Touch Pad

Discussion in 'The Projects Forum' started by Art, May 10, 2015.

  1. Art

    Thread Starter Distinguished Member

    Sep 10, 2007
    Hi Guys :)
    A few years back I made this part of a circuit which worked well as a metal touch plate (ie. no moving parts button).
    It was not an analogue pin... I was timing the discharge of the capacitor via the resistor.
    The variable resistor must have been adjusted to what proved to be the best range to sense with my program.


    I am wondering if this is the best place to connect the touch plate.
    It worked well, but I don’t remember the physical situation ever changing, like using it outside of the same room.
    Maybe I was helping earth it? I don’t know if it would change the value of the capacitor to touch the other side of the resistor.

    I would like to have a go at a whole 16 key keypad with a micro that has 16 digital IO pins.
    any ideas appreciated.
    Cheers, Art.
  2. DickCappels


    Aug 21, 2008
    To address your thoughts on grounding: Yes, if you ground it you will probably get more predictable behavior. A common human body model assumes that a person has on the order of 2,000 pf to earth. If you ground your circuit then you can expect to see that entire 2,000 pf or whatever it really is. If the circuit were battery powered and floating then you end up with the series capacitance of that 2,000 pf and whatever your circuit has to ground -which might vary a lot.

    One other thing: Whether you ground your circuit or not, you are putting your poor little PIC at risk by not using electrostatic discharge protection on the input pin. True, each I/O pin on a PIC has some ESD protection already, but it is good form to add some external robust protection. To wit:
  3. MaxHeadRoom


    Jul 18, 2013
    The circuit shows it is earth grounded already?
  4. Art

    Thread Starter Distinguished Member

    Sep 10, 2007
    All of the ground symbols should just be vss. It was a battery circuit with no real ground connection.
    DickCappels, it looks like the document is suggesting a diode clamp on the IO pin, I can do that.