[sdiy] Weird hobby of the week: Chip-collecting!

[harrybissell]

That would have the same problem... One stage of the S/H in one BBD
would not map to one in the second BBD.  Each pulse would be longer, or
shorter.  Fidelity would suffer (hah I made a funny, like fidelity and bbd might
be said in the same breath :^)

It would be better, I agree (or think I agree, without thinking too deeply) to have

dividers from a single clock than from several free running clocks... especially if
you want the taps to remain a constant ratio apart. You could sweep the master
clock and keep the ratios the same.

H^) harry

Senso wrote:

> >imho that would SUCK...
> >
> >(no seriously :^)
> >
> >I think that running the BBDs at different clocks is a recipe for suck-dom.
> >Problem is you need to recover the analog signal at each output then re-inject
> >it into the next input. Any residual clock may cause beating which
> >could reflect
> >
> >into the audio range (unless the filters are really really good). And the
> >sampling
> >apertures will not match, so you will be smearing the individual
> >samples a lot.
> >
> >I'm not one for series connection of BBDs unless its with a
> >synchronous clock...
>
> And how about one clock with a collection of (switchable) dividers?
>
> Senso