Keithley 199 -- FIXED!!!!

[alm]

U50-53 are just three terminal linear regulators. The corresponding ground is on pin 2, except for the 79xx, then it's on pin 1.

[staze]

U50-53 are just three terminal linear regulators. The corresponding ground is on pin 2, except for the 79xx, then it's on pin 1.

I don't see a 79xx... I see a 7805, and two 78M15's. though it could be what's attached to the heatsink that they don't actually SAY what it is in the parts list. Just "heat sink, ass'y"

[alm]

Are you claiming that U50 is a 78M15? That's unlikely to work. What are the voltages on its pins?

U53 is the +5V digital regulator, an LM2940 according to the schematic.

[ftransform]

I like the keithley 199. I got a GPIB card recently so I will find out how it works with the computa. I should start buying all the 199's so I can destroy them and be left with the last one.

[staze]

Okay, so, V51 reads 5.007V (good) V50 reads -15.046V (good), V52 reads 15.123V (good). All less than 1% out. 

On my board, V51 is a 7805CT (gnd pin 2, output pin 3), V50 is a LM79M (gnd pin 1, output pin 3). V52 is a LM340T15 (gnd pin 2, output pin 3).

V53 is a LM2940CT and outputs 4.9805V (gnd pin 2, output pin 3).

I'm not seeing any ripple  (a few mV) on any of the regulator output, so that's probably not it.

So the power supply would seem to check out (I don't have an ESR meter, yet), but capacitors check out as C59 1.49mF, C63 720uF (should be 680uF), and C65 as 722uF (should be 680uF). First time I measured C65, it rose to about 780uF or so. Tested again, was the previous number, 3rd time, it was 1000uF. It is in circuit when testing, but this seems rather inconsistent. C63 and C65 are seemingly connected in series, but testing between C63 and C65 results in 345uF (which I think makes sense). So I have NO clue there.

I guess the next step is go through the 3V diag, or give it a simple 1Vpp sine into the input, and trace where that goes. If there are any other suggestions, I'd be very appreciative. Boy I was just hoping it was something obvious...

Oh, and alm, I do think I hear some of the relays on power up... VERY quiet, and tinny)... guess I just missed them before. That said, I don't think I hear any when switching between DC and AC (but that might just be me).

[alm]

If there is no significant ripple, than I wouldn't worry about electrolytic caps. This is a linear power supply, and it's not like you're trying to track down a noise problem. I would start with the diagnostic. Tracing the signal to the ADC is kind of confusing, because the multiplexer continuously switches the ADC input between zero/ref/zero/signal. The diagnostic mode lets you just choose one of those. I suggested 3 V DC because DC involves the least extra circuitry, and the 3 V range does not involve amplification or an attenuator.

[staze]

If there is no significant ripple, than I wouldn't worry about electrolytic caps. This is a linear power supply, and it's not like you're trying to track down a noise problem. I would start with the diagnostic. Tracing the signal to the ADC is kind of confusing, because the multiplexer continuously switches the ADC input between zero/ref/zero/signal. The diagnostic mode lets you just choose one of those. I suggested 3 V DC because DC involves the least extra circuitry, and the 3 V range does not involve amplification or an attenuator.

Okay, though I'm not sure I understand what it's telling me to look at/for.

There's 4 modes. And yeah, it's zero, zero, cal (guessing that's reference), and signal. So I should run like a 1V DC signal into the Hi/Lo inputs, put it in debug 3V DC, then in N01, look....? It tells me which relays/switches should be closed, but I'm not sure that helps me much. =/ Sorry, again, pretty new to this. Software and computer hardware I can diagnose, this stuff is still a bit new.

Thanks!

[alm]

Read chapter 5 of the manual (principles of operation), especially the part about the mux and the ADC. The ADC will measure the zero, reference and signal in quick succession to calculate a reading. Measuring the three signals (relative to analog ground) after the mux will tell you whether the problem is before or after the mux. TE1/TE2 looks like a convenient place to measure it. If all signals are correct, then the problem has to be in the ADC or digital logic.

