Questions for experienced Fuzz Face builders

[9Lives]

I've built a butt load of pedals by now and I never in my life cared a whole lot for ANY fuzz pedal. In my process of building the openhaus I got a few extra transostors from steve at smallbear that came with specific resistors. I figured what the hell. It there has to be some reason SO many ppl love these. I put it together and I was surprised how much the simple design grew on me. In my experience pedals like these sound very gated and muddy. I HATE gated sounds. This was almost like a glassy overdrive that wasn't quite as smooth. I actually perfer the sound WITH a buffer in front. Don't nail me to hard...

so I started digging threw my transistor drawer and found some ac188 2n1309, ac128 and started learning the process of figuring gain leakage and all this wonderful stuff.

my question is to those who have build SEVERAL FF. When biasing I used the online calculator and I always come up short with the voltages. 2-3v for Q2 collector. Sounds muddy and gated. Grrrrrr. When I start tweaking resistors to get 4.5 it just doesn't sound the same as my finished build. I've been obsessing over this for 2 days and in need of some explaining.

Before anyone goes through the trouble of posting links (as much as I appreciate it) I've read them all. And I can't get a clear cut answer. Even when I use a trimpot it seems to compromise the circuit. The gains are 111 and 86 roughly both with 100 leakage. Am I doing somthing wrong?

[midwayfair]

This is hardly limited to Fuzz Faces: it's the reality of working with the wild inconsistencies of Germanium.

You will not be able to exactly duplicate the sound of another Fuzz Face being without extraordinary lucky. You would need identical characteristics of two different sets of transistors. Unlikely. Do you know the exact characteristics of the set you used previously? How close are they to the ones you're using now? What were the types of transistors? There can be a big difference between the breakup characteristics of TI vs. AC vs. RCA vs. OC etc etc.

Some things that might help:
Those other resistors from Smallbear are part of an overall biasing network. Are you tweeking those to produce the same voltages on ALL pins of both transistors, or simply trying to hit 4.5v on the collector of Q2? There's a big difference, and there's a set of target voltages for each pin from units known to sound good. Luciferstrip has posted them many times on DIYstompboxes. Some transistors simply won't hit these target numbers. You should at the very least measure the voltages on every transistor pin in your finished model and use those as your targets. That's apparently the sound you like: so you need that exact data as a reference.

Next, use your ears in addition to your multimeter. The simple fact is that some fuzz faces will sound better with 5.5 volts on Q2c. You'll certainly avoid gating in those situations.

[badgerific]

I always build FF with bias trim pots or normal pots because I like having the option of being able to adjust it easily so I can get more gated or bigger sounds from it. I'd recommend doing that in the future

[9Lives]

in my original build I used 2n1306 or 9.. Not positive. They were an NPN set and I didn't take the time to test them honestly bc I was pretty sure I'd hate it lol. And no I'm not using the same resistors but I'm trying to find that magical method to come up with that "biasing network" you speak of. With out it, it's garbage as far as I'm concerned. I figured that I wasn't hitting those pin voltages bc 4.5 collector for q2 sounded like shit. It's like a murder mystery figuring this out for me at this point. If you change one resister you gott change them all it seems like. I havnt messed with the 100K feedback. I'm assuming that one is pretty universal? So do you know what I'm trying to do besides hit the collector voltage? What I might do is break open the first one and get the readings and start trying to match and see if it sounds similar. But this is also a difference between 2n and AC.

[midwayfair]

in my original build I used 2n1306 or 9.. Not positive. They were an NPN set and I didn't take the time to test them honestly bc I was pretty sure I'd hate it lol. And no I'm not using the same resistors but I'm trying to find that magical method to come up with that "biasing network" you speak of. With out it, it's garbage as far as I'm concerned. I figured that I wasn't hitting those pin voltages bc 4.5 collector for q2 sounded like shit. It's like a murder mystery figuring this out for me at this point. If you change one resister you gott change them all it seems like. I havnt messed with the 100K feedback. I'm assuming that one is pretty universal? So do you know what I'm trying to do besides hit the collector voltage? What I might do is break open the first one and get the readings and start trying to match and see if it sounds similar. But this is also a difference between 2n and AC.

Ah.

The feedback resistor should not be altered.

