Clucking Boss Touch Wah TW-1

[strungout]

I have been given a TW-1 to fix. It works well. The problem is it 'clucks' like a chicken when the strings are hit hard. Soundclip: https://soundcloud.com/user-165425177/bosstwah-1-clucking-duck/s-p5Kog51lEZM

Schematic: Assuming Q* = Q4 and the other to the right of it Q5...

Some voltages:

Supply 9.10V

LM324N
3.97
3.33
3.54
8.63
3.98
3.43
3.97

4.04
4.04
4.05
0
3.44
3.37
3.37

Q1 (2sc732)
E 2.82
C 3.26
B 8.63

Q2 (2sc732)
E 0.67
C 0.09
B 4.35

Q3 2sc732)
E 1.27
C 2.88
B 8.84

Q4 (2sa1015)
E 4.60
C 0
B 3.94

Q5 (2sa1015)
E 4.60
C 1.61
B 4.08

D6
A 3.59
K 4.04

D7
A 4.04
K 3.45
I'm thinking it's the resonance that goes awry? Where should I start looking?

I messed with the trimmer, but it sets up the sweep range.

[ElectricDruid]

Does it respond to the Peak or Sensitivity controls? Can you prevent the "cluck" by tweaking those?

[strungout]

Tom: Yes, the Sensitivity and Peak pots do work as expected.
I played some more with it. At about 12 o'clock max, on 12V, it won't 'cluck'. And, I thought it did on both, but it only clucks in the downward Drive position.
I'm using a 12V regulated supply. The Boss ACA adapter (suggested adapter) is 12V unregulated and it lowers the voltage down to 9V, as I've read. Feeding it 9V ( I think it lowers the voltage to 8V, with a 4V bias), it won't cluck as much, even with the peak knob turned all the way clockwise.

[Rob Strand]

If the cluck is very quick it's probably something after the cap C15.  Perhaps the up/down switch.

If the cluck seems to be following an envelope then perhaps something before C15.   As a test to narrow down the problem you could  solder a larger cap across C15 to see if you can slow the cluck down, although that test in itself might affect the behaviour.   It could be C15 itself or even a noisy sensitivity pot.

[Mark Hammer]

There is a trimmer on-board that adjusts the sensitivity of the envelope follower.  My guess is that it is set too high.  The sweep doesn't seem "gentle" or smooth enough.

[strungout]

Rob: I added a 10uF cap in parallel with C15. It slows down the decay, but seems to have no effect on the clucking. So, I guess that means the problem might be upstream.

The cluck happens when you hit the strings hard. I'll be playing and pick a little too hard and it'll cluck. It sorta follows the envelope. Seems to happen at the max sweep point. My first idea was that it has to do with the resonnance, but it also sounds like there is oscillation, just for a moment... I don't know, I'm just guessing. There's definitely distortion going on in the cluck, though.


Mark: Yeah, that's the first thing I tried, hoping that's all there was to it. The trim works, but no setting gets rid of the cluck.


I'm gonna try and change the Sens pot and the switch. See if that does anything.

[Mark Hammer]

When anything with a sidechain, be it autowah, compressor, limiter, or noise-gate - is described as "distorting", sometimes what we're really discussing is unwanted envelope ripple.

As I understand it, the "12V" Boss adaptor measures 12V with no load.  If you have one, just for the hell of it, use a 9V regulated supply.  Consider lifting one end of D1 when doing so.

[Rob Strand]

Rob: I added a 10uF cap in parallel with C15. It slows down the decay, but seems to have no effect on the clucking. So, I guess that means the problem might be upstream.

I presume the cluck goes upward in frequency since what would make it more evident in the Down mode.    For the cluck to go up the opto must be drive on in a spurious manner.   In UP mode the signal has already driven the sweep up so the fault might still be there just you can't hear it.

Check the solder joints from C15 upto the opto, including the trimpot, the opamp and the opamp power pins.   A noisy trim pot VR3 could cause problems as well.

If it were mine I'd first look at the waveforms from C14 upto the opto and try to see abnormal behaviour.   After that I'd put in a constant tone like a sine, disconnect the wiper of the sensitivity pot, then manually feed DC into C15 via resistor to control the envelope.   Another way to manually control the envelope is to  take a resistor and connect it to where C14 connects to R23 then connect the other end ground via a push button switch. .  That lets you fake the envelope by pushing the switch.  You will need to tune to value of the R23 to fake small level signals and large level signals.   You can also check the DC voltages with and without the button pressed.

Worst case would be the opamp has failed and is doing something weird.

[Rob Strand]

Q4 (2sa1015)
E 4.60
B 0
C 3.94

Base at zero might be something to look into.  It should vary with VR3.

I think your B and C labels are swapped.

[strungout]

EDIT 2: I started looking at the filter section and probing around. When I connect to pin 1 of IC1 (on layout. It's pin 8 (LM324) on mine) and hit the string I get clucking, repeating and slowing down as the signal decays. The other pins, I either get nothing. Pin 7 of the LM324 gives out a glitchy high pitched noise. When I remove my probe from that pin, I can hear it sweep a little. Maybe it time to check if replacing the quad fixes things.

EDIT: When the pushbutton switch is open, it clucks when I hit the strings. Thing is, it does it WITHOUT the sens wiper connected. So, there's no signal going to the Envelope detector. That would mean the problem is in the filter section, wouldn't it?


Mark: Ripple, right, le mot juste! Had tried it on regulated (my 12V is regulated too), but since you suggested lifting D1, I tried it again. Without D1 the cluck is worse than with it.


