Keithley 197 Repair

[staze]

I don't really see the point of replacing parts before you've fixed the problem, it's not like putting them back is going to help. What you can do is cut the V- pin of U107 close to the body, but leave a little metal sticking out of the body. You can then easily resolder this connection later on. If this doesn't help, remove VR101, Q127 and Q128 again. If that doesn't help,  go searching the PCB (is easier than the schematic) for other parts connected to -V. For example a short on the PCB, or some other component that we missed in the schematic.

Mainly replaced it because I couldn't get the original back in (the legs were all bent at the factory to be lower profile, but it made for re-installation to be rather problematic.

I've traced the -V rail pretty well at this point, and those components seem to be it (save S101). I don't see any shorted/broken traces, and the resistances work out. I'll see if I can snip that lead (it's in a bit of a bad location). Any harm in snipping it, yet leaving Q128 attached (it would have -18V on it's collector, but nothing on it's base).

Thanks!

Thanks!

[alm]

Leaving Q128 should be fine as long as it's not shorted .

[Amarbir[Lynx-India]]

Well,
     This and The Older 199 Thread Is Pure #$%#%#% I Love It Guys  .You Are Making Me Buy These And Get It All The Way To India For Repairs .I Hate You Buggers  .Lets See How You Crack This One ,Would Be a Awesome Effort Like Last Time  " Best Of Luck "

[staze]

Bingo. Cutting the negative lead (pin 4) on U107 (op-amp) brought the negative rail back up (down). It's only at -12V, but that's certainly better than it was.

Know of a good replacement, or should I just try to find a replacement? They don't seem to be a very common part. Mouser has them, but shipping and all that.

Any thoughts before I order the part? I'm going to need to no matter what, since I pretty well mangled the pin trying to cut it (my flush cutters aren't really small enough to get in that space.

Though, just realized that I should have used an X-Acto knife, or the like, to cut the pin... probably would have mangled it less. Ah well.

[alm]

Given its function and location, I can't see this component being very critical for either correct operation or accuracy. Almost any op-amp that accepts this supply voltage would probably work fine in this position, at least for testing.

[staze]

Given its function and location, I can't see this component being very critical for either correct operation or accuracy. Almost any op-amp that accepts this supply voltage would probably work fine in this position, at least for testing.

Makes sense. I don't think I have anything that'll take that range. I'll just order the replacement, and swap out the current op-amp for a 8 pin socket.

Here's hoping!

[staze]

So, removed the dead op-amp, and installed a socket. Tried dropping a jellybean LM741 in there, and no love (not sure if that LM741 likes single supply or not). So, guess I'll wait for the true replacement from Mouser. Hopefully that brings it back to life.

Does it make sense that without that op-amp, the voltage across VR102 would be +6.2V rather than the stated -6.4V? Guessing it would be because Q128 isn't pulling it down without the op-amp driving it's base.

[alm]

Something could be pulling up the -9 V rail (or -6.4 V rail) through its ESD protection diodes. With VR101 disconnected, feeding 9V from a current-limited bench supply to the emitter of Q128 should restore the proper voltages, I think.

[staze]

Something could be pulling up the -9 V rail (or -6.4 V rail) through its ESD protection diodes. With VR101 disconnected, feeding 9V from a current-limited bench supply to the emitter of Q128 should restore the proper voltages, I think.

Any idea what I should limit the PSU to (current wise)? 50ma or so?

[staze]

So, good news and bad news.

Good news: opamp was bad. Doesn't pull the negative rail up/down to ground anymore.
Bad news: something is still wrong. The -6.4V reference is a +6.2V, and the -9V rail is up around -4.6V. The -V rail at the op-amp is at -12.8V, but at the filter cap, it's at -16V or so. So something is still pulling stuff down/up. Guessing this is due to the -6.4V reference being all wonky, and wondering if something broke when I had the opamp out (or the 741 in) and powered things up. Not sure why that would be though.

Any thoughts?

[staze]

interestingly, I'm seeing about 2.5V on the collector of Q127 (when in volts mode), or about 4.8V in ohms (which seems normal). Am trying to figure out where that voltage might be coming from.

