Stupid Pedal Trick: Photon Face (a si fuzzface that don't suck) model B

[pinkjimiphoton]

the full-gospel si fuzzface. mods detailed. output pot 100k, output cap .047uF
e to ground q1, 220r  e to "top" of fuzz pot q2, 47r, 100r between fuzz pot and ground

fat and philthy, and cleans up just like the ge ones and responds pretty much the same as well. couple extra parts and huge diff.

any npn's will work. i use 'em around 215hFE/.650ma

if ya use say, 450hfe, double up the e resistors. you'll have to play with the bias probably to keep the tone right tho.
this is bone stock other than the changes detailed.

[rutabaga bob]

Next up - the switchable version!  Stupid Face to Photon Face!

[pinkjimiphoton]

be easy enough to do. tried the stupid face live last nite, thru a friend's rig. it needs to be a little darker, so i gotta adjust the input cap and output pot a little bit. sounds great with buckers or single coils cranked, but gets a little "too clean" when rolled down with the guitar volume.
we'll figger it out.

[thermionix]

i use 'em around 215hFE/.650ma

You have Si transistors with leakage?

[pinkjimiphoton]

i think that's the amount of current draw by the transistor in this case. they are definitely silicon.
i've gotten to really love these little component tester gizmos i get on ebay, plug 'em in and it tells ya all kindsa stuff.
but in this case, i guess it takes about (i'm assuming again) .656 ma to hit the hfe of the transistor?
beats me. just what the meter says.

i love this thing. http://www.ebay.com/itm/401148685804?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[thermionix]

Oh I see.  My transistor testing experience is limited to the RG circuit thing.  I never saw leakage on Si, so I actually quit testing them for it.

[pinkjimiphoton]

the little tester thing is really handy for like, 12 bux. i use it to match fets too, diodes etc. it shows the leakage and orientation of ge's really well too. pop it in, and it tells ya the pinout, the gain, leakage if applicable etc. really handy, bro!

[thermionix]

Roger that.  I looked at the Peak Atlas things, kinda pricey.  $12 is much better.  So, have you boxed yours up?

[pinkjimiphoton]

well, the last one, no... it lasted 2-3 years. but this one i was going to, but they didn't send me the one i ordered with the battery holder.
so... maybe

[Digital Larry]

Hi Jimi,

F00kin brilliant!  I'm mostly into DSP but recently built a Rat-like circuit and messed with clipping diodes and what not.  I don't go for distortion in the digital world for the most part.  As a guitarist, I mostly play acoustic, and to be honest I think that most distortion circuits sound good, unless they respond suddenly and harshly and don't decay in a smooth way.  Playing through it is the real test, at the point where you stop turning knobs and just play for 5 minutes you know you're on to something.

I'm all for circuits that avoid unobtainium parts!  It looks to me that the big change over a Ge FF is the emitter resistors which provide feedback, reducing the gain.  I think that the feedback from the second transistor to the base of the first helps keep it turned on, so that the difference in Vbe from Ge to Si is not such a big deal.  But the current gain differences between Ge and Si ARE significant and you have compensated for that.

Do you have any other insights into this strange magic?  I think I will build one up in the not too distant future.

DL

[pinkjimiphoton]

look into the schizoid face. my lame videos sound makes it hard to tell, but the schizoid is a combination of this and the earlier stupid face one. you can footswitch between them.

adding a series resistance between the 22u cap and ground will affect the circuit in a noticeable way too. everybody biases with the c resistors, and usually trimmers.

that will make it fuzz, and get you a working pedal, but imho the only way i've found is to get emitter resistance added to the circuit. it doesn't have to be a lot. in fact, most of the time the e resistors i see being used in other circuits are way too big for a fuzzface. they work, and you get fuzz..
but you loose the feel and controlability. by adding just small resistances you can limit the gain of the transistor in a softer way than c biasing. the combination of the two end up working together to not only limit the gain of the q, but bias it into a useful range where it will actually sound good.
you wouldn't think such small resistances would matter. but they do. i can absolutely hear a difference in q2 between the 47r i spec and a 44 r (two 22r resistors in series ) or 54r (two 27r's in series). that's not much of a change, but makes the q react differently. using standard values i tried up to 1k and didn't like it. below 47r it's a little too fuzzy and the volume knob on the guitar reacts differently. above that, and you get a glitch in the sweep of the volume knob on the guitar. there's a real sweet spot. but i found that i could vary the gain and limit it pretty easily... tho it loses the tone and ya gotta compensate with other stuff after a while.
i wanted it as close to stock fuzzface for this one as possible in component values, cuz they DO affect the tone of this circuit.
but say you need to figure out the r's for the e's with a gain twice as high? simple. go 470r and 100r if you got q's around 400. go 1k and 220r if ya got gain around 800. no real headscratcher math involved, it actually works like that. but you do have to compensate by changing the c bias. but it's pretty close to just doubleing it in this case, which was the other reason i thought somebody may like it.
its stupid easy to implement. probably wrong from an EE standpoint.
i'm just mr opposable thumbs, a monkey with a breadboard. all i give a shit about is if it sounds good.

i haven't used my GE fuzzfaces live now in almost three weeks. the only comment i've heard is my guitar sounds a bit cleaner with a bit less murk and noise. good enough for me.

rock on brother

[Kipper4]

Based on what schematic Jimi please?

