Debugging a pedal that stopped working after not being used for a day

[swever]

A very very strange thing happened to my keeley 4-knob compressor build. I had tested the board after populating - it worked fine. After that I had boxed it and tested it for a few days - everything worked fine. Then I put it aside and started another project. Just now I wanted to try the 4-knob in conjunction with a freshly populated ea tremolo. To my surprise the pedal did not work when engaged. I checked for the usual culprits - shorts, bad soldering etc - but everything seemed to be fine. I desoldered it from the the box and attached a new power connector and jacks directly to the board - still not working properly - I only get very faint output (both "sustain" and "clipping" cut the output when set to minimum) with no compression.

I used the storyboardist's layout which is apparently based on this schematic

This is how my pcb looks currently. There are no solder bridges or anything like that (they could not have appeared by themselves while everything was boxed up anyway). I used 2n2222a transistors instead of 2n3904's. Most of the caps I used are ceramic disk. Electrolytics seem to be oriented correctly.

I tried using a different CA3080 chip (that also worked fine before this mystery happened) - the same result.

I started "audio-probing" the circuit and noticed there is no audio at any of the IC's pins. I guess that's a sign of something but I can't really understand the schematic very well. Which places should I check?

The voltages are the following:

Battery: 7.68V

Q1:
E:2.16V
B:0.56V
C:2.87V

Q2:
E:1.58V
B:2.11V
C:5.26V

Q3:
E:0.00V
B:0.00V
C:6.26V

Q4:
E:0.00V
B:0.00V
C:6.50V

Q5:
E:6.26V
B:6.26V
C:0.02V

IC:
1:0.06V
2:3.47V
3:3.47V
4:0.00V
5:0.66V
6:2.10V
7:6.89V
8:0.06V

This whole thing drives me insane tbh. It's the first time in my experience that a build works fine and then dies by itself while not being used or moved.

[EBK]

Sometimes mechanical tension can work itself out slowly, causing a failure as it reaches equilibrium. My first guess is bad ground at the input jack, but I haven't looked at the pic yet.   

[swever]

I removed the board from the box and replaced the jacks. I am 99.999% sure its not a mechanical issue.
Must be a compnent failure.

[thermionix]

Q5C

[EBK]

Q5C

You sunk my battleship!

[stringsthings]

7.68V is pretty low for a battery.  I'd try a fresh 9V.

[swever]

Right! Q5 voltages were measured wrong. My bad. Re-measured with a fresh(er) battery it and its the following:

Battery:
8.37V

Q1:
E: 2.57V
B: 0.57V
C: 3.18V

Q2:
E:1.99V
B:2.51V
C:6.26V

Q3:
E:0.00V
B:0.00V
C:8.14V

Q4:
E:0.00V
B:0.00V
C:7.42V

Q5:
E: 7.60V
B: 8.14V
C: 8.18V

IC:
1:0.06V
2:4.08V
3:4.08V
4:0.00V
5:0.66V
6:2.51V
7:8.15V
8:0.00V

Still not working.

[Mark Hammer]

Um, you didn't happen to wire up the input jack wrong so that the battery never disconnects, did you?  I'll confess to that error in past.

[rezzonics]

The biasing voltages on Q2 look Ok with 0.5V between B and E. But Q1 seems wrong. The rest are also wrong. B and E 0V is not good

[duck_arse]

yeah, Q1 base is too low - check the 470k's and the electro. also, Q3 and Q4 collectors are joined, how come yours are showing different measures?

[swever]

Um, you didn't happen to wire up the input jack wrong so that the battery never disconnects, did you?  I'll confess to that error in past.

Nope. I re-wired jacks and battery connector when started debugging.

The biasing voltages on Q2 look Ok with 0.5V between B and E. But Q1 seems wrong. The rest are also wrong. B and E 0V is not good

yeah, Q1 base is too low - check the 470k's and the electro. also, Q3 and Q4 collectors are joined, how come yours are showing different measures?

Q3 and Q4 collector voltages were the same - just a measurement error.



I desoldered the 470k's and measured them outside the cirucit - they were okay so I soldered them back. Replaced the 1uF electrolytics with new ones. Replaced Q1 with a new 2n2222a. Now I do get a slightly overdriven output just below unity with all the controls set to maximum.

The new voltages are:

Battery:
8.34V

Q1:
E: 1.45V
B: 1.87V
C: 6.68V

Q2:
E:2.02V
B:2.53V
C:6.11V

Q3:
E:0.00V
B:0.00V
C:8.14V

Q4:
E:0.00V
B:0.00V
C:8.14V

Q5:
E: 7.70V
B: 8.23V
C: 8.23V

IC:
1:0.01V
2:4.33V
3:4.33V
4:0.00V
5:0.58V
6:2.45V
7:8.24V
8:0.00V


I guess I gotta keep replacing trannies one by one starting with Q3 and Q4, which look the most suspicious.

[swever]

It seems to have come back to life after I replaced Q3 and Q4. Surprisingly their voltages did not change much. I replaced the remaining two transistors just in case.

I am wiring it all back into the box now. I have a feeling I lost some output volume compared to "before" - will have to test it with bypass wired.


Here's the voltages I just measured:

Battery:
8.34V

Q1:
E: 1.45V
B: 1.87V
C: 6.74V

Q2:
E:1.94V
B:2.46V
C:6.25V

Q3:
E:0.00V
B:0.00V
C:8.19V

Q4:
E:0.00V
B:0.00V
C:8.19V

Q5:
E: 7.66V
B: 8.19V
C: 8.20V

IC:
1:0.01V
2:4.26V
3:4.26V
4:0.00V
5:0.62V
6:2.46V
7:8.20V
8:0.00V

[EBK]

Well, now that it's working, make sure you use that pedal every day!   

[swever]

Well, now that it's working, make sure you use that pedal every day!   

 

In fact it is a really good pedal compressor - the best I tried yet. To my ears, it's much a better sounding, efficient but delicate at the same time and more tweakable compressor than both orange squeezer and bearhug. I'm really liking it, so I may be using it often indeed.

I boxed it up again and made sure that it works like it did before.

What worries me is that I have no clue at all what was the reason. If it wasn't  just bad solder somewhere wich is very unlikely, that would mean that at least two transistors died. I'd love to find out why and how it happened and how to prevent it.

[swever]

I just realised it could be that the pcb's ground plane was shorted to the power plug's positive terminal for some time.
Could this have killed the trannies?