Ran at 12v accidentally, now not working.

[Locrian99]

Hey guys,

Before I spend the next couple hours looking for a bad solder spot or a bridge.   Is there any reason this would get damaged by running at 12v.   Built it, plugged it in to my test box and it was working.  Though I had almost a thump in time with the lfo.   Turned it off went to swap out transistors in q3/4/5 for some lower gain ones, did that and all I get now is a clean signal through it.   Q1 and q2 look good collector voltage of 9.x and 4.3x on those.  3/4/5 nothing.  The 4.7k resistor in the next to the diode that runs down to feed q3 and q4 collectors has 9v before the resistor on the lead and 0 after.  I'm assuming that's grounding out and that is where my issue is I just wanted to double check that running at 12v wouldn't have blown a cap.   Several of the electro's are 16v, and all film are 50v.   Thanks.  Here's the vero link and picture the schematic is pnp and links to the other forum. 

http://dirtboxlayouts.blogspot.com/2021/02/park-auto-wah.html?m=1

[Locrian99]

Well back to square one.   It's just a constant ticking, the effect comes through and sounds good.  Speed knob increases the rate of the ticking etc.  On the layout page there was a similar complaint and he was using some higher hfe trannies.   Originally I have 5 pn2222a in and they all are 250-275ish hfe now I've got two 2n2222as in here for q1 and q2 right around 200hfe, and q3-5 are 2n4123 all around 100 hfe. Voltage at q1 collector of 9.x, 4.5 at q2, q3 fluctuates between 1.3ish to 1.8ish, q4 1ish to 1.5ish.   Then base of q5 moving around .5.  Any ideas would be awesome.   I also tried all 5 with 2n3904.   

[duck_arse]

awwww, you know the rules by now! THE circuit diagram. parts references.
Q1 C -
     B -
     E -
voltages, then Q2, Q3, and so on. photos of your thing.

[Locrian99]

I can post pics of it when I get home didn't take any of the board.   The layout I used and the schematic link are in the first post.   Thought I posted the relevant voltages for what the issue is, oops if I need more.   In doing some more research today I wonder if the issue with the lfo is related to it stopping working until I knifed out all the rails.   Wondering if one of the dc filter caps has a dodgy solder job and that is allowing the lfo to bleed through.   Unfortunately I haven't found much on transistor based lfo it appears to be mostly op amp stuff about lfo bleed through.   

[Rob Strand]

It's just a constant ticking, the effect comes through and sounds good.  Speed knob increases the rate of the ticking etc.  On the layout page there was a similar complaint and he was using some higher hfe trannies.   Originally I have 5 pn2222a in and they all are 250-275ish hfe now I've got two 2n2222as in here for q1 and q2 right around 200hfe, and q3-5 are 2n4123 all around 100 hfe. Voltage at q1 collector of 9.x, 4.5 at q2, q3 fluctuates between 1.3ish to 1.8ish, q4 1ish to 1.5ish.   Then base of q5 moving around .5.  Any ideas would be awesome.   I also tried all 5 with 2n3904

Are you saying the pedal is working 100% but you have an annoying tick artifact?

The problem is probably not going to be fixed by changing the parts.  Could be a layout issue but more likely to be a design issue.

If any parts will blow with 12V it will be Q3 and/or Q4 but it's not very likely the way power is fed to the circuit via R11.

Q1 and Q2 should bias with Q1.base=0.6V, Q1.collector = Q2 emitter voltage + 0.6V = 5.1V roughly.  Q2 emitter voltage will sit within 4.5V +/- 1V.  Not much change with different transistors except very low gains.

Q3 and Q4 the base oscillate from a negative voltage upto +0.6V.   The collectors will oscillate from 0V to 4V or so.

Q5 is where the tick is likely to come from.   Some tick is present by nature of the way Q5 works there is an inherent varying "DC" level which follows the LFO.  The trick is how to tame it.  Some delicate maneuvering by increase R17 might help.   Increase to the point where the tick is decreased by the effect isn't lost.   Increasing C11 can also help but you don't want to go too high unless the sound is OK at high speeds.  (You can also ride a fine line with C11 where there is tick at low speeds and not at high speeds.)

Going further would mean modding to circuit to add a biasing network to Q5's base.   There are some other similar circuits like tremolos out there that do this.   The idea is to stop Q5 from completely turning off, thus minimizing the size of voltage glitch on the collector of Q5.

