Keeley 4-knob compressor: Distorted and extremely gated

[Gravizt]

Problem experienced
1. A lot of distortion from circuit. Redused slightly by adjusting c2. lipping pot.
2. Heavy gating effect. No signal comes through unless the guitar is played vigorously. Notes die out very suddenly. Lighter playing can not be heard

Pictures
Drawn in red mark to indicate which corner is which on front/back pictures
Component side
http://imgur.com/2CyRKdH
Solder side
http://imgur.com/fLzN12i
Layout
http://imgur.com/a4dNnU2
Schematic
http://imgur.com/jGcT5PA

Voltage Measurements
Supply voltage: 9.56

IC1
1: 0.16
2: 4.95
3: 4.95
4: 0
5: 0.19
6: 7.45
7: 9.56
8: 8.84

Q1
C: 7.03
B: 2.86
E: 2.50

Q2
C: 6.65
B: 7.40
E: 6.64

Q3
C: 9.49
B: 0
E: 0

Q4
C: 9.49
B: 0
E: 0

Q5
C: 9.52
B: 9.50
E: 8.93

Measuring around the base of Q2, it seems (in my limited knowledge) that the surplus voltage has to come from the IC. The sides of R9 and C5 NOT connected to Q2 measure around 3.5 volts. I thought maybe the extra voltage could be dumped in through these, but it seems thats not the case. The only other point connected there is pin 6 of the IC, which I figure is whats dumping the extra voltage. No idea why though.

Audio Probe
I've followed the signal from the input to the IC with a probe. The signal arrives relatively inscathed at pin 2 and 3 of the IC, but comes out terribly distorted at pin 6. Not sure where to go from here.

Part Substitutions
IC: CA3080EZ
Transistors: 2n5088
Polarized 1µF caps are supposed to be tantalum. I've used electrolytic.
Unpolarized 1µF cap in the top left replaced by 100n ceramic. It was a polarized 1µF, but I canged it for the highest value non-polarized I had to see if it made a difference. It did not. The pictures are taken before this switch.
500K Reverse Log sustain pot replaced by 500K Linear.
150K Linear attack and clipping pots replaced by 100K for testing.
2K Trimmer should be "Set dead center". I am using a 25 turn 2K trimmer set to 1K.

Measures Taken
Double checked component placement. Found and corrected one error.
Double checked polarity of all the parts against datasheets for the components I used.
Replaced Q2. No difference measured.
Replaced IC. No difference measured.
Reflowed all solder joints.
Checked for solder bridges with DMM's continuity tester. None found.
Cried.
Wrote this post.

Additional information
I made one wrong cut. At pin 5 of IC1. Bent the pin and used extra solder to fix. DMM confirms continuity of the joint.
The socket for the IC is a 16pin socket cut in half, because I had no 8 pin IC's.
All parts substitutions made on the basis of availability at the time of building.
No attempt has been made to match transistors.

Really hoping someone can offer some advice on this. I am fairly inexperienced, and completely lost as for what to do next. I can see that the base voltage on Q2 is way high, but I cant figure out why. I don't know if its the output from pin 6 on the IC, or if theres a bad component somewhere.

[Agung Kurniawan]

Its seems i cant help you much my friend. I have breadboard it and i get success(you can see my last post entittled 'Problem On Breadboarding CA3080/LM13700 Comp')
Just check the solder side, may be some get 'break' up or maybe there's still some dust on component pin wich block the solder to connect the solder side of the pcb. Or maybe just try using LM13700 to subtitute the CA3080. Hope this helping

[PRR]

Welcome.

Where did you get the '3080? There are a lot of fakes on the market. Wastes a lot of people's time and energy.

Bodging '3080 pin 5 is not good. That pin 5 is one of the few pins in chip-world which can be blown with careless connection.

R17 R18 make a 2.9V supply (check that!). Pin 6 of '3080 should bias via R9 to very near this voltage. One common sign of fake '3080 is that pin 6 will not do what R9 tells it to do.

Your detector T3 T4 and follower T5 are showing voltages expected for no-signal, and Sustain pot appears to be trying to pull '3080 pin 5 up (but not getting the 0.5V-0.7 needed to wake-up a healthy '3080). So I'd focus on '3080, bad chip or un-found wiring errors.

[Gravizt]

I sourced the chips from Ebay, so I guess they could be bad. These ones.

http://www.ebay.com/itm/152181529095?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

So I'm getting 3 volts between R17 and R18. Pin 6 is sitting just over 6, so I guess thats no good.

Regarding pin 5 and the sustain pot, I saw some strange stuff that I I couldn't understand there before. Any help understanding whats happening there would be great.

Lug 3 of the sustain pot is seeing 7.6 volts. With the pot turned all the way up, so zero resistance, I'm seeing that voltage getting to R15, which in my case is 27k.The voltage on the other end of that, by pin 5, is 0.17v.

If I set the sustain pot to 27k and measure the voltage through it, I find that after going through 27k of resistance from the pot, its sitting on 3.9 volts. What I don't get is why 7.6 volts through 27k resistance from the pot becomes 3.9 volts, while the same voltage, when run through 27k resistance from R15, becomes 0.17 volts. I understand how the chip doesn't see a difference in where the resistance comes from, but thats kinda whats throwing me off. Has it something to do with how the chip works? Dark magic? Or just something fairly simple that I'm completely missing?

[PRR]

> Lug 3 of the sustain pot is seeing 7.6 volts.

I assume this is also T5 E, which was 8.93V yesterday.

I did not see the 27K for 15K R15 substitution.

I also can't make sense of your numbers, to the point that the schematic is annoying me (connect-the-dots) and I'm tired of computing various values.

Look. T5 E should be nearly 9V, within 10%. '3080 pin 5 should be under 0.6V, zero within 10%. Using a fat crayon, we have 9V and zero V, feeding a series string which may be 27K+27K or zero and 27K.

Bottom right box:



Assuming the 9V & zero assumption, we DO expect about 4.5V one way and 9V the other way.

Pin 5 of '3080 "should be" 0.4V to 0.6V for any useful current. Even over a 1,000:1 ratio. 9V through 527K to 27K is only a 20:1 range. Pin 5 of a healthy '3080 should change very little as Sustain is turned from 500K to zero (plus fixed 27K). The actual voltage on pin 5 is not much use. However a value of 0.19V or 0.17V is just too low to be right.

Hmm... if you change from 27K to 27K+27, 2:1 ratio, we expect about 20mV 0.020V change of diode voltage. Is that how you are reading 0.17V vs 0.19V?? If so, there is a chance the '3080 is fine and your meter has a problem reading low volts.

[Gravizt]

You're correct in that its T5 E. Sorry about the change there. I forgot to check the battery I was using. With a fresh one it reads the same as in the OP.

I've gone over everything from the picture you posted, and right up to pin 5 everything is a match, within a few hundreths of a volt.

With turning the sustain pot, I'm measuring a low of 0.05 volts and a high of 0.19 volts at pin 5.

There could very well be an issue with my meter. Its a cheapo chinese one.


Sorry if there is anything more from your post I didn't respond to. I'm having some problems following all this. Thank you very much for your help so far.

Edit: I found another schematic. A couple of values are different, but other than that it seems to be the same. Its a bit easier to read, at least for me.

http://imgur.com/FG8xnM0

[Gravizt]

Bump