RE: Building the impedance measurement circuit

dang74 — Wed, 14 Aug 2024 00:09:55 GMT

You could maybe use the DDS to sweep frequency. Then attach a resistor in series between the DDS output and your sensor and observe on an oscilloscope. At lower frequencies the series resistor and sensor would form a voltage divider. So you will be able to determine the sensor's resistance. Somewhere in the sweep the amplitude will reach a peak. This is the resonant frequency. With the series resistor, sensor resistance and resonant frequency all being known at this point, you can determine the capacitance. If I am not mistaken at resonance the series resistor will be equal to the parallel combination of the sensor resistance and capacitor reactance. So use that as the basis to figure out the capacitance.

RE: Building the impedance measurement circuit

shabaz — Sat, 10 Aug 2024 03:37:29 GMT

This is non-trivial, not sure how you expect to design and build this if you're completely new to circuit design!

You could buy your way out of the problem, there are LCR modules that support the measurement range, but you're looking at a cost of approx $1k USD. For example, this module supports up to 1 Gohm apparently; but strongly suggest you discuss with the Analog Devices before buying it, to be 100% sure it will meet your needs: https://at.farnell.com/en-AT/analog-devices/admx2001b/messmodul-evaluationsboard/dp/4305964?ICID=I-RP-STM7REC-0

Otherwise, you'd have to resort to looking at reference designs, those are frankly your best guides if you plan on studying and spending a significant amount of time (months) designing and building and coding and testing. Nearly all solutions will entail op-amp circuitry, high-res ADCs (20 bits or more) and signal processing software. Quite a lot to learn for a beginner.

Example partial reference designs: TI https://www.ti.com/tool/TIDA-060029 (doesn't support 30 Mohm as-is, so you'd need to try to improve it (e.g. use lower-noise op-amps and lower-noise circuitry, which is likely part of the limitation).

Also, it won't be fully automatic, you'll need to run calibration initially, and then from time to time (e.g. if the wiring changes).

You should also realize it won't be really compact, if you're using prototyping boards like a DDS board, Arduino, etc. You may as well just buy an off-the-shelf LCR meter, for under $1000, and physically, it won't be an order of magnitude larger. And you'll save many months of work (easily 6 months if you're completely new to it).