Diodes for an octave effect?

[ibanezts808]

If I were to take a standard diode distortion and add another diode, so I would have 2 diodes one way, 1 diode the other way, would this create an octave effect?  I'm still learning, so I'm almost positive that I haven't provided enough of a desctription of the circuit to say whether you could or not.

[niftydog]

if only it were that easy. I think I know where you're coming from though.

think of a sine wave. The frequency is the reciprocal of the time inbetween two adjacent peaks (or two adjacent troughs).

Doubling this time halves the frequency.

No amount of diode clipping will have this effect. Even though you might clip the peaks AND the troughs, the time inbetween them remains the same thus the frequency does also.

[Alex C]

 
(here if image doesn't load.)
If this is what you're talking about, then no, an octave effect will not be produced.  Rather, this will cause what is called asymmetrical clipping.  

Diodes are indeed involved in most analog octave effects.  
Check out this thread, in which I asked a similar question.  It contains an amazing post by R.G. explaining the topic.

[Mark Hammer]

The "classic" method for delivering an octave effect is to take an original and an inverted version of the input signal and "chop off" half the waveform of each version with diodes.  Since the diode only conducts in one direction, that means that if you were to view each waveform on a scope, instead of going up and down symmetrically, it would go up, then go splat against ground/Vref without a complementary negative-going half-cycle.

If we the inverted copy of the original and do the same thing, something interesting happens.  You know that negative-going half-cycle we just chopped off with the diode?  Well, because the signal is inverted, that means it is on the "good side" of the diode, and not the bad side.  So, what gets chopped off is actually something we already have fromthe other copy of the signal.  When these two versions of the signal get combined in equal measure, instead of having a signal that goes up then down, we have something that goes up, back to ground/midpoint then back up again instead of going down.  Wth two peaks occurring during the same time period when we would have had just one, the frequency is essentially doubled.

So, to go back to the wrong turn in your query.  Yes, the diodes are crucial.  In this application, however, their function is not so much to clip the level, as they do with a fuzz, but to separate the two half-cycles of the input waveform, in much the same way that some recipes call for the cook to separate the yolks and whites of eggs so that they can both be processed separately for a different result than simply using the whole egg at once.

[niftydog]

mmmmm, egg yolks... [transcendent gargle]

Nice analogy Mark. You're making me hungry.