A fundamental problem of exponential VCOs?

[Paul Perry]

At 05:30 PM 10/07/97 +0100, Haible Juergen wrote:
>Hi!
>
>I spent some thoughts on beating between a pair of exponential
>(V/Octave) VCOs.
>
>I found that I run into one problem quite often: When I detune the VCOs
>so much
>that I get a decent beat rate of 0.2 to 1 Hz (which I find very
>pleasant), the VCOs
>sound very detuned in higher octaves, *regardless* of the scale

If the problem is to have to exponential (though I think of them as
logarithmic) VCOs that differ by say 0.2 to 1 HZ and track with this
relationship over a number of octaves, then one has to exceed the precision
of any analog bases exponential convertor ever made.

But there may be a solution, using a frequency synthesis method where one is
generating a LFO by beating two VCOs, with one VCO the normal externally
controlled type of VCO, and the other driven by the normal CV plus a
correction factor which is generated by comparing the actual and ideal LFO
frequencies.

Now this may give a *too consistent* beat relationship but this can be
overcome by judicious very low frequency noise injection into the LFO
feedback correction path.

This sort of thing is done in communications engineering all the time (well
not the noise injection!). I have often wondered why nobody bothers to use
ultra low frequencies to make lfo's more interesting....if anyone does, use
a 1/F generator & please let's know how you get on.

Paul Perry Melb Aust