Harmonic Percolator - make one!

[Morocotopo]

Paul, don´t despair. Have you tried another guitar? As I said, I discovered that what was feeding back in my case were the guitar pickups themselves, not even the strings, and when I tried a higher Hfe Q2 the thing was in constant oscillation. And this circuit seems very prone to feedback, so it might be that the problem is not there but provoked by it...

[Mark Hammer]

I hate this stupid circuit. I've rebuilt it and I can still only get it to squeal at various frequencies depending on how the knobs are set. No matter what transistors I use it just oscillates at a higher or lower frequency depending on what PNP transistor I use. When I play my guitar thru it, it sounds good if you ignore the constant oscillation.

I don't get it. I guess I'm just stupid for even trying.

Silly question to ask, considering your extensive experience building pedals, but is it possible you've connected something that, in the drawing, is actually not connected?

[Paul Marossy]

Paul, don´t despair. Have you tried another guitar? As I said, I discovered that what was feeding back in my case were the guitar pickups themselves, not even the strings, and when I tried a higher Hfe Q2 the thing was in constant oscillation. And this circuit seems very prone to feedback, so it might be that the problem is not there but provoked by it...

It does it with no signal at all going to the circuit. If I can't use it with my guitar, there's no point in even using the thing. Don't know if it matters at all, but my guitar has only humbuckers on it.

Silly question to ask, considering your extensive experience building pedals, but is it possible you've connected something that, in the drawing, is actually not connected?

I don't think so. I've used two different schematics, both with the same end result. I must be doing something wrong, but I don't know what. 

[Morocotopo]

Paul, I know you have experience in building, but I´ll respectfully suggest some things that come to mind. Probably you have already tested them, but, sometimes another person´s look can be helpful.

- Input wire shielded/far from other wires?
- Defective trannys?
- In the Giblet schem, C3 seems to be connected to Q1´s base, but actually doesn´t (no connecting dot where lines cross)
- PNP/NPN right?
- Too high Hfe in one of the trannys?
- General PCB/breadboard shielding, like for example laying it on top of a grounded aluminum sheet?
- Power? batt, power supply? Feedback trough the power/ground to other circuits/amp/etc? The thing doesn´t have power filtering at all. Perhaps try a R/C there?

Well, that´s whart comes to mind now. Maybe it will help.

[Paul Marossy]

- Input wire shielded/far from other wires?
- Defective trannys?
- In the Giblet schem, C3 seems to be connected to Q1´s base, but actually doesn´t (no connecting dot where lines cross)
- PNP/NPN right?
- Too high Hfe in one of the trannys?
- General PCB/breadboard shielding, like for example laying it on top of a grounded aluminum sheet?
- Power? batt, power supply? Feedback trough the power/ground to other circuits/amp/etc? The thing doesn´t have power filtering at all. Perhaps try a R/C there?

Moving around input wire does nothing, it's not shielded. I've tried multiple transistors, they just vary the oscillation from crazy to insane.
Yes, PNP & NPN
I've kept the Hfe as low as possible in Q1 & Q2, still oscillates no matter what.
It's not currently shielded in any way. It sounds like it's trying to pick up some RFI, but I don't think that is the cause of my oscillation problem.
On my first build, I tried the capacitor/resistor on the power supply, still oscillated like crazy. I've only powered it with a battery so far.

If I pull Q2 out of the socket, the crazy oscillalation stops. Current Q2 has an Hfe of about 100. 2N5089 sounds the same, just a little more gainey. If I use a really low Hfe in Q1 position, no fuzz of any sort. IIRC, the oscillation will stop if either pot is almost at zero. Harmonics pot has the largest affect on the amount of oscillation. 

[midwayfair]

- Input wire shielded/far from other wires?
- Defective trannys?
- In the Giblet schem, C3 seems to be connected to Q1´s base, but actually doesn´t (no connecting dot where lines cross)
- PNP/NPN right?
- Too high Hfe in one of the trannys?
- General PCB/breadboard shielding, like for example laying it on top of a grounded aluminum sheet?
- Power? batt, power supply? Feedback trough the power/ground to other circuits/amp/etc? The thing doesn´t have power filtering at all. Perhaps try a R/C there?

