Harmonic Percolator - make one!

[pinkjimiphoton]

on mine also, it's set to short the fuzz to ground in bypass.
but the voltages i posted are what i had...
first set was what it had when i fired it up, with q2 backwards and the knobs set randomly (as in: not set at all)
second set was with trannys "right" and knobs pegged.
third set was with q2 inverted and knobs pegged.

pretty large variance in voltage, huh?

weird circuit!!! really really weird!!

[Jazznoise]

The pics aren't great, but they're as reasonably as I can get under artificial light. Behind the 2N3904 is indeed a cap. The two resistors behind that are a 100K and a 680k wired in series for a measured 770K.  Red is input, black is ground, pots are off screen. Vero is sillily oversized because the boxes I have are slatted and thus it's easiest to simply make the board too big and sliiiide it in.

I don't see anything wrong. My solder isn't the most appealing but there's no cold joints or shorts. It's also because I've resolder to test resistors for being deffective (my feedback resistor for the 128 reads 6K in circuit and the standard 250ish out. No short across the base and collector. This pedal is starting to give me a cold joint in the soul!

Any observations or advice recommended. I'd even just take a hug right now! 

[garcho]

well, those broken pic links ain't helpin' 

This pedal is starting to give me a cold joint in the soul!

smoke a cold joint and don't beat yourself up

[Jazznoise]

You win again, dropbox

https://www.dropbox.com/s/kj7s2w2pphedclh/Picture%20018.jpg

https://www.dropbox.com/s/f2cth218r492i3g/Picture%20017.jpg

A hot shower beats a cold joint any day!  

[duck_arse]

I've taken my circuit off the breadboard for an octavia, but that's a different story. my perko voltages were so far off that I was too ashamed to write them on the circuit.

every single time I put this on the breadboard, I'd get the transistors around wrong because of the weird circuit. like pink, I had it going backwards, and looking interesting on the cro (if difficult to drive), but then reboarded it, and did it proper.

I'd suggest you breadboard it with different parts if you can, see what you get then.

and the "20017" page only gives error

[pinkjimiphoton]

this circuit is just plain weird.
i reccomend just getting it to work, and do whatever ya gotta do to make it sound good!

[Jazznoise]

Fixed that link.

It is definitely beyond weird. Looking at the caps I've used I'm wondering if the two little white fellows aren't giving me undue greif. If no one spots an unprecedented level of stupidity by yours truly I think I may have to shelve this until the end of the academic year.

[pinkjimiphoton]

try replacing the cap, maybe.
is q2 the right way? i've made the mistake more times than i can count.
and on mine, having that transistor backwards yields a completely different animal

[Jazznoise]

I've re-orientated them both. All 4 permutations are equally noiseless and miserable.

I agree, Jimi, I'm thinking swapping them might be a good port of call. The DC values don't seem unreasonable, but it's the inability to pass audio that makes me think the AC impedance must be totaly messed up.

[pinkjimiphoton]

i've had a few builds laid low by bad caps or resistors, i mean brand new ones.
i've noticed particularly on "standing up" components, sometimes they look fine but are "decapitated" internally.
so far almost every circuit i've had trouble with, it came down to one side of a resistor being broken off (but still looking connected) or bad electros that just plain don't @#$%ing work.
don't give up!!
fwiw, if ya have to, build the jerkulator tim escobedo drew up, it sounds pretty much the same as the perc does.
and it's WAY easy.

[garcho]

Do you etch PCBs? I can give you a layout if you'd like. There's got to be a number of them already online. Ferric chloride isn't expensive, and a little PNP blue paper can go a long way. Worth looking into anyway.
Have you checked out any of the layouts made by DIYSB forumites? A lot of people around here have built it with much success, and I'm sure many of them have done it with vero/perf. Have you compared your layout to anyone else's? I misread the schematic the first time I tried to breadboard the Percolator, and realized it when I looked at someone's PCB.

