The germanium Fuzz Face was the first iteration of the famous Dallas Arbitor Fuzz Face pedal that was originally released in 1966. I think that this pedal's longevity is due in part to its very simple design. There are four resistors, three capacitors, two potentiometers, and two transistors...that's it. The original resistors were 1/2-watt carbon-composition with a tolerance of 10%. The input capacitor was a 2.2uF electrolytic axial lead. The capacitor from the wiper lug of the fuzz pot to ground was a 20uF electrolytic with axial leads. The 0.01uF capacitor that connected to the third lug of the volume pot was metallized polyester. The original versions of the Fuzz Face were equipped with PNP Germanium transistors. There were at least three types used: NKT275, AC128, and SFT363E. These three transistors are quite hard to come by these days. There was a problem however with these early germanium pedals, the transistors would fail if the ambient temperature got too hot, this however wasn't an issue with the later silicon pedals.
Quite a few months ago when i was building the silicon Fuzz Face pedal Jon Clift sent me 2 old germanium transistors. These transistors are from the 60's I believe and it's thought by Jon that these transistors would be good built into a Fuzz Face pedal. A while ago I built a silicon transistor based Fuzz Face pedal, and said that in the future I would make a germanium based fuzz, well this seems like an opportune moment to build a germanium fuzz and try the transistors Jon sent me. The transistors are old Sony transistors, one has a gain of about 30 and the other about 50. They're both NPN. The only information Jon could find about the transistors on-line says that they are germanium. It looks like that's right - the base-emitter voltage using the meter's diode test is 490mV in one case and 530mV in the other, which is lower than you'd see with a silicon device.
For completion, here are Jons observations from when he tested these germanium transistors on his scope, reproduced from the Silicon Fuzz Face blog post.
The germanium transistors seem to work ok. Here are the waveforms. The yellow trace is the input - that's just a 200Hz sinewave coming from a waveform generator. I'm then feeding it through a 4k7 resistor to simulate the winding resistance of the pickup coil before the 2.2uF capacitor. The blue trace is the output. I didn't bother with the 500k level control at the output, so that is the level at maximum. I increased the size of the 330R resistor to 2.2k because the signal level was so low.
Here's what the same circuit does with 2N3904 transistors. [Note that the vertical scales are different.] These are silcon, with a reasonable fT, though the gain is middling (about 220, or something like that). This is back with the 330R.
This time there's enough gain to precision clip. The 1k adjustment pot changes the slope of the edges (which affects the high frequency content).
Jon then made a further post, as he'd made an incorrect assumption with his initial posts around the output capacitor................
I made a wrong assumption with the above posts. I assumed that the output coupling capacitor would pass the signals unchanged driving the 500k pot, but it doesn't. The time constant is so short that it's acting as a differentiator. So the traces above, which show the signal at the junction of the 330R and the 8.2k resistors, don't represent the output of the pedal.
I've added the 10nF cap and a 330k resistor to ground (to simulate the pot and an assumed 1M input resistance to whatever it feeds) to my breadboard and here's what the output looks like with modern transistors. I've superimposed the waveforms for both extremes of the 1k pot so you can see what the adjustment does.
The main disadvantage I see with this circuit is the way that it's so dependent on what is feeding it and what it is driving. The base feedback clobbers the input waveform to some extent, so the sound will partly depend on the guitar that's driving it, how many coils it has and how they're wired and so on, and the small coupling cap at the output means that the waveform shape also depends on the input resistance of whatever is being driven.
Jon also had some old silicon Mullard transistors which he also sent me. These may be the subject of another blog later on, but for now onwards and upwards........ Lets build this pedal and have some fun.
For this build I'll be using the same schematic and possibly PCB layout I used for the silicon Fuzz Face. I'll also wire it as per the silicon, as these germanium transistors Jon sent me are NPN, not PNP as most germanium transistors are that are used in fuzz pedals. i'm really interested to hear how this will sound once finished. Depending on the pin outs of the transistors will dictate what board I use, I don't want to do too much twisting of legs so will use the board layout that will suit the transistors. The issue 1 board was created for the early germanium transistor Fuzz Faces, the issue 2 board was developed for the later silicon fuzz Faces. You can see that the base and collector are swapped around on the 2 boards, so the transistors likely will suit one of these boards. Either way I'll follow the schematic for the silicon transistor Fuzz Face as the transistors are NPN the polarised capacitors will be correctly orientated on the silicon schematic. If the transistors were PNP the polarised capacitors would have to be turned around on the issue 2 board.
Once again, as before I'll etch my own board. I've found it easy etching small boards, but when you get into the realms of boards around the 3" x 3" size and larger distributing the heat across the board evenly is a lot harder. I generally end up with some tracks missing, etc which require a bit of precision Sharpie work. Looking at the paperwork that Jon thoughtfully sent with the transistors he's advise me that the pin outs are CBE so I'll use the Issue 1 board, as there'll be no leg twisting involved. On with the etching.
I use the toner transfer method, that is I laser print a reverse image onto photo paper, then iron the image onto the PCB. After that I soak the PCB in boiling soapy water until the paper loosens and then peel the paper off carefully. Once it's been cleaned you end up with the board in the 3rd picture...... Then into the Ferric Chloride to etch it. I normally use 50% ferric to 50% water... Not sure if that's right, but it works for me. I have known people etch in 100% ferric but i'm not sure whether that'd be too strong or not. Once the board has finished Etching i'll drill it and then start populating it.