ross compressor debug! help!

[jstack]

So I've been searching all the other ross comp debug topics. Seems none of them are having the same troubles I am 

1. Ross Compressor
2. http://www.aronnelson.com/gallery/main.php/v/MarkMs-Gallery/album15/album76/ROSS_comp_LAYOUT.gif.html
3. I followed it exactly
4. I used 2n5088 instead of 2n3904 transistors
5. Negative ground
6. The output is compressed, but it makes a horrible fuzz distortion sound  The intensity of the fuzz increases with the compress knob, as if it was a gain knob =/

Thank you VERY MUCH for looking. I'm still pretty new to electronics, I really appreciate the help this forum brings!

It seems as though all of my voltages are pretty far off 

Q1
C 7.16
B 2.05
E 1.65

Q2
C 5.96
B 3.38
E 2.85

Q3
C 8.82
B 0
E 0

Q4
C 8.72
B 0
E 0

Q5
C 8.75
B 8.75
E 8.17

Q1
P1-unused
P2-4.12
P3-4.15
P4-0
P5-0.68
P6-3.36
P7-8.8
P8-unused

D1
A 0
K 8.66

D2
A 0
K 0

D3
A 0
K 0

I have used Mark's layout, but referenced the schematic from fuzzcentral, I hope that wouldn't have caused this problem 
These are the voltages that it should match, from fuzzcentral: http://fuzzcentral.ssguitar.com/ross.php

Q1
C 7.98V
B 2.10V
E 1.65V

Q2
C 7.28V
B 2.87V
E 2.34V

Q3
C 9.5V
B 0V
E 0V

Q4
C 9.5V
B 0V
E 0V

Q5
C 9.5V
B 9.32V
E 9.0V

IC1 (CA3080E)
Pin 1 = N/U*
Pin 2 = 4.92V
Pin 3 = 4.91V
Pin 4 = 0V
Pin 5 = 0.633V
Pin 6 = 2.76V
Pin 7 = 9.5V
Pin 8 = N/U*

[MikeH]

Your battery is a bit low- I'm not totally sure if that has anything to do with it, but I like to use a nice fresh one for debugging.  I think that might also be cause for a couple of inconsistencies I'm seeing in your voltages.  For instance the collector of Q3 and Q4 as well as the base of Q5 should all be the same, but they're a bit different.  Could also be your volt meter.  Other than that  though the voltages look to be along the lines of 'correct'. 

Do you have an audio probe?  Just takes a minute to make one, and that could help you track down the source of the fuzz, for starters.

[jstack]

Thanks for the suggestion of an audio probe! This thing looks very helpful!

[MohiZ]

It's probably just his volt meter. The voltages are close enough to be "the same". Mine also shows slight differences in situations like this, although the voltages must be the same since they're directly connected together. I agree that it's a good idea to rule out the case of low battery, although I doubt that's the real culprit. +1 for the audio probe, that can be used to kind of narrow it down since all your voltages seem O.K.

There's a 2K trimpot in the schematic. Have you tried to tweak that? Also, have you checked for solder bridges and bad solder joints? Re-check the polarity of the electrolytic caps, and the orientation of the transistors.

One thing worth noting is that 2n5088 has massively more gain than the 3904. I hope that's not the cause of your problems, seeing that both the builds you linked use the same emitter and collector resistors. You could try using smaller collector resistors for Q1 and Q2. This would reduce the amount of gain without affecting anything else. Although I'd imagine others would've had this problem as well if this were the reason.

[R O Tiree]

First thing that occurs to me is that there are a fair number of places where we can add up voltages and get "9" or so, and we aren't. For example, at Q1, we should get around 9V if we add the collector and emitter voltages. We get 8.9, which is not far off but, as we progress down the list, we get 8.8V at Q2 doing the same trick. (This is a by-product of having equal collector and emitter resistors, BTW). We are only getting 8.75V at the collector of Q5 (that should definitely be 9) and 8.66V at D1 (again, that should definitely be 9).

This suggests to me that the battery is on its last legs and, as jstack was working his way through the test points, the voltage was gently declining.

In addition, I'd endorse what MohiZ says about the trim pot. Stick it in the middle and then adjust around that point. 1% Metal film resistors are pretty damn good and the readings on pins 2 and 3 of IC1 should be within a gnat's doo-dah of each other, not 30mV apart. Having that trimpot mal-adjusted will definitely give you grainy distortion.

Next point is that the Vb is about 0.5V too high. I'd check and re-check the 56k and 27k resistors for value (de-solder one leg and measure them). I seem to remember an answer Mark H (?) gave a while ago about the 3080 not being happy with too high a Vb in this circuit, because it runs out of headroom.

Lastly, I'd get a really strong magnifying glass (a jeweller's loupe is invaluable for this) and have a really critical look at the solder-side of the board. It only takes one tiny little whisker of solder or a minute piece of copper that wasn't quite etched away to short 2 components and you'll be there all week trying to find what you did wrong, only to find that it is all, in fact, correct but the short is throwing things out of whack. A dentist's pick is invaluable for getting rid of these little copper/solder whiskers.

