digital control of analog component question...

[Fraser, Colin J]

> -----Original Message-----
> From: Harry Bissell [mailto:harrybissell at prodigy.net]
> Sent: 13 January 2000 02:44
> To: The Proteus
> Cc: synth-diy at mailhost.bpa.nl
> Subject: Re: digital control of analog component question...
> 
> I believe that the major difference between the DCO and the 
> VCO is that most DCO's
> produce a waveform that starts over from the beginning of the 
> wave with each change
> in frequency...  sort of an unintended reset function. Most 
> samplers work this way.
> 
> VCO's respond with a continuous waveform... even a jump 
> between octaves is a smooth
> transition... and glide etc is always smooth, not a series of steps...
> 
> And some circuits are halfway inbetween... but few... very few...
> 
> VCO's... because I can taste the difference....

I've been thinking about DCOs for a polysynth design.
For true analogueness I want to avoid using digitally created waveforms, and
also to allow the period of each cycle to be altered after the cycle has
begun.

The Juno 106 has a neat trick for creating it's waves. A timing capacitor is
charged at a rate set by a DAC that is approximately right for the osc
frequency, but the reset pulse for the cap comes from a digital timer.
This gives a smooth 'analogue' ramp, with a precise digitally controlled
period.
Any error in the charging rate set by the DAC results in an amplitude error
in the wave - not a big problem.

To allow the period to be changed mid cycle, I'll use a phase accumulator
oscillator to generate the reset pulse for the timing cap (reset when the
accumulator overflows).
This will allow the frequency control for the DCO to be adjusted at any
time, with the same end result (I hope) as a true analogue osc.
Depending on the rate I can get the frequency control data into the
oscillator processor, audio rate FM mught be possible, but I'm not sure what
aliasing or other artefacts this might produce.


Colin f