0
\$\begingroup\$

I am looking to do post-grad research study next year in the sensory field for sensing gases like hydrogen, nitrogen or ammonia (not decided yet). So trying to understand what is out there and how best to approach the problem of measuring capacitance (never done this before). I have been evaluating different ways of measuring capacitance from these planar electrode (required to use this particular sensor, photo below) using a microcontroller such as a Arduino/Nucleo board as this needs to be portable/field deployable. I have been told these planar electrodes have a capacitance of approximately 9pF in air, computed using B&K Precision LCR meter.enter image description here

I need to be able to sweep through different sinusoidal frequencies ranging from 100 Hz to about 1 MHz which I can generate using AD9951 (direct digital synthesis IC) with amplitude of 1 Vp-p and send the measured capacitance values to an IoT cloud.
Since I am not allowed to go back to university just yet, I have no components or equipment to test performance of different circuits.

I have come across many papers where they calculate the impedance of the sensor using EIS (Electrochemical Impedance Spectroscopy) method but I am looking to measure capacitance instead. So, what I have found so far:

  1. Using dedicated ICs such as FDC2214 for capacitive measurement (normally used for touch sensing application), capable of sensing maximum of 250nF.
  2. LC tank/resonant circuit with the planar electrodes connected in parallel (as these electrodes act as a capacitor, the capacitance changes as the dielectric changes) and using an OPAMP such as LM311 to measure the change in frequency as capacitance on electrodes change.
  3. Using operational amplifier such as LM311 in the integrator or differentiator style configuration to measure changes in frequency at the output when swept through different frequencies. A reference circuit is shown below (pulled from Google).

enter image description here

I am just looking for a confirmation if I am on the right path and suggestions on how efficiently I can measure capacitance with sinusoidal signal using microcontroller. Some of my thoughts on what I have found.

  • The issue with LC tank circuit is that it generates its own resonant frequency and computes capacitance using a fixed resonant frequency. So, sweeping through range of frequencies would not possible (unless I am wrong).
  • Using FDC2214, the implementation would be easy but I'm not sure at this stage if the planar electrodes would exceed is maximum capacitance sensing of 250nF. Would this still be a viable option if the sensor does not exceed the ICs limit?
  • Using Op-amp LM311 as the integrator configuration although I don't know much about this configuration at this stage.

Also, what I would love to know is how LCR meters compute capacitance with different voltage levels and range of frequencies to sweep from. Maybe someone can point me in the right direction. I hope this question is not too broad.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Since you mention Nucleo - many STM32 parts (F0 come to mind in particular) have a touch-sensing peripheral built in. This is basically a capacitance measurement by another name. I am not sure if this would be useful to you, in particular I can't find the typical capacitance of such a sensor, but do give it a look. \$\endgroup\$
    – jaskij
    Sep 28, 2021 at 12:29
  • \$\begingroup\$ @JanDorniak, thank you for mentioning that. I wasn't aware. I will certainly have a look. \$\endgroup\$ Sep 28, 2021 at 20:22
  • \$\begingroup\$ You're welcome. There's also some cheap-ish (~1$) ICs from Microchip built for the same purpose. Touch buttons essentially form a capacitor between the electrode and the finger, and it's capacitance is then measured. As I wrote earlier, I'm not sure about the typical capacitance, but single digit pF sounds about right. \$\endgroup\$
    – jaskij
    Sep 29, 2021 at 12:01

1 Answer 1

1
\$\begingroup\$

Can't go past the HP Impedance Measurement Handbook: https://assets.testequity.com/te1/Documents/pdf/keysight/impedance-measurement-handbook.pdf

This is a typical arrangement that you will find in some precision LCR instruments that measure up to 1MHz. It's clever for many reasons, but one is that both ends of the DUT are at a low impedance. Ideally, you need two ADCs to sample the voltage at the high end of the DUT and the output of the op amp. The complex ratio of these times the feedback resistance gives you the impedance. One challenge would be to keep the virtual ground a virtual ground at 1MHz. One refinement is to measure the voltage at the inverting input and correct for it.

enter image description here

Alternatively, there are precision capacitance measuring ICs developed by several manufacturers, but I'm not sure they cover the frequency range you want.

\$\endgroup\$
2
  • \$\begingroup\$ Thank you for sending this. Although it's quite informative, it focuses on impedance and I am looking for capacitive measurement unless I have missed it. Correct me if I'm wrong. \$\endgroup\$ Sep 28, 2021 at 7:52
  • 1
    \$\begingroup\$ That's how the precision instruments that measure capacitance at different frequencies work - they measure some circuit parameter like impedance (or admittance, or reflection coefficient) and deduce the capacitance. I've edited the post to include the typical arrangement in some precision LCR meters that work up to 1MHz. \$\endgroup\$
    – Tesla23
    Sep 28, 2021 at 9:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.