[staze]

okay, but still not sure I get the process.

I'm assuming I have to input a signal of some type. TE1 and TE2 look like posts that several transistors are hooked to.

So, I hook the digital board back up (leaving the analog shield off), plug it in, and power up. I put it into debug mode for 3V DC. I set up a ? signal on my FG, and run it into the hi and lo inputs on the front, and then with either the scope, or my DMM, check TE1/TE2 (going by chapter 5, it looks like TE1, as the 3V input is routed through K1 (which I can't get to it's contacts easily), to Q30 (which hooks to TE1) for said signal relative to the lo input? Is that right?

As a side note, I note that the board isn't actually grounded, seemingly. It's all isolated from the chassis/mains ground. I'm assuming this is normal, but thought I would check. I also note some of the ceramic disk caps are looking a little worse for wear, but I haven't checked their capacitance yet, so no idea if they're actually bad, they just look a little flakey (the yellow ceramic seems to have chipped off in places).

[staze]

So, going by what I posited above, I set my FG to output a 5khz, 1Vpp sine wave, and put that into the hi-lo inputs on the front of the unit. I then put the unit into DC volts, on the 3V range, and with my scope probe ground hooked to the "lo" input, probed TE1 and TE2. Both show the 5khz wave shifted up 1V, and slightly reduced in Vpp (about 780-800mVpp). So that would seem to indicate... it's the ADC or Digital side?

Where would be best to look at this point?

Sadly, I think I've also hosed the calibration data for the 300mV and 3V DC ranges... =/ not quite sure how... but now both read "overflow" even when there is no input. Guessing I must be doing something wrong in the calibration attempt. I've giving it -300mV (or -3V), then +300mV (or +3V), then 0 volts (well, I'm turning my PSU output off for that one). Maybe this is why you need a real calibrator... =/ But until I can get a successful calibration, I can't be sure it's not just bad calibration data that's making these things not work. It seems really friggin' unlikely there'd be 6 of these things for sale, by the same seller, that all died the exact same way unless it was something other than random failure.

[alm]

I would see what the reference and cal/signal zero look like (N=1,2,3) at TE1. I would use a DC, not AC, signal to test the VDC function. Then I would check the ADC counts output (while not in the diagnostic mode) on both sides of the opto-coupler (see theory of operation, table 6-13 and schematic). I would change the input voltage and observe how the pulse chain changes if you change (eg. invert, or change by a factor 2) the input signal.

Attempting calibration if you suspect the ADC circuit is broken is a bad idea. The power supply was probably not the problem (although switching it off is not a good way to get 0 V, you should short the inputs), but the ADC was feeding the MPU garbage information. After repair, you will need to send it out for calibration or have access to accurate 300mV and 3V sources to bring it within spec. I would probably switch to the 30 VDC range for testing, although it involves more components and is therefore less ideal for testing.

[Smokey]

If you don't have any need for the scanner card, I need a replacement one for my 199.  Send me a message and we can work something out.

[staze]

I would see what the reference and cal/signal zero look like (N=1,2,3) at TE1. I would use a DC, not AC, signal to test the VDC function. Then I would check the ADC counts output (while not in the diagnostic mode) on both sides of the opto-coupler (see theory of operation, table 6-13 and schematic). I would change the input voltage and observe how the pulse chain changes if you change (eg. invert, or change by a factor 2) the input signal.

Attempting calibration if you suspect the ADC circuit is broken is a bad idea. The power supply was probably not the problem (although switching it off is not a good way to get 0 V, you should short the inputs), but the ADC was feeding the MPU garbage information. After repair, you will need to send it out for calibration or have access to accurate 300mV and 3V sources to bring it within spec. I would probably switch to the 30 VDC range for testing, although it involves more components and is therefore less ideal for testing.

Just getting back to this.

So, putting in a 2.50V DC signal from my FG. Put the unit in Diag for 300V (3V, and 30V both give overflow) DC. Checking TE1 at N 1,2,3,4 all give nothing. zero voltage (this is with my scope lead ground hooked to the Lo input, and probing TE1). TE2 gives the same. I'd almost wonder if a relay is stuck open/closed... or a bad FET, or something. 

