fuzz face - gain not working

[zoo]

I just finished assembling my first build.  I chose the Dallas Arbiter (http://www.montagar.com/~patj/fuzzface.gif) because it looked easy enough.

The problem is the gain knob (I'm assuming the 1K pot) has no effect on the sound.  The 500K pot does work (volume knob), so I basically made a volume control, which is pretty lame.

I made a couple of changes to my build, for availability-of-parts reasons.  The 20uf cap is replaced with a 22uf cap, and the transistors are NTE158, not AC128 (however the NTE is supposed to be an equivalent).  And the whole thing is assembled on a breadboard, if that matters.

I've checked the directions of the polarized caps, all grounds, and the wire connections, and even made sure I made it correctly.  The battery works, no lines are crossed, and I've unassembled and reassembled this thing three times, just to make sure I didn't make it incorrectly.

If anyone can help me, that'd be great.  If you have anymore detailed questions that need to be answered, I will get to that to the best of my ability.  I have an uneasy suspicion that I'm a total bonehead and it's obvious why it doesn't work.  Feel free to flame me (after having told me the solution!) if that's the case .

-z.

[Dragonfly]

Refer to the thread stickied at the beginning of this forum called "what to do when it doesnt work", and report back with the pertinent information.

AC

[zoo]

Here are the voltage readings for what I have (after reading Dragonfly's reply):
I wasn't sure exactly what to measure, just put the red lead anywhere I could.  The number on the left is the resistor/cap/tran value (the link shows the schematic) with the values on the right showing the left/top - right/bottom (depending on whether the resis/cap/tran in scheme is horiz/vert).  I hope that makes sense.  The DMM was set at DC 200mV.  Most of the values changed from beginning of the measurements because I guess the battery started dying.  It started at a cool 9.4 volts, and heated up (literally) to 6.1 volts, so hopefully my info isn't too skewed.

470: 2.7 - 2.0
33K: .8 - .6
8.2K: 2.0 - .9
.1uF: 2.0 - 2.0
500Kpot: 2.0 (the center and ground weren't measured)
Q1: E - .6 ; B - .7 ; C - .7
Q2: E - .7 ; B - .7 ; C - .9
2.2uF: .7 - .8
100K: .7 - .8
20uF: 6.7 - 6.7
1Kpot: 6.7

Hopefully that helps.

-z.

[Pedal love]

The gain required to get this overdrive, is produced when the 22uf cap completely bypasses the 1k pot. So when the capacitor is connected to the wiper on one side and the other side is connected to ground this effect is reproduced. I can almost guarantee to you that either the 1k pot or the 22uf capacitor is causing the problem. The capacitor is either not connected properly or the pot is bad. Check these carefully.pl

[Peter Snowberg]

The voltages you posted look a little strange and the battery heating up is a bad sign. I would say there's something else going on here.

The meter should be set to the 20V range for these readings. Where did you have the black lead for this testing? Even though this is a positive ground circuit, connect the black lead to the circuit "ground" and probe with the red. The voltages reported should all be negative numbers. I'm assuming you have a digital meter that will take the reversed polarity.... if not, you will have to place the red lead on the circuit ground and probe with the black lead.

The cap voltages you posted for the 22uF (20uF: 6.7 - 6.7) shows that something is not connected correctly. One side of the cap (and the 1K pot too) should read 0 volts.

[guitar_199]

A couple of particularly odd things jump out at me:

1) according to the votages as sent,   Q2e = .7v     100k(right end) = .7v  and the top of the 1k pot = 6.7v
Somthing's up there because they should all be the same point electrically.  Looks like Q2e may be tied to 100k(right) but the two do not appear to be connected to the top of the fuxx pot( 1k ),

2) top of the 33k = .8v   470ohm(left end) = 2.7v.  Same case as above, those two should be the same voltage but they aren't....looks like they may have a connection problem.

3) the 20uF cap tied to the fuzz pot seems to have 6.7v on both ends.  The bottom is tied to ground so, if your meter was properly referenced to ground, you should have 0v on the bottom of it.

4) (using the voltage drops as they are listed) 1.1v is dropped across the 8.2k ( 134uA ) and .2 v is dropped across the 33k ( 6 uA ).  That is a total of 140 uA of current accounted for and that does not add up to sucking a  battery down like that.

Here's what you might want to do:

Take a real close look at the connection between the 33k and the 470 ohm  You should have three things tied to that node:  the top of the 33k, the left of the 470, and the battery minus.

Take a close look at the area of Q2e.   There should be three things tied together:  100k(right end), fuzz pot top( Ik pot), and they should be tied to Q2e.

Take another close look at the grounding throughout.  Essentially, 6 things should be connected together here: bottom of fuzz pot( 1k ), bottom of the fuzz cap ( 20 uF ),  bottom of the volume pot ( 500k) and the ground from the input and out put jacks, and the battery plus.  If you have connected any power swithing methods, try unhooking that and run the raw battery wires to the proper places.

