keithley switches".

[Shock]

Just to add...

In that videos comments it's mentioned they have silver contacts on that model he is repairing (so oxidization is possible). Some people were reporting that the solder joints may become intermittent due to stress as well.

I normally test intermittent joints by gently tapping (then pushing) on or around the component/pcb with the back of a screwdriver handle while looking at the display (or scope or multimeter measurement) in an attempt to see if I can reproduce it. You can also try visually inspecting the joints.

Here is a video mentioned in the replies, it's possibly a similar ganged switch but gives you an idea of what these could be like inside.

[zrq]

Thanks for your comments. In my case, it got perfectly fixed by applying some KONTAKT 60 plus into the switches.

[Greybeard]

I hope you have carefully and completely removed the remains of Kontact 60, otherwise it would later destroy the contacts, as it is very aggressive. It's so aggressive that it even eats through its own tin can over a period of years.

[sintoodio]

Thanks for your comments. In my case, it got perfectly fixed by applying some KONTAKT 60 plus into the switches.

Hey ZRQ, Good Morning, I sent you a private message, could you reply?

[zrq]

Thanks for your note. It seems KONTAKT 60 would require washing as described on its SDS

KONTAKT CHEMIE Kontakt 60 works in many cases even without after-treatment. However
this increases the risk that corrosion damage will appear again after a short time.

and it appears KONTAKT 60 plus doesn't require washing, only simple wiping and application of a protective film is recommend.
Anyway, I'll watch out for any sign of corrosion, thanks.

[alm]

I believe TUNER 600 is what's usually recommended as the Kontakt product for cleaning more delicate contacts.