Question about R8, MXR Blue Box

[tsr]

Can someone tell me what R8 does in the MXR Blue Box circuit and how it might cause the circuit to behave differently with different guitars/pickups?

I built this circuit according to the Tonepad layout using 470k for R8.  As is, I had a continuous sputtering even though all intended notes had long decayed (using a G&L F-100 with stock MFD pickups).  Hand muting the strings was the only way to prevent the sputtering.  When I reduced the value of R8 to approx 100k, the unintended sputtering stopped on its own (without having to mute the strings).  However, when I then used a FAT Telecaster, the sputtering returned.  I had to decrease the value of R8 even further to eliminate the unwanted noise.  I now have a 500k pot subbed for R8 to account for these apparent differences between these two guitars.

I would really like to understand what is going on here so any comments would be appreciated.  I am using a 4558 for IC1 and a 270pf ceramic cap for C3 (schem. calls for 300pF).
Thanks.  


TSR

[Mark Hammer]

R8 helps to set the gain in the first op-amp stage.  Gain in a non-inverting op-amp is set by the ratio of the feedback+ground resistor (R5 + R8 in this instance) to the ground resistor (R5).  Here it is set to 471 (give or take the tolerances of the two resistors).

That stage needs a lot of gain in order to drive the envelope follower formed by that little subcircuit you see with the two diodes and the resistor+capacitor to ground.  The envelope follower is part of a simple noise gate, formed by the two transistors around the Blend control.

Essentially what happens is this.  When the input signal is sufficiently high to result in relatively error-free octave division by the 4013 (meaning that the signal is above the required threshold for the flip-flop to do its thing), the two transistors are "turned on" so as to permit the fuzz and octave signals to reach the output.  If the input is insufficient and likely to result in flip-flop sputtering, then the envelope follower does not (or in theory SHOULD not) drive the transistors hard enough to allow that sputtering to be heard.

It is not at all unreasonable to replace that 470k resistor with a 220k resistor and a 1meg pot in series to permit adjustment of the input sensitivity so as to compensate for a wider range of input signal levels.  This would permit the existing noise gate circuit to function optimally.

Incidentally, I find it conspicuous that the two transistors involved are set up to respond identically to the envelope signal.  Although the octave side needs to be able to stay silent during the "sputtery parts", the fuzz side should not sputter at all.  In that instance, the gating/signal-fading on that side is solely to eliminate the noise stemming from ultra high gain, as well as to make the fuzz and octave always start and finish at the same time.

The question is, do they always NEED to?  If they don't, is it worth tinkering with the collector resistors (R13, R14) to make the octave fade out before the fuzz?

[tsr]

Thank you very much Mark.  I appreciate the thoughtful answer and discussion.

TSR