One other thought, on the votages you posted the top voltage only on the fuzz pot and it is the same as the top of the fuzz cap.  You did not post the voltage on the bottom.  It should be 0v just like the fuzz pot.  If it, too, is 6.7v then neither of these may be properly connected to ground but that would seem weird since,   the top ( by voltages) does not seem to be connedted to  Q2e and if the bottom is not 0, then it would either not be connected anywhere ( then where is that 6.7v coming from??) or something is wrong with your ground.

Just some thoughts on the top of my conically shaped head!   Good luck and come back for more or to let us know that you got it working.

[zoo]

I got it fixed (sorta).  It seems the major problem that was afflicting me was the ground connections.  The 22uF cap was not grounded properly (as predicted by everyone who replied, which makes me feel stupid), creating all these weird voltage values and a hot battery!  The battery was also grounded incorrectly, because I got the two connections of the stereo jack switched.  It took me a while to figure this out, spending a good amount of time probing with the MM.  But now it's kicking and fuzzing.  There is some crackling here and there when turning the knobs, but I think that's just the loose connections.  Fidgeting here and there cleans it up a bit.

The only problem left is the volume.  It's not loud enough.  With gain and volume cranked to maximum, my amp (Fender Champion 110) whispers from volume 6 and down.  Basically, with 10 on vol, gain, and amp vol, the loudness is equivalent to just the amp (no effects) on vol 2.  And all this on the clean channel (the dirty is pretty gross sounding).  Anyone have any ideas about this??

-z.

[guitar_199]

Two things....

Let me refer you to a neat place for Fuzz Face research (if you have not already been there).  Go to
http://www.geofex.com
Look in the top left corner for Technology of in a menu......click it.
then, look down the list for Fuzz Face.....click it.
This is a great article on the Fuzz Face and the history thereof.  Somewhere down the page shows the Fuzz Face with Roger Mayer or Jimi Hendrix mods.  Note that it changes the 470 to 1k, the 8.2k to 18k, and the 1k fuzz pot to 2k.   To the side it says that it increases the output level and gain.  This might be something more to your liking.

One other note, a lot of people replace the Q2 collector resistor with a trimpot of appropriate value.  This allows you to more accurately set the operating point of Q2 for transitors that vary in gain.  This might be something you want to consider.  Like if you keep the 8.2k, use a 10k trimpot.  If you go the Roger Mayer mod route, use a 20k.  Usually folks start by setting Q2 collector to 1/2 of the measured battery but then, play with it an dial it in.  In my experience the tone and sound you get out of this thing varies with this bias adjustment and......as it is designed.... it will not be correct for every transistor that you put in there....becuase the transistors themselves vary pretty widely.

Just a couple of things to consider....

Bob

[aron]

The only problem left is the volume.  It's not loud enough.  With gain and volume cranked to maximum, my amp (Fender Champion 110) whispers from volume 6 and down.

There's more miswiring somewhere. This is a good time to re-measure the voltages on the transistor legs.

[zoo]

I remeasured the voltages on the whole build (http://www.montagar.com/~patj/fuzzface.gif) hoping there would be some answers about the low volume.  Here are my values (grouped by junction this time):

33K, 470 -            9.36V
470, .01uF, 8.2K -  9.36V
.01uF, 500K -        9.36V
33K, Q1C, Q2B -    .51V
Q1B, 100K, 2.2uF - .08V
Q2C, 8.2K, -          5.67V
100K, Q2E, 1K -     .42V
1K, 22uF -            .42V
all grounds have been checked to be at 0V

and for easy reference (suggested by Aron):
Q1              Q2
E - 0V         E - .42V
B - .08V      B - .51V
C - .51V      C - 5.67V

Both knobs (the 500K and 1K) are turned to maximum.  I just looked and I realize the pots are 1K-A and 500K-B (they are supposed to be 1K-B and 500K-A).  Is the incorrect linear/log action of the pots causing the problem?  I don't have spare pots to test with yet, but thought I'd give you guys as much info as I can.

The only thing possibly wrong that I could see is that there is no voltage drop across the 470 ohm resistor.  I hope this info cranking will solve the problem.

-z.

PS  If I later decide to do the Roger Mayer mod, what differences should I be looking for?  Just a gain/vol boost?

[petemoore]

Is the incorrect linear/log action of the pots causing the problem?
  No, if the pot says 1k and is working...
  You'll get between 0k - ~1k adjustable resistance through a Linear or audio 1k pot, but the taper will be different, the taper is only a problem to how touchy the knob gets toward one end of the travel [ie touchy adjustment].

[guitar_199]

(( this is painfully long, but I hope it will help))

I could be wrong  but something seems a little off on the Q2-E voltage...just seems a little on the low side to me.  I hope someone will chime in if they know off the top o'there heads what it should be.  The Q2-C voltage is a little on the high side so that could have something to do with it.  If Q2 is a little bit underbiased (as in not turned on quite enough) then the collector voltage will be a little high and, due to the lower current flow, the voltage across the fuzz pot would be a little low.

