Using a dedicated functionality on a processor, I know how to detect touch on a metal pad.

My question is doing it on a pcb in a nice looking way.

  1. I could just expose few large pads (which are metals) and connect them to ports on the processor, it's nice but I want something else.

  2. Can I put the pads on the bottom of the pcb, and draw the buttons right on the top of it so I can sense without touching directly on the exposed pads ?

My concern is that if it works with a single button, then I will not be able to differentiate between multiple because the board is a uniform shape, and if I touch on point A it's like B in terms of capacitance ? ( or at least almost identical because of the noise. )

  • 2
    \$\begingroup\$ It's a bit hard to tell without knowing what the "dedicated functionality on a processor" does - for capacitive touch sensing, the touched surfaces should typically be non-conductive (i.e. not exposed copper pads) but isolated. \$\endgroup\$ – Marcus Müller May 30 '17 at 10:42
  • \$\begingroup\$ Thanks can you explain why should they be isolated ? if I take out a wire from the processor and touch it, it works, so my logic says I should just expose a pad on the pcb, no ? \$\endgroup\$ – Curnelious May 30 '17 at 10:44
  • \$\begingroup\$ because capacitive implies exactly that: isolated. \$\endgroup\$ – Marcus Müller May 30 '17 at 10:45
  • \$\begingroup\$ So how is it working when I just connect exposed wire to the processor, and NOT working if it's a wood ? isn't that means that it should be an exposed metal? \$\endgroup\$ – Curnelious May 30 '17 at 10:46
  • 1
    \$\begingroup\$ Having an exposed conductor for people to touch shorted to a processor pin sounds like a good way to break the pin and or processor. \$\endgroup\$ – Colin May 30 '17 at 10:53

If you are using a multilayer PCB, you could place 2 Pads on the most bottom and the second most bottom layer. If you connect them to an OPamp, you should be able to detect changes in the charge distribution on the pads. Since nobody is really neutral/chargeless, a person touching the "painted" pad on the influences the two pads differently, which causes a peak at the output of the OPamp. Using a lowpass on its output and maybe using a schmitt-trigger before processing the signal. Since the Coulomb-potential scales with 1 over r^2 the potential diffenrence between the two pads should not be that much, that it works from far away, also you may scale it yourself, hence you should even be able to place multiple "buttons" on a single pcb.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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