Tonefiend Forum

Great question! That circuit has always been a bit of a mystery to me. (Did you hear my Blender clone in the video comparison here?)

That pedal has a LOT of different settings — is there a particular one you dig? Or do you like the whole shebang?

Q1 and Q2 form a gain stage with frequency shaping. [Note: see disclaimer several posts later]. Probably an LTSPICE simulation of this section would be the easiest approach to understanding exactly what it does.

Q3 and Q4 and surrounding circuitry form a full wave rectifier which gives you your octave. Q4 also has a lot of gain so its output may contribute to clipping. The first 10 uF cap into the back to back diodes shifts the asymmetical rectified waveform to clip the back to back diodes around zero volts.

The rest of the passive components and Q5 form more frequency shaping and gain. Again, LTSPICE is the quickest way to understand what this circuit does.

With those sections isolated you can now ask yourself what you want to do. Change the frequency response pre/post octave? Change the gain?

Try a different rectifier circuit?

Remember that full wave rectifier octave circuits work best (most octave sound) when the input waveform is symmetrical. This means you either pick the string at the octave node point or you filter off a lot of the high frequencies (or both).

It's just electronics, you can change or replace anything you like. I can't theoretically determine whether anything you would do would sound better, so you'll have to test with your ears.

I did a SPICE analysis on the first stage and it looks like it has a gain of about 20-something dB and a low cut at about 100 Hz or a bit below. You can increase the gain by making the 47k resistor larger and modifying the feedback network, though that will also affect the DC bias point and clipping behavior. I'm pretty sure the first stage clips in general as I had to reduce the input voltage to about 100 mV to keep it from clipping during simulation. Anyway the things to think about with any gain stage are the input and output impedances, voltage gain, frequency response, and clipping behavior.

Well, when you find someone who knows what they're doing, let me know.

Hahah I was wrong! The first stage is actually a half wave rectifier with gain for clipping. I hadn't noticed that the little feedback network around the transistors also provides the base bias path for Q1. When the output voltage drops, Q1 actually turns off instead of being always in a linear zone as a straight gain stage would be.

Here's a simulation of the output (green) of the first two transistors and the input (blue), a 220 Hz sine wave at 0.1V, 0.5 V, and 1.0V.

Notice that with higher input levels, the output looks more square-wavey due to the clipping. As the input gets smaller (0.1v is about the "sustain" voltage of a guitar) then the output is more asymmetrical, just clipped off on one side.

Uploaded with ImageShack.us

Since you like the sound of this circuit with a boost or compressor in front of it, that would be keeping the output in a more square-wavey zone. So why not start with a square-wavey front end stage like a fuzz circuit? No harm to breadboard it. I think it will greatly change the character of the tone, especially as single notes decay.

OK I have simulated a bit more of the circuit, and it is pretty wild how this thing actually operates. I will put some pictures up later when I have more time.

When hit with a strong enough signal, the output of the full wave rectifier stage puts out a narrow pulse twice for every cycle of the input signal. As these pulses are not equally separated in time, you're not getting a true "octave" effect - there is still some fundamental. In fact, I don't know what it is exactly without doing a spectral analysis. Nevertheless, the half wave rectifier circuit is critical for setting the position of these pulses in the overall cycle.

Also, as you vary the strength of the signal, the width of the pulses coming out of the full wave rectifier changes. The higher the level, the narrower the pulses. I would guess that this is going to sound something like an analog synth doing pulse width modulation - a bit of a "flangey" sound following the dynamics of your playing. The amplitude of the rectified signal is heavily clipped and so it has very little dynamic range.

So, while using a "fuzz" type front end will give you a more consistent level going into the full wave rectifier, it's also going to change the spectrum a lot. I'm pretty sure it's going to give a stronger octave effect. Depending on the dynamic range of the fuzz, it may also reduce this PWM aspect of the sound.

You may find the best solution (assuming you really like its current sound) is just to fix another gain stage or compressor in front of the existing half wave rectifier. But then again, who knows, maybe putting a fuzz circuit in place of the half wave rectifier is just the formula that the music world needs today for some re-invigoration. Maybe they are both so swell that you will want to have both sounds available at the flick of a switch.