As far as the drop across the 470 ohm, I've never measured that voltage but here would be the way I'd look at it:
  The voltage across 470 + 8.2K would be 9.36v - 5.67v = 3.69v.

  3.69v across 8.67K (  8.2k + .47k ) would give a current of about 426 uA.

  Now, take that 426 uA and multiply it times the 470 ohms and I would expect the 470 ohm to drop
  0.2 volts not much when you look at the rest of the voltage in that leg.

  Also, if you look at the voltage drop from Q2-B to Q2-E:   .51  - .42 = .09 v.  That really isn't enough base bias to get that Ge transistor turned on.

And, as I look back at something else...... that 426 uA collector current through Q2.... when you multiply it times the 1k of the fuzz pot, you get .42v dropped, and that is exactly what you have.

IMPORTANT NOTE:  This is another thing that raises some suspicion.  Your Q1 voltages look a little weird you might want to poke around in here closely.  The reason I say that is....... Q1-E is, of course, 0v since it is tied to ground.  Q1-B is .08v over that.  That is NOT enough to turn on a Ge transistor so this one should be OFF.    If Q1 is OFF there will be no ( or darned little) current flowing through the 33k resistor which translates to no ( or darned little ) voltage dropped across the 33k, which means you should actually have pretty close to battery voltage at Q1-C.   But you dont...... you only show .51v on the collector.  With .51 measured on the colector and the rest dropped across the 33k that says you are pumping a lot of current through Q1 even though the base bias looks like it should be OFF.   I'd be wondering about Q1 right about now.

The magic here is that the bias on this thing is somewhat "self-seeking".  When you first hit it with battery

a) Q1 will be off which will have Q1-C up in voltage which is applied directly to Q2-B which means,

b) Q2 will be trying to turn on HARD.  This will cause Q2 to start conducting current,  but....

c) Q2 emitter current going through the Fuzz Pot ( 1k ) will start raising the voltage  at the Q2-E/Fuzz Pot junction.  This voltage rise is fed back to the base of Q1, which means....

d) Q1 will start turning on, which means current will start flowing, which means current will start pulling through the 33k resistor, which means voltage drop will start to come up across the 33k, which means voltage at Q1-C HAS to got down, which means....

e) voltage (bias) on Q2-B will go down which means Q2 will MOVE TOWARDS turning off. As that happens the Q2 current will slow its initial climb, which means the voltage across the Fuzz Pot will slow IT'S initial climb, which means the Q1 base bias climb will slow down.

This "magic race" keeps going (albeit very quickly) until the circuit hits a spot where it balances....meaning Q2's current, develops enough voltage at Q2-E that is fed back to Q1-B, such that is biases Q1 JUST right so that the current through Q1, drops enough voltage across the 33k, that the voltage on Q1-C, which is fed over to Q2-B, will hold Q2....rigggghhhhhttttttt where it is!

And then....... Q1 and Q2..... sit there and hold each others throats.

What we'd like to see is, when this "balance" occurs.... Q2-C at about half the battery voltage...or there-abouts.  Your tastes might prefer a little variance.

Again, what I think that I see as a problem in yours is that Q1 is not biased on ( Vbe on it is only .08v) BUT it appears to be turned on pretty hard (  Q1-c is low, the drop across the 33k is high ).  If Q1 is hosed, sucking a lot of current, then Q2 will never get the bias that it needs to "start the race" above.

I think that's where I would be focusing.... around Q1 ( resistive paths on the board, 33k is not 33k, maybe Q1 is smoked....just some ideas....).


Oh, by the way, when you get this thing working, and the bias stablilizes more correctly, measure the Q2-C votage.  If it is not in the middle area of the battery voltage........ THIS is why a lot of folks put the 10k trimpot in place of the 8.2k resistor.    Adjustment of the trimpot allows you to "change the bias point" and this will let you adapt to transistors whose gains are not "precisely to spec".

As far as the Roger Mayer mod.... I've never done one so I don't know.  But I will say, my stock (home built)  Fuzz Face circuit has enough gain to blow hell out of my amp.  I mean that baby roars.....but EVER so smoothly.

Good luck, buddy.  Get back with what you find next.

[zoo]

I got it to work!  Thanks for the long explanation.  I tried applying your ideas to debug the board.  I moved some resistors around, replaced some, and replaced the transistors as well.  I almost lost my hearing turning it on (cuz the vol/gain/amp vol were still set at 10) but I think it's coming back to me.  Unfortunately, because of my lack of step by step debugging, I don't know exactly what went wrong.  Might have been a bad transistor or something, and also those unchanging voltages where something else I think.  Stupid me.  I think I'm going to start looking into that website (from guitar_199) to mod this thing and see what it can do.  But yea, this thread can finally rest in peace.

-z.

[guitar_199]

Good!  I'm glad you got it working.  That pedal is really sensitive to it's knob settings and, one time, you adjust it one particular way........and     whoooooop, there it is!!!!! Heaven!!!!!
Have fun using it!

Bob