[jstack]

Thank you all so much for your replies!
As many of you have said, that battery was dying (forgot to check it before I plugged it in 
I replaced it with a new one and rechecked all my voltages, this is what I came up with:

Q1
C 7.6
B 1.9
E 1.7

Q2
C 7.4
B 2.4
E 1.9

Q3
C 9.4
B 0
E 0

Q4
C 9.3
B 0
E 0

Q5
C 9.3
B 9.3
E 8.7

Q1
P1-unused
P2-4.5
P3-4.4
P4-0
P5-0.67
P6-2.4
P7-9.3
P8-unused

D1
A 0
K 9.1

D2
A 0
K 0

D3
A 0
K 0

I checked both the 56k and the 27k resistors, they are both the correct values (all of the resistors -trimpot are metal film). The trimpot, works as it should but there is a large amount of nasty distortion no matter it is 
I have checked pretty much every trace there is on this board and still can't find any bridges. I will look again tomorrow when its light outside...
The audio probe is working pretty well, but since I'm pretty new to this its a little difficult to source the fuzz. It seems as though its coming from pin 6 of the IC, but I'll have to investigate it better 

[R.G.]

Thank you all so much for your replies!
As many of you have said, that battery was dying (forgot to check it before I plugged it in 
I replaced it with a new one and rechecked all my voltages, this is what I came up with:
...

I checked both the 56k and the 27k resistors, they are both the correct values (all of the resistors -trimpot are metal film). The trimpot, works as it should but there is a large amount of nasty distortion no matter it is 
I have checked pretty much every trace there is on this board and still can't find any bridges. I will look again tomorrow when its light outside...
The audio probe is working pretty well, but since I'm pretty new to this its a little difficult to source the fuzz. It seems as though its coming from pin 6 of the IC, but I'll have to investigate it better 

The voltages don't look all that bad.

Check the resistance between pin 5 and the emitter of Q6 with power off. I think you may have an open circuit that leaves the gain turned up all the time.

[jstack]

Thanks for the help RG, did you mean Q5?
The resistance between pin 5 and the emitter of Q5 is: 48.6k
I redid all the solder joints around there and on the rest of the board as well. Still nothing 

[R.G.]

Thanks for the help RG, did you mean Q5?
The resistance between pin 5 and the emitter of Q5 is: 48.6k
I redid all the solder joints around there and on the rest of the board as well. Still nothing 

Yeah, Q5. If you can't run the sustain pot up to over 450K or so by turning the knob, it's hosed.

The way this thing works is that the voltage the base of Q5 drifts up, pulled up by R20, unless signal into the bases of Q3 and Q4 cause them to pull it down. High voltage on Q5 base means high voltage across the sustain pot and R19, which means lots of current into pin 5, and therefore, high gain. Signal into the bases of Q3 and Q4 pull down on Q5 base, lowering the voltage at Q5 emitter, lowering current into pin 5 and lower gain.

I suspect you have either too little resistance between Q5 emitter, or an open between Q3/Q4 collectors and the base of Q5, never pulling it down when there is signal.

[jstack]

Sigh, this one may have beaten me 
I replaced Q5 and R20, measuring them for tolerance before installing
Then measured the resistance between both Q3 and Q4 collectors and the base of Q5, both read 0.0


Also, if it helps, the 'dirtyness' of the fuzz is related to the input signal. If you slam on it its much nastier than if you play normally. While playing very quietly, its just a little bit distorted.

Should I just pull the chip out and redo this whole project?

[R O Tiree]

Hmmm... Actually, now you mention it, I had a Ross build that distorted badly and I came to the exact same conclusion in the end. Replaced the 3080 and it was fine after that. I'd forgotten that... sorry.

It's worth trying, but do the solder-side inspection first, just to be sure. In addition, check what RG said - with power off, measure the resistance between Q5's emitter and pin 5 of the 3080. As you turn the sustain knob, you should get readings from 27k up to about 500k or so.

The new Vb figure you have is much more reasonable, BTW. While you're doing the solder-side inspection, have a really good look at every joint to make sure that they are nice and shiny and no crystalline ones, with a "concave" meniscus attached to every leg of every component. Vb dropping by a whole volt after you de-soldered/re-soldered those 2 resistors might indicate that there is something else amiss. Your pin 2 and 3 voltages are now 100mV apart, and that's very wrong.

[MikeH]

I built a ross a long time ago that was doing this too.  I had a solder bridge somewhere that was mucking up the gain.

[R.G.]

...Your pin 2 and 3 voltages are now 100mV apart, and that's very wrong.

Yes, it is. But it's quite difficult for beginners to measure more accurately than 100mV with voltage-to-ground and an unknown voltmeter. I deliberately left that kind of thing out of the "what to do" for that reason.

Pin-to-pin measurements are needed sometimes, but I pick those up on the second pass after I get the other voltages. Generally, if it messes up high impedance pin voltages, it messes something else up too.