I should tell you that PNP germaniums tend to sound different from NPNs. They will have brighter clean tones, sometimes a little harsher and more scooped, but they tend to sound softer when distorting. I don't know why this is, and I don't have a spectroscope to prove it, but I was very happy to learn I was not the only person that thinks so. I also happen to think that the RCA 2N13xx NPNs sound spectacular despite some consistancy issues, and Smallbear has also said the same thing ... but that no one wants them because they're hung up on getting continental transistors.

Here are Luciferstrip's voltage suggestions:

        e   b    c
Q1 -   0, .1, .6
Q2 - .45, .6, 4.5

Edit: link http://www.diystompboxes.com/smfforum/index.php?topic=98002.0

The dude knows fuzz, so I'd say these are every bit as important as what RG Keen says ... if you can get to these and you don't like it, it's okay. You might just not be a fuzz face guy. There's no shame in that! I'm not a Tonebender or Muff guy, but I like Fuzz Faces, you know?

One thing I'd definitely do if you're aiming for "perfection" is solder up some pots to use as variable resistors -- full sized, accurate ones -- for those four important resistors. (This of course is if you think it's worth the trouble to get a perfect sounding unit.)

I will note one other thing. It's about the gating, because I'm not sure from your description if what you're hearing is always the transistor's natural transition from saturated to clean or actual gating where the sound goes from VERYLOUDDISTORTED to practicallyoff. I know you use active pickups -- and you prefer a buffer in front. Part of what's going on here is that the fuzz sound is always going to gate on some level between input volume thresholds, especially in the bass frequencies. It might be more noticeable with your setup than with low-output single coils, which are known to sound more natural with a Fuzz Face. It's just the way the distortion in it works. Hit it hard and it breaks up. Back off and it's clean.

What happens if you hit the target voltages and you're still not happy? Well, it's time to start looking at other sections of the fuzz. You can increase the mids and the overall output by messing with the 470Ohm resistor. (You can also just add the '69 mids pot.) You can change the bass going to and from the circuit by changing the input cap. You can add the pregain control (I usually have to change the pregain and back off the fuzz knob some to get the sound I want even when I want full-on fuzz). You can reduce the voltage to the entire circuit and rebias Q2c to 1/2 the total supply voltage -- this will give you a darker sound that will saturate more easily, but may occasionally result in some gating or "transistor sound." You could charge pump the thing. You could try a hybrid. You could build a Cosmo instead.  

You can also get a more saturated fuzz face (and less gating of any sort) by going the silicon route. Play with Brian's "smoothing" cap in the Hipster, find a really low gain silicon for Q1 (PM me if you want the one I used in my Hipster demo ... I'll donate it to the cause, or you can be really adventurous and use a FET ) and you can get it to sound every bit as good as a germanium.

I know, I know, this seems like a lot of work for a circuit that you aren't in love with. It's like a puzzle, though, right? And it's why people who are really into fuzz don't want to trust buying a FF from DIY guys, because it's actually work to turn it into a great sounding unit.

[9Lives]

that's why this circuit is so fun to me even tho I'm not big on fuzz. It's so touchy. I perfer the sound of a good overdrivin amp. Like the klon deal. I don't use the active pups that much. I enjoy using pedals more. The npn ff I made sounds more over drivin. I'd say with the rotation of the fuzz pot I have 75% cleaner tones  and dimed out I get the fuzz nastiness. I'm not fond of that. The npn is so smooth and perfect. I'm wondering if I just got freakin lucky on this one.

Now. Back to the PNP. I have a 2n1309 for Q2 gain 111 leakage 100 q1 gain 92 (ithink) leakage 101. I actually managed to hit every pin almost perfect. Q1 .7  .09  0   Q2  4.55 .57 .44   that's real close. I still notice a big difference. NPN is so lush sustain dimed out and clean glassy treb on lower settings. This is with the buffer. W/o it, it just sounds a damn mess to me. Both of em lol.  Opiniomn are like assholes tho..  The PNP is more fuzzy at all settings. I don't get that cleaner sound out of it. Like it's some what breathing. Hard to describe. That and my son won't shut up and I can't think straight.

does the value of the resisters matter at all as long as you get the voltages? I had to got down to 15k for Q1 collector. Do I have to be closer to 33? Could the breadboard and the transistor itself cause these difference?

