popping on diode clipping switch

[Mark Hammer]

The premise is that the diodes provide a viable path then it's only the diodes and cap.  When there is a big resistance i series, then the diodes are essentially ineffective (i.e., it's like there's an open circuit instead).  So, the toggle is intended to render each diode pair in circuit by bypassing that 1M resistor.

But, but, but....that's what you were trying to do in the first place!  That's true, but with nothing to let the cap drain off, reconnecting a diode pair would result in a similar rush of current, much like a "floating" input or output cap on a pedal.

Of course, now that I reflect on it, the 1M resistor will only allow some of the excess charge in the cap to drain off, because once the charge drops to a value below the forward voltage of the diodes, the cap is as good as disconnected and floating.  And, since you're using a pair of LEDs, that's around 1.5V just sitting there waiting for a chance to move those electrons.  My guess is that switching over from the LEDs to the silicons would get you a bigger thump than the opposite direction.

RG's solution is a better one (which is why he does this for a living and I just do it to make annoying noises at one end of the house).  The 1M resistor in parallel with each diode pair provides a drainoff path which does not depend on the residual charge in the cap reaching some threshold.  My guess is that if you did a kind of compromise between what I suggested and what he suggested (i.e., 1M in series and in parallel with each diode pair), you'd get something a little quieter than my original suggestion, but not as quiet as what he eventually comes up with.

Taking a few steps back, though, were starting to get to a point where building two different versions of the single transistor circuit, with different diodes in each, and selecting between them, will involve fewer components than building one and getting it to switch diodes quietly.

Just a thought.