[sdiy] 2164 tempco in VCF applications

Neil Johnson neil.johnson71 at gmail.com
Tue Nov 6 18:46:36 CET 2018

Hi Rutger,

> I've been impressed with the performance of recent VCO designs (yes, you David!) using 2164 for temperature compensation. I've been wondering if people have attempted similar things with respect to VCF applications?

For VCAs and non-critical VCFs you can use an external temp.co to
compensate the gain control.  See the SSI2164 datasheet.

>When using integrators in a VCF based on the 2164 you already have an exponential control law of the inputs, so wouldn't need a full exponential current source as in a VCO.

Indeed, and with the right scaling you can get a good approximation to 1V/oct.

> Can you still apply the same principle for making a VCF temperature stable?


>Do you need that degree of stability in a VCF (yes, different opinions are welcome)?

Depends.  If your filter can self-oscillate and you intend on using it
as a pitch source (rather than just a randomish tone) then you'll
probably want it to track as well as your other system's VCOs.  In
which case the same temp.comp scheme should be used otherwise they
won't track and the VCF tone will drift compared to the VCO tone(s).

> And what would you do if you already used up all 4 sections of a 2164 in the core of the VCF, meaning you don't have another 2164 available in the same (isothermic) package?

You'll never get better than using sections on the same chip to
compensate for temperature as anything external to the chip won't
track the exact temperature of the die.


More information about the Synth-diy mailing list