Interestingly, I AM seeing that signal on K3 and K4 (on both sides). According to the diagnostics table, these should NOT be closed.

Checking 6-13, I'm having a bit of trouble on two points.

1. Where are all these referenced to (guessing digital ground), which should be the shared ground but I'm having a hard time finding a good grounding location. It would appear the pin closest to the 5V digital regulator that goes from the main board, to the digital board, is ground for the digital board, as it's connected to Vss (pin 1) on U8.

2. I can't find U3B, which is supposedly where to measure the A/D counts. I see U3, but U3B is no where to be found (though, I'll fully understand if it's right infront of my face).

Anyway, with all that, here's what I'm seeing:

U8-pin2: 1ms (good)
U8-pin3: 11ms (should be 15ms)
U8-pin6,7,8: 5V, pulse when button pressed.
U8-pin9: 0V (no pulse when buttons pressed)
U8-pin19: lots of pulses from 0 to +5V. Seemed to have some "runts" in there as well (much shorter duration than "normal" pulses).
U14 pin 4: 22ms 5V to 0V pulses. No variation.
U3B: Unknown (can't find it). U3 (incase it's a typo) pin 13 gives a 3.84197Mhz 0V to 5V square.
U8 pin 12: 1.92098Mhz 5V clock.
U8 pin 13: 960.491Khz 5V clock
U8 pin 14: 480.246Khz 5V clock
U8 pin 15: 240.123Khz 5V clock
U8 pin 16: 240.123Khz 5V clock
U10 pins 34 and 35: 2.0032Mhz 5V clock
U10 pin 37: 5V
U5 pin 9: 0V to 5V pulses, 1ms apart. Getting some definite noise, pre-pulses, runts, etc in there.
U5 pin 19: 5V
U5 pin 18: 2Mhz square 0V to 5V.
U5 pin 28: 3.72V

As a note, I had to re-close the "scanner" jumper (since the scanner wasn't attached) to get the A/D counts to work, as well as the 2Mhz square out of U5 to be steady.

I'm not exactly sure the U5 oddity is anything significant since I believe that's just the GPIB driver. But it behaving oddly might be a symptom. Or, maybe what it's doing is normal.

And that's it. not exactly sure what to test next. Nice to get some readings, but a bit frustrated that nothing is jumping out here.

As for measuring on each side of the optocouplers... they're UNDER the digital board... a bit hard to measure. =/

Any thoughts? Thanks for all the help!

[staze]

Correction. The first time I did my measurements of TE1 in Diag, I had not shorted the scanner jumper.

Now in 300V or 30V DC, in diag, I get 2.5V DC (my input signal) on all N settings. So, about the same as before, just that I'm actually seeing a signal now. So it still seems something isn't switching correctly. =/

[staze]

So, interestingly, Pin 18 on U8, which is tied to Pin 5 on the header between the analog and digital board, is labeled as "clock". Checking this with the scope gives a pretty bad looking "clock" signal.

Also of note, Pin 37 on U10 is not a +5V ±5%, it's seemingly either clock, or TTL of some sort. It's just a square between 0V and 5V, at about 510Khz. And pins 38 or 39 on U10, which according to the schematic should be an 8Mhz clock, are no where near that. No idea if that's just the impedance of the probe stopping the clock, or what.

The whole thing seems very inconsistent. Sometimes it does one thing, sometimes another. Seems like it could still be something funky with the power rails.

[staze]

As luck would have it, Dave covers multi-slope integration measurement in the TDT this week... which I THINK is how the 199 does it's measurements.

Still doesn't help much in what's wrong with this meter... but.

