Keithley 2010 Repair

[Cyclotron]

To see more details, it is the "Shift" key.

So now I'm down to only one fault remaining.  305.2 BIAS CURRENT

[KungFuJosh]

I don't know if its appropriate to start a new thread. The K2010 I have was claimed to have faults with reading DCV when I bought it, seemed fine during tests but I had pushed the front/rear button a couple times.

I've now put new caps in it and the new LED display cleaned the parts and boards.  I also upgraded the firmware from 08 to 15. 08 may have been the original? It's old.

Now I get errors when I run a self-test, but they are inconsistent and seem to be related to my messing with the front/rear switch.

The errors I'm getting most consistently are 305.2, 600.1, and 600.2.
Sometimes I get others like a 400.2 once.  But it is getting fewer as I mess with it and time goes by.  I'm guessing that the front/rear switch has some corrosion.
I thought Josh said you could push a button and it would display some text for these errors, but I can't find that post again.

Clean all the contacts with IPA really well. Then let the meter run for hours to warm up.

I'm assuming you have enough toys to run through the calibration (at least for DC), so you might want to do that before trying again. I think firmware changes like that can lose the cal data.

[Cyclotron]

I don't know if its appropriate to start a new thread. The K2010 I have was claimed to have faults with reading DCV when I bought it, seemed fine during tests but I had pushed the front/rear button a couple times.

I've now put new caps in it and the new LED display cleaned the parts and boards.  I also upgraded the firmware from 08 to 15. 08 may have been the original? It's old.

Now I get errors when I run a self-test, but they are inconsistent and seem to be related to my messing with the front/rear switch.

The errors I'm getting most consistently are 305.2, 600.1, and 600.2.
Sometimes I get others like a 400.2 once.  But it is getting fewer as I mess with it and time goes by.  I'm guessing that the front/rear switch has some corrosion.
I thought Josh said you could push a button and it would display some text for these errors, but I can't find that post again.

Clean all the contacts with IPA really well. Then let the meter run for hours to warm up.

I'm assuming you have enough toys to run through the calibration (at least for DC), so you might want to do that before trying again. I think firmware changes like that can lose the cal data.

At this point, I think I may have not completely dried the board after its last wash. I thought it was but as it sits and warms it does seem to be getting better.
I backed up all the CAL data and checked it again against the backup.  I am going to let it sit over night running.

I can probably hack my way through a 7.5 cal with the stuff I have. That's really at my limits though. I do 6.5 and below all the time but 7.5 is pretty much where I peg my limit and I'm probably overestimating myself (Dun and Kruger and all).

We will see how she feels tomorrow. Thanks for the comments and blazing this trail.  I am digging the display. Can't wait to see the new lense on it.

[KungFuJosh]

At this point, I think I may have not completely dried the board after its last wash. I thought it was but as it sits and warms it does seem to be getting better.
I backed up all the CAL data and checked it again against the backup.  I am going to let it sit over night running.

Self-oven is a new feature.

I can probably hack my way through a 7.5 cal with the stuff I have. That's really at my limits though. I do 6.5 and below all the time but 7.5 is pretty much where I peg my limit and I'm probably overestimating myself (Dun and Kruger and all).

It doesn't need to be perfect, just good enough to sanity check/pass the self tests. Once it's all good, you should send it out for cal anyway.

We will see how she feels tomorrow. Thanks for the comments and blazing this trail.  I am digging the display. Can't wait to see the new lense on it.

The lens was supposed to be there today, but I'm guessing tomorrow is more likely for Mr. Tron's package.

Be careful taking out the old one. I used a very thin metal guitar pick, and it worked great without any real force. I basically just wedged/cut the space between the lens and the double-sided tape all the way around the lens (front side of panel).

[Cyclotron]

The lens was supposed to be there today, but I'm guessing tomorrow is more likely for Mr. Tron's package.

