Reply to thread

Your bridge references to the unexpected values, which seem to not make sense.   500 uF for 500 pF nominal results and 500 kOhms for 10Meg nominal results.

[ATTACH=full]193100[/ATTACH]

Note that in the Wheatstone bridge and any other bridge with a voltage source, one must use balanced references equal to the device under test to null for good accuracy. (otherwise, the impedance is a complex product and sum)

Only in the case of current sources will the impedance be proportional and linear to voltage.

I suggest that you justify your setup as I do not understand your readout or do as I suggested for impedance measurements with AC constant current (CC) to measure voltage as done in RLC meters at different frequencies.

The Howland CC is one method for AC rms or DC current.

It is also unclear how your sensor functions and there must be some generative response to stimulus that is unclear and nonlinear.

Show a polar plot or a log sweep to 1MHz for example with a fixed RC in this range for calibration.