these kind of measurements are plagued with
 noise of all sorts, so usually they are done in
 locked-phase (PLL). There are circuits that
 PLL for you - you need to read a bit about
 the principle and then really pay attention
 at the producer's applied-notes.

--- Andrew O'Grady  wrote:

>
> We have a device with which can be modeled as three parallel plates, with
> the middle plate moving back in forth via a spring, affecting the
> capacitances between it and the other two plates. We were hoping to measure
> differential capacitance O(1-100fF), while at the same time actuating the
> middle plate via electrostatic force using a large DC bias 50 - 200V. We are
> looking for a circuit or sensor we can buy that will do this differential
> measurement (probably via AC current) while allowing us to actuate the
> middle electrodes. The only similar sensor we have found that you can buy is
> sold at microsensors.com , however this uses a
> reference DC voltage around 1V to make the measurements, thus not allowing
> actuation using a DC bias.
>
> We would like to avoid having to build our own circuit board to do this, but
> if it comes down to it, how complicated would this be for someone with a
> mechanical engineering background?