Biasing the Fuzz Face

[Mark Abbott]

I have a quick question to ask about biasing Fuzz Face pedals. I know from circuit diagrams and R.G.Keen articles on the Fuzz Face that Roger Mayer used a 2K pot to get the second transistor's collector voltage to half the Vcc. Another way to do this would be to reduce the collector resistor.

My first question is has anyone experimented with these two alternatives, and what do they prefer?
(I will try this myself, though I'd like to get a bit of feedback prior to doing this.)

Speaking of feedback, from my noodling with a Fuzz Face circuit in Multisim (electronic workbench/Pspice type of program) I found that lowering collector resistor resulted in extra current travelling through that path.
Upon reducing the value of the 100K feedback resistor (from the emitter of Q2 to the base of Q1) I found that the operating point of the first transistor could be altered. This resulted in the mark space ratio being more even.

The obvious result of reducing the collector resistor would be that the voltage gain of the second transistor would be considerably less, but given the gain on tap from a Fuzz Face, I'm not sure anyone would notice this.

For a friend of mine, I rebuilt a Dunlop Jimi Hendrix Fuzz Face using 2N3904's and the reduced second collector value. I compared it to my Fulltone 69 fuzz and I found the gain sounded quite similar, the overall character of the Fuzz Face was retained in the pedal though there was an increase in high frequency content which sounded like a slight edge to the tone of the pedal, the result was that the rebuilt pedal with Si transistors did sound better than the Fulltone pedal (which I'd describe as the Fuzz Face Jimi Hendrix should have owned.)

Not wishing to reinvent the wheel does anyone have any tips on Fuzz Face pedals?
(I do recall Joe Gagan did a lot of work with them.)

Thanks for your help.

Yours Sincerely

Mark Abbott

[hendriko]

i dunno when you were planning on building yours, but i'll be starting a RM modded FF early next week (as my first pedal)
so, if you wanna wait till then, i could tell you what difference the pot makes (i got a 1k (original FF) and a 2k laying around here)

cheers

[m_charles]

I believe you said you liked the high-end "edge" on your Si build for your friend, but if you want to get even closer to the ge tone, or just experiment, put a .022-.047uf (I like .033uf) in parallel with the 470 ohm (you can push this resistor up tp 1K2 for more output if you want). You can also put a .001uf from Q1's gate to ground. I got these mods from j. philpot's site, and one other I can't remember.

I also highly reccomend a 1k reverse log as opposed to the other fuzz pot options.

have fun!

chuck

[Mark Abbott]

Dear Hendriko

Personally I would recommend trying out both the 69 fulltone circuit and the Mayer circuit, and see which one you like. I have played both pedals and while the RM pedal does sound very Fuzz Face like, to my ears the 69 Fuzz is the better pedal.

I have now re-read R.G.Keens articles on the Fuzz Face, RG talks of the evils of the Si FF, but not of getting it to work for you. From my experience transistors like the 2N3904 work well as they have a relatively low gain of 100 to 200.

I did check out the Schaller fuzz clone which is  FF clone of sorts and this is the housing my FF clone will live in, as I really hated the Schaller fuzz. Oh yeah, if you own a Schaller fuzz and the noise is too much, try and get hold of copper adhesive tape and stick that to the inside of your plastic enclosure. While there does appear to conductive paint inside the enclosure it doesn't appear to be enough to stop the noise.

I'll continue plodding with Multisim and see what I can find.

Thanks for your assistance and replues.

Yours Sincerely

Mark Abbott

[Gus]

A story.
  Back about 95 I started to play with effects.  Shortly after that I got on to the web. Now before the web I had worked on TSs and FF circuits for friends.  Then I found the web and Lepers, R.G.s and DMZ.   I started to buy and build FF circuits back then a friend had different FFs and I got to trace a number of them.  I have a page in one of my notebooks with a circuit and lists of the different values used in the FF circuits.  The Vox circuit at R.G.'s site is a good sounding Si transistor one and the circuit I like is very close to it. 

  After building  with Ges I decided I wanted to use Sis.  I built a FF type with almost all the resistors variable(one thing I see here is people often don't use fixed lower value in series with the pots)   For example the feedback use a 33K and 100K for 33k to 133K.  I do like the sound of Ges so I did some thinking and math to come up with the resistors around the first transistor to drop the openloop gain.  I also played with the Ic range by ear IMO it changes a little with different transistor dies and masks or what is called a process number.

  I also found I don't like cap(miller feedback) collector to base as much as a passive filter.

