Ross Comp Clone dies a slow death.

[eliktronik]

I finally got around to looking inside the damn thing, but can't find anything wrong. This effect was the first I ever built, full of less than perfect solder joints and the like. But, once I got it working, it worked for quite some time. Slowly, it started to get quieter and quieter, until it finally died one day. Now, quite some time later, I finally opened 'er up, and can't find anything wrong. Everything that's supposed to, gets 9v, and all the grounds look good. Does anyone have any ideas as to what could be wrong? Is there a part that would cause the effect to become quieter and then die? Thanks in advance.

[Paul Perry (Frostwave)]

The only 'fade away' components I know, are batteries and electrolytic caps.

[eliktronik]

Hmm, is there an easy way to test electrolytic caps? What could make them go bad? Thanks.

[petemoore]

reverse supply overvoltage and age
  replacing with a known god cap is probably the easiest way to check ... tha I know of

[eliktronik]

reverse supply overvoltage and age
  replacing with a known god cap is probably the easiest way to check ... tha I know of

Well, I only used the effect for a few months, so I don't think it's age. Maybe I reversed the voltage, and damaged them at one point... Anyone else have any ideas?

[Nasse]

I bet there may be a bad solder joint somewhere. If you can listen the signal at various points or take voltage measurements you may find it. Or just trust your luck and preheat the ugliest again

[Mark Hammer]

Occassionally, lead wires can fracture on the inside even though they appear good.  There is also the possibility that the trimpot on your board is kaput.  If you have a meter than can read AC signals less than 500mv (and preferably below 200mv) try reading the appropriate test points in the envelope follower just to verify it is actually detecting your inpt signal.  For that matter, there are a bunch of test points where you should see something like a signal.  E.g., is there really a signal reaching the OTA?  Try testing the circuit with a probe at the emitter of Q1.  You should get something at least 30mv or so when you strum.

[eliktronik]

Well, I've gone back and re-soldered some joints, measured all the pots, and checked all the wires/connections. Now on to the harder stuff. Could you elaborate on how to check for signal at Q1? My voltmeter has a 200mv scale and I tried to check the input signal, and got nothing. I don't exactly have a very good grasp of the electronics involved w/ guitars. Thanks for helping out.

[Mark Hammer]

(Note that all components are referring to the Tonepad layout document)

Set your meter to the 200mv-AC range, set your guitar down in a safe position with it plugged into the pedal, touch one probe to ground, touch the other to the emitter of Q1, (which should be the leftmost pin looking at the flat side with the pins pointing downward) or the emitter side of the 10k resistor that is tied to it if that's easier to reach....and strum.  You should see the meter respond, albeit sluggishly.  In fat, you can do this at a number of places just to make sure you still have signal continuity.  For instance, touching the pad where the input wire reaches the board confirms that at the very least the input signal has safely made it from the input jack to the switch to the board.  Finding that you have detectable signal at the emitter of that first transistor confirms that a) the transistor is oriented properly and isn't burnt out and b) the input cap and resistor are fine.

What some folks like to do is use a audible signal tracer, rather than meter.  This is used following the same logic, but is essentially a mini-amp with probes instead of a plug/socket.  The nice thing is that unlike a meter, you don't have to look at it.  The not-so-nice thing is that the sound level may be too low or too loud at times (i.e., you can use your meter at all times of the night or when the baby sleeps, but maybe not your signal tracer).

Figure that you start off with something like a 30-50mv input signal when you leave the guitar up full and strum all 6 strings with a pick.  That signal should be about the same, maybe a wee bit hotter at the emitter of Q1, since Q1 serves as an input buffer.  If it is lower at the emitter than it is at the base (middle) pin where the signal hits the transistor, then something is amiss with the transistor.

If that signal is fine, then the next major audio stop-off is the output of IC1, which is pin 6.  There should be something there every time you strum.  It may not be audible at the output jack because of the pot setting but it is always there at the OTA output.  After that, check the emitter of Q2, which is where the output level pot gets its feed from.  If there is something at IC1-pin 6 but othing at Q2 emitter then something is wrong *after* the OTA.

Another major test-point will be the envelope follower.  Here, you can check a bunch of spot but a sensible place is the junction of the Q3 and Q4 collectors (should be rightmost pin looking at the flat side).  Q5 serves as a kind of mixer for the outputs of these two transistors so you should get something at the emitter of Q5 which is *at least* what you see at the Q3/Q4 junction (which is also tied to Q5 base).

