Questions about capacitive sensing

Question

I'm trying to use the FDC1004 capacitance-to-digital and I have a few questions:

1. Why do we (our fingers or another part of our body) act as the other conductive plate - don't we have a very large resistance? The image below shows the finger grounded but why does it show that - isn't the resistance between us and the ground extremely high?

2. How does changing the finger from the picture above to a metal plate affect the capacitance reading? Since the metal plate is more conductive and has a greater effective surface area will my capacitance readings be higher?

3. I was thinking of putting some sort of compressible material on top of the sensor and then adding a metal plate on top of the compressible material. This way I will create a large capacitor - I would have two conducting surfaces separated by a dielectric. If I put an object on top of this setup the dielectric material will compress, causing the distance between the plates to decrease and increasing the capacitance reading. I want to use this behavior to measure the force/weight of an object but I'm concerned that if I try to measure the force of someones finger pushing down (or something else conductive) it might drastically change the results since i am introducing another 'conductive plate' into the system.

In the picture above, the blue line would be someones finger coming to push down on the force sensor - wouldn't the finger start to change the field lines and cause significant changes in capacitance before any compression starts to occur ?

Answer 1

Since our body fluids with water have a dielectric constant , Dk of 80 compared to air , it is the change in capacitance by touch that conducts, not the resistance, when place between tiny electrodes.

Almost touching might be 1pF and this has a 15 pF range.

Answer 2

First of all capacitive sensing doesn't rely on anything about body resistance. It works even with insulation in between. In fact, with glass it works better given it's dielectric properties (there is an appnote for the FDC converters working thru 10mm of glass, for example). That aside we are reasonably grounded (on the order of kiloohms), otherwise people wouldn't get electrocuted so easily; rubber boots help for that!

More or less everything is capacitively coupled with the ground: the earth is quite a big electrode and, nominally, a capacitance of about 4pF is assumed between some conductor and earth, unless electrostatic shielding measure are made.

Remember that the FDC as a sensibility down to femtofarads. There's a lot of noise but it is reasonable to get some picofarad of signal out of that (the FDC also has an active shielding system, but that's part specific). So, normally you have your touch sensor which is at nominally 4pF to ground (more or less); that's the baseline level and you calibrate from it (in fact baseline level calibration is crucial in working with wet surfaces and similar issues).

When you go near the sensor with your finger you are essentially putting your body capacitance in parallel to the baseline one. The read value starts raising at some centimeter, if you tried the eval board. The fact that it triggers at touch is not because of conducivity but simply because the signal threshold is configured that way. There are systems kept in low power and the wake up is given by the hand going simply near the sensors; that's a good way to handle the power saving.

Also, if you look at the electric field formula, it attenuates with the square of the distance: the touch condition is much more easily detectable.

As for the last question: yes, the sensor will pick up the finger just nearing the elastic material. Unless said material is conductive and possibly grounded, in that case it will not. But in that case the capacitive sensor will work between the plate and the elastic material (it has its uses, however; some pressure sensor work that way).

A similar but completely unrelated technology to the last example is inductive sensing: a cap sensor will look for the electric field from the finger (or the diaphragm); an inductive sensor uses a coil and the conductive diaphragm alters the magnetic field. It's completely different but quite important too (if not more, considering all the inductive proximities in use in automation)