You can find the result of this work in the distortion section of the 3 tran , rocket, and my take on FF types in schematics.  The cool thing with it is most any Si you drop in will work I have used lower gain 2n2222a to MPSA18s different tones but all bias up.

I could use a 2kgain  pot but I like the bias changes I get with using a 1K gain with a 5K bias adjust I am not a person who thinks there is only one bias point of 4.5V at q2 collectors

I have an older GE(the company) transistor manual from 1964 this has all the math for the FF type circuit in it.

[hendriko]

Personally I would recommend trying out both the 69 fulltone circuit and the Mayer circuit, and see which one you like. I have played both pedals and while the RM pedal does sound very Fuzz Face like, to my ears the 69 Fuzz is the better pedal.

i'll put in values for all types of FF mods (the original/RM/Smallbear/fuller) and see which combos i like best
cheers

[mac]

I have an older GE(the company) transistor manual from 1964 this has all the math for the FF type circuit in it.

Can you email me a copy? I've wrote a program that emulates the dc bias of a FF, and a pdf with the dc bias solution. I'd like to compare.

http://geocities.com/guitarfxs/fuzzface_dc_bias.pdf
http://geocities.com/guitarfxs/bias/main.html

mac

[Mark Abbott]

Dear Mac and Gus

Thanks for the info, I too would like more info on the circuit to wake up the grey matter. 

mark_a_abbott@yahoodotcomdotau

Thanks for your help.

Yours Sincerely

Mark Abbott

[Gus]

I will look for the manual

The distortion circuit I like is not a true FF because of the external emitter resistor

When I read about Si clones I often wonder why I don't read about people  building the VOX circuit, like the one in  R.G.s writeup.

[Mark Abbott]

Dear Gus

I will give it a go and see what it sounds like. Until you mentioned it, I never realised that it is an optimised circuit for Si transistors. RG speaks about Ge transistor mostly in the article, and by the time I got to the Vox part I had felt that I've gotten the point he is making. The things that are a little off putting regarding the Vox circuit is the 0.0033uF going the the volume control and the 47K across the 500K volume control. Obviously someone will pop up and say, that is to pass frequencies of ???K.

I do want that FF vibe from this pedal though.

Thanks again for your assistance Gus.

Yours Sincerely

Mark Abbott

[mac]

I built my Si FF with 2n2219 (110,117) based on the original FF circuit. As the transistors have the recommended gain, I did not consider to use a lower feedback resistor like the Vox, 47k. Just added a 100pf from C2 to B2, a 0.047uf across the 680R (470R original), a variable cap at the input and a 100k in series also, in case I use HB or a wah. That's it. Sound nice.

Now I'm planning to replace transistors or build a new one using 2n2369, and without those pico farads caps from C to B, and the cap in paralell with the 470R. I noticed that they introduce some weirdness. Instead I'll experiment with a small cap across the feedback resistor or simply a cap across the volume pot.

I guess that higher gain transistors need a lower feedback resistor, and possibly a small resistor at Q1 emiter, ie the YAFF. Some say that bc108 sound very good this way. Never tried.

mac

[Mark Abbott]

Dear Gus

Thanks for your help. The transistors that I intend using have a gain of 150. I haven't had much luck at finding Si transistors with low gain, though I supoose that isn't a big deal as I can simply alter circuit values to get the desired gain.

The Vox circuit is quite interesting, it has the Mayer thing, (or perhaps Mayer borrowed the Vox circuit change.) There is the 10K collector resistor instead of the 33K and of course the lower fedback resistor.

Yours Sincerely

Mark Abbott

[brett]

I am not a person who thinks there is only one bias point of 4.5V at q2 collectors

THOSE, children, are words to live by. 

What's good bias, and what's not good bias?  What musician would seriously ask a DMM to tell us these things?

[MartyMart]

I know that some of us dont like trimpots, but replacing the classic 8k2/33k's with
20k and 50k trimpots, or a combination of trimpot/small r works well for me and allows
fine tuning of the bias at collector.
I recently bought some good quality trimpots, which are around $1 each and have had
no issues at all with them if carefully soldered.
I really like the Roger Mayer FF version and the Vox too.

MM.

[m_charles]

just curious,

what are people's complaints about trim pots??

[mac]

My Si FF is a little lower than vcc/2, (0.2-0.3v). I trust my ears more than my DDM.

Marty, ever tried to set ib1 = ib2 ? 