This is not an exhaustive list, but gives you some idea of the logic to employ.  Essentially you look for major junctures in the circuit and signal path that let you know that a) things made it safely through to that point, and b) an appropriate change in signal level is present, indicating that everything is doing what it is supposed to be doing.

Hope this helps.  It's a nice compressor, so we're just as impatient for you to start enjoying it as you are yourself.

[analogmike]

HI,

One more easy thing to check is the tantalum caps if you used any. They can tend to die. Hopefully you have a cap setting on your meter, that will tell you if they are dead.

[eliktronik]

Mark, I can't thank you enough for posting a detailed writeup. I'm starting to get a feel for this.
Basically, I tried to test Q1 at the emitter, and got about 90 mv. Does that sound right? With the guitar silent, the meter would read 88mv or so (no signal), and when I strummed it would slowly go up to about 91mv.
Anyway, thanks to everyone for responding, I'll check everything out and see if I can find the culprit.

[Mark Hammer]

I forgot to mention a few things so don't thank me just yet.

1) If you are doing your testing near either a flourescent fixture or television/monitor or anything else that produces EMI/RFI your AC readings can get screwed up, so try to make sure the only AC sources your meter probes are picking up come from the guitar itself as much as possible.

2) There are assorted DC biases strewn throughout most circuits.  Series caps will, of course, block these DC voltages.  So, when looking for AC signals, if you have a suspicion that a DC bias might be at work, try tomeasure on the "far" side (i.e., after) of a DC blocking cap, rather than the "near" side.

In principle, 91mv sounds about right though you ought to get very little reading when no guitar signal is present.  In complete conflict to my own advice, I often end up testing things under a flourescent light in front of a monitor in my home office, and even there I rarely get readings above 15-20mv with no input signal.  Just moving away from the monitor can easily drop this to a couple of mv, so it sounds to me like something is amiss at your front end.

[Jay Doyle]

eliktronik,

Following up on what Mark said, it does sound like something is wrong with the input stage.

Above someone stated that reverse voltage could damage an electrolytic (or tantalum) cap, though I think that you may have misinterpreted that as a reversage (I don't think that is a word but...) of the battery polarity. In fact it means that if a polarized cap, represented by the + and - signs on either end in the schematic, is inserted in the circuit the wrong way, that reverse voltage now present (where the - side of the cap is now at a more positive point than the + side) will slowly break down the cap and can do all kinds of weird things. By my count there are 6 caps that this can happen to. Though from your explanation I would suspect the 1uF cap coming off of the emitter, though there are others that could cause the symptoms you have stated. The 1uF cap off of the split 470k resistors biasing the input, the 1uF cap on the collector of the input transistor and the 1uF cap on the 27k resistor in the bias string could all be culprits as well. Though I would check the first one off of the input transistors emitter first.

If a reversed cap isn't the problem, post the voltages on the pins of the input transistor, the pins of the 3080, and the pins of the output transistor. I highly doubt that it is a problem with the envelope follower as no matter what happened here the IC would still pass the signal. The only thing that COULD happen is if you installed a too low value resistor feeding pin 5 of the IC, though I think that this would just kill the chip dead quickly instead of strangling it.

Try the above recommendations and get back to us with pin voltages if nothing works.

Hope this helps,

Jay Doyle

[eliktronik]

ca3080 (I checked the dc voltage):
Pin 8: 0
Pin 7: 8.7v (v+)
Pin 6: 2.72v
Pin 5: 0
Pin 4: 0
Pin 3: 4.5v
Pin 2: 4.5v
Pin 1: 0

Note: I replaced the battery with a fresher one after doing this test...

Q1 (input) AC voltage with no input:
Collector:~175mv
Base:~175mv
Emitter:~150mv (leftmost when looking at the flat side, pins down)

Note this is in a room, where I get 70 mv between the contacts of the voltmeter when they're not touching. And that's with the monitor and desklamp off. The only other lamp around is halogen...

Q2 (output voltage):
All legs approx. 160mv, with the collector reading the lowest.

Does this mean anything to anyone? Trying to read the ac voltages is really a pain, and I'm not very confident in my findings....

I've also made a new discovery: If I hold the sustain pot, while hitting a string, the signal comes through very faintly/distorted....

Next I'm gonna test the pot/connections and go through with a signal tracer...

[eliktronik]

I'm really mostly concerned with the fact that I can hear the faint, distorted signal coming through if and only if I touch the shaft of the sustain pot with my hand... Does anyone know why that's happening? Is the shaft of the pot connected to any single terminal?