Be careful taking out the old one. I used a very thin metal guitar pick, and it worked great without any real force. I basically just wedged/cut the space between the lens and the double-sided tape all the way around the lens (front side of panel).

I've already had the OG out and polished the paint off. The recommended method is probably not the best given the results of using acetone on the lens is uneven. I just used some 1k grit paper to remove the paint then 2k - 5k and polish and it looks fine. I am looking forward to the gold KungFuJosh look though.

[KungFuJosh]

I agree, but not everybody has the option to replace the lens, or maybe they want to keep the original logo. It's probably better to remove all the paint anyway. There's no need for it without the VFD.

[KungFuJosh]

I replaced R359 with a 5k pot with 1k resistors on each end. I replaced VR122 with a 1N4578A-1.

Last night, I tweaked it with the DAQ6510 attached to relative stability (I like the graph on that better). Then I let it run overnight, and connected the 34465A across VR122. I let it run for a while with small adjustments until I got it around 7uV PP on the short side (5 minutes or so), and a little higher if I let it run a long time.

The attached screenshots are how it performs after that adjustment, but not substantial warmup time.

Now, I'm going to let it warm up a lot longer, and see how adjusting the R359 pot affects measurement stability. I'll attempt to get it as stable as possible, within the limits of adjusting R359's value. If I get that to a reasonable stability, I'll pull out Funky R359, measure it out of circuit, and replace it with the closest resistor value.

Thanks,
Josh

[KungFuJosh]

You can clearly see how important R359's value is. Adjustments directly affect stability and drift over a fairly broad range. I haven't gone too crazy with the pot. It's a 21 turn pot, and the screenshot is probably within 2.5 turns at most.

[KungFuJosh]

Working with the tiny pot was a little tricky, especially to make or track small changes. Thankfully, the decade resistor box I ordered arrived yesterday. After cleaning it up, I connected it in place of the pot (after measuring the pot as a starting point).

I let it warm up overnight, and it was already pretty close to the 34465A, though I think it can be improved further. Let's see what happens.

Thanks,
Josh

[KungFuJosh]

Current settings before I tweak again. Already looking pretty good. Certainly better than before.

[KungFuJosh]

Well, this decade resistor was definitely the right tool for the job.

I went up and down 5Ω on the rheostat, and that made a big difference (bad in either direction). I worked down to +/- 1Ω and found that I already had the best/repeatable value.

The attached screenshots are the same settings as the decade photo above.

Room temp about 25C, 64% RH. Gross.

Thanks,
Josh

ETA: Now that I have the K2010 repeatably more stable than the 34465A at short and longer time periods, I'll see if I can get that again without the decade box.

[KungFuJosh]

Looks like I will probably use a 3.01k + 1.05k in series. I'll select a few to see if I can get closer than the nominal value.

Right now I'm thinking YR1B1K05CC, and YR1B3K01CC.

Why axial? Well, let's just say that replacing that resistor at R359 1000 times (before I got the decade box) didn't do anything good to the old pads on the PCB.  Why bodge an SMT resistor when small THT will save the extra trouble?

Thanks,
Josh

[KungFuJosh]

First, the good news. I got the resistors in, and managed to find a pair that was exceptionally close to the target value. All previous fails are now gone. Everything is warming up now to see if stability lives up to expectations set by the previous test.

[KungFuJosh]

...and the mildly annoying news: now there is fail 500.2 "4W SENSE."

I'm guessing/hoping I just need to clean the contact pins where I probably touched them too many times removing/reconnecting the internal leads. Or maybe I did something stupid and didn't push them on all the way, or whatever. We'll see.

[w.v.s.]

Good to hear! I only read good news. Your sentence seems to suggest that bad news might be following. Any reason to be pessimistic? O.K. the question is answered in the meantime.