Moving around input wire does nothing, it's not shielded. I've tried multiple transistors, they just vary the oscillation from crazy to insane.
Yes, PNP & NPN
I've kept the Hfe as low as possible in Q1 & Q2, still oscillates no matter what.
It's not currently shielded in any way. It sounds like it's trying to pick up some RFI, but I don't think that is the cause of my oscillation problem.
On my first build, I tried the capacitor/resistor on the power supply, still oscillated like crazy. I've only powered it with a battery so far.

If I pull Q2 out of the socket, the crazy oscillalation stops. Current Q2 has an Hfe of about 100. 2N5089 sounds the same, just a little more gainey. If I use a really low Hfe in Q1 position, no fuzz of any sort. IIRC, the oscillation will stop if either pot is almost at zero. Harmonics pot has the largest affect on the amount of oscillation. 

is your PNP oriented correctly? Try flipping it around. They should be connected at the emiters.

[Paul Marossy]

Yes, I checked the orientation of the PNP already. It'll do it if reversed too.

[Morocotopo]

Don´t know if it´s on breadboard, but, I -think- a bad part (r, c)? A pain to check, but...

[Paul Marossy]

Don´t know if it´s on breadboard, but, I -think- a bad part (r, c)? A pain to check, but...

I don't think it's a bad part. I suspect that I am going to have to monkey around with the value of R2 connected to Q1.

Can someone post the voltages they have on the pins of Q1 & Q2?

[Morocotopo]

Paul in my build mentioned above, I get:

Power: 9,12V

Q1 - 2SB175A - Hfe: 60 - Leakage: 147 uA
C: 1,36
B: 1,43
E: 1,54

Q2 - 2N3904 - Hfe: 167
C: 3,34
B: 1,40
E: 1,56

My R1 is 2M8. The rest is the same

[pinkjimiphoton]

http://www.diystompboxes.com/smfforum/index.php?topic=99596.msg874231#msg874231

paul, check my voltages earlier in the thread, the link give hfe and voltages.

too high gain...i mean, anything over 100, and it WILL feedback...even by itself. it's gainy, so it seems to oscillate.

which method did you use to construct? my lame vero in this thread is verified, and works great.

just for now.. try adding them two caps, and see if it nukes the oscillation.

if you wanna pm me your snail mail addie, i can send ya some lo gain pnp's if you need them.

you really gotta get this circuit going...it's simple, but tricky. the "consensus" schem is not right, i don't think...as built, it sounded bright and quite shrieky..about 2 octaves too much high end. to my ear, it didn't sound "right"...but them two caps brought it into the albini ballpark i think.

i used this thing last nite, first in line, and tell ya what.. holy SH....err..COW!  it makes a great "fuzzface helper" and i used it all nite. worked great into the fuzzface, the sho, the klon and even the diphonizer. i was impressed.

i had a phatter tone than i've had live in a long time, and the control from the guitar was pretty sick..roll the tone knobs down, and catch some subharmonics...randomly and unexpectedly.

with my strat and the ff after it, i could get inSANE sustains...pretty much endless..

and feedback is different....not what i was used to, but it seems to catch octaves, 3rds 5ths, 6ths etc. pretty cool

you gotta get this thing workin' paul... you'll love it!!!

hey, just for shitzngrinnz, try the ge in q1 backwards. i know it sounds crazy, but it just may work...if the ge is leaky enough for the cct to be happy, it probably doesn't matter.

[Paul Marossy]

Paul in my build mentioned above, I get:

Power: 9,12V

Q1 - 2SB175A - Hfe: 60 - Leakage: 147 uA
C: 1,36
B: 1,43
E: 1,54

Q2 - 2N3904 - Hfe: 167
C: 3,34
B: 1,40
E: 1,56

My R1 is 2M8. The rest is the same

Thanks, that will help. I'm guessing that probably the Hfe on Q1 is too high and/or I'll have to mess with that Q1 emitter resistor some.
Why did you use so high a value on R1?

[Paul Marossy]

http://www.diystompboxes.com/smfforum/index.php?topic=99596.msg874231#msg874231

paul, check my voltages earlier in the thread, the link give hfe and voltages.

too high gain...i mean, anything over 100, and it WILL feedback...even by itself. it's gainy, so it seems to oscillate.

which method did you use to construct? my lame vero in this thread is verified, and works great.

just for now.. try adding them two caps, and see if it nukes the oscillation.

