simple touch keyboards

[egasimus]

Hey all,
I would like to etch myself a "keyboard" on a piece of copper clad. I prefer not to go the stylus route (although it's always an option) - so is there a reasonably simple to build circuit for a touch keyboard? I'm not afraid of AVR microcontrollers (don't have a programmer for PICs, though), but where's the fun in using one when we have CMOS
Cheers,
Adam

[ericcoleridge]

short answer is no, there's no simple, easy touch keyboard.

If you do some research you can find some complicated difficult ones

[egasimus]

how 'bout something based off this?

It's as simple as it gets - gotta experiment and see what it's good for.

[MadScientist]

[LFLab]

Well, this is not too hard to do, two opamps and touch sensitivity not just a switch
Probe wah clone, look for the schematic with two opamps

edit: and the one with two bjts is also not too bad.

edit2: actually, the one i posted is not touch sensitive but is proximity sensitive, duh! You can tune the sensitivity to only respond to touch.

[egasimus]

yeah, I breadboarded a probe wah a while ago, and the 4093-based (I think) proximity detector worked like a charm (although I couldn't get the wah to work as expected and I tore it all apart). thanks for reminding me about it, I might put it in there as a separate controller.

I'll hopefully try some of these stuff soon - I've got two metal cookie boxes which beg to be turned into synths - maybe one drone/sequencer and one keyboard-playable?

[LFLab]

Never tried the CMOS one myself, built the control section with the two opamps (no wah section), works like a charm. Never did anything with it though, just built it to try it out, hooked up a piece of copperclad as an antenna.

[richardc64]

I'm always fascinated by the willingness to put one's body in direct electrical contact with active circuitry -- low-voltage, static-sensitive or not. See this:

http://www.paia.com/ProdArticles/touchsw.html

The advantage of this type of touch detection is that since it relies on changes in capacitance, the contact need not -- and indeed SHOULD NOT -- be a bare conductor. It could have a coat of polyurethane or some thin insulating material and still operate.

Rather than tinker with the time constant of each touch plate, varying the clock rate, via R2, should provide a Sensitivity adjustment.

As for circuits that never made it off the breadboard, I've attached one of mine. Resistor and cap values were missing from my original hand-drawn schematic, so experimenting will be necessary. In operation, as long as D is low when Clock goes high, Q will be low. When point "T" -- which should be insulated -- is touched, the rise and fall of D are delayed. When they're delayed long enough that D is high when the rising edge of Clock arrives, Q goes high. To get 8 more contacts only requires another 8-bit latch. 74Cxxx is getting hard to find, but 74HCxxx should work.

There's a similar Ian Fritz circuit here somewhere, but I couldn't find it.

---

touch374.gif
Description:
Filesize:	4.55 KB
Viewed:	3340 Time(s)

---

---

touch_timing.gif
Description:	highly exaggerated
Filesize:	6.62 KB
Viewed:	137 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[egasimus]

Very interesting! I'm hopefully gonna etch a keyboard pattern on a piece of copper-clad tomorrow as I build my drone box, and later experiment with some of these circuits on the breadboard. I was even considering a hex keyboard - this would be pretty awesome, but hard to do; perhaps your circuit would be the way to go.