I'm just going to go back through the voltage checks and see what's within spec, and what's outside. I also note, according to the schematic, all the buttons should be "out". I'll do that, and see if that changes anything as well. But obviously something else is blown, and maybe that took out the opamp to begin with. =/

[staze]

So, still issue with negative rails all around. For some reason, the negative rail seems to be at -19V on Pin 12 of the power switch, but once it gets to U107, it's at -12.8V, and there's a fair (about 400mV+) ripple on that rail on U107. I would think that should all be smoothed out by C127. 

Other negative rails are as described before. So... I have next to zero clue what is going on. Part of me wants to replace VR105 with a LM7915, but I'm not 100% on how I would wire that. Could VR101 just be toast?

[staze]

I keep looking at U104 as a possible, as well as the Q103/Q127 pair.

U104 is interesting because the resistance between pin 3, and pin 6 is only 36KR. Which, looks like it should be higher than that since I don't see how they're connected. Also, pin 4 to pin 7 is only 7.5KR. Which also seems wrong.

The interesting part of the Q103/Q127 pair, other than why the heck they're there (I know, you said, but it still doesn't make a whole lot of sense), is why there's +5.6V on the base/collector of Q103/Q127 respectively. That's leaking in from somewhere... but hell if I can tell where. But this could all be a big while goose chase... and the real issue is staring me in the face.

I'm off to bed.

[staze]

well, not Q127.

so... try pulling VR101 again? All the positive rails look perfect. U101 is the only thing I can see on the -6.4V rail... but it doesn't seem right that it would be flipping the -6.4V to +6.2V, unless it's ALL screwed up and taking that 2.5V on it's pin 5, plus it's +10V on pin 7, and dumping it ALL into it's negative rail.

I'm having a hard time finding anything that says what one should expect, impedance wise, between the rails of an opamp, or the input and output, etc.

Thanks!

[staze]

U101, for interest, shows the following on it's pins.

1. 6.308V
2. 6.308V
3. 5.8V
4. 6.304V
5. 5.864V
6. 6.125V
7. 6.125V
8. 9.995V

Now, it LOOKS like a dual op-amp, but the manual lists it as an IC, and I can't find anything with the numbers on the chip. What's interesting is it seems to be a bit "off", when it comes to input/output. Guessing that's just because the negative rail is not right, so it's slamming into that limit (though, it seems to be using the lowest voltage available for it's negative rail, rather than the real negative rail (since the output on pin 3 is lower than the true negative rail).

[staze]

actually wondering if the integrator might be the issue now. Looking at the "ripple" on the negative and positive rails, it kind of looks like integration. Will fire up the scope tonight, tomorrow, or next week (as baby allows) and take a look. But please, if you have other ideas/suggestions to check, let me know. I'm kind of bouncing around, I know. I'd cut all the negative rails on devices (one by one), but I haven't figured out a good way to do it (x-acto knife didn't work very well, and flush cutter just mangled the lead).

Thanks alm!

[free_electron]

TLC27L2

i have a 197 so i opened it to see what mine says.

[staze]

TLC27L2

i have a 197 so i opened it to see what mine says.

Sorry, U101 (by the relays) is a TLC27L2 for you? I know U107 (by the input jacks) is a TLC271CP in mine. Any chance you want to check what you get on pin 4 of U107? Or the emitter of Q128?

[staze]

For interest, this is what my -9V rail looks like at this point (measured on Pin 4 of the integrator). But, obviously this isn't the problem, but may be a symptom.

[staze]

So, interesting symptom...

Put the unit in diagnostic mode, DC, 2V range. Shorted in inputs for good measure, and looked at pin 6 of U102 (the multiplexer).

My understanding from the block diagrams is that the mux flips between signal, ? (says 0V for 20V range, doesn't say what it is for 2V range), the 2V reference, and 0V). Great.

Problem is, on every diag mode, pin 6 sat at -2.08V or so. No change. So, it would seem that maybe U103 (which seems to be what controls the Mux's FETs) isn't working? Is this just because the -9V rail isn't working, or could it be that U103 is the cause? It's just a jellybean LM339, so cutting a pin wouldn't be that big a deal, but it does seem odd that NOTHING would happen.

