Fuzz Face Revisited

Background

I already tried making a fuzz face before, but back then it didn’t fit my playing and my gear very nicely so I sold it. However, after discovering love for Les Paul and more vintagy tones, I decided to give fuzz another chance. I liked the way Bonamassa fuzz sounds on youtube clips and even more I liked the idea of Russian germanium transistors because they are cheaper an according to fellow DIY-ers much more consistent than what’s left from the western world germanium transistor supplies. So I got a standard Fuzz Face PCB made using the nice and compact Tonepad layout and ordered a bunch of Russian (and some Serbian!) PNP transistors to try. Tonepad layout can work for both PNP and NPN circuits, but all the transistors I got were PNP, so I wired it in PNP mode.

fuzz_schem

The circuit is deceptively simple: two transistors, four resistors, three capacitors and two pots. But huge variations in transistor parameters and sensitivity to component choice make all the difference between an exceptional and bland fuzz build. I omitted the 10ohm (or 100ohm) resistor between fuzz pot and 22uF cap that prevents oscillations because I couldn’t fit it on the board. It turned out that I didn’t need it because there’s no sign of noise or oscillations. Other than that, Bonamassa is pretty much standard Fuzz Face with Russian transistors.

Transistors
Here’s what I’ve got:

  • Russian MP39B [datasheet]: germanium PNP, relatively low hFE (20-60) used in Dunlop Bonamassa Fuzz at Q1 position. Out of 20 that I bought, there was not a single one leaky or even close to what’s considered a leaky transistor. Current leakage was in the range of 20uA.
  • Russian GT308V (ГТ308В) [datasheet]: germanium PNP, don’t confuse cyrillic В with latin B because it’s actually equal to latin V. Russian transistors are sorted by hFE, but using cyrillic alphabet: A (A), B (Б), V (В), G (Г)…The V variation has hFE ranging 80-150 and is used in Dunlop Bonamassa Fuzz at Q2 position. I got 50 of these and less than 10% had leakage in the neighborhood of 100uA (which is considered great for AC128) and the rest had much less, less or close to 50uA. Only one of transistors was in the 110-120 range, all the others were lower with majority having gains around 100.
  • Russian 1T308B [datasheet]: germanium PNP, similar to the previous but lower hFE (50-120) and 1T instead of GT (ГТ) indicates that they are made for military which means higher quality. Being higher quality, none of them were leaky. Also, even by looking at them side by side you can tell that 1T is made better than GT. The case is nice and shiny, while GTx develop some sort of patina. Probably doesn’t affect the performance, but it’s and indicator that more effort is put in the 1Tx. I noticed the same thing with Russian military tubes (-EV suffix), they look much better than regular ones. Most of them had low gain, around 50-60 or so.
  • Serbian Ei BC179 [datasheet]: silicon, PNP equivalent of BC109. I was looking for a PNP silicon transistor to mix in hybrid circuits together with Germanium transistors and BC109 seems to be a popular choice when it comes to silicon NPN fuzzes. I looked up PNP transistors with the same specs and BC179 was the answer. I bought these in a local store for around 30 cents each and they seem to be made in the city where I live which is awesome. They probably don’t make them any more because Ei is bust, but it’s still cool. They are not rated for gain (A/B/C are standard suffixes for BCx transistors), the one I ended up using has hFE 120 which is considered perfect for Q2 position. I measured it only after realizing that it sounds perfectly, so theory is confirmed in practice.
Parts
  • NKK 4PDT bypass switch. I use it to bypass the fuzz and kill the battery at the same time. Since it switches DC there’s an audible pop, so bare that in mind. I don’t play live so it’s not a biggie. It’s easier for me to switch it off rather than having to unplug the input jack.
  • Cliff UK jacks
  • One CTS 500K log pot for volume and one Alpha 1K lin pot for fuzz level (that’s what I had in my bin). If you can, use a reverse log pot for fuzz level because all the magic happens in the last 20% of the pot travel. Reverse log will spread the usable range across the whole travel of the pot.
  • Sanyo solid organic polymer 22uF electrolytic. I try to avoid standard wet electrolytics, and these should last longer and have lower ESR.
Construction

As indicated before, the board is PCB, but I used gold plated pins as sockets for pretty much every component on the board, except for the 100k resistor and the 10nF output capacitor. That allows me to change transistors, bias them and play with the input capacitor. Also, I can use jumpers to select PNP or NPN operation in case I want to experiment more with silicon transistors. Since pins are very thin at the base, I epoxied pins and all wires to the board as strain relief. It should help keep them in place after dozens of components swaps.

So how do those transistors sound?
Very differently, especially Si compared to Ge, to summarize:

  • Ge sound warmer and clips much softer than Si, producing somewhat hi-fi sounding distortion. Silicon transistors sound much harsher when distorted.
  • As you go higher with hFE, the sound is more fuzzy and harsh.
  • When playing with two Si transistors there are noticeable noise artifacts when note decays. At some point as the note dies off some weird almost digitalized noise occurs. It’s kinda annoying.

So going from mellow to harsh and clean to fuzzy, I would sort the available transistors as follows: MP39B < 1T308B < GT308V < BC179

What about the input capacitor?