[midwayfair]

that's why this circuit is so fun to me even tho I'm not big on fuzz. It's so touchy. I perfer the sound of a good overdrivin amp. Like the klon deal. I don't use the active pups that much. I enjoy using pedals more. The npn ff I made sounds more over drivin. I'd say with the rotation of the fuzz pot I have 75% cleaner tones  and dimed out I get the fuzz nastiness. I'm not fond of that. The npn is so smooth and perfect. I'm wondering if I just got freakin lucky on this one.

Now. Back to the PNP. I have a 2n1309 for Q2 gain 111 leakage 100 q1 gain 92 (ithink) leakage 101. I actually managed to hit every pin almost perfect. Q1 .7  .09  0   Q2  4.55 .57 .44   that's real close. I still notice a big difference. NPN is so lush sustain dimed out and clean glassy treb on lower settings. This is with the buffer. W/o it, it just sounds a damn mess to me. Both of em lol.  Opiniomn are like assholes tho..  The PNP is more fuzzy at all settings. I don't get that cleaner sound out of it. Like it's some what breathing. Hard to describe. That and my son won't shut up and I can't think straight.

does the value of the resisters matter at all as long as you get the voltages? I had to got down to 15k for Q1 collector. Do I have to be closer to 33? Could the breadboard and the transistor itself cause these difference?

The value of the resistors doesn't matter, though there is a small possibility of very large deviations affecting the serial resistance of the whole pedal, which can change tonal response at the output and occasionally necessitate a different output cap. (Find CJ's recent post on a useful tone calculator ... you could maybe measure the serial resistance of the whole circuit and match the PNP's to the NPN's at the output with a resistor in series with the volume pot.) If you were WAY off, I would say it was indicative of a transistor being inappropriate for the circuit. You aren't that far off. It's not like you had to use a 330K or something.

As far as 75% of the fuzz pot being cleaner tones, use a reverse taper. It's a bias control and you can test where the "fuzz" starts happening by sticking fixed resistors of varying values in there. You'll find it's usually somewhere south of 500Ohms (and really kicks in around 100Ohms), so a 1KC is a better taper in any circuit like this (e.g., the SHO's 5kC pot), since you'll get the last little bit of resistance for a lot of the pot's travel. You can splurge on the final product and get a 500OhmC, or retaper a 1K. Like I said before, I (and many other people) back off the fuzz control a little.

I can say this about the buffer: The fuzz face I use on my board always falls after a compressor (at unity volume) and I set it up to sound good that way. The Silicon Fuzz Face I actually bothered boxed up (on the Hipster board) is voiced to sound good after another pedal through a bigger smoothing cap and really low gain Q1. So I don't subscribe to the "can't have a buffer" rule. Some just sound better that way to some peoples' ears. That's why MXR puts it on a buffer on it. Heck, Skreddy gets tons of work building Fuzz Faces with good input sections for a reason. Some "rules" are just dumb if the don't get you a pedal that sounds like what you want. If yours sounds better with a buffer, stick a decent buffer in front of it when you get around to soldering.

That said, when this is all done, you might want to just build a Screwdriver if you want this family of sounds in your stable. It's got Fuzz Face character on command but works much better as an overdrive. Everyone should have a real honest-to-god Fuzz Face built and boxed, but that doesn't mean you have to like it and use it.

[marauder]

I've had good results recently using this calculator recently to get the correct resistors for a fuzz face.  Much better than with trimmers. 

I've built 4 or 5 with varying gains, and they all sound great, not gating, decaying nicely to clean.  I used RG's method to calculate the real gains of the germanium transistors, never tried it with silicon.

http://www.diystompboxes.com/analogalchemy/emh/emh.html

[9Lives]

ok guys, I f*#%d around with this thing until 4am determined to re create the sound of the first one. I just can't accept a pedal on my board on luck. I have to know how it works (is that weird?) what I found is that on my GE NPN version (all american npn) I had the bias WAY up on the trimmer. 7v. I backed off it some just bc I felt like I should.. I put it at 6.5 and left it bc that's where I like it. I also figured out the resistor network and got those pin voltages damn near perfect. That is w the 15K resistor. This is probably due to my transistors being on the higher leakage side. 300uA I think may be giving them to much credit. Never the less I found that the normal bias of a fuzz face just isn't my fav sound although I found this little adventure very fun and will continue experimenting and searching for better transistors. I thought the ac188 was the mackdaddy of the transistor world? Mine are fairly high leakage w gain from about 80-110.. Pretty typical? Is it ok to set the bias this high or am I damaging them? I think that's all ineed to make this a "case closed". Thanks your your help everyone.

