Repairing HP E3610A

[morrone]

Hi folks,

I have an older (1992 I believe, says 9207 on the mobo, and other components look like they are from '92) E3610A.  The last time I turned it on the voltage display quickly went over scale, and it let out some of the magic smoke.  I could use some help with the repair.

It looks like someone did some minor repair in the past and replaced C7 and C13, but they bodged in some axial caps.  I replaced those with nice new radials for good measure.  Trimmer pot R20 was clearly replaced as well (three inline legs bent to fit the three offset legs holes in the board).  I replaced that as well, making sure to set it to the same resistence value as a starting point.

Not working yet.  But I didn't really expect it to work yet because I see a trace running from Q2 that has the solder mask burned off.  The trace still seems to be intact, fortunately.  I have a new 2N4036 on its way to replace that, but its taking the slow route to get here.

In the mean time, I did some more probing to try to figure out if that is going to be my only problem remaining.  But I'm not really sure yet, and thats where I could use some tips.

First of all, on the 2A range I see an output voltage of 26.72V, on the 3A range I see 16.47.

The voltage control does not appear to be working.  I tried the CC settings, and the current does seem to track, at least at lower amperages.  I was reluctant to turn the current setting up too high with the voltage unregulated and a trace on the pcb already smoking earlier.  But my assumption is that the CC block is working.  Seem reasonable?

I checked all test points and it looks like the display power supply voltage is good, and all of the voltages in the "Reference and Bias Supply" block are good (+12V, +5V, -12V).

Next I checked voltage between TP2 and TP4, which is right after the CR2 bridge rectifier.  On the 2A range I see 27.33V and on the 3A range I see 17.03V.  For kicks I checked the voltage across C1 (and feeding CR2) and see on the 2A range 20.89 Vrms and on the 3A range 13.26 Vrms.

What do I check next?  Does it seem likely that Q2 is the main culprit?

Actually, staring at the circuit, I see that Q2 is controlled by U1 in the "Voltage Error Amplifier" section, which is where the voltage control connects.  U1 is an op amp, so maybe I can learn something by probing around that.

I clipped my ground lead to TP6, which is also the positive Output, and the reference for the +12 and -12 that are feeding the op amp.  Sure enough, pin 7 has 12V and pin 4 has -12. 

Since this is an op amp, it should try to keep pins 2 and 3 equal.  But that is not what I am seeing.  Pin 2, the negative input, is at 0V.  Pin 3, the positive input is -0.7V, and varies with the voltage control knob.  Output voltage on pin 6 is -10.10V regardless of the voltage control knob setting.

So my question is this: Is the U1 op amp dead?  Or are the settings just fixed that low because Q2 is fried and the output voltage is unresponsive to U1's attempts to correct?

[just_fib_it]

What do I check next?  Does it seem likely that Q2 is the main culprit?

If CC mode is working then Q2 can't be the culprit. Q2 is controlled by both the voltage and current error amplifier circuits (via diodes CR4 and CR5). Basically whichever error amplifier pulls Q2 down harder wins.

Since this is an op amp, it should try to keep pins 2 and 3 equal.  But that is not what I am seeing.  Pin 2, the negative input, is at 0V.  Pin 3, the positive input is -0.7V, and varies with the voltage control knob.  Output voltage on pin 6 is -10.10V regardless of the voltage control knob setting.

This sounds as if the op amp is working, pulling down the output because the noninverting input is negative. Maybe CR5 it the problem? What is the voltage at the base of Q2?

[morrone]

I am seeing -0.65V on Q2's base.

[megajocke]

It looks like someone did some minor repair in the past and replaced C7 and C13, but they bodged in some axial caps.  I replaced those with nice new radials for good measure.  Trimmer pot R20 was clearly replaced as well (three inline legs bent to fit the three offset legs holes in the board).  I replaced that as well, making sure to set it to the same resistence value as a starting point.

I have some recollection of a service bulletin affecting at least those two capacitors. The values were changed to suppress some switch on/switch off transient if I remember correctly. It might have been another model though.

