Circuit Details

Verdict

Reference

The ´Zoid Face (click to enlarge). 2022 amendments in green A silicon Fuzz Face (hereafter: FF) following an original design by PinkJimiPhoton[1] (hereafter: PJP). As different to the early designs that use germanium transistors, it uses silicon. The core circuit has been tuned by PJP to achieve better noise performance and stability over their ancestors while retaining similar tone and distortion behavior - where he succeeded in my humble opinion. Being a two-face (pun intended) unit (hence the term "schizoid"), it has a laughing and a sour face on the front, whose eyes flash up dependent on the selected mode. I named them "Jekyll and Hide" reflecting a stringent (incidentally: more modern) and a bodacious, evil (more classic FF) mode. Red indicates bypass mode. I added a few other tricks from my pocket besides input buffering to circumvent all sorts of sound degrading and impedance issues usually associated with such devices. The prototype. Sadly, a self made PCB´s plus hand wiring is in the light of today´s professional appeal almost deemed a bit amateurish. But this is tweakable, reliable and unique. A commercial PCB is none of that. It was professionals who built the Titanic, but it was amateurs that built Noah´s Ark.

I was traversing the web and stumbled over Brother Photon´s works^[1]. I was intrigued by the schizoid face, which is a  fuzz face with two modes (a polished one and an evil one) and got in contact, since recently they blur the images on D.I.Y. forae. This has semi-improved meanwhile. After building some GE devices that come wrapped with some inevitable hiss, I set out to build this one.

From past experience I know fuzzes are sensitive to driving impedance as shown in one of my earlier articles, it makes the guitar´s volume knob behave differently and makes the FF sound different. Not enough with that, the guitar signal is loaded down and loses clarity progressively. A buffer, which is normally called for rescue, is said to make the FF sound hard, but without, some other pedals will simply not interface with a stock FF.

It is not true that the FF interacts with the volume control on the guitar, in a way that, say, a high capacity cable may form a resonance circuit together with the pickup when the volume is fully up.

• True is, that the volume pot´s law (and thus its smooth and even volume reduction functionality) get´s altered by seeing a heavy load (faster volume decrease per rotational angle).
  
• Also true is, that the FF´s input is current driven, which if driven by a much smaller impedance than a pickup would represent, sounds different.
• Also true is, that the signal loses treble when the volume is dimed, but this has nothing to do with this circuit.
  

I personally like to eliminate imponderabilia like those wherever I can. I am sure Hendrix would have done so, if just the technical aspects were as known and mature as they are today. Indeed we do not know what his engineer, Roger Mayer, actually installed into those cast iron shells. And I am sure he keeps some sweet little secrets.

So I like to buffer my faces. Fortunately, this is straight engineering applied. Even more fortunate, it is easy.
It turns out that this FF unit, GE or not, sounds and behaves much like old GE units you can hear on the internet.

Copyright Notice: The circuits and concepts depicted here, particularly the artwork, are the sole intellectual property of persons known as PinkJimiPhoton (hereafter: PJP) and Aquataur Music. They are destined for personal use and for educational purposes. Any commercial exploitation thereof is thus at least immoral. Any re-hosting is prohibited.

Back To Index

---

Circuit Details

updated

Buffer (click to enlarge): the input buffer is a straight copy from a Joe Gagan pedal, which is by itself a straightforward emitter follower. Any buffer would do by the way. (Note: there is an updated schematic available, V1.4) Now, as I explained elsewhere, a fuzz face does not like to be driven with a low impedance. We talk about a few hundred ohms or even less. Fortunately, a 10k resistance or thereabouts in series with the buffer will yield exactly the same drive conditions as a typical pickup will exhibit. On the board, I have provided a jumper that connects either the raw or the buffered signal to the FF´s input for quick A/B checks, but did not even bother using it since it sounded great from the start. Input bandwith: I added the ubiquitous anti-pop resistor in front and a low pass filter that kills any RF interference. The signal bandwidth into a electronic device should always be restricted to the bandwidth of interest. Any pedal should have that and you would be surprised, how many designers of commercial units oversee this, for the sake of saving ... almost nothing worth the mention. Hendrix went mad over this during a concert, but what has changed over the decades?

