Sv: [sdiy] Digital VCO update

[Eric Brombaugh]

Michael Bacich wrote:
> So how do you select the waveform?

I use a 3-bit binary input to the PSoC. Yes, you could use a couple of 
8:1 analog muxes to select offsets and gains.

> Your 5kHz filter, while useful for cleaning up the saw and square 
> edges, won't help much in regard to the stepping issue at slow LFO 
> rates, unfortunately.

That's true.

> It'll be interesting to find out how that stepping is (or isn't) going 
> to sound.  If your DAC gives you 245 pitch steps per octave, that 
> means that you have about 21.3 steps per semitone.  If you put that 
> into an unattenuated 1V/octave CV input on a VCO or a resonant filter, 
> I'm guessing that it might be pretty easy to hear those steps at slow 
> rates.  Just a hunch -- I'd love to be proven wrong.
The oscillator's output DAC is at least 9 bits with a 10V range after my 
filter/buffer/offset.  This gives a stepsize of 0.019V which is about 51 
steps/octave, or about 4 steps/semitone. If I can really get 11 bits 
(which I'm not at all certain of), then it's more like 204 steps/octave, 
which is more like 17 steps/semitone. Either way, the resolution is 
extremely coarse and by your estimate probably audible.

The PSIM and the new group-effort computer voltage source that forked 
from it are using 12-bit DACs over a 10V range - they're getting more 
like 34 steps/semitone and users seem pretty happy with the results.

After I button up the DVCO project I'll try a few experiments with a 
DLFO and we'll see how well it works. It doesn't me anything but time, 
and if that were an issue I wouldn't be doing this hobby. :)

Eric