pnp fuzzface with negative ground trick

[pinkjimiphoton]

jack orman is the guy who created this brilliant take on the fuzzface and it sounds freekin great, classic fuzz face tone. usually.

but sometimes it will oscillate, some cases it will motorboat, too and can sound fairly ugly... and worst, in my case on about 6 out of 10 attempts had horrible hum or oscillations that would change pitch as the guitar was turned off with its volume. the fuzz sounds and reacts perfect til ya turn it all the way down... then HHHRRRRRONNNNNNNNNNNK mixed with insect noises crossed with mating beefalo... juggalo? whatever

rg reccomended a big freekin cap. i reccomend three of them.

to get all this crap gone i added rg's suggestion,  and had to go to a 2200uf electro across the rails. no more motorboating. still oscilates n hums with guitar off.

added a snubber, 3.3n across the rails, and a 47r resistor in line from b+

most of the insect noise now gone, still acts the same like it should, til ya turn it all the way down on your gtr. vrooooooooom. fudge me.
try messing with the circuit with different combos of r's c' and rc networks.
usual tricks had little effect, and to get rid of the noise with the usual 100=470p across the b./c  junction robbed it off fuzz and highs way too much.

so i stuck a double roach clip wire on a big ole freekin cap, connected the other side to ground, and started feeling my way around the circuit to see what would nuke the noise. 33n.

most places, did nothing. but from q1 b to ground,  it reduced a good bit of noise. still hummed with the guitar off.

so i tried it from q2b to ground, not bad... then q2c to ground, BINGO.
dead quiet, very little tonal change other than less noise, and hum almost gone with the guitar way down.

this led to more 'xperimentalations, like, and i settled on a green poly 100n chicklet reading 104.4n on my meter from q2c to ground.

wango, muthafunka.

no more rfi, no more motorboats, no more hiss or hum and dead silent almost when the guitar is rolled all the way back.

so if you run into this issue, this may help.

peace out!

[Fenton Bresler]

Hi there!

Long time lurker here. I haven't posted or moved past the breadboard as of yet. I have gleaned what information I can from here and the net but I'm still very much new to electronics. So hello all!

From what I understand, bypassing Q2's collector with a capacitor is reducing the overall gain, is it not? I guess this might solve the motor-boating issue, but at the expense of having a reduced amount of gain on tap. Please correct me if I'm mistaken!

I could never get this circuit to work properly without a big a$$ cap across the power rails, but then I came across a slightly different solution to yours.

When building a negative ground PNP Fuzz Face (still only on the breadboard), I have had great success bypassing Q2s emitter resistance to "ground" ground rather than to the positive rail. So with the positive side of the cap on the emitter and the negative to the lower rail. Seems to solve all the issues I was having.

When looking at the circuit again, it kind of makes sense to me seeing as the outside of the coupling caps are not referenced to the positive rail either and are reversed in polarity. Is shunting the AC from Q2's emitter to the positive rail what's creating all the problems in the first place? This could explain why using a big enough capacitor across the rails seems to work, it is shunting all this dirty AC that's now on the + rail to ground. So why not send it straight to ground rather than going though that extra step?

I look forward to hearing everyone's thoughts. And thank you Jimi for bringing this up.

Best regards

EDIT: Added schematic

[pinkjimiphoton]

no idea man, monkey with a breadboard here!

WELCOME TO THE FORUM, FENTON!

[PRR]

> I could never get this circuit to work properly without a big a$$ cap across the power rails

Big/fat caps are a cornerstone of good audio. "Big" in context of a FF is not many pints or bucks. What's wrong with that?

> ...great success bypassing Q2s emitter resistance to "ground" ground rather than to the positive rail.

That is the "proper" connection.

R.G. has explored this territory and teaches "no one way"-- there is always some connection which "should work" but doesn't. So be prepared to be flexible.

[R.G.]

No one way. Huuuh. Shoulda trademarked that, I guess. 

Classical Nyquist stability theory teaches that you can transform an oscillating setup to a non-isolating setup by cutting open loop gain so that the gain dips under one at a frequency lower than the frequency where the total phase shift (forward gain plus additive shifts) hits 360 degrees. It is -posible- that this fix manages to slug the gain down by that much. Opamps commonly use single-cap based "dominant pole" compensation. In a 741-ish opamp, this is a single cap of about 30pF. In audio power amps, there is often a single cap from collector to base on the main voltage gain stage. A cap from collector to ground will do that too, it just doesn't get a boost from the Miller effect.

So this is very plausible in my mind. In fact, I did try something similar in some of the "hard cases". If it works for you, do it, and go with &deity.

My gripe is that a fix that was great in one case would not work in the next one. If it works for you, do it.