[midwayfair]

ok guys, I f*#%d around with this thing until 4am determined to re create the sound of the first one. I just can't accept a pedal on my board on luck. I have to know how it works (is that weird?) what I found is that on my GE NPN version (all american npn) I had the bias WAY up on the trimmer. 7v. I backed off it some just bc I felt like I should.. I put it at 6.5 and left it bc that's where I like it. I also figured out the resistor network and got those pin voltages damn near perfect. That is w the 15K resistor. This is probably due to my transistors being on the higher leakage side. 300uA I think may be giving them to much credit. Never the less I found that the normal bias of a fuzz face just isn't my fav sound although I found this little adventure very fun and will continue experimenting and searching for better transistors. I thought the ac188 was the mackdaddy of the transistor world? Mine are fairly high leakage w gain from about 80-110.. Pretty typical? Is it ok to set the bias this high or am I damaging them? I think that's all ineed to make this a "case closed". Thanks your your help everyone.

You won't damage the transistors. 7v is Rangemaster/treblebooster voltage ... so you've basically created a Rangemaster with a feedback stage. That's why it sounds like an overdrive instead of a fuzz.

[9Lives]

hellz ya. So I'm going to put that bias trimmer as a pot. That will be a good mod for me. Thanks

[Vallhagen]

I've had good results recently using this calculator recently to get the correct resistors for a fuzz face.  Much better than with trimmers.  

I've built 4 or 5 with varying gains, and they all sound great, not gating, decaying nicely to clean.  I used RG's method to calculate the real gains of the germanium transistors, never tried it with silicon.

http://www.diystompboxes.com/analogalchemy/emh/emh.html

A little bump on this thread.

I havent bothered much so far, for getting into the universe of Fuzz Faces, but i guess now is the time, as a friend actually asked me to build him one. So, as i type, I'm browsing around the web for do's n dont's etc... The only suitable transistors i have at hand is 2N3904, so i guess i will start my breadboard work from there.

Question: The EMH calculator in the quote ^^ refers to a "Beta" value, that could be measured with a DMM. What exactly is that Beta value and how do i measure it? Leakage?

Cheers!

Edit: I guess i found the answer myself... sometimes wikipedia is helpful: Beta=Ic/Ib, correct?

[marauder]

Beta is hfe, no leakage on silicons.  I've never used that calculator with silicon, only germanium, so it will be interesting to know your results.

[Vallhagen]

Beta is hfe,

Yea, of course its hfe... sometimes the answer makes the question almost too obvious Thanx for clearifying.

***

Another "moron of the day" is that i just spent an hour measuring that hfe getting ... strange results ... til i found that i used 100k base resistors instead of 1M... hell well... I even made a Spice model to see wtf was happening, before i happened to put my eye on those ... color rings.

***

Anyway; with those hfe/beta measured on a few 2N3904, i build a circuit with:
hfe(Q1)=187,
hfe(Q2)=213,
... (and R4=100k, gain pot 1k),  gives me
R1=15k
R2=4,7k
R3=220

... thats the calculated results. I guess u were more curious about the sound results though. Ill be back!

[Vallhagen]

Some voltages after breadboarding the calculated resistor values (prev post):
Battery: 8.48

Q1:
C 1.58
B 0.65
E 0

Q2:
C 3.38
B 1.58
E 0.90


When i change to components as in the Hipster schematic (including trimpot, and go for Q2c==4.5V ), i get:

Q1:
C 1.42
B 0.63
E 0

Q2:
C 4.57
B 1.42
E 0.75

...and i am not really sure how to "translate" those numbers to the reality of good fuzz faces. In the first place, its pretty far from (actually VERY far from):

Here are Luciferstrip's voltage suggestions:

         e   b    c
Q1 -   0, .1, .6
Q2 - .45, .6, 4.5

Edit: link http://www.diystompboxes.com/smfforum/index.php?topic=98002.0

...but those values are for GE transistors. Is it relevant to compare them to a silicon based circuit?

Also, i play with pretty high hfe values on my transistors. If i understood things right, hfe for Q1 is recommended to keep low.

If my circuit sound "good" so far? Nah, not really. But if i shall compare the two setups (as above), i prefer the first one, with 3.3V on Q2c.

Now; bedtime in sweden...

Cheers