[sdiy] VCO Drift

[James Patchell]

The capacitor will only give you offset errors in your tuning.  These can 
be easily taken out with the tuning (pitch) knob.

The drift contribution of the capacitor, if say it was 100ppM/C, will be 
about 1/6 of a cent per degree C.

If you want to control the temperature of the capacitor, do what you said, 
but pot it in thermal epoxy.  You need to minimize the thermal lag between 
the heater and the sensor to make the loop as tight as possible.

And before spending a lot of money on 10ppM resistors, be sure you do an 
error analysis on where the critical paths are.  Measure the oscillator 
very thoroughly and identify the drifts you are trying to 
eliminate....there are two major ones that must be dealt with...offset and 
span.

You don't need to buy a lot of expensive parts to make an expo VCO that has 
reasonably low drift.


At 11:17 AM 7/29/2006 -0700, Tim Ressel wrote:
>Yo,
>
>Had a thought. I'm using the heater servo method on my
>VCOs to keep the expo convertor in line. That means
>the polyprop cap is probably the next biggest in terms
>of tempco. I wonder if the heater servo trick would
>work for it as well?
>
>Something like: wrap a 1-watt resistor to the cap with
>(what else?) Kapton tape, add a diode (or transistor
>in diode mode), and entomb in styrofoam. Servo the
>whole thing for temperature with the same curcuit as
>the expo. Keep it at 60C or so.
>
>I figure the rest of the VCO is just resistors, and
>those can be gotten in tempcos down to 10ppm without
>breaking a sweat.
>
>So like, how nutz is this idea??
>
>--Tim (nailing jello to a tree) Ressel

         -Jim
***************************************************************
http://www.oldcrows.net/~patchell
http://www.noniandjim.com/
***************************************************************