With Si the condition is, if ic2=vcc/2:

R2/R3 = vcc/2/vbe1*(1 - R4/R3/hfe2)

and

R1 = 2*hfe2*R2*(1 - vbe2/vcc)/(hfe1 + 1) - hfe2*R3/(hfe1 + 1)

R1=q1 collector resistor; R2=the sum of both q2 collector resistors; R3=the drive pot; vbe1=q1 base-emiter diode forward drop; R4=the feedback resistor; hfe2=q2 gain.

For example, if hfe1=90; hfe2=120; the drive pot R3=2k; R4=100k; vbe1,2=0.6v; vcc=9v then:

R2=8.75k; R1=18.9k; ib1=ib2=0.0042857ma; vc2=vcc/2.

Note that R4/R3/hfe2<1 or hfe2>R4/R3, that's why I used a drive pot of 2k.

With Ge leakage must be considered. Some more algebra.

Curious result I've posted long ago. I discovered this accidentally playing with R1 to find the best value.


mac

[Mark Abbott]

Regretably I must plead complete ignorance regarding the gain of the transistors and this circuit. I built up the circuit using Multisim (Spice) and the difference in Beta don't really make that much difference to the outputs that I am viewing on the virtual CRO.

I dare say that I am missing something, but what have you guys found to be the difference in tone between the first transistor having a beta of 80 and a beta of 120?

I seem to recall being taught years ago that the external components will dicate the gain of the circuit. Though it has been too long to remember this lesson.

Thanks for your assistance

Yours Sincerely

Mark Abbott

[jaytee]

If you look at the fuzzface circuit the 100k feedback resistor is feeding base current to Q1 to bias it from the emmiter of Q2. However, most people will have the gain pot at max all the time or just have a resistor in there. As far as signals are concerned there's no negative feedback or emitter resistors in either transistor because of the bypass cap accross the gain pot. So there's nothing to tame the gain of the transistors. At least that's what it looks like to me. If you put an emitter resistor in for some local negative feedback it will bring the gain down of the transistor at the same time it will make it more linear. So I can't see that it would sound the same as just having raw transistor gain.

[Gus]

jaytee

    Yes an emitter R is not the same as a low gain transistorat the first stage.  Yes at max gain things change.  Some people do use the gain at less then max   

  That is  why this what looks to be simple OLD two transistor circuit is so much fun to play with and learn from.

Also lets say it is an NPN neg ground the closer the transistor collector is to gnd the higher the gain.

Look in the Art of Electronics a "simple" one transistor stage at collector at 1/2 VCC has a gain of about 20  X supply voltage  ONLY for small signals the closer to gnd (sat) the higher the gain,  the closer to supply the lower the gain.  So a Triangle wave (like a diagram in TAOE) will be bent.  The gain changes in a dynamic way on both sides of the signal.

In the treble booster writeup at R.G.'s the collector is set closer to power supply 7 something VDC.

  Now the FF type has two stages with the collectors going in different directions so the gain are changing in a complex way.  So how one sets up the two collectors "starting points" idle DC voltages and Ic and then base current needed at different hfe etc
Read  macs's posts

Open a bad pot and look at howe it is built.  MOST POTS are not made to pass DC via the wiper to the resistive track.

[zbt]

in my experience with silicon transistor BC108, BC109, BC183, all VC measurement are below 4.5V.
while some make 8K2 as bias with 10K trimpot, I just curious about it's volume

If we look 8K2 and 470 as voltage divider like volume in rangemaster, then why silicon version use 330, my though is to lower the volume.
while trim below 8K2 might get 4.5V, it would increase the volume.

470/(8200+470)*100% = 5.42%
330/(8200+470)*100% = 3.86%
470/(6800+470)*100% = 6.46%
470/(5600+470)*100% = 7.74%

with bias like this also make a different volume level.

Also why is below 4.5V, probably its because Q1 is sleep (off).

VC = 4.5V
IC = VC / (R470 + R8200) = 4.5 / 8670 = 0,000519031
IE = IC
VE = IE * RE = 0,000519031 * 1000 = 0,519031142V

isn't it VE is supposed to be greater than VBE Q1 0.7V


      hfe   VB   VE   VC
      
Q1 BC108C    66   0.640   0       1.582
Q2 BC108C   109   1.582   0.949   1.113

Q1 BC109C    51   0.604   0       1.596
Q2 BC109    105   1.596   0.964   1.012

Q1 BC183     76   0.528   0       1.756
Q2 BC183L   212   1.756   1.091   1.118

in my measurement, even with VE greater than 0.7V, I feel Q1 still lazy.

while learning my old book about direct coupled, the circuit it is just alternative where RF 100K
connect to collector of Q1.

So I tried this bias circuit.




It works!

Correct me if I'm wrong, please test if the sound is pleasing.