The voltage control does not appear to be working.  I tried the CC settings, and the current does seem to track, at least at lower amperages.  I was reluctant to turn the current setting up too high with the voltage unregulated and a trace on the pcb already smoking earlier.  But my assumption is that the CC block is working.  Seem reasonable?

What do you mean by "the current does seem to track"? Did you measure it with an external multimeter and a low resistance/shorted load? If you didn't -- then don't do that just yet! 

If that wasn't what you did and you meant that the current display changes when twiddling the current knob with the "CC set" switch or whatever it's called activated then it's very likely that your pass transistors Q1,Q3,Q6,Q7 (probably two installed) are shorted and Q2 blew open as a consequence.

With negative voltage at the base of Q2 there shouldn't be any output voltage.

[morrone]

I have some recollection of a service bulletin affecting at least those two capacitors. The values were changed to suppress some switch on/switch off transient if I remember correctly. It might have been another model though.

Yes, thanks, I spotted that in another thread too.

What do you mean by "the current does seem to track"? Did you measure it with an external multimeter and a low resistance/shorted load? If you didn't -- then don't do that just yet! 

If that wasn't what you did and you meant that the current display changes when twiddling the current knob with the "CC set" switch or whatever it's called activated then it's very likely that your pass transistors Q1,Q3,Q6,Q7 (probably two installed) are shorted and Q2 blew open as a consequence.

Yes, sorry, I should have been more clear.  You are correct that I just see the current display change when I twiddle the current knob while holding CC Set.  Don't worry, I didn't want to try checking the current while the voltage appears to be unregulated.

With negative voltage at the base of Q2 there shouldn't be any output voltage.

I just checked the voltages around Q2.  This is a PNP transistor, and I found that the emitter is at -8.07V relative to the base.  For current to flow in a PNP, the base needs to be at least 0.7 volts lower than the emitter, so I can see that Q2 should not be passing current right now.

The circuit diagram shows the FETs Q1, Q3, Q6, Q7 for the E3612A rather than just the two Q1, and Q3 NPNs of the E3610A and E3611A.  The base of both NPNs is controlled by Q1, correct?

Ok scratch that.  I was going to try to reason out how Q2 controls Q1 and Q3, but I found something very strange along the way.

R3 is connected to +12V on one side and the emitter of Q2 on the other.  One side of R3 reads 12V, the other 1.1V.  Now the strange part: the voltage at Q2's emitter pin a very short trace away reads -8.75V.

I think I just need to take out Q2 and see what is happening under there.  Maybe the trace is now broken right under Q2 where I can't see it.

[morrone]

It looks like the trace is ruined after all.  I'm measuring over 1 Mohm from that the pad for Q2's emitter to the pad on R3.

See the attached image.

I cleaned the area after removing Q2.  It looks to me like the pads are nearly gone.  And pin 1's (emitter) connection to the trace looks bad.

I wonder if the C7 and C13 leaked at some point, and the electrolyte corroded the pads.  You can see a bit of corrosion on the pad in the upper left of the image.  That is just a post of the heat sink.  But that is also near the C13.

Now I suppose I'll need to investigate PCB repair if I want to go any further.

[morrone]

Good news!  I drilled out the corroded through hole for the Q2 emitter, fixed the trace, and then installed a new eyelet in for the through hole.  Continuity was good, so I installed a new Q2.

I now have voltage!  I measure pretty darn close to what the display says on a multimeter.

One thing seems odd though, both the 0-8V and 0-15V ranges top out at around 17V.  Does the E3610A always do that?  Or do I have another problem to track down?

[atsthng]

One thing seems odd though, both the 0-8V and 0-15V ranges top out at around 17V.  Does the E3610A always do that?  Or do I have another problem to track down?

Both ranges top out around 17V in mine.

[georges80]

One thing seems odd though, both the 0-8V and 0-15V ranges top out at around 17V.  Does the E3610A always do that?  Or do I have another problem to track down?

Yep, as stated, perfectly ok. The voltage 'range' is just what is guaranteed for the maximum current of the unit. i.e. it doesn't set an actual calibrated maximum.

Good job on the repair. I like the E36xx series and have 4 of them myself (one bought new and 3 freebies). Decent performance within their spec and quiet being fanless.

cheers,
george.