What in fact IS a Capacitive Sensor?
Even electronic engineers who use capacitors and know how they work have difficulties imagining the function of a capacitive sensor with a person being one of the electrodes.
Imagine two conductive plates left and right with the left one grounded. The left plate will maintain the 0 V potential. The right one, electrically insulated, has 0 V at the beginning, too. Now a wire connected to a positive voltage touches the right plate which will take on the positive voltage. But what happens inside the right plate? There is a short current pulse pulling that many electrons out of the plate that the plate takes on the potential of the wire. (Electrons have a negative charge. So, putting them into the right plate would cause a negative potential.) Obviously, more electrons have to be moved to put the right plate to a higher voltage. The same holds true if the plates are larger. It is less obvious though, that more electrons have to be moved also, if the plates are positioned closer to each other. In the same way, fewer electrons have to be moved if the plates are more distant to each other. Once the right plate has taken on the voltage no more current flows.
If the right plate is grounded its potential must become 0 V. Which means that the same number of electrons has to be put into the plate that has been taken out before. Applying an alternating voltage to the right plate therefore will cause an alternating current. The two plates are now called electrodes and the left plate is represented by a person or a part of it, say a hand. The right one is the sensor electrode.
If the electrodes get closer the alternating current raises and vice versa. Electrically speaking the impedance (resistivity under AC conditions) raises with departing electrodes and vice versa.
In our technology, the current is not being measured. The alternating voltage to the right electrode is applied through a very high impedance (a very small capacitor). This is a capacitive voltage divider. The voltage of the right electrode – the sensor – will decrease when the left electrode – the person – approaches and vice versa. This voltage is measured and indicates the presence of a person.
If the right electrode is “charged” to a certain voltage, a certain current has been flowing for a certain amount of time. Since the expression “current” means a (very large) number of electrons per second, current times time is charge (i x t = q). This charge is the elementary charge of one electron times the number of electrons. The arrangement of the two electrodes can be seen as a capacitor. The “capacity” of a capacitor is the charge necessary to change its voltage by one volt. The unit can be written as Amp times seconds divided by volts. There is a special unit, the Farad (after Mr. Faraday).
Daily life gives an impression of volts and amps as well as one can imagine what a second is. One farad (1 F) is not very useful for most applications, though. Imagine two electrodes beeing 1 mm apart. How large would you expect them to be to build a 1 F capacitor? The answer is: about 100 square kilometers (39 square miles)! Imagine a square of 10 km by 10 km (6.25 miles by 6.25 miles). That is not handy, plus we want to detect a person in a larger distance than just 1 mm. Therefore a more convenient unit is the picofarad (pF) or one farad divided by a 1 with 12 zeros. A capacitor with 1 mm distance between its electrodes has an area of about one square centimeter to represent about 1 pF. The bad message is still that a person is to be detected in one centimeter of distance at a roller mill or 30 or more centimeters (12 inches) with collaborative robots. The good message is that a person is larger than 1 square centimeter.
The Sensor in its Surroundings
It is good to know that a capacitive sensor works best if both electrodes are comparable in size. If only the left electrode would be enlarged on the above figure – a large person in front of a small sensor – the sensor signal would not increase adequately. If only the sensor area is increased, the person (or his/her hand) would not evoke more change of the sensor signal. Even worse, the sensor would pick up much more noise from the surroundings. The size of the sensor should equivalent to the smallest area (e.g. a hand) to be detected. The following image shows different situations.
Our video shows a model of a robot with two sensors. One is located on the surface of the upright part the other one on the horizontal one. When the hand approaches the end of the sensor with its smaller surface the hand is only detected at a shorter distance. The hand below the horizontal sensor shows only a very small area towards the upright sensor.
There exist no “rays”, by the way. The voltage used is extremely low and the frequency harmless. Even a radio receiver in close vicinity would have problems detecting our signal. The “aerial” is much too small to transmit the frequency in use. It performs like static electricity in spite of the alternating voltage.