[codeboy2k]

I'm trying to wrap my head around everything you've tried so far

yes, the table 6-13 is the digital side, so all the voltages in that table are referenced to the digital ground, which would be pin2 of the 7805 marked U53. Or any ground pin on an IC would work too, in a pinch.  It shouldn't be that hard to find a digital ground.  What I often do when I am beginning a troubleshooting session is to solder a short black wire to a digital ground plane near the regulators, and a short BROWN wire to the ground plane near the analog regulators.. this is my DGND and AGND references, and I clip everyground of my scope or DMM or FG there   If I ever need to solder an earth reference I will use  a GREEN or GRN/YEL wire for that. This is just my convention, perhaps a best practice, to save time in the future clipping grounds all the time.

U3 is a dual 4-bit counter, so there is two parts inside.  If you have found the IC labeled U3, then you have found both U3A and U3B.

U10 is the MC6809 processor and Pin 37 is a /RESET line, so you should see it simply pulled high during normal operation, you should not see a clock there. Make sure you are probing at x10. I don't like the loading of a x1 probe, I almost always use x10.

you won't be able to measure the clock by probing pins 38 or 39 across the crystal. But I don't think your processor is a problem, it's clearly working, as evidenced by the fact that  the display is working, the front panel operations seem to basically work, and you can enter the various diagnostic N modes. So don't chase down anything in the digital section yet, you should stick with the analog side and verify the analog voltages, the reference voltages, and the signal paths as explained in table 6-11 for the various ranges. Stick with the 3VDC range first, and verify that path through the relays and muxes and that it reaches the input buffer amplifier U46.  In the 3VDC range, the buffer amplifier U46 should be configured for x1 gain. Verify this. The microprocessor switches the input buffer amplifier between x1 and x10 gain using  U45A.  Table 5-1 show the buffer amplifier configuration for each range. Verify this buffer amplifier is at the proper gain setting for each range. Verify the signal is appearing at the output of the buffer amplifier. This is the first point before the integrator. Verify the signal from your generator is appearing at the integrator input on the gate of Q1.  Verify the 3.84Mhz clock as used by the counters on page 4 of the schematic. print off the schematic and mark of signal paths that you have tested and known to be good.. use a highlighter, it can help here. You'll get a clear picture of where you are and where you need to go.

The integrator is made up of Q1/U19/C32 on page 4 of the schematic. Figure 5-6 shows a simplified schematic.  Put a scope on the output of the integrator, does it look like its "integrating"   looking at Figure 5-6 you can see the operation of the integrator , comparator and how the clock is gated out to the ADC counts.  Mr. alm already said, you should check the ADC counts at both sides of the opt-isolator and verify that it is changing pulse train with varying inputs.

All this needs to work. This looks like a very serviceable machine, actually.  I haven't bought one yet, but I am still considering it. After studying the schematic, they have written the service manual with a fully explained operation guide, detailed diagnostic procedures, built in diagnostic modes, and of course a complete schematic. And it's all discrete parts -- definitely a very serviceable machine.

[staze]

codeboy2k,

thanks! I'll have to give this a all a shot... and yeah, I never use 1X (though they are dual probes).

Once I'm home and can take a look, I'll probably have questions, because just looking at the schematics and parts list, I can't see where the gate actually is on Q1, or for that matter, where's the output of the integrator. As serviceable as this thing seems through the manual, it's actually a bit annoying since the digital board sits above a good chunk of the analog board, so probing is a bit troublesome.

When I'm searching for my input signal, should I just clip my probe ground to the input Lo, or should I clip to the analog ground? Guessing the answer is, both, in that when looking for signal, clip to "low", when looking at rails, clip to analog ground. Do you see anything wrong with inputting just a straight 2.5V DC signal for my test?

Also, looking at the schematic, I see say, "VOLTS OHMS HI" on page 1, it heads through the switch, then off to R13 and Q6, then over to Q30, then down and around. So I should be checking, say, the source of all those FETs to see my signal, maybe the C30 cap as well. As alm said, TE1 and TE2 look like they're in that signal path. And I have checked that, and the signal appears. So, the signal is making it to those points. So I think we're good to that point.

So then we head to U46, and check pin 6 (are they clockwise or counter clockwise from pin 1) to see if it's the same signal (or x10 if I'm at 300mV), and I'd assume also check pins 4 and 7 to see if I'm getting the ±15V rails.

