-/+ swings of signal treated differently

[petemoore]

  I was thinkin' again..uh oh...
  Ok With that prelude, anything goes...
   First of all how hard is it to split the negative and + swings of a source?
  Can it be done cleanly or with a nice fuzz fairly easily?
  See, I was thinking of takng the + swings, and 'treat' them to something interesting, then have the _ swings 'treated' also, but differently...
  Anything like this? I guess it's hard, or offers limited returns for some reason.

[bajaman]

I was thinking of that idea to here in NZ.
Should be possible by half wave rectification and then mixing the two  + and _ swings at the output.
I may just get round to it sometime!!!
eg. use silicon clipping on one half and germanium on the other.
Good luck!

[moosapotamus]

Somone else was thinking the same thing, too...
http://www.maindragmusic.com/js/two_face.html
Not sure what or how he did it. But, I'd guess that the bipolar power supply is good clue.

~ Charlie

[petemoore]

  I was thinking...
 Envelope Control 'set high' for _ swings / Env Control 'set low' for + swings or...
 An inductor for mid notch or boost on _ with + having a bass peak.
 I don't know how 'reasonable or feasable all this is though...
 Just shaping the -/+ swings isn't too hard...as far as levels of clipping like in signal/ground clipper...and of course the _ and + clipping is a place where some voicing could be done with HP/LP or whatever filters on the opposing clip diodes....could just be the long way around to the same thing.

[petemoore]

  And it actually looks really good too !!!
  Nice batch of knobs to dail .. and I notice the one switch is labeled "Split / Link"

[brett]

Hi Pete
The RM Octavia and other octaves split the signal and invert one half (thereby doubling the peaks).

In overdrive and fuzz circuits, there are several circuits that treat the pos and neg cycles differently.  The boss DS-1 and tubescreamers with the "asymetric" distortion option have much less clipping on one side of the signal (achieved by having 2 diodes in one direction and only 1 in the other.)  I personally really like this effect, and find it very "tube-like".  Even on a Distortion plus, Bosstone or a Rat can be improved by having uneven clipping.

Finally, there's the fuzzface.  Because the emitter of Q1 is connected to ground, and the base sits only 0.3V or 0.7V above ground (depending on whether the transistors are Ge or Si), the negative side of the signal gets clipped much earlier and less than the positive side of the signal.  I suspect that this is one reason why everyone really likes Ge transistors - they will give more asymetry.

Keep digging Pete - you're a legend, and there's still a million great ideas out there.

[Paul Perry (Frostwave)]

If there was room for one more switch, a cool thing to have would be an inversion switch for one of the fuzz outputs before the recombining stage. (only need one transistor & a single pole switch for that).
And if there was room for one more pot, I'd like a blend pot, adding the original signal to the output, in a variable amount from -1 to +1 times (so you could scoop out some of the fundamental).
And the next TWO switches would bypass one or both of the half wave rectifiers.......
Seriously though, looks a great idea!

[amz-fx]

In overdrive and fuzz circuits, there are several circuits that treat the pos and neg cycles differently.  The boss DS-1 and tubescreamers with the "asymetric" distortion option have much less clipping on one side of the signal

You can take this a bit further by using a "warp" control as shown in my article at:

http://www.muzique.com/lab/warp.htm

Also, if you have my CD, the Fet Muff shows a method of clipping where gain stages in series each work on one side of the signal swing, and can be switched out for different effect.

regards, Jack

[petemoore]

  Ying/Yang Box...
  Always good to see what your'e better half is up to...

[R.G.]

There is a circuit that came out of radar setups called a polarity separator. It's really two half-wave precision rectifiers with opposite polarities. The output of one is the positive half cycles, the other is the negative half cycles. That would give you the ultimate freedom.

What really happens after that is that you have two new signals. A half wave rectified sine contains the original fundamental and all harmonics at varying amplitudes. If your signal was symmetrical, you'd get similar signals to process. But musical signals are often not symmetrical from top to bottom, so you're processing non-identical (in terms of harmonic structure) signals. Which is good for your purposes.

When you then do whatever you do to them, you're really processing different information,  but info that's somewhat time-aligned on a cycle by cycle basis.

It might be a good way to produce stereo sounding similar-but-different signals for two different outputs.