Rob: yes, they were. My bad. I fixed it. Added D6-7 voltages as well.

I don't have a scope. Might be something I'd want to get... But if I understand correctly, I can still learn some stuff with the push-button switch. I'll try and do that.


So, I replaced the sens pot and trimmer, and removed the switch and hardwired it into the down position: still clucking. While I was doing that I found that R49 (4k7 going from the sens pot to ground) was dead. Not the problem either, though.
I've checked the solder joint, reflowed some. Looks ok to me.

[Mark Hammer]

It hasn't been mentioned so far, but as autowahs go,the TW-1 is one of the few whose circuit attempts to be an actual envelope-controlled wah-wah circuit, with an inductor, rather than merely just another bandpass or state-variable filter.

As such, it is well worth repairing.

[strungout]

Mark: Absolutely. It does sounds killer!


And it is fixed! Rob had an inkling the IC might be the issue. Probing the pins and what I was getting supported that, I thought. I removed the IC, put in a socket, as I don't have any LM324 at the moment, and used a TL074. No clucking at all.

I'll test everything once I get the right chip.


In the meantime thank you guys for all your suggestions and ideas!

[PRR]

> I don't have any LM324 at the moment, and used a TL074.

I wasn't going to say (biting my tongue all night) but the '324 can be a truly awful audio amplifier. I don't see where you hear audio through the '324 in this rig, but I still dislike it. Above 50uA (about 100k in 9V systems) you can have a 0.6V crossover glitch.

[Rob Strand]

And it is fixed! Rob had an inkling the IC might be the issue. Probing the pins and what I was getting supported that, I thought. I removed the IC, put in a socket, as I don't have any LM324 at the moment, and used a TL074. No clucking at all.

Good news. 

I'm with PRR, since the IC is passing audio changing to a TL074 sounds like a good idea.   My only area of concern with the swap is if the signal level in side chain (IC2) exceeds the common mode range the TL07x's can do weird stuff.  Particularly the IC2b.

FWIW, I had an idea to test the opamp this morning.  Use the trick with the added resistor where C14 connects to R23.   Except now make is variable over a wide range, unfortunately you might need to go to 1M to 10M for a full test; check open as well.   This way you can adjust the resistor an watch the voltage on C14.  It should increase monotonically then hit some limit.   What you don't want to see is when you continue to decrease the resistor past the point where the output limits that it the output starts to drop (ie. reverses direction and moves away from the limit).  Do the same at the output of IC2b (the opamp going to the transistors) for both up and down modes.   Obviously the signal *direction* will change with the different switch settings but once the opamps hit the limit you don't want to see the output reversing direction as you continue to decrease the added resistor.   The idea is is you see a reverse direction it could mean a potential to cluck.

[strungout]

> I don't have any LM324 at the moment, and used a TL074.

I wasn't going to say (biting my tongue all night) but the '324 can be a truly awful audio amplifier. I don't see where you hear audio through the '324 in this rig, but I still dislike it. Above 50uA (about 100k in 9V systems) you can have a 0.6V crossover glitch.

But, aren't all mojo parts?


I take the point. I also have an LF347, if the wide bandwidth is useful. The guy I'm repairing this for wouldn't care about the IC, really. He'll just be stoked it works right. It's socketed so, if someone down the line wants mojo, they can easily swap it.

FWIW, I had an idea to test the opamp this morning.  Use the trick with the added resistor where C14 connects to R23.   Except now make is variable over a wide range, unfortunately you might need to go to 1M to 10M for a full test; check open as well.   This way you can adjust the resistor an watch the voltage on C14.  It should increase monotonically then hit some limit.   What you don't want to see is when you continue to decrease the resistor past the point where the output limits that it the output starts to drop (ie. reverses direction and moves away from the limit).  Do the same at the output of IC2b (the opamp going to the transistors) for both up and down modes.   Obviously the signal *direction* will change with the different switch settings but once the opamps hit the limit you don't want to see the output reversing direction as you continue to decrease the added resistor.   The idea is is you see a reverse direction it could mean a potential to cluck.

I'm not sure I want to mess with the pedal any more XD But, this is cool, I think I'll breadboard this circuit and try it.

[Rob Strand]

I'm not sure I want to mess with the pedal any more XD But, this is cool, I think I'll breadboard this circuit and try it.

If you are happy it is working there's no problem leaving it as is, no need to mess with it.

I was just suggesting a test if you wanted make 100% sure it was OK, or, you wanted test the old opamp.    The original idea for test was to check the opamp in-circuit using only a multimeter.   When I checked the posts you had already found the problem!

[PRR]

....But, aren't all mojo parts?...

Just because it sucks does not mean it has mojo. A '324 can be just plain annoying. Since it does not seem to be the "original" plan, I would drop the thought.

[Rob Strand]

Just because it sucks does not mean it has mojo. A '324 can be just plain annoying. Since it does not seem to be the "original" plan, I would drop the thought.

It's probably a real unit.  Here's a pic with a quad,  JRC 3403A,



Looks like a LM324,
https://www.datasheetq.com/datasheet-download/762278/1/JRC/NJM3403A

Hmm, fine print says improved version with no crossover distortion.

Unfortunately I don't have my Boss list with all the versions and the years/order they appeared.   There's a surprising number of variants and the schematics for versions other than the common ones aren't so easy to come by.

[strungout]

I'm gonna keep it as is. Plays good, sounds good!