[sdiy] Hello, and some questions

[Andre Majorel]

On 2013-05-01 20:10 -0500, John Ames wrote:

> Also, some questions about VCOs: I know they have a reputation for
> tuning instability, but I've gathered that it's better now than it
> used to be. Is that just thanks to higher-tolerance parts being
> available at lower costs now, or are newer designs more stable? (Are
> fancy newer parts like tantalum or polymer capacitors any good for
> audio?)

I'm not exactly an expert but since nobody else bit...

Sawtooth core VCOs work by converting the control voltage to a
current and charging a small capacitor with said current. A
comparator compares the voltage across the capacitor with some
fixed voltage (E.G. 10 V). When that threshold is reached, a FET
shorts the capacitor, bringing the voltage across back to 0 and
starting a new cycle.

A VCO can go out of tune because
1. the control voltage drifts with temperature,
2. the exponential voltage to current converter drifts with
   temperature,
3. the value of the capacitor drifts with temperature,
4. the threshold voltage drifts with temperature or power supply
   load.

#1 is easily taken care of by using op amps with low offset and
low temperature coefficient.

#3 and other reasons mean that you have to use a polystyrene or
NP0/C0G ceramic cap.

#4 is avoided by using a proper voltage reference instead of a
resistive divider.

#2 is by far the worst offender. The fix is usually to use a
so-called "tempco" resistor with a temperature coefficient of
about +3350 PPM/K. Or doing the exponential conversion with an
SSM2164/V2164 (David Dixon's speciality).

http://home.comcast.net/~ijfritz/sy_cir2.htm
http://home.comcast.net/~ijfritz/sy_cir4.htm
http://home.comcast.net/~ijfritz/sy_cir7.htm
http://home.comcast.net/~ijfritz/sy_cir9.htm
http://houshu.at.webry.info/200908/article_1.html
http://houshu.at.webry.info/200908/article_2.html
http://www.uni-bonn.de/~uzs159/expo_tutorial/
http://www.uni-bonn.de/~uzs159/rbe.html

-- 
André Majorel http://www.teaser.fr/~amajorel/