[pinkjimiphoton]

i think its definitely frequency dependent, only seems to work around 100n, which i'm guessing is the proper cap for the frequency the noise is at.

i told mike piera about it yesterday, its now an analog approved mod

[pinkjimiphoton]

works with noisy hybrids and si ff's too

pnp or npn, ge or silicon. adds a lot of cream to the tone, too.

[Fenton Bresler]

no idea man, monkey with a breadboard here!

WELCOME TO THE FORUM, FENTON!

Hi Jimi, thanks for the welcome! Definitely looking forward to learning as much as I can from you guys.

> ...great success bypassing Q2s emitter resistance to "ground" ground rather than to the positive rail.

That is the "proper" connection.

Is it? This topic has been beaten to death and I can't seem to find a single mention of the neg gnd PNP Fuzz face circuit with the bypass cap connected the way I suggest anywhere at all. I've only ever seen it as per Jack Orman's schematic on http://www.muzique.com/lab/fuzzface.htm (- side of cap on the emitter and the + to the positive rail, i.e directly in parallel with the emitter resistance.)

[PRR]

> Is it?

Historically, yes. Here's an example of a straight-amp from 1959. Note how the emitter bypass caps are connected. This is a negative-ground (+ hot) circuit using PNPs.

[pinkjimiphoton]

well, this being a fuzzface, the emitter resistor <the 1k fuzz pot> is partially bypassed already by the 22uF cap, right?

[Ben N]

My gripe is that a fix that was great in one case would not work in the next one. If it works for you, do it.

Truly spoken like an engineer.

[pinkjimiphoton]

so far, seems to work great on ALL fuzzfaces i've tried that were noisy, ge, si, pnp, npn. i tried it on other circuits too, it works with ic chips too. think of it as a decoupling cap, all it really is, really

[dbwbr]

sorry to resurrect an old thread - but i had a question about this:

so far, seems to work great on ALL fuzzfaces i've tried that were noisy, ge, si, pnp, npn. i tried it on other circuits too, it works with ic chips too. think of it as a decoupling cap, all it really is, really

what exactly is the trick here? the facebook screen grab suggests a cap on Q2C, another quote suggests "great success" with a capacitor at Q2E.

i don't mind the occasional stray oscillation, but i am trying to focus on the ground hum issues.

[percyhornickel]

sorry to resurrect an old thread - but i had a question about this:

so far, seems to work great on ALL fuzzfaces i've tried that were noisy, ge, si, pnp, npn. i tried it on other circuits too, it works with ic chips too. think of it as a decoupling cap, all it really is, really

what exactly is the trick here? the facebook screen grab suggests a cap on Q2C, another quote suggests "great success" with a capacitor at Q2E.

i don't mind the occasional stray oscillation, but i am trying to focus on the ground hum issues.

Here you can see how the negative ground "trick" works:

http://www.tonepad.com/getFile.asp?id=82


I have build the mine using this pcb layout with negative ground and works great.

I made silicone and germanium npn versions instead, so 22u is backwards in the mine and ground is negative.

For npn you just have to use the first image, connect the ground where the red line is, connect the 22u backwards and connect the first small jumper above with ground instead (just like the second image).

Other builds suggest using a small cap over the Q2c (like axiss face) which helps with the hum / rf noise and getting warm the sound. This cap changes the tone a little bit and that´s why I never use this mod, your taste...   

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

Saludos

P.H.

[pinkjimiphoton]

sorry to resurrect an old thread - but i had a question about this:

so far, seems to work great on ALL fuzzfaces i've tried that were noisy, ge, si, pnp, npn. i tried it on other circuits too, it works with ic chips too. think of it as a decoupling cap, all it really is, really

what exactly is the trick here? the facebook screen grab suggests a cap on Q2C, another quote suggests "great success" with a capacitor at Q2E.

i don't mind the occasional stray oscillation, but i am trying to focus on the ground hum issues.

q2c to ground bypass with 100n cap or so. go with the smallest cap you can to nuke the oscillation.

let us know how ya make out

peace!
pjp

[cctsim]

What has "always" worked (in 20 or so builds):

1.   220 Ohm resistor on the emitter of Q1 bypassed to the ground with a 1000uF electrolytic cap or two 470uF in parallel (depending on space).

2.   Lug 1 of the Fuzz potentiometer directly on 9V (connected before the 220 Ohm resistor).

3.   For the 22uF cap:  (+) electrode on Q2e and (-) on ground.

These tricks come from the red Dunlop fuzzface.

[pinkjimiphoton]

connecting 9 v directly to a pot can smoke it tho. caveat emptor.

[cctsim]

It will only smoke it if there is a DC connection between Lug 1 & 2 or 2 & 3. In this case, Lug 2 is only AC connected to the ground.

Anyway, connecting 9V to Lug 1 of the fuzz control is normal in pnp fuzz faces.