The stock value of 2.2uF is chosen to pass pretty much all audible frequencies through. Lower values may be used to tame the bass and shape the sound of the fuzz. Letting lower bass frequencies go through a fuzz makes it sound more woolly, while reducing bass makes it sound more focused and sharp. Both can sound good in their own right. I experimented with values ranging from 10nF up to 1uF (didn’t have 2.2uF to try). Lower values make the fuzz sound almost like an overdrive pedal because it removes all the woolliness from the sound. Values as low as 220nF may be used if there’s too much bass, but it also depends on the amp and pickups. I ended up using 1uF as an in-between choice. It has some of the woolly qualities of old fuzzes, but it’s not too muddy.

And about bias?

I started off with values measured in the original Bonamassa fuzz and that’s pretty much where I finished after fiddling with bias voltage. Q1 is biased very low, around -0.8V and Q2 biased little higher than the midpoint, around -6.5V. Higher bias gives a tad more output and more fuzz. If your build is too fuzzy you may want to back down on Q2 bias before cashing out money for another transistor. Also, I tend to play quiet in the apartment, people playing louder would probably want something warmer. Everything sounds warmer when played quietly 🙂

What did you end up using?

Choosing the right transistors (and everything else) is very much a matter of taste. Any combination will work, but it’s up to you to decide what works best for you and yout guitar and amp. As far as my choice, I was totally blown away by hybrid between a Russian Ge transistor in Q1 and BC179 in Q2. It truly offers best of both worlds. Thanks to the Ge transistor there’s warmth and softer clipping, but silicon transistor brought in clarity and a bit of an edge and aggression. In a way, it combines Joe Bonamassa with Jimi Hendrix and a bit of Eric Johnson Fuzz Face. It still responds nicely to playing dynamics and guitar volume control. Also, there’s no annoying decay noise! Every Ge/Si mix I tried sounded good. Using MP39B will yield a warmer tone with less fuzz. For me, the winners are 1T308B in Q1 for humbucker equipped guitars (used my Les Paul mini) and GT308V for single coil equipped guitars (tried with my Strat). Both are brighter than MP39B, but GT308V has more gain and works better with lower output single coils. It felt like 1T308B and MP39B were lacking fuzzy goodness with a Strat. On the other hand, GT308V may have too much gain when used with higher output pickups, so sound may be too distorted and very woolly. Once again, I urge you to try as many combinations and find the one that works the best for you.

Useful Links

R.G. Keen’s
Technology of Fuzz Face is a must-read if you’re considering building or modding a fuzz face. It has tons of information on the circuit, how to test and select transistors, some mods, etc. Highly recommended read.

Comments
21 Responses to “Fuzz Face Revisited”
  1. ON says:

    Hi, as earlier mentioned, the schematics is wrong. Q1 emitter should be connect to ground. And its collector to 33k. That is why called as grounded emitter or common emitter stage. Both transistors emitter and collector is exchanged, pls fix it.
    🙂

  2. Jake says:

    Hi,
    Where is the ground on the circuit and where does it connect to?
    Thanks

  3. Jake says:

    Hi,
    Where is the ground on the circuit and where does it connect to?

    • Bancika says:

      both jacks are grounded to the back of one of the pots. Circuit ground goes to one jack’s ground.

  4. Csaba says:

    I have just made this schematic on breadboard. For Q1 I used MP42B for Q2 1T308B. I like how it sounds, but seems like it needs a booster after the efect. The bypassed signal is much louder than the signal fron the fuzz. How can I fix that?
    Thanks

  5. Darren says:

    The transistors in your schematic are up-side-down. Collectors should be on top, emitters on bottom.

  6. Brian says:

    Hi! I am trying this combo out in one of my pedals and I’ve found that most of my BC179s are extremely high gain, like 400-550. I also couldn’t find a spec sheet for a 109 that made it seem like 120ish was normal. Thoughts?

    • Bancika says:

      Could be, I might have gotten a black sheep, but it works well 🙂

      • Brian says:

        So I think it’s because I got the BC179C designated trannies. They did not specify this on the order, what a bummer. It looks like the A version should test around 120. Oh well! Got any of your 179s laying around you’d sell? 😀

        • Bancika says:

          I don’t have any left. I could buy you a few if you want, but shipping from here would cost more than trannies 🙂

          • Brian says:

            Hah! Well, where are you located? And if I can’t find any B179A’s, even online, I’m out of luck anyways! What do you think post would be?

      • Steve says:

        Thanks for posting this comprehensive comparison of these transistors. I’m just about to start on my build using these Russian transistors.

  7. Petter says:

    Just got to say that is one neat wiring!
    Also, very interesting reading as I recently had a russian friend import some GT108G:s for me. I have to try them out soon.

Leave A Comment

  • About

    The idea behind this site is to share my experience with Do It Yourself approach to guitars, amplifiers and pedals. Whether you want to save a couple of bucks by performing a mod or upgrade yourself instead of paying a tech, or want to build your own piece of gear from scratch, I'm sure you will find something interesting here. Also, this is the home of DIY Layout Creator, a free piece of software for drawing circuit layouts and schematics, written with DIY enthusiasts in mind.