Converting 'touch' circuit, How does salvaged 'touch' IC work ?

[RJSV]

Hey, uhh, my project I would like to force a couple of 'Touch Switches'. The orig had cheezy fan, 4 V, and nice diffused LED.
  But how, does it detect, read that it oscillates (), Huh? Sounds like OSC would drain batt fast.
 
  Any rate, maybe 10k up, and 10k connected down, test that at 2V, there.
  Haven't gottaclue

[ledtester]

What do you mean by "force a couple touch switches"?? A schematic or picture would help.

Touch switch chips use dedicated circuitry to keep power consumption low -- e.g, the TTP223 consumes < 13 uA. With careful programming you could might be able to achieve the same with a microcontroller.

Are you trying to reuse touch switches you've found in some device? It might be easier to use some off-the-shelf modules -- e.g.:

https://youtu.be/XBHgJi6smls

[Neilm]

The oscillator type touch switches I have seen and researched in the past require critical layout for the touch pad sensor. If you salvage an IC without knowing the details your switch sensing capabilites could be horribly inaccurate and unreliable. It may be acceptable for an ON/OFF switch but if you had several switches they could interfer with each other causing phantom touches.

I have seen touch switches done with a bit of metal connected to a sensing circuit through a 10 Mohm resistor (needed as the device was connected to the mains). The metal was covered with a label and when the user touched the marked area the sensing circuit triggered.

[RJSV]

 Hello, short reply, the Touch-toggle is a shelf retail FAN , with a nice LIGHT. Runs on 3 cells, recharge-able gets that to 3.6 (V).
  Suspect a custom IC, but it had some 'metal', a thick foil, one for fan.
  So, thanks, will study the TTP223 for ex.
  Uh, can a high-Z scope feed that (fast) osc waveform,  for salvage, a pet hobby here, thanks

[RJSV]

More details, the little Fan / lamp has some motor switch and so I expect a mid level current, maybe 300mA.
 Input, how about a 2N2222 switching a cap. ? To get one of those touch inputs to toggle one or the other.
  Plus the flat LED lamp is sharp looking.
  Sorry, needing tripod.
  The circuit, that has 2 foil SQ, is a bland, dual on-line, no drama.

[ledtester]

So you are trying to trigger the touch switch electronically, like from a microcontroller GPIO pin?

[RJSV]

Yes, electronic trigger and a rough standard, 0 to 3 V range, or at least, tolerant inputs.  That means, a motion sensor, to a 2N2222 brings the 6 VDC unit out of the picture, only driving the xtsistor base.
  Plus, while a 2N2222 and capacitor provided one direction, perhaps a 'reverse' diode (pointed up, like a typ coil has. That's because would not expect transistor to conduct, backwards.
  This way, looks like a chance? That an osc will ecounter a pretty-near cap. to me

[ledtester]

Most touch sensors operate by detecting a change in capacitance, so you might try electronically connecting a capacitor to the touch sensor. You'll have to experiment with capacitor values. My guess is that it would be < 1uF. You'll also have to experiment with connection methods -- relay, bipolar transistor, MOSFET transistor, etc.

And then there's these ideas:

https://electronics.stackexchange.com/questions/19435/triggering-a-capacitive-sensor-electronically

https://youtu.be/gSifTiR_e2E

[RJSV]

Enclosed picture shows, the thinking is to have that little control cause a 'capacitor-like' (participation in circuit).
  So, any downward flow goes thru the transistor, but the other direction is possible. That is only approx, however, but forgetting the diode has voltage drop, the 'typical' touch circuit could still sense enough to (decide to toggle the lamp channel).
   Now, in view is R1, that allows 12 V modules to directly interface, with caution...
  My best est is, the dual-in-pack IC is small, might be one output, at I=20 mA other ch max of 300 mA.
The input, will be triggered / toggled, by the likes of push-button, a CD-4017, a garage light (motion), at 6VDC, etc, including AUDIO, that's, from now on, something worth trying, mixing modes...
Heck, I need to try, use of LED, to trick motion detector, that way there is 8 second timer!
(The garage wall mount motion sending would light up, for 8 seconds, when activated.)
  Prob start R1 at 10 k, with gain roughly 40, gives 20 mA or so, but reduce R1, slightly as needed.