Then there's the integrator. So, I look for the signal on the drain (or source) of Q4 (if I can reach it), then on G2 (if I can find it) of Q1 (maybe check the -15V on the source for Q1, and the +15V off R56/R55), then on Pin 2 of U19, and ideally, then a sloped line on pin 6. Can also check the rails on 4 and 7. Gotta say, the integrator is not what I'd normally look at and say "oh yeah, that's an integrator". 

K... all that seem reasonable? I'll give results in a few hours (once I'm home).

So, all that down, how do I check the 3.84Mhz clock? I know I can't probe it directly because it loads it down too much and stops it... so...?

Still have the nagging "there are 6 of these that all died in the same way". So either it's a design flaw (unlikely), or user error (what would blow if you pushed 500V into these things? Maybe a FET or an opamp?

Will get back to you with the results.

[staze]

To answer my own question, the Volts/ohms Lo input is tied to analog ground. So, I just clipped to that.

So, starting out, I powered the unit up, and on TE1 and TE2, with a 2.5V input, they read 4.5V. Hmm. Power cycled the unit, and they now read 2.5V. So SOMETHING isn't working reliably. I also note that when I powered it up the first time, the LED on the front (guessing it signifies trigger) pulsed about 2 or so times a second. Now, it pulses about 4 or more times a second. Power cycled again, and now TE1 and TE2 read 0V.

Power cycled again, got the expected reading (2.5V) on TE1 and TE2.

So... thoughts?

What would cause that?

Anyway... the below is done outside diagnostic mode.

U46:

The rails are +15V, and -14V. The non-inverting input is a bit higher than my input (about 2.9V). The inverting input is the same. And checking, it would appear the resistance between pins 2 and 3 (inverting and non-inverting) is basically 0R. Can't say if this is normal or not.

At first, I saw seeing 10.6V out of U46 at the 3V range, but after letting the unit sit a bit, I am seeing 3.4V at the 3V range, 10.6V at the 300mV range, and 0V at the 30V and 300V range. So that would seem relays aren't closing/opening.

On Q1, Gate 2 (which I THINK is shown here http://www.datasheetarchive.com/indexer.php?file=DSA00411750.pdf&dir=Datasheet-023&keywords=MP841&database=user-highscore as pin 7) shows 1.2V. But, checking it's "rail" I don't see a -15V on any of the pins. So I'm really quite stuck on that.

U19's output MAY be integrating... it's kind of hard to tell. I do see a +15V rail on pin 7, and a -14V rail on pin 4. The inverting and non-inverting inputs appear identical (I think this is correct). The output from pin 6 looks like about a 2.8V "noise" then a slow drop down to -80mV, then a pretty quick jump back up to 2.8V. See attached photo. Also interestingly, in the 300V range, in the 4.5 digit mode, I actually get a "useful" reading. It quickly jumps between 1.76V and 2.12V, and occasionally -95V or something. If I turn on the input, the reading drops down to .70V and .35V, still occasionally jumping to -95V. But in 5.5 digit mode, I get nothing useful. Also interestingly, the trigger pulse is consistent in the 5.5 digit mode, where in 4.5 digit mode, it pauses occasionally.

So... I got nothing.

[codeboy2k]

first things first... recheck all your voltage regulators and reference voltages too, there is a -6.4V reference, a -10V reference, a -2.8V reference and a -4.3V reference. make sure those are all ok.

For diagnostics, stick with the 3VDC range for now, try to get just that one working..

> and the signal appears [ed. at TE1 and TE2] . So, the signal is making it to those points. So I think we're good to that point

Yes if you're seeing your signal at TE1 and TE2, that is just before the input buffer, so your good there.  But you say power cycling you get different values at TE1 and TE2, and I presume you are still putting 2.5VDC on the input?  Just make sure every time you power cycle it still is starting up in the same mode, 3VDC range. If you power cycle it and the display says you are in the 3VDC range, and you are supplying 2.5VDC, but are seeing 4.5 volts at TE1,TE2, then don't just turn it off again, check where is that coming from??  That bus that is the input buffer (TE1,TE2) is fed from the sources of just 5 FETS.  If you have 4.5 volts on the TE1,TE2 bus, then just check the drains of each of those 5 FETS, Q30,Q32,Q33,Q34,Q35 for that 4.5V.  also, use a scope, not a DMM, is it 4.5VDC or is it oscillating and your DMM is just showing some average DC voltage?  These are diagnostic questions you should be asking.

