[sdiy] EDN: Shift-register based Quadrature oscillator

[Harry Bissell Jr]

I've seen it done better as well... with a 1/4 sine
table and they invert and mirror the same data... uses
fewer resistors for arbitraty precision.

Why would there be a phase error with equal RC filters
???   It would be different from the input phase but
the outputs should still be equal

H^) harry

--- René Schmitz <uzs159 at uni-bonn.de> wrote:

> 
> 
> rdrake wrote:
> > Hmm, this looks interesting...  why didn't we
> think of that?
> 
> Actually this is fairly old. I've seen it done
> before with 4017s. Though 
> not with multiphase, but thats straight forward.
> 
> The trouble starts when you try to get 16bit
> accuracy out of this.
> Besides the smoothing filter he uses is a source of
> phase error
> between the individual phases.
> 
> Cheers,
>   René
> 
> -- 
> uzs159 at uni-bonn.de
> http://www.uni-bonn.de/~uzs159
> 
> 
>