[KungFuJosh]

Thanks. I remember touching the pins with my bare hands and immediate thought "Hmm, I should clean that." Then I forgot all about it. It would be nice if that's all there is too it. We'll see (eventually). 🤷

[Kleinstein]

The decribtion does not looks like that the error 500.2 would be just a cleaning issue. It is about getting essentially a zero reading from sens low. So leakage should not be an issue, but more like too much offset or switch / relay not connecting. AFAIK one can run the tests one at a time and this way get the reading as more info than just pass / fail.

I would more suspect maybe a front rear switch contact problem or similar.

[KungFuJosh]

The decribtion does not looks like that the error 500.2 would be just a cleaning issue. It is about getting essentially a zero reading from sens low. So leakage should not be an issue, but more like too much offset or switch / relay not connecting. AFAIK one can run the tests one at a time and this way get the reading as more info than just pass / fail.

I would more suspect maybe a front rear switch contact problem or similar.

Thanks, that is a good point. If I step through the test I can press Shift to get the value.

On the bright side, even with only 10 minutes warmup, the K2010 already won the 2 minute stability test.

[KungFuJosh]

The 500.2 fail value is -5.608751. 0.0001 is the limit. 

[Kleinstein]

Such a high value looks like a open input / not working switch, cold solder joint or similar. A first suspect to test is the front rear switch. One could also test with the 2 wire short at the back.

[w.v.s.]

What would happen if the sense-cables were swapped? Testing with the rear input set is probably the easiest thing to do first.

[Kleinstein]

With a 4 wire short one would still get close to zero voltag for both sense high and sense low. it depends on the design of the 4 wires short, but the wrong way would be still well below 1 ohm, maybe 10 mohms.
A loose connection at the 4 wire zero could be an explaination too.

[KungFuJosh]

I think a loose connection makes the most sense.

I cleaned the contact pins, reflowed the PCB connections, and reassembled.

I can't say for certain, but I think maybe the sense low wire connection to the socket that slides on the pin might be angry after being removed/reinstalled so many times. Just in case, anybody know a good part number for that?

Anyway, no more errors now...until I reinstalled the front/rear button, and touched the sense wires, and now it's back. 

Thanks,
Josh

[Cyclotron]

I think a loose connection makes the most sense.

I cleaned the contact pins, reflowed the PCB connections, and reassembled.

I can't say for certain, but I think maybe the sense low wire connection to the socket that slides on the pin might be angry after being removed/reinstalled so many times. Just in case, anybody know a good part number for that?

Anyway, no more errors now...until I reinstalled the front/rear button, and touched the sense wires, and now it's back. 

Thanks,
Josh

Josh, I had my front/rear switch off the board and apart and discovered that it would "hang" sligthly when returning to the front position. I honestly have no idea how I fixed it because I took it apart a couple times to see what was causing it. In any case, it started working smoothly again on last reassembly so I left it alone without much of an answer. 

No idea if this is related but maybe something to poke at.

[KungFuJosh]

Josh, I had my front/rear switch off the board and apart and discovered that it would "hang" sligthly when returning to the front position. I honestly have no idea how I fixed it because I took it apart a couple times to see what was causing it. In any case, it started working smoothly again on last reassembly so I left it alone without much of an answer. 

No idea if this is related but maybe something to poke at.

I do the "poke it in and out" repeatedly thing to avoid that. I think the (what I assume is a) crimp connection is compromised on the end of the wire that connects to the socket pins inside the meter. ETA: Now that I think of it, you also have to make sure the button snaps on to the switch correctly. Sometimes it sits at a weird angle if you don't snap it on properly.

This issue with the connectors also caused 400.2 to show up again (likely what caused it previously too).

I pulled off, reinstalled, and adjusted the sense connectors. All errors are gone now and it's fully assembled. 

When I first adjusted the sense pins this time, it switched the error from one pin to the other. Then I adjusted again, and all errors are gone.

So, it's safe to say, if you're going to disassemble & reassemble a million times, be more careful than I was with those connectors.

I will buy some more crimp connectors for those pins because I will likely / eventually replace the LM399 with an ADR1399.

Thanks,
Josh