Having -14V on the -15V rail on U46 seems off.  Is the power supply bad? -14V is not terrible, but since the -15V rail is a 7915, I know it should do better than this. So when you see -14V there, check the 7915 voltage regulator U50.

> The non-inverting input is a bit higher than my input (about 2.9V).

Thats far too much higher, I would think.. They should be equal, that's the opamp way... more importantly is the output of that opamp...
The output of the input buffer amplifier should be + or - 3.03V max. Since you are giving it 2.5VDC at the input terminals, you should see 2.5VDC there at the U46 buffer amplifier output. however, it's not that simple, because ....

> At first, I saw seeing 10.6V out of U46 at the 3V range, but after letting the unit sit a bit, I am seeing 3.4V at the 3V range,
> 10.6V at the 300mV range, and 0V at the 30V and 300V range. So that would seem relays aren't closing/opening.

Why do you say the opamp output being too high is an indicator of a relay problem? on the 3VDC range there is only 1 relay, K1 in the signal path. oh I see, you're talking about the 0 VDC on the other ranges... stick to 1 range, get it working first... understand how it works and the path it takes, don't jump around.  Try to get the 3VDC range actually reading correctly. 

Q1 is an MP841, that spec sheet seems correct, they show an LS841 equivalent and yes, G2 is pin7 viewed from the bottom in that drawing K1.
During a non-diagnostic mode, ie. during the normal ADC cycle, the input at G2 will be switching between AGND and the SIGNAL. This switching is done by U45B/Q4/Q17.

You can probably try to select a specific N mode of diagnostics to hold that signal at either one or the other for a check.

Note that the signal at the G2 gate is NOT the signal you input. Because the input signal can be bipolar +3.03V to -3.03V, there is an offset added to the signal before it hits the integrator, this offset will level shift the bipolar +- 3V input to a negative input only. This is done by the voltage divider formed by R58/R61/R62 and the switch U42C. I haven't been able to figure out the exact voltage shifts... maybe 0-3V becomes -4.6 to -3 and -3 to 0 input becomes -6 to -7.2... I think.. just not sure of the switch configuration when the input is positive or negative. If I am correct then your 2.5VDC input should be at the top of R58 on Sheet4-D1.  The bottom of R58 should be about -3.3V ... all referenced to analog ground.  This -3.3V should appear at the gate G2 in a pulse as it cycles through the ADC capture..check with a scope there.

> U19's output MAY be integrating ...
nope. It looks like it's oscillating. you should see a steady series of ramps up and down.  That area is a big problem, it needs to be focused on and fixed. but also check the other signal areas I mentioned above.

[codeboy2k]

If you want to check the digital clock on that analog section, then check : ( i hope I got these right )

U43 pin 4 should be 3.84 MHz
U17 pin 3 should be 1.92 MHz
U17 pin 6 should be 768 KHz
U17 pin 7 should be 384 KHz

U18 pin 4 should be 96KHz
U18 pin 5 should be 48KHz
U18 pin 6 should be 24KHz
U18 pin 8 should be 1.5KHz

U8 pin 3 should be either 48Khz or 1.5Khz or depending on fast or slow mode (4.5 or 5.5digits)

maybe some of those will help you verify the clocks...

But generally just focus on getting the DC 3V signal correct at the front end.. what you are seeing there just doesn't seem like its right to me.

[staze]

codeboy2k,

Thanks very much, this is good info. I'm swamped today and this weekend, so I'll be getting back to this Sunday night, or Monday evening, which sucks. I really wanted to stay in the groove of figuring this out.

