stupid pedal trick: fuzzy balls overdrive

[Chuck D. Bones]

I've been looking at the schematic PJP posted on 3/22 and I think I see a typo.  Shouldn't there be a resistor between the collector of Q2 and +9V?  Also, the diode network after Q3 has some superfluous parts.  These diodes will never conduct: D4, D6, D8, D9.

Cheers!

- Chuck

[pinkjimiphoton]

hi chuck,=
yeah, it looks weird, but nope, directly to b+ bro! i just drew what i'd built. i WILL check to be sure i didn't drop the ball, which i do frequently

the diodes are whats in there, and tho it looks weird it really seems to matter to it tonally. or maybe its fuzz narcosis. i played with it quite a bit on the breadboard to see what it would do, to me, it sounded great like that. its almost an adaptation of a passive fuzz i saw somewhere, i think the guy called it a black scorpion or something.

why won't they conduct?  d8, 9 make half of a hard clipper with d5, d5 and d4 are a part of the bridge.  you are likely right, lol, but i think it made a diff in the sound of it,

think of it as a diode hard clipper made of two diodes in series on each side to ground from the signal. at the mid points of the two clippers are two other hard clippers to ground. the combo of different diodes i think comes into play, as the si conducts before the si does, so its fairly asymetric.

i'll trust ya being right, but would love an explanation of why they shouldn't work. i'd have to look at my original drawing for it to see wtf i did

i'd think if anything d6 and 7 wouldn't work along with d8 and 9. but i am just a monkey with a breadboard, sooo

[Chuck D. Bones]

PJP,
Here's what I see when I look at the diode cluster.  D4 & D7 are in parallel.  D4 is Si, D7 is Ge.  D7 has a much lower voltage drop, so it will hog all the current to itself.  Try taking D4 out of the circuit and it should sound the same.  Same logic holds with D5 & D8.  D5 has a lower voltage drop, so it will hog all the current.  Now let's talk about D6.  For D6 to conduct, there has to be a source of current.  It can't be D3 because D3 & D6 point opposite directions.  It can't be D4 or D7 because D4, D6 & D7 are in parallel.  The same logic applies to D9: there is no source for the current to flow in D9.  The only sources for AC current to drive the diodes into conduction are Q3 & R10.  At very high frequencies (way above 20KHz), the diode's capacitance comes into play and then they can all conduct.  But we won't hear any of that unless we are bats.

As for C4 connecting to +9V, there can't be any signal there because C2 & C11 short all AC to ground.  According to the Vero layout, C4 connects to Q2's emitter.  BTW, Q1 & Q2 are swapped between the schematic and the Vero.

In the immortal words of my boss & mentor Bill Hagen: "The circuit works the way it's wired, whether you think it's wired that way or not."

Don't get me wrong, I respect what you are doing and your circuits are amazing.  But when I see something that I don't understand, I speak up.

Cheers,

- Chuck

[pinkjimiphoton]

well, if anything, the vero ... the first one... is a copy of what i found in the pedal.
its entirely possible i screwed the pooch on this.

its a bitch going back 5 or 6 years after the fact to see what the hell i did. i lost most of my notes with that hard drive crash, so...

i'll look at it.

but the diodes i think are a little weirder than ya expect. there's been a few examples of weird shit i've run into.

looking at the diodes...  i beg to differ. i discovered a few years back that in a big muff circuit, a ge diode in parallel with a silicon one works as a noise gate. the ge should conduct before the si, and therefore the si shouldn't matter, but it DOES matter still. try it sometime.

the ge has to be in parallel for it to work, its in parallel with d4 as marked on the eh boards.
i've beeen trying to get an answer to why it works as it does, cuz it seems it shouldn't.

my current theory is the ge conducts before the si and is therefore putting the clipped signal out of phase with its antiparallel twin, and thus cancelling the noise. but beats me. monkey with a breadboard.


as for the vero, no, the vero matches the schematic. q1 has the 10k c resistor, 2.2m b to ground, and 10k trim from e to ground. so the q's aren't reversed. q2 has c directly to b+, and r7 os the e to ground resistor.

now the diodes may be a possibility for sure, but the rest works as drawn. i ended up adding the third stage simply cuz it had no volume. hard to guess all these years out, but when we started messing with it, we put together a whole bunch of different possibilities, this was the first one i had come up with. so my memory is sketchy as hell anyways... but i looked over the board pretty good and compared the schematic and the vero to it... the first vero is exactly what's in the video. the second vero is combining all the p2p shit onto it all as one board.

you don't always need a collector resistor for a q to work.

that said, who knows what the hell q2 actually is, when i had removed it to test, it showed up as a diode, not a transistor. so i gotta assume its leaky as hell, as it was one of the earliest ge's commercially available.
sometimes them old ge's will not show up as a transistor on a tester but instead as two diodes. thats what this reads as... so maybe it IS two diodes, can't find any data sheets or references to the unit on the web. i bought it from either surplus sales of nebraska or the transistor museum, i forger which, but it was a lonnnnnnng time ago.

that particular unit sounded best in this cct, so thats what i went with.

the diodes i can play with some and see what happens i guess. pretty lazy tho. it works, sounds good, so... i am not inclined to @#$% with it too much

one thing i've learned as a fuzzmonger is that which EE's say can't work often does. EE's are taught to minimize distortion and maximize potential.

i'm a monkey with a breadboard intentionally trying to @#$% sound up. so these rules don't always apply in fuzzy logic land.

like my bias clipper from a lifetime ago. ya wouldn't think it worked, but it did.

or the one where i used a ge transistor in the gain loop of an opamp with a switch to power the ge's collector from b+ so it would go from a clipping diode to a pseudo boost while clipping.

i know that probably doesn't make sense til ya actually try it, but like i said... i've learned with fuzz, all bets are off most of the time.

btw, the final stage is for all intents the "booster" add on output we found undocc'd in the ludwig phase II in some units. that shouldn't have worked, either

maybe the diode drop of d1 is what allows q2 to work... or maybe, q2 is actually two diodes. beats me chuck!!

please don't take me wrong, i am stoked to have ya explain this shit. i don't understand 99% of what i do anyways

[Chuck D. Bones]

Right on.  I'm not your typical EE because I came up building & modifying guitar amps and pedals as a teenager.  So I'm no stranger to distortion and uneven frequency response.  Along the way I've worked on circuits where transistors were deliberately installed backwards to exploit a little-known feature.  I've built digital circuits using analog parts and analog circuits using digital parts.

These days, I tend to analyze the $h1t out of a circuit before I ever build it, and I'm usually surprised when I get around to listening to it.  How a circuit behaves in simulation, how it looks on a scope and how it sounds can be three very different things.  They are all informative, but it's only the last one that matters in the end.  I was recently messing with a variation on the Expandora circuit and that was a real learning experience.  But I digress...

I'll think about the diodes some more and get back you.

Cheers,

- Chuck

[pinkjimiphoton]

and i'm one of them weirdos that does exactly all the stuff ya mention too

backwards transistors and all

[PRR]

> you don't always need a collector resistor for a q to work.

Your plan shows signal coming from the battery.

Either the battery sounds better than you do, or it isn't really wired that way.

[pinkjimiphoton]

works for me

i'll look again. tomorrow. beat up tonite.

[pinkjimiphoton]

just looked at this




and am betting that somehow i missed the 39k resistor that seems to be on most of the layouts. maybe with that there would make more sense,  its a bit of a nitemare cuz everything is really squished onto the tagboard and wires all over the place cuz i had a tendency towards too much wire still back when i built this.       

i'll look tomorrow. peace!