The second one down (npn) here?

http://fuzzcentral.ssguitar.com/fuzzface.php

Thanks Rich

[pinkjimiphoton]

hi rich,
nope, not #2, the first one. with the 470r resistor, not the 330r one.

the actual fuzzface (the original pnp ones) used a 33k and 8.2k c resistor, the other two were 100k and 470r. 1k fuzz pot, 500k volume.

the caps are 2.2u input, 22u fuzz and .01 (some say .1) for the output cap.

in this case i use all the original values except i add 220r as an e resistor to q1 and 47r to q2. this limits the gain available, reducing the 215hfE q's i'm using to standard fuzz gains, or at least standard enough where the original part values really shine.

ALL the components in a fuzzface matter. i wanted one that could be silicon, clean up right, fuzz right, feel right and react right, be impervious to temperature and sound consistent. tall order. 300 + fuzzes later, this is what i got.

it's closest thing to a germanium i've found out of all the variants i've tried, including most of the really popular ones.

this circuit is only half of it. do a search for the schizoid face, it has the entire project there. it's meant to have a 3pdt switch that changes the input caps and location of the ground on the 22u cap from either directly tied to ground or floated above it with a 100r resistor.

for input caps, photon face is 2.2u. stupid face WAS 47n, which can be great with dark humbuckers, but i've revised it up to 390 or 470 n now, doesn't roll off quite so much bass. the 47n sounds almost like a transparent overdrive with the guitar down.  but 470n sounds the same, but with a substantially deeper low end.

this circuit was designed for npn q's at 215hfe (of which i have a huge bag of).

there's more info on the other forum, as this seemed a subject of agravation to some folks here when posted.
i didn't ever say it was the best fuzz out there as some seemed to take it... i said it's a si fuzzface that don't suck, and i stand by that.

the thread about it contains video of the whole thing, with both this circuit and the other in one box, and compares back and forth to my blue germanium fuzzface with the nk275's in it (i prefer my older DA red one, but that lives with my pedalboard).

you can hear in the video, the only real difference is that my circuit doesn't have as much noise or emg/rfi interference as the original, and a slight bit less mud at lowered settings. full up, i doubt anyone could tell the diff. the full project includes the happy/sad face graphic, and the "eyes" change... 2 blue for the stupid face, 2 green for the photon face, or they all turn red in bypass. stupid and useless but fun to look at.

i've been using these live now for the last three weeks, including a studio audio/video recording session and it worked tits. check it on the breadboard, and see what ya think. if ya have a ge model, try a/b'ing and see. the video is hard to tell cuz the program i use only allows mp3 compression up to 128k which is horribly low.

ANYWAY the combo of the two options is worthy of a footswitch, as you can switch between two very valid flavors on the fly... either the more overdrivey stupid face, or the old-school sludge of the photon face.

hope ya diggit rich. peace bro

[pinkjimiphoton]

sorry, just to be clear... the first schem on the fuzzcentral page would be it, just reverse the electros and use npn's, and add the mods i listed and you should be fine. to use higher gain q's, say in the 400 range, double the e resistors values.  800 range, double it again.
that said, at that point it probably will not fire up right without the c resistance being changed to achieve bias.
i feel it starts to sound less like a fuzzface if ya monkey with the values too much... there IS a bit of magic to the particular components of the stock circuit imho, my thing here is just a way to make it work with the least circuit molestation possible.

again, everybody sets the bias via the c resistors. but that doesn't affect the gain the same as limiting the gain with e resistors will. and if they are too big, the sound and feel goes out the window.

[Kipper4]

"any npn's will work. i use 'em around 215hFE/.650ma"

sorry after this comment in post 1
I assumed it was the npn version

[Kipper4]

Thanks Jimi.

[pinkjimiphoton]

totally welcome rich, i hope ya diggit. yeah, i just used the standard ge but set up for npn, slight molestation via the e's seemed to make it work. never liked the 330r or 1k suggestions i've tried there. i like the 470. i think it has something to do with the way the fuzz interacts, but i might be trippin

[Digital Larry]

its stupid easy to implement. probably wrong from an EE standpoint.
i'm just mr opposable thumbs, a monkey with a breadboard. all i give a shit about is if it sounds good.

No shame needed.  I AM an EE and I can tell you that I never learned anything like this in school because there are few applications other than guitars and whatnot where distortion is a desirable result.  So I look on with fascination as I try to figure out how a lot of these things work and why they sound good.

DL

[Ice-9]

Great stuff Jimi, sounds great and nice to see you ding 'stupid Pedal Trick' again. 

[pinkjimiphoton]

thanks guys, i hope if you build it you diggit, and i hope ya make it better than i did