I am using a scope for pretty much everything except continuity/resistance testing. I'll stick with the 3V range from now on... I have noticed while working on all of this, the relays seem to be working better, or at least, I can hear them now when switching between ranges/measurements, but that's neither here nor there.

I'll see if I can check any of this today. I'll recheck my VR's, but here https://www.eevblog.com/forum/chat/bench-multimeter-(another)/msg247316/#msg247316 is what I got last time around.

The annoying thing about the reference voltages is that they produce one thing, then divide it down, so finding where that division end point is... anyway, will see what I can find, I think last time I checked, the -2.8V was bang on, and therefore so was the -6.4V.

Will see what I can get to before I have to do, you know, work. =)

[staze]

So before other stuff takes me away, I did check the voltages on the transformer output. I can't see which pin is which, but there appear to be 3 taps off the secondary side. One of these gives 9.35V between what I'll call pins 5-6. One gives 19.100V off pins 3-4. The last gives 15.15V off pins 1 and 2. All are RMS AC. Also appears there's a 4th tap on the other side that gives 8.1V RMS. The others I must be measuring wrong, as they don't match up with what's going into the regulators. The ±15V regulators are getting 23V or so... oh.... filter caps. duh.

The 5V analog regulator is still giving 5.0065V DC. The -15V regulator is giving -15.043V. The +15V is giving 15.125V. So I think those are still good. Those are all with the meter as I didn't have time to fire up the scope.

Digital 5V gives 4.9794V. And now that I'm looking, that 4th tap seems to be for the digital regulator.

Anyway, more to come after the weekend. Thanks for all the help. If you think of anything else, let me know. I guess, hypothetically, what's the likelihood this is all caused by one broken component vs multiple failures?

[staze]

U17
Pin 1: 3.84Mhz.
Pin 3: ~1.62Mhz (fluctuates a bit)
Pin 5: ~650Khz
Pin 6: ~325Khz

Interestingly, I've found the Rigol DS1052e isn't a great multimeter. Where my Fluke says the -15V regulator is putting out a nice clean -15V, the scope at first said -14.4, but then as you zoom in, it gets to be a nice clean -15V. So it could be that the integrator and A/D buffer is getting that -15V as well. But, I'm not sure.

That said, something would seem to be up with my voltage rails, as I would think with NO input at all, U46's (input buffer) output should hover around 0V... instead, it's around -.1V. With my 2.5V input signal, I'm getting out 3.06V. The input (non-inverting) from the input buffer (read on TE1/TE2, and pin 3) read 2.56V (which is what the scope seems coming from my FG directly at the inputs to the DUT).

U46's rails read as +15V (on pin 7), and -15V on pin 4 (and the case, as it should be). Pin 2 of U46 reads as that same 2.54V. With no signal, Pin 3 (input) reads 0V (basically), and Pin 2 reads that -.1V previously indicated (so that's where it's coming from).

-10V reference checks (-10.058V). -6.4V reference checks out at -6.440V. The -2.8V checks as -2.8729V. There's a +5VR marked on the analog board, and that checks as +5.0835V. The -4.3V reference, though, is a bit hard to find. It's on CR16, which is right in the middle of some resistors on the bottom, it takes the -6.440V reference, and takes it down to a theoretical -4.3V, but in my case, it's -4.1746V (which seems a bit low).

Anyway, that's all for tonight. More tomorrow once I have some more time. At least I'm getting some use out of my test gear... my wife doesn't think it's quite so silly now. =)

[staze]

Interesting tidbit... on a whim I was fiddling with the unit listening to when relays would flip, etc... and I switched it over to ohms, and clicked auto, and it went to overflow on the Megaohms scale. Okay... so I shorted the inputs, and low and behold, it's reading 1ohm or less... though it occasionally flips to like -49R for one reading then back again. So something is still wonkey there (since, you know, negative ohms?)

Un-shorting the inputs makes it go back to overflow on the Megaohms scale. But interestingly, testing with a 4K7 resistor resulted in it clicking between K and M ranges, and never reading anything. So yeah, still doesn't work. Ignore this. =)