Circuit to test capacitive touch?

[groinksan]

Newbie alert! Newbie alert!!!

I'm looking to see if there is a simple circuit I could build that will basically detect the touch of a finger or other capacitive material, and then convert that to a signal that lights up an LED.

Some background information... Back in the 1980s, Keytronic used capacitive technology in its keyboards. It detects a key stroke by using two contact pads on a PCB that acts as a sensor, along with mylar material that acts as a capacitor. Each key on the keyboard has a foam pad with a layer of mylar material that hangs over the contact pads. When a key is pressed, the hammer pushes down the pad, so that the mylar material touches the two contact pads. When this happens, the capacitive reactance between the two contact pads drops. The circuit detects the drop, and somehow converts this to a simple SPST switch-like signal that can then be used for various things like, in the case of the keyboard send a keystroke to a computer.

One article that covers this, probably more accurately that my description is liked below:
http://www.eetimes.com/document.asp?doc_id=1272727

The article shows a simple schematic. But, it is more theoretical than practical, and the capacitance measuring circuitry discussed isn't illustrated to a level where I can just go to Digikey and get the parts to build the circuit.

I'm in the process of restoring Keytronic keyboards by making replacement foam pads. And, just one keyboard has almost 100 foam pads, so the task of replacing these pads can be a lengthy process. I'm planning on maknig literally thousands of capacitive foam pads. The problem is testing these foam pads is that I've been using an existing PCB from a keyboard, connected to a computer as a test tool. And, the longer I continue testing the foam pads this way, the more likely I'll accidentally break the PCB and possibly even the computer.

I'm looking for something much simpler to use. I'd like to find a simple circuit that allows me to drop a foam pad on top of two contact pads, and then light up an LED, letting me know the foam pad works.

So, if someone can point out a very simple circuit I could build to test these capacitive foam pads, I'd really appreciate it!

[Rerouter]

Look at a sample and hold circuit, (people use the adc's in micros to work as quick and dirty capacitive touch sensors)

you first charge up the system to your vcc voltage with a pullup, you then disable the pullup and switch in a discharged sampling cap, the voltage will then divide down by the increase in capacitance, this correlates to voltage vs capacitance, and could be pulled off with not many components if you wanted to do it discreetly,

(timer chip to give your sampling clock, a decade counter to sequence it, 1 mosfet for the sampling switch and 2 op amps, one to hold the sampled voltage, and the other as a comparator for the correct voltage threshold)

[mazurov]

If you need to use cap.touch in a keyboard discrete sensor circuits won't scale very well. MCU-based ones are much easier to scale. Both PIC and Atmel make micros with cap.touch peripheral, which (in PIC) is a current source you can connect to a pin and then measure charge time of a pad etched on a PCB, hence the name "Charge time measurement unit" AKA CTMU. With 100-pin PIC24 you'll be able to have a pin for each key plus any PC keyboard interface you require (AT, PS/2, USB, etc.).