This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

FDC2212: Datasheet calculation verification

Part Number: FDC2212

Goodday!

I am currently designing a humidity sensor module that is based on a capacitive sensor element (MK33-W). The full range of the capacitor is approx. between 250 to 370 pF, where 0,45pF/%RH is the measurement output, the rest is manufacturing inaccuracy of sensor.

Combining properties from both the sensor and the FDC datasheet my calculation came to set the CHx_SETTLECOUNT to 0,0956 (rounded up to 1). This number seems quite small, and I am not sure, if my values are selected correct:

Vpk = 1,6V

Csens = 300 pF sensor + 10 pF parasitic

Fclk from external source = 4MHz

Fref (CH0_FREF_DIVIDER set to 400) = 10kHz (recommended frequency for MK33-W)

IDRIVE set to minimum = 16uA (settlecount is inversely proportional to idrive, higher current leads to smaller settlecount)

The required output resolution should correspond to 0,1pF.

I also suppose that a selection of 1mH + 27nF LC tank (around 9,6 kHz oscillation freq.) is suitable for single ended configuration.

My calculation is theoretical without actaul measurement, because I am looking into different solutions to convert capacity to digital value. I would like to appreciate some guidelines to define the accuracy output of my system. Temperature compensation with calibration is considered, so I am mainly interested in the ability of the FDC component.

Best Regards

Istvan Szeker

  • Hello Istvan,

    The 10kHz operation, while supported by the FDC2212, will require a sensor with a very large amount of fixed and parasitic capacitance. The large amount of capacitance can be driven by the FDC2212, but it will dilute the sensitivity of the measurement. Another concern is the variation of the parasitic capacitance due to temperature shifts and other changes; unless you have a way to compensate for them, you may not be able to accurately measure the desired signal.

    For example, a C0G cap has a 30ppm/degree C specification. For a 50nF cap, you will have a shift in sensor capacitance of 1.5pF with 1 degree of change, which is equivalent to a 3% RH shift. This does not include the change from the fixed inductor.

    You may need to use an analog switch to connect and disconnect the humidity sensor to obtain an offset which you subtract from the measurement with the sensor.

    With a large amount of fixed capacitance, you will need to increase the RCOUNT setting. The fref divider should be much smaller than 400; setting it to ~20 (to have an effective Fref of 2MHz is fine). Note that this will proportionally increase the measurement time.

    I recommend increasing the sensor frequency to above 10kHz - if you can make it >70kHz the sensor will be much easier to implement, and also more accurate.
  • Hello Mr. Overhauser,

    Thank you for the very fast reply! I have now changed the design, so the oscillation frequency is around 80kHz for the LC tank, via 2MHz reference (fref divider 25). There are a lot of limitations for my circuit to be done, so it is a bit difficult to find the optimal components. The temperature range is expected to be between -20 to +100 °C or even more, while the output accuracy has to be excellent. Even with a 7.5nF cap i will get a 0,5%RH drift for 1°C change. Also, I have to use a fairly large inductor, that will add inuaccuracy to the system in another way.
    I think I will keep an eye open to find a more optimal measurement solution!

    Thank you again for the detailed answer you gave me!