analog storage sequencers

[Eli Brandt]

brad sanders wrote:
> Yes - except you can easily increase precision by *scaling* voltages.
> Like I said: it takes no more than 1Vp-p input, so the voltages have
> to be scaled anyway. Increase voltage division to give 1 octave
> tracking and you have MAX freq. inaccuracy of a few cents.

Restricted to one octave, it's 1200/64 ~= 20 cents, so something like
+/- 10-cent error.

> Also, the steps aren't discretized; the voltage in each cell is
> continuously variable (think of a moderate accuracy, infinitely
> sustainable S/H circuit). These cells are cycled through at a fixed
> (adjustable) rate. Given the higher sample rate (say, 11KHz) a 1KHz
> low pass filter on the output means the output would be a weighted
> "average" of the last several storage cells read.

Good plan, if the inaccuracy shows up as whitish noise.  Actually, it
might be fun to leave the noise in, for a little FM of the carrier
oscillator...

-- 
   Eli Brandt
   eli+ at cs.cmu.edu