Beyond that is layout issues (sorry I haven't checked yet).  The ground side of C8 should be near the emitters of Q3 and Q4 and no audio should pass *through* this ground path.  The ground side of C8 then wires back to the input supply ground.  Ground on Q5's emitter should be on the audio ground.  Keep the Speed pot wires away from the audio as well.

[Locrian99]

Ok that is helpful.  The effect is coming through there's just a tick in the same speed of the rate led that can be altered by the speed.  The depth and tone seems to have an effect on the tick as well.    I tend to think it's not a layout issue with placement of the components  because the place where I got it has several comments about it working fine, one mentioned the lfo tick issue out of several.  That said I do think the speed pot wiring may have some issues with where the speed pot wires are going so I'll check that out.   I want to say the collector if q5 was at 0v, I will need to double check this when I have a chance.   Also I think the voltages you mentioned.   I am wondering would a bad joint at those caps or resistors still allow audio through and the effect but not be filtering out the tick or at least decreasing it.  Thanks again for the info

[Rob Strand]

I want to say the collector if q5 was at 0v, I will need to double check this when I have a chance.   Also I think the voltages you mentioned.   I am wondering would a bad joint at those caps or resistors still allow audio through and the effect but not be filtering out the tick or at least decreasing it.  Thanks again for the info

It's perhaps not a simple as it seems.

The DC voltage that appears on the collector of Q5 is a very small voltage.  It's in the order of say 3mV to 5mV (0.003 to 0.005V).   It's hard to measure reliably without making an effort.   The problem is that voltage varies as the LFO waveform cycles up and down.   To make matters worse the collector of Q5 is connected to the input of the audio amp.   The whole way the circuit works is it is injecting junk into the audio!    It's a inherent flaw which common to all tremolos and autowahs which use a transistor as control element.    I should point out this isn't a fault of the transistor, it's how all transistors work.   So the whole idea of the circuit has inherent flaw.

If the transistor is allowed to turn fully off then the LFO could swing the Q5 collector waveform from 0V to 4mV (say) which in relative terms is a large amount of junk to be injecting into the audio (keep in mind guitar signals are small).   If the transistor isn't allowed to turn off or doesn't turn on as hard then the swing of that voltage can be reduced to a minimum.   This is where tweaking R17 and adding biasing mods can help.

Another thing that makes the tick audible or not is how quickly the transistor is turned on and off.   If the transistor switching waveform is slow the junk can be confined to frequencies less than 20Hz (say).  So even though it is present you don't hear it.   This is where playing with C11 can help.    A very good mod is to use two caps, say a C11 and smaller one across the base but you have to get just the right value.  (C11 has an additional effect in that it can stop the transistor turning off as the base waveform is prevented from getting to zero)

There's other tricks like filtering the audio path however this works best if the above causes are managed first.

The whole idea is to try to tweak things so the tick is kept below an acceptable threshold.

Having said all that, it's still possible you have a completely different issue.     That's the problem with dealing with inherently flawed circuits it makes debugging more difficult because you have a bigger list of suspects to check.

[Locrian99]

If I remember correctly when I recorded this I was just hammering on not actually picking so the guitar volume is probably closer to unity with the noise.   But here's the sound.

https://drive.google.com/file/d/1VkRusPd8vzE62nt3PBIzds-aVzqsuYW1

[Rob Strand]

If I remember correctly when I recorded this I was just hammering on not actually picking so the guitar volume is probably closer to unity with the noise.   But here's the sound.

https://drive.google.com/file/d/1VkRusPd8vzE62nt3PBIzds-aVzqsuYW1

Unfortunately, I can't access it because you have to log into Google.

It's probably easier to just try playing with C11 and R17.

A very simple test to see if it's coming from somewhere other than Q5 is to temporarily short out C11.  That will kill effect but it should still pass signal.   If everything else is OK (other than the Q5, C11, R17 part of the circuit) then  shorting C11 should stop the ticks.

[Locrian99]

Thank you so much you pointed me to exactly where the problem was.  I resoldered r17 and c11 and it's perfect.

[Rob Strand]

Thank you so much you pointed me to exactly where the problem was.  I resoldered r17 and c11 and it's perfect.

Cool.  That was lucky, it would have been tricky to find otherwise!