Every Harmonic Percolator I make dies ;(((

[DelSpanisho]

Well, not every one. The only one I've successfully created was with the Madbean "Pepper Spray" board, but I sold it soon after making it, since I didn't like how it sounded without any sort of tone control. I use 4 distortion pedals in sequence, and I NEED tone controls for each one of them, or else my signal just gets over-saturated with fuzz and it becomes non-listenable.

I still liked the special flavor of distortion it offered, though, so recently I've dedicated my time to modifying the circuit to fit my needs. I wanted to add:
-A tone knob
-A voltage starve knob
-An additional knob to further control the gain (a control between "gain stages" between each transistor, maybe? I'm not sure if that's the right term, maybe something like the "Pinch" knob on the Woolly Mammoth)

Here's the schematic I cooked up:


I removed the 100pf resistor most schematics of the H.P. use, I added a switch between asymmetrical LED clipping and mosfet clipping (my perf layout doesn't have the fet's, though, and the LED's aren't in the picture), and copied the tone circuit from some other pedal (I really can't remember which one to be honest, I used one of the first I found on Google). 








When plugged in, this perf board get's no signal at all, even if it's not plugged with a 9v cord (ie. bypassed). I've prodded it with an audio probe, and some low, muffled signal stays around the base and collector areas of the 2N404A, but that's it, there's no other evidence of the signal going through the circuit. I tested that all the ground points are connected too, it doesn't seem that anything is in the wrong place. 

Frustrated yet not easily defeated, I decided to quick whip up the Sabrotone verison of the Harmonic Percolator. No luck there either. 









I replaced the 100nf cap with a radial cap I had leftover from my Pepper Spray percolator, the diodes with an asymmetrical pair of LED's, added a starve knob, and the 100pf cap is just a jumper wire. I get a bypassed signal, but no signal when it's turned on. The audio probe picked up absolutely NOTHING at any point in the circuit (not counting loud BUUUUUZZZ sounds at some points, there was no sound of my guitar's signal), and I'm at a loss as to what I did wrong.

I want to learn from my mistakes, really I do, but it seems like whenever a pedal I build doesn't work the first time around, I can never figure out what it is I've done wrong. I've scrapped countless projects just because I got so angry and flustered at my inability to isolate problems. If any of you can point out what I'm doing incorrectly, I'd really appreciate it.

[anotherjim]

I think you need to take a deep breath and get your designs working on a solderless breadboard first. Going straight to PCB isn't a good idea unless you have a circuit that you're completely familiar with or have good reason to be confident in.

Other than that, you need to break it down to smaller chunks. Get the HP on it's own working before you add your extras (which are good ideas). Voltages? Check Q1 and Q2 have forward bias between Base and Collector. Pot 3 as you have it is capable of shutting Q1 off as I read it.

The tone control is a Big Muff type (though it must have existed before EHX used it). As it is, it's versatile, but has a mid range cut which might take the bubble and squeak out the of the HP - there is a modded version of this tone circuit which adds a "body" control that might be better to tune into the HP's "Thing"

[Buzz]

Is that solid core wire you are using? Braided wire takes a lot more punishment.

[DelSpanisho]

I think you need to take a deep breath and get your designs working on a solderless breadboard first.

Ugh, I would if I could. As convenient as it is, I find breadboards really difficult to use; I struggle with "translating" a schematic into a breadboard layout, and perf allows me to lay out every part almost exactly like it's placed in the schematic. 

Pot 3 as you have it is capable of shutting Q1 off as I read it

Yeah, I really didn't know what I was doing when I added it. My new schematic incorporates it in a different way, I'll upload it soon.

The tone control is a Big Muff type (though it must have existed before EHX used it).

Huh, really? I can see the resemblance now. I'll give that "body" addition a try.   

Is that solid core wire you are using? Braided wire takes a lot more punishment.

I find braided wire to be much more brittle and hard to work with. You have to twist the end of each wire, and even then there's no promises it won't all undo itself and leave you with shreds of spaghetti all over the place.

[wildebelor]

I find braided wire to be much more brittle and hard to work with. You have to twist the end of each wire, and even then there's no promises it won't all undo itself and leave you with shreds of spaghetti all over the place.

You have to tin braided wire!

[anotherjim]

"Ugh, I would if I could. As convenient as it is, I find breadboards really difficult to use; I struggle with "translating" a schematic into a breadboard layout, and perf allows me to lay out every part almost exactly like it's placed in the schematic. "

I know, it challenges my mental mapping too
But, from scheme to breadboard to prove it - then from corrected scheme to perfboard isn't too painful with practice. Even better if you have enough parts to leave the breadboard version in working condition so you have a reference to check voltages against for the built version.