2022 changes here:
Transistor choice: from my axis face silicon project I had some PN2369A transistors left over (hfe=60), that appeared to work best in pos1. For pos.2 a 2N2222  was lying around (hfe=190) which turned out to work great. Some other, lower gain type, appeared more trebly and less smooth. This is really a matter of trying, but as explained on Fuzzcentral and in R.G. Keen´s Technology of the Fuzz Face, the magic lies in the gain factors. Even a medium power transistor like BD139/140 does work, those have lower gain.

I used a b-c cap of 22pF (Styroflex), 47pF appeared to big. The bias pot appears to settle at 3.5kOhm for greatest clarity.

Bias Emitter Resistors: I fiddled around with these, and ended pretty much with slightly different values for the hfe values chosen. This resulted in a value of 120 Ohms for the first stage and 82 Ohms for the second stage.

More insight^[3] revealed that the above choice was working albeit being generic, but it was not what PJP was driving at (reply #9). I finally populated the circuit with exactly the values PJP designed for it (look at schematic V1.4).
The obstinate collector won´t bias with the values given. I tweaked the trim pot until I achieved optimum symmetry on the oscilloscope. I concede that I am not sure if I tried hard enough to get the thing working with the emitter resistors. 

PJP recommends the following biasing procedure:

"(...) [using a buffer] in this case, its [the FF circuit] preset to stay fairly sweet, but i agree, won't necessarily clean up completely. if ya wanna do that, what i do is replace the c resistor on q2 with a b10k trimmer, then turn the guitar down to where it just kicks in, maybe around 1 on the guitar knob... then adjust the bias for the "cleanest" sound i can get. usually that gets you a nice overdrive around 4 or 5 and crank it right up to its full fuzzy goodness." -Pink Jimi Photon: on freestompboxes.org[2]

Supply Tricks (click to enlarge): I tapped the LED´s supply after the polarity protection diode. Very many designs supply the annunciation LED(s) right off the battery, which in case of an accidentally reversed supply polarity may mean the death of the LED. Take into account that some LEDs, such as the RGB types I used, can only withstand reverse voltages up to 5Volts! Supply filtering: I also made R/C filters akin to valve amps, decoupling subsequent stages. I learned this trick with guitar poppa, this is cheap but optional.

Bypass:  (click to enlarge): A stock true bypass switching scheme (as depicted and used) grounds the board´s input during bypass to stop noise buildup on the output that may corrode the bypassed signal (red arrow). Note that PJP originally deviated from this, because he used the switches´ pin to activate the red LED´s during stand by, which is reverse to what may be customary, but eminently valid as an artistic expression. It looks way cool.

LED face frontplate:  (click to enlarge): The pin that during "effect on"  provides a ground line to the LED(s) (signal ENA, short for enable), is fed to the second switch (Photon Face / Stupid Face as he called it), which in turn powers the green LEDs resp. the blue LEDs selectively. Since at any time during "effect on" one of these lines is low (both high in bypass) it was obvious to make a diode-wired or type circuit to drive a transistor that illuminates the red LEDs in this case. Voilá, full bypass is restored.

• Axis Face Silicon (an all silicon FF that claims to sound like GE)
  
• Tarpit Meltdown by Joe Gagan (Nine Volt Nirvana), which I built earlier (yet to be documented). This is a hybrid FF at its heart, with a GE transistor in pos. 2.
• Brontoboost (NVN too, GE rangemaster type)
  

All three FFs can´t deny their common heritage and are obviously in the same ballpark, with the Axis Face Silicon being a bit colored. The ´Zoid Face however is very flexible and sounds totally uncolored. An extension of the guitar. I´ll demonstrate that in the sound samples. It cleans up reasonably well, maybe faster than the GE type and has very many nuances of timbre you can adjust with the guitar volume.

The Brontoboost is different since it is built on a different basis. So GE alone is not all. It is the FF dirty goodness.

First, if you are using any FF type circuit, a bass cut facility on the guitar is strongly recommended. This is not stock on most guitars, but very helpful for overdriven sounds. It helps to dramatically clean up distorted tones. Low end gets in the way of the bass guitar anyway.

