si fuzz face - 'high end taming capacitors'

[jg24]

Hi, couldn't find an answer to my question with a search.  Read in RG Keen's tech of the fuzz face article that you can add small caps to the transistors to get softer distortion - and as I'm building a si fuzz I thought it'd be a good idea to add some.  Thing is the article is a bit inconsistent about where you put the caps (at least to a beginner like myself).

At one point it says you can connect 10-100pf caps from collector to base of the transistors (I'm guessing both... I think I've seen a photo of this done to a wah pedal to get rid of radio interference?).

Elsewhere it says you can add a 10-100pf cap from collector to base on the 2nd transistor OR a 100-680pf cap across the collector resistor of the 1st transistor (am I right in thinking this is the 33k resistor in the traditional circuit?).

There's also a schematic of the JH2 dunlop fuzz face - which uses silicon trans of course - and that has a 47pf cap across the 68k resistor (the 100k resistor in the traditional circuit).

So what I want to know is - which is right or best or are they all possibilities?  Can anyone enlighten me?

Cheers, Johnny

[John Lyons]

Yes, those all will work.
Take a look at this one as well.
http://fuzzcentral.ssguitar.com/axisface.php
the .033 in red is one of the ways you mention.
You can touch the cap to the board to hear the difference.
If you turn up the amp to hear the amp/pedal hiss you can hear
what the cap is doing.
Try the cap connected across the base and collector
and then across the collector resisistor or 1K (etc) to hear
the differences.

John

[Gus]

http://www.diystompboxes.com/pedals/3tran.JPG
http://www.diystompboxes.com/pedals/gusFuzzFace.gif

note the Q1 emitter resistor and lowpass filter and date

[jg24]

Thanks John and Gus.

Man, and I thought a fuzz face was going to be a fun, straightforward build if I avoided the whole can of worms that are germanium transistors for the time being!    There are so many mods and variations on the theme you end up with too many options!  I liked the idea of a two knob pedal - probably going to end up with five now.

[mac]

my ge ff has 5 knobs, vol, fuzz, bias, input cap, input resistor.

instead of the taming capacitor you can try a dark ge at q2. you get some ge mojo and it will not suffer thermal runaways.

also try a small resistor from q1 collector to q2 base. this will reduce gain  a little but will tame highs. you can restore gain by using a slightly higher gain transistor at q1.
also try a 47k instread of 100k as feedback resistor. this will send q1 collector voltage down a little, more to ge levels.

mac

[jg24]

Thank you Mac.

By dark ge do you mean low or high gain and/or low or high leakage?  I would have to find an npn ge as I'm building a negative ground ff.

Yes it was the temperature thing and problem of sourcing decent ge trans that has put me off building a ge ff for now.  I plan to - but after trying a si pedal first to get an idea of what I might like.  I've not had the chance to use a ff yet.

The 5 knobs I decided on including are vol, fuzz, bias, contour/boost, and battery sag... and maybe a switch to change the stock 2.2uf input cap to something smaller - .1uf or less - for a bass cut (I read the pedal can by muddy sounding?).

By small resistor between q1 collector to q2 base do you mean less than 1k?

The 47k feedback resistor is interesting - I had wondered why the dunlop jh ff schematics have 68k here (sometimes with a 47pf 'taming' cap)...

So many options!!! 

[jg24]

note the Q1 emitter resistor and lowpass filter and date

I didn't get at first why you mentioned to note the date Gus - I guess you came up with the high end roll-off cap on the volume resistor (the 330-470R in the traditional circuit)??  And the 47k feedback resistor?

Why the q1 emitter resistor incidentally - is that a bigger factor in smoothing the sound of silicon transistors than parallel caps?

Edit - I just found the YAFF page which answers the emitter question 'the resistor induces some local feedback into Q1 which reduces the gain... By using a small amount of degenerative feedback, the gain is lowered and the response of the circuit is more useful for milder distortion effects'.  Do these 100ohm resistors work better than small caps across the collector and base of the transistors?  Could you use both - or is that overkill?

[mac]

By dark ge do you mean low or high gain and/or low or high leakage?  I would have to find an npn ge as I'm building a negative ground ff.

i mean a ge that has a low freq transition, ie that has trouble passing highs.
2sd352, 2sd72k, even ac187.
2n388, 2n1114 on the other hand are very bright, at least the ones fro TexInst I have.

Yes it was the temperature thing and problem of sourcing decent ge trans that has put me off building a ge ff for now.  I plan to - but after trying a si pedal first to get an idea of what I might like.  I've not had the chance to use a ff yet.

temp runaway depends on q1 ONLY. heat q2 only and you"ll notice a very small drop at q2 collector.
also leakage is a problem at q1. for exemple if you bias an all ge you can exchange q2 for a silicon and q2 collector will almost be the same. put leaky transistors at q2, although i do not know if there is a relation between leakage and future faults.

By small resistor between q1 collector to q2 base do you mean less than 1k?

<10k-22k

The 47k feedback resistor is interesting - I had wondered why the dunlop jh ff schematics have 68k here (sometimes with a 47pf 'taming' cap)...

see
www.diystompboxes.com/biascalc
you can play with resisotrs values and trnasistors.

mac