Simple touch probe crashes AVR

I built a single touch probe that shorts one of atmega pins to ground. When I test it with a piece of wire things work fine, but when I attach everything to my repstrap triggering of the probe kills the sanguino. I've tried adding a 100k ohm resistor to the path, but so far with no success.

What am I doing wrong? How should one design such a probe?

... maybe you've received static high-voltage?

When moving or rubbing/pushing plastic components, you can easily accumulate some kilovolts, so you should apply some safety against this too ...

Viktor

VDX Wrote:
-------------------------------------------------------
> When moving or rubbing/pushing plastic components,
> you can easily accumulate some kilovolts, so you
> should apply some safety against this too ...

This happens with just metal components, I'm running the probe against some PCB substrate. I think it has something to do with the spikes generated by the sudden close of the circuit. Maybe a few caps and current limiting resistors will help?

... with long wires and closing through the housing you have an antenna, which receives all the ambient frequencies - e.g. the power line or nearby HF-sources like stepper-drivers or such.

Try something with shorter wires and/or twisted pairs and parallel capacitive or serial inductive filters ...

Viktor

And the problem is solved. Turns out Sanguino has a slightly different pin layout than the atmega644p that it houses. Notably one of the GND pins was actually VCC. No wonder things didn't work they way I expected them to..