[sdiy] Temperature stable lin-exp converter with a CA3086 or CA3046

[Tim Ressel]

Yo,

Leave it to Mr. Blacet. It would be fairly simple to
program a uP to compensate for Tc drift. There are two
basic schemes, both with pros and cons. First is where
the uP measures the VCO input and output, and deduces
and corrects drift. The second involves characterizing
the drift, and programming the uP to measure the
temperature said temperature and compensate. In theory
you would not even need a tempco resistor. Just
measure the temperature and lookup the correction
factor, which can take into account the expo drift AND
the drift of the rest of the circuit. Might be worth
looking into.

--TR

--- John Blacet <blacet at blacet.com> wrote:
> The whole "how stable should it be" regarding VCOs
> is an interesting
> discussion. On the one hand, you don't want absolute
> lock-step digital
> stability in an analog oscillator, as being just a
> bit drifty makes them
> sonically much more interesting to use. But, you
> don't want to be
> reaching for the Fine Tune knob more than a few
> times a session; that
> gets annoying.
> 
> The whole technology limitations of analog guarantee
> that absolute
> stability is really unobtainable, but you can get
> close enough to make
> nice analog VCOs. The best commercial VCOs go a bit
> beyond the customary
> temp comp resistor approach with techniques to "mop
> up" residual drift.
> Some synths of old, for example seem to auto tune
> via uP on a periodic
> basis.
> 
> I wonder about a PIC based processor in a correction
> loop? Say, just a
> nudge now and agian to make the CV to freq ratio
> stay "close". You could
> even program in bits of subtle "misbehaviour".
> 
> Regards,
> -----------------------------
> John Blacet
> Blacet Research
> http://www.blacet.com
> 
> 


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com