Yes, that's probably part of my problem, Q1 Hfe is too high. I built it on perf and triple checked everything.

Which two caps are you referring to? I didn't see anything about that...

[Mark Hammer]

I built mine on perf, and all those little parts was-a-huggin' each other.  I used sockets for the transistors.  VERY helpful for nailing one's optimal tone...or at least the oscillation thats in the same key as the tune.      The machined socket pins are real nice, but you can't always rely on perfect contact with the device.  Sometimes you need to tin the transistor leads to thicken them up enough.

[pinkjimiphoton]

http://www.diystompboxes.com/smfforum/index.php?topic=99596.msg874231#msg874231

paul, check my voltages earlier in the thread, the link give hfe and voltages.

too high gain...i mean, anything over 100, and it WILL feedback...even by itself. it's gainy, so it seems to oscillate.

which method did you use to construct? my lame vero in this thread is verified, and works great.

just for now.. try adding them two caps, and see if it nukes the oscillation.

Yes, that's probably part of my problem, Q1 Hfe is too high. I built it on perf and triple checked everything.

Which two caps are you referring to? I didn't see anything about that...

damn, dude, you need to read the thread 

use a 1.5n chiklet between q1 c and ground, and up to a 2.2n cap from level pot wiper to ground.
wiped out most of the oscillation/noise/earwigs.

what about paralleling a couple ge's for q1? should be able to get a lower gain that way, maybe?

[Morocotopo]

Paul in my build mentioned above, I get:

Power: 9,12V

Q1 - 2SB175A - Hfe: 60 - Leakage: 147 uA
C: 1,36
B: 1,43
E: 1,54

Q2 - 2N3904 - Hfe: 167
C: 3,34
B: 1,40
E: 1,56

My R1 is 2M8. The rest is the same

Thanks, that will help. I'm guessing that probably the Hfe on Q1 is too high and/or I'll have to mess with that Q1 emitter resistor some.
Why did you use so high a value on R1?

Paul, in Madbean´s Pepper Spray schem some values are different that the ones in Giblet´s schem. That R is 3M9. When I first listened to it, 750K didn´t sound too good (too little distortion if I remember right) so I experimented with that and 2M8 seemed to give the best sound. Nothing more than that.

[pinkjimiphoton]

I built mine on perf, and all those little parts was-a-huggin' each other.  I used sockets for the transistors.  VERY helpful for nailing one's optimal tone...or at least the oscillation thats in the same key as the tune.      The machined socket pins are real nice, but you can't always rely on perfect contact with the device.  Sometimes you need to tin the transistor leads to thicken them up enough.

i buy bagfuls of the machined 6 pin dip sockets,  i cut 'em in half with a dremel, instant GOOD sockets...them SIP strips people sell suck, cuz once ya swap out a few parts, their connectivity goes down hill, and yes, you do have to tin 'em.

the machined ones are best, imho

[midwayfair]

http://www.diystompboxes.com/smfforum/index.php?topic=99596.msg874231#msg874231

paul, check my voltages earlier in the thread, the link give hfe and voltages.

too high gain...i mean, anything over 100, and it WILL feedback...even by itself. it's gainy, so it seems to oscillate.

which method did you use to construct? my lame vero in this thread is verified, and works great.

just for now.. try adding them two caps, and see if it nukes the oscillation.

Yes, that's probably part of my problem, Q1 Hfe is too high. I built it on perf and triple checked everything.

Which two caps are you referring to? I didn't see anything about that...

My Q1 is ... um, about 40hfe. Even 70 was too high for me. So go loooowww. Sockets!

[pinkjimiphoton]

i tried a bunch of different trannys, and anything over about 45 or so was simply too much.

i have to wonder if the guy from interfax designed this so he could use cheap garbage transistors that would be too low gain/leaky for other circuits.

i tried up to 400hfe, and fuggeddaboudit. even 100 was too damn high.

[Paul Marossy]

damn, dude, you need to read the thread 

use a 1.5n chiklet between q1 c and ground, and up to a 2.2n cap from level pot wiper to ground.
wiped out most of the oscillation/noise/earwigs.

what about paralleling a couple ge's for q1? should be able to get a lower gain that way, maybe?

I glanced through it. I can't sit here and read the whole thing while I am here at work. I'll see if these "snubber" caps help anything. Thanks...