Octature osc?

[Paul Perry]

At 03:09 PM 4/09/98 +0100, jorgen.bergfors wrote:
>I would like a three phase LFO. Any ideas how to make one? I already have one 
>that use all pass filters to get 120 degrees  phase shift, but it only works 
>right for a very limited frequency range.
Imagine a simple rc network that produces 2 stages of 120 deg shift.
Perhaps you could use switched capacitor/resistor techniques to track.

Actually, if you were using the barbaric sine generation method based on
weighted summing of the outputs of a counter, then it would be easy to 
get 3 phases.

But, if I had to do it, I think I would try to generate 3 triangles with
the required phase relationship, then shape them to a sine by overdriving 
a 3080 (if 2% or so is good enough). Hint: think about triangles with 3
times the frequency of the LFO you want.

That phase shift oscillator (where you use a couple of c and r to change
phase by 180 deg to make feedback positive) would give different phase 
at various points.... perhaps something could be done like this, with the
tuning done by switched resistors.

A problem with LFOs is that when they run slow, any discontinuities sound
terrible if you are modulating pitch with them. In this regard, lfo wave purity
is more critical than that of audio oscillators.

paul perry melb aust