[deadastronaut]

+1 . breadboarding isn't that difficult, and saves you building something you either don't like, or need to rip apart and tweak/mod to your taste..

it will save you time and parts in the long run..


i tend to have the schematic open in a graphics prog, and mark off bit by bit as i go along...as it can get a bit confusing when you have 3-4 paths going off in different directions..

it also helps if you get distracted by a cat, a wife, a knock at the door etc....after a while you'll get to the stage where 'common' configs are noticable and becomes a lot easier on the brain..

good luck.

[DelSpanisho]

Here's my updated schematic:



I'll give it a try on breadboard when I have some free time.

[DelSpanisho]

Okay, SO, after several tries I've gotten the schematic on breadboard to get a signal, but there's no distortion. The LED's don't light up, and all I get is a low frequency farty tone when I turn pot #2 up. Any suggestions? I can upload pictures if needed. 

UPDATE: Upon further examination, this oscillation noise is prominent even if the circuit isn't getting any power; I can remove the 9V plug from the breadboard entirely and I still get that noise, although it gets higher and higher in pitch as time goes on. Flipping around Q2 changes the sound of the oscillation, and causes the clipping LED's to light up a little, but still no actual "distorted guitar" signal comes through.

[anotherjim]

Ok, I suggest breaking it down into a smaller chunk.

Q1 is a fairly basic PNP amp, so try that part first.
Going by your latest schematic, try removing Q2 and R4.
Place a jumper wire to the Q2 emitter and join to Q2 collector (simulating a Q2 turned full ON).
You should now be able to get useful signal from the Collector of Q1.

[DelSpanisho]

Q1 is a fairly basic PNP amp, so try that part first.
Going by your latest schematic, try removing Q2 and R4.
Place a jumper wire to the Q2 emitter and join to Q2 collector (simulating a Q2 turned full ON).
You should now be able to get useful signal from the Collector of Q1

Tried these out, got no change in the sound. Hopefully I followed your instructions correctly! 

I also drew the breadboard layout on paper then used that drawing to redo the breadboard. Now pot 2 does nothing. LED's don't do anything either. 

I'll post my drawing once I'm off work in 6 hours.

[DelSpanisho]

Disregard the wrong wiring of the output jack at the top, the ground lug is connected to ground on the breadboard.

I'm seriously starting to think my schematic is just not functional. I based it off the schem in the madbean pepper spray pdf, but maybe I should re-do it with a different Percolator schematic.

I already have a PCB created for this schematic, I'm seriously considering sending it over to OSHPark and having them make a few boards, and then attempt to make this pedal from that board. If it doesn't work, I'm only out $15.

[idiot savant]

flip that 2N3904. 

[pinkjimiphoton]

check my photonic juergulator project. gain controls "between" the stages won't work. it's not really "stages"... the two transistors work if i remember correctly as a unit in there.

swtc may work, but percs don't have a huge output to them; very freakin weird circuit. i found in my experiments, the lower gain q1, the LOUDER it tended to be. go figure.

i would also advise starting on a breadboard. can save ya a lot of heartache, and this circuit is so dang easy it shouldn't give you this kind of grief.
you can make a percolator outta damn near anything and get it to fuzz.

i would use stranded instead of solid wire... solid wire often breaks internally and makes ya crazy.

i would also check your voltages on things... suspect resistors and caps if nothing else makes sense, sometimes the damn leads break internally and even tho they look connected, they aren't. particularly if ya may have heated 'em up a little too much inadvertantly.

you'll get it. the perc is a cool circuit, and very forgiving once ya get used to it making little sense.

[anotherjim]

flip that 2N3904. 

+1


Looking at your breadboard sketch, have you connected the power strips on the left hand side to the those on the right? Breadboards I have don't do this for you.

[duck_arse]

going right back to your perfboard - marking all the pads you are going to solder with felt-tip pen or similar is a very bad idea. all that ink junk will get boiled into the solder as impurities. and don't use pencil on pcbs, as you'll be drawing resistors all over.

have you posted us any voltages on your transistors yet? I'm not sure the perc at its beat will have enough oompff-f to drive series led clippers. you might need something like an lpb or single muff stage as a driver before the tone control, as well.

it's sad when a pedal dies ....

[Gus]

Post the transistor voltages.

Then get rid of the diodes the starve control and tone control.
 
How does it sound "stock" without the diodes?

The circuit does have two stages the 47uf controls the power supply interaction between the stages.

The circuit is a little tricky to get to work like in the video people sometimes use as a reference.

The C to B resistors are part of the gain setting and biasing of each stage.