[duck_arse]

yr board's way too big. you should always use the smallest, meanest bit of vero you have, just to make everything easier.

I never use sockets. have you thought of them being the problem?

[garcho]

I never use sockets. have you thought of them being the problem?

what kind of problems have you had with sockets? I've never had any myself.

Germanium transistors don't like too much heat, be real careful if you're not using a socket.

[pinkjimiphoton]

sip sockets like most people use imho are a pain in the ....

i get 6 pin ic sockets, and dremel them in half.
the prob with sip sockets is once you use them a couple times, they don't grip very well.

a tip from dave w. is to use them to find the trannys ya want, then CAREFULLY squeeze them just enough to crack the plastic, then solder the transistor to the sockets.

the ic sockets tho  seem to be a bit more robust usually.

[duck_arse]

what kind of problems have you had with sockets? I've never had any myself.

that's the trick. I've never used them, so I've never had a problem with them. sort it out on the breadboard (or stick in a switch).

[Jazznoise]

I hear you on the big board, but it's mostly unused space to make the mounting of the board easier.

No detectable shorts and the socks are fine - everything reads as it should under the gaze of my watchful DMM. I'm the same as Jimi, I take DIP sockets and I cut them up with my snips. Just never saw the neccesity of buying SIP sockets.

[duck_arse]

any sign of output yet?

[duck_arse]

so I wired up my perko in its jig, works first go. the value of diagrams, eh? anyway, now it's built, I'd better add my "harmonic percolator voltages" to the pot ....

using (q1) 2sb156 hFE=43 leak=100uA, Rc1=22k, Rb1=220k, (q2) 2n3565 hFE=290 (maybe), Rc2=22k + 50k trim, Rb2=150k/680k, V+=9V11, I get:

with Rb2=150k -- Vc1= 1V88 ~ 3V79 -- Vb1= 1V85 ~ 3V80
     Ve= 2V08 ~ 4V48
     Vc2= 2V79 ~ 5V31 -- Vb2= 2V61 ~ 5V01

and with Rb2=680k -- Vc1= 1V78 ~ 3V48 -- Vb1= 1V70 ~ 3V13
    Ve= 1V95 ~ 4V09
    Vc2= 3V13 ~ 5V62 -- Vb2= 2V45 ~ 4V51

(the 150k Rb2 is mounted on a header, for user selectable long or short sustain. user can also decide their own bias preference.)

with my resistor values different to everyone else's, my voltages are not surprisingly different to everyone else's. to my ears, the clipping diodes are masking any differences the biasing might produce, except for an increase in volume as the Vc2 gets near V+/2.

which doesn't look right, because the diodes will keep clipping at the same level, no? must be one of those "things" that sounds like more volume, but isn't. anyone have a word for that apparent volume increase besides volume?

now to hammer it into that bloody salad bowl .....

[LucifersTrip]

I used a 2N3565 (hfe around 150) and a 2SB33 (hfe 52-54)

The voltages, however, are as follows (battery = 9.7V):

Q1: E-1.58  B-1.49  C-1.39

Q2: E-1.58  B-2.12  C-3.24   

Finally breadboarded and my voltages are very close to Mark's

Q1 (GE 100 hfe / 50uA leak)
Q2 (SI 100 hfe)

9.69V supply

E        B       C
1.71,  1.60,  1.43
1.71,  2.39,  3.24

Sounds good, but not nearly good enough to finalize...time to experiment. There seems to be a sh*tload of possible hfe combos.

[hellwood]

for those of you less than impressed with this circuit, dig through your junk drawers for your leakiest dud transistors before you give up. Germanium for Q2 makes all the difference too...I was seriously about to scrap this thing after what I thought was an entire day wasted experimenting. Now I cant stop playing with it.

Q1 2n404 (610 hFE/6.1 leakage)
c 1.60
b 1.68
e 1.8

Q2 2n1302 (125 hFE/1.25 leakage)
c 1.88
b 1.86
e 1.81