Secondly, never run a FF into a clean amp. It will sound brittle and awful. Try it. Either you push your amp over the proverbial onset of clipping through sheer volume, or put a mild, tone-shaping overdrive of your choice after the FF, which emulates exactly this. The overdrive, as insignificant as this may be,  does something beneficial to tone. For the following sound samples I use exactly that. You may want to read how to use your fuzz right.

So bear in mind, any recording you hear on the web, always reflects the whole sound chain, which is where of course those guys are cheating you. There is nothing constant and thus comparable.

Sound Samples
The recordings have been done using the following setup and no further processing:

• Strat with Bill Lawrence (Wilde Pickups) microcoils, neck pickup
• J-FET buffer directly after the guitar
  
• ´Zoid Face :-)-:     Max volume (when guit= full) is set slightly louder, so that the following stage(s) get thumped.
  Equal loudness will be met when guit vol = dimed and FF cleared out. Just like in real life situations. Fuzz control= 12
  
• Peavey Bandit 112 silver stripe, clean channel, Resonance on, T-Dynamics fully down, volume=9, all others=12
  
• Wampler Euphoria Overdrive, setting = natural. Gain = 1. Slight crunch appears when guit. vol = full and heavy playing.
  
• Rode M3 mic @ 20cm distance
  
• Recording device: Focusrite PC interface into DAW
  

(Names may be copyrighted by the associated copyright holder, no association with any of them)

A feq quick and nasty takes. No exercise in timing or else. Crank you speakers.

clean guitar into clean amp, (after pause) with overdrive at the onset of crunch. Hardly any mangling by the overdrive, yet it is felt. Guitar volume = full.

 

stupid face (blue- Jekyll side) :-( , volume at 15 o´clock (18 being max), note long clean decay. Cleans up well.

as above, volume at 17, last measures full throttle.

photon face (green- Hide side) :-) , full volume, second half with bass cut on the guitar.

• (...)
  
• ´Zoid Face :-)-:   (...)
• no overdrive
  
• Peavey Bandit 112 (...),  all controls=12 (loud!)
  
• (...)

crazy jimi jam on the photon face (green- Hide side) :-) , riding the volume, slight bass cut. Note the way the amp gets fatter when the loudness increases. This is straight into the amp without sound shaping overdrives.

A few tests regarding buffered / unbuffered and an alleged similarity with Tonebenders MK2:(...)

• Buffering as specified
• Clean Boost as specified
  
• ´Zoid Face :-)-:   Stupid face = Jekyll face, all other settings as above.
  
• Peavey Bandit (...),  volume=9, all others=12
• Wampler Euphoria Overdrive (...)
• (...)

FF stock (no buffers). Guitar= full volume (18 o´clock), half ways to 15 o´clock. Typical altered volume pot behavior. Note: this is not the fuzz face, this is the volume pot who´s law gets changed into somethink like super- logarithmic. Note also the typical loss in treble due to the cable capacity.

FF core driven by  a low impedance buffer (built in buffer disabled). Guitar= full volume (18 o´clock), half ways to 15 o´clock. Standard volume pot behavior restored. No loss in treble. Gets into Tonebender territory.

FF driven by built-in 10K buffer (and cable buffer). IMHO sounds and behaves not different than unbuffered, except pristine signal and functional volume pot travel.

FF alone, no buffers. Halfway FF thumped down with a clean booster in front. Hard drive. (Signal amplitude of booster way to big to be recorded). Note how little the volume increases. Surprisingly not too much change in tone. Gets definitely into TB territory.

SHAME. I fell for the age old trap.

"Louder sounds better" - age old axiom and amongst audio circles well known wisdom

I turned the  Fuzz control down and lo and behold... apart from a little difference in loudness, both variants sound largely the same! Explanation: the 10k resistor with the 2k input impedance forms a voltage divider, kind of an input attenuator. The FF stage gets hit way harder. Another theory up the swanee.
Back To Index

• May 3, 2022  updates marked in green
  
• Feb 6, 2020  update section on Mk-II switch
  
• Jan 22, 2020  fixed switching pops caused by LED inrush current
  
• Jan 20, 2020  minor functional updates
• Jan 19, 2020  first release