Foolish octave question

[ethrbunny]

I have been reading about octave pedals and as they say, a little knowledge is a dangerous thing. The 'green ringer' on tonepad appears to be a full wave rectifier. The octave-up sound is the negative side of the signal 'flipped up' to simulate a doubled sound. Truth?

So what does a half-recitified signal sound like? Would you get the octave-down equivalent of the above circuit?

The obvious answer being - trrrrrryyyy it... just curious to know what the 'smart person' answer would be as to why its not done already.

[Jaicen_solo]

I thought about this for a realllly long time. It's possible to get a similar sound by mis-biasing a transistor so that it can't swing rail to rail.
Basically, as you'd imagine it sounds like a ripped speaker. I think the Bazz Fuzz does something similar actually.

[Mark Hammer]

I have been reading about octave pedals and as they say, a little knowledge is a dangerous thing. The 'green ringer' on tonepad appears to be a full wave rectifier. The octave-up sound is the negative side of the signal 'flipped up' to simulate a doubled sound. Truth?

So what does a half-recitified signal sound like? Would you get the octave-down equivalent of the above circuit?

When you have a bipolar NPN transistor with equal-value resistors from collector to V+ and emitter to gnd, the outputs of the collector and emitter are identical but opposite polarity.  This is referred to as a phase splitter, and is found quite frequently in octave-up boxes.  If the phase-reversed outputs of the transistor are combined in selective and clever ways, one can end up with an apparent doubling of frequency.  The cleverness often consists of "shaving off" one half cycle of the waveform/signal with a diode.  When the *same* half cycle of two phase-reversed copies of the same signal are combined, the "dip" of one version is replaced with the "peak" of the other.  That's how you can get double the number of "peaks" in the same time period and an apparent doubling of frequency.

So what happens if you don't shave (or produce) BOTH copies of the waveform/signal?  Simple.  You end up with a waveform that is exactly what you fed in, except that:

1) one half cycle is essentially squared off to DC by the diode that prohibits AC of that "direction" from passing, and

2) Since you have a full waveform in one direction (for one half cycle), but the other half cycle stops abruptly at ground or Vref, you end up with a waveform that is essentially half the amplitude of what you fed in.

Incidentally, when discussing envelope controlled filters, I often mention what I like to call "undersignalitis".  This is the phenomenon whereby the envelope-follower signal that the instrument generates is insufficient to make the control element enough to produce an audible filter sweep.  Since simple half-wave rectification results in a waveform that is approximately half the initial amplitude, you can easily see how this happens, and why it can sometimes be necessary to increase the gain in the follower gain stage.

If a little knowledge is a dangerous thing, then I'm heading off to the police station to register myself as a lethal weapon! :lol:

[Jason Stout]

The quick answer is unfortunately, no. The half wave rectified signal still cycles with the same frequency as its source, just without a positive or negative half cycle.

[ethrbunny]

I guess if you start with a half rectified signal and then amplify it you would get the correct amplitude on half of the wave and something less on the other half. Would be a strange sound indeed.  

What if you half-rectified it again?  Ran it through an opamp follower with gain of 2 and a diode?