[pinkjimiphoton]

i thought the main purpose of the 47u was to be a "neutral" ground point between the two transistors as a tsikli pair or whatever ya call it bro.
it goes to the common point of both transistors, and floats and isolates them above ground, doesn't it? and establishes leakage current to a point to introduce some feedback in the common emitter or collector or whatever , i don't have it in front of me and don't remember the exact circuit..
doesn't seem to be a lotta gain there the way it works, to my eye it's 2 transistors workin as one stage, but i am often wrong
i mean, a push/pull amp stage may have two whatever tubes/transistors as amps, but it's still one stage.
the perc does funny things when you play with it.
a starve control i don't think is gonna help much. i messed with it, made it blatty and not work.
messing with the resistors and turning 'em into pots was interesting, but i found i had to float the pots with resistors to get
reasonably useful ranges.
for a tone pot, i'd just use the guitar. but it really depends on how ya wanna tune 'em in most of all.. i never saw the need, or had the output for a tone pot other than on escobedo's variant.

[DelSpanisho]

First off, thanks so much for all of your comments! I'm surprised there seems to be this much interest in my shenanigans.  

I took Gus' advice and stripped off all the fancy add-ons, just leaving the bare, "original" circuit...IT WORKS!

On my Ampeg PF-350, there was very shrill, high-pitched noise if I crank both knobs all the way and place my guitar too close to the speakers. I didn't get that effect with my Univox 1086 tube amp head, which is odd.  I thought this might be due to the 100pf cap that I removed from the original circuit. I assume it's to cut out radio frequencies, and I wanted this pedal to be as haywire as possible, so I committed it from my schematic. I tried putting one back into the breadboard, no change, still an annoying siren pitch.  

After I played around with it for a bit, I re-added the tone, body, and starve controls. I couldn't hear any difference with the starve knob, so I removed it. I definitely like having the tone controls; my build is extremely treble-y and harsh, and raising the bass and mids a bit helps even things out.

I tried 1N34A's for clipping, and they lowered the volume and "squished" the sound a bit, and I removed them. I tried 5MM red LED's, but I couldn't hear a difference between LED's and no LED's. They never lit up either. There's 1N914's in there now, I believe. Can't hear much of a difference with them in, but whatever.

flip that 2N3904.  Wink

  Figures I got the pinout incorrect. Reminds me when I was building the Pepper Spray; I thought that the grey strip on polarized caps denoted the POSITIVE lead, not the NEGATIVE lead. Imagine my surprise when one of the caps exploded after the pedal was plugged into the power supply.  

Looking at your breadboard sketch, have you connected the power strips on the left hand side to the those on the right? Breadboards I have don't do this for you.

I hadn't when I started, but they're connected now.

going right back to your perfboard - marking all the pads you are going to solder with felt-tip pen or similar is a very bad idea. all that ink junk will get boiled into the solder as impurities. and don't use pencil on pcbs, as you'll be drawing resistors all over.

Hmm, makes sense. I'll make sure to leave the sharpie out in the future. I've never used pencil on the copper/metal part of any build (at least I don't remember doing it), so I'll keep that habit.

How does it sound "stock" without the diodes?

Pretty good, as I said earlier, very treble-y, harsh, and LOUD. I had it plugged in as the last pedal in my chain, and when I turned on my other distortion pedals the resulting sound was kinda bitcrush-y, but not all that great.  

I tried replacing R1 (51k) with a 25kB pot, and it indeed could raise and lower the overall gain of the circuit. Turned all the way up, it was harsh and distorted like usual, but all the way down I got a clean tone. Kinda neat, I might try putting a higher value pot in there soon.  

Now for the voltages. I set my meter to "9V" DC, hopefully that's the right setting:

2n404a
----------
C-3.60
B- 3.65
E-3.80

2n3904
----------
C-5.3
B-1.58
E-3.75

Anyone have suggestions for diodes I could use? I'm looking for something that won't lower the volume, and'll really distort the signal. I can make it an asymmetrical pair of diodes if needed.  

uou might need something like an lpb or single muff stage as a driver before the tone control, as well.

Hmm, never made a booster before, I'd love to give that a try. Would it go before the clipping diodes, or after? I'm assuming before. Also, what would be the most distorted booster circuit I could use? FET's? Something like the AMZ? Or maybe with an opamp, like the LM308?   

EDIT: I'm planning on making a few of these, so I'm looking for more germanium PNP's. 2n404a's are scarce on the 'bay, but there seem to be plenty of eastern european ones available. Stuff like this, this, and this. Would these work? IIRC, Q1 can be pretty leaky and still work great.

[duck_arse]

the final stage of a muff is there to make-up all the losses inherent in the muff-style tone control, and to provide a consistent load to the tone control. so it would be best there for that tone control. boost before the diodes will provide more signal for them to clip. you might be better with the germ diodes and a pot instead of the usual single resistor, to listen what happens there. the perc is not a wild kinda distortion I don't think.

and, your transistor voltages don't look right. the base of the 2n3904 should be higher than the emitter. do you still have resistance between the emitters? here is a thread, with loads of stuff in it, might be helpful on a few counts .....

http://www.diystompboxes.com/smfforum/index.php?topic=55987.0