David Salter wrote:

> Florian,
>
> My understanding (peter correct me if I'm wrong) is that any VCO design
> uses a primary oscillator (core) whose output will be a waveform
> (Triangle or sawtooth for example) and any other types of waveform are
> constructed from this primary waveform using waveshaping circuits,
> rectifiers etc.

Ok, Then "core" is what I thought about it:

>> The basic waveform the bare VCO is producing (as the Curtis
>> CEM3340 does
>> triangle, the moogish VCOs like the A110 does saw)


> It seems that whatever waveform you start with will impart a timbral
> characteristic on the other waveforms.
>
> This is quite odd when you think about it as a sine wave is a sine wave,
> a waveshape that produces the fundamental with no harmonics.
>
> However no waveshaper is perfect and that's why if you look at two sine
> waves from different VCO's on a scope (with enough resolution) they are
> never perfect just close. It's the imperfections that alter the sound
> characteristic.

If a sine is produced via a waveshaper, it always is based on a
triangle. In some oscillators the triangle is the core-waveform, in
others (most) cases the triangle is derived from the core saw.

In my book there are two drawings, which show the output of an FM'ed VCO
on the different Waveform-outs. Drawing one shows the output you would
expect if the FM would be the core-waveform of the VCO; drawing two
shows the real output, that is produced by an core-saw VCO followed by
waveform converters. I don't remember the page at the moment.

Florian