AW: A fundamental problem of exponential VCOs?

[Haible Juergen]

	>Has anybody tried this - one exponential converter, driving two
oscillator
	>circuits? In other words, split (or mirror) the exponential
current into
	>two integrator/comparator circuits, and then alter some
parameter to make
	>the oscillator sections slightly different. Some thoughts
include:
	>
	>* Alter the discharge trip point voltage slightly (for
sawtooth-based
	>designs)
	>* Alter the schmitt trigger trip points (for tri-square
designs)
	>* Vary the integrator cap values (they'll be different anyway,
by some
	>amount)

All three methods would mean *multiplication* with some factor,
and not addition of some offset.
The equation for charging a capacitor is 
I * T = Q = Vsaw * C
So with f = 1/T we have 
f = I / (Vsaw * C)
So without adding a constant term to I, you cannot do anything
in the oscillator core. It must do it in the V/I converter.
(This was for zero discharge time. But a constant discharge time
will affect the upper range, not the lower range, so there's no help
either.)

JH.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 1991 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/19970711/43019e9d/attachment.bin>