[timb]

Cool thread and very timely! I just scored a 197a on eBay for $85USD, looks pretty clean from the picture. Should be here end of the week!

I couldn't resist at that price. If it works, I'll calibrate it with my Tek 4030 as a transfer standard. A DMM with microvolt resolution is always good to have around! (At least that's how I justify these type of purchases.)

[staze]

hopefully it does work! I do wish I could get mine running. My Keithley 197 is great (though I could use to have it calibrated).

Just need to figure out what's f'ing with the negative rail.

[alm]

Anything connected to the -9V / -6.4V rail may act dodgy as long as that rail is out of spec. You can either cut pins / tracks to isolate sections, or try to figure out where the current is going. To pull up the negative rail, a fair amount of current needs to be flowing from a more positive rail. Maybe you can find this point? The more positive the negative rail, the closer you are to the culprit. Of course if the negative rail is positive, ESD protection diodes may start conducting and make life interesting. I would consider powering the -V or -9V rail from a lab power supply again (make sure you disconnect the zener). This will allow more current to flow. You may then be able to use voltage drop (the more positive the closer to the culprit) to find the culprit. If only you had a high-resolution DMM . It might be something stupid like a short somewhere on the PCB.

If the -9 V rail is that unstable, then either something's drawing a serious amount of current (does any part feel hot?), or the -9V rail has a high impedance. Does the same happen to the -6.4V rail? Labeling the circuit around U107/Q128 (where these rails are generated) with voltages and studying that may provide some clues. Resistors are convenient gauges to indicate current direction and quantity. Also check with a scope: is anything oscillating? What's the cause and what's just an effect? Any way to verify the operation of the -9V / -6.4V circuit without the load connected?

[staze]

Anything connected to the -9V / -6.4V rail may act dodgy as long as that rail is out of spec. You can either cut pins / tracks to isolate sections, or try to figure out where the current is going. To pull up the negative rail, a fair amount of current needs to be flowing from a more positive rail. Maybe you can find this point? The more positive the negative rail, the closer you are to the culprit. Of course if the negative rail is positive, ESD protection diodes may start conducting and make life interesting. I would consider powering the -V or -9V rail from a lab power supply again (make sure you disconnect the zener). This will allow more current to flow. You may then be able to use voltage drop (the more positive the closer to the culprit) to find the culprit. If only you had a high-resolution DMM . It might be something stupid like a short somewhere on the PCB.

If the -9 V rail is that unstable, then either something's drawing a serious amount of current (does any part feel hot?), or the -9V rail has a high impedance. Does the same happen to the -6.4V rail? Labeling the circuit around U107/Q128 (where these rails are generated) with voltages and studying that may provide some clues. Resistors are convenient gauges to indicate current direction and quantity. Also check with a scope: is anything oscillating? What's the cause and what's just an effect? Any way to verify the operation of the -9V / -6.4V circuit without the load connected?

alm,

Thanks very much. I'll check this once I get my "lab" back (mother-in-law in town). I kind of think the -9V rail looks like integration from the integrator (which makes me look at anything that could act like an integrator that's on that rail (U107 (just replaced), U102, U110 (the actual integrator))... but maybe that's just me. Nothing feels hot except R132 (since it's dropping like 7V at this point), Q123 (guessing that's why it's in a TO-220 package), and Q122 (a FET that drives the +10V rail, so maybe that's a symptom of the current draw).

I do note that the -6.4V rail is completely reversed to +6.2V, which makes me wonder if something that's flipping it is getting both the +10V rail, and the +2V reference (which makes me look at things like U102 and U101).

Hadn't thought of just looking for where the most positive part of that negative rail is.

I don't know of a way to remove the load on the rails... I could pull VR101, but that would kill the whole shebang except the -V rail itself (well, and whatever U107 and Q128 would do with their remaining -V connection.

Hmmm... grumble.

[alm]

Are you sure you didn't configure U107 as an inverting amplifier?