Hello all,

As some of you may remember, I encounted a micro stall on my board when pnumatic pistons were switched on.
After many, many tests I realized that the spikes that appeared on my board were induced and not conductive (ie via the supply).
When the lead length from the board to the Pistons were reduced the amplitude of the spike also reduced. I think that these leads were acting as transmitters and the board was acting as a reciever. Please correct me if I am wrong.
The reason the PIC board operates fine is that its input operating range is 2 to 5V and the 89S51 is 4 to 5.5V. This may also be the reason my micro stalled and the PIC did not.
Again, please correct me if I am wrong.
When my board was opperating close to (but not in any way connected to) the pistons while they were being switched by the PIC board, my board still stalled after a while.

My plan of action from now is to redesign my PCB layout, and provide my micro and digital cctry with a stable supply using a charge pump, a diode network to clip the negative spike and clipper network to clip the spike at 5V.

My problem is at the frequency of the spike ie 10Mhz, capacitors have a very low resistance and therefore provide a path to ground when connected in parallel to the supply output. I do not understand how the cap can be used as a pump to provide a stable supply when it is providing a path to ground at this frequency.
Also for a 5V clipper cct I need to use a 5V cell in series with a fast switching diode. Due the application, the use of the a cell is not an option, is there anything else I can use instead of a cell, a Voltage reference chip perhaps? Is there are any other method I can use to Clip the spike at 5V.

And finaly, at 10 MHz will the voltage regulator (LM7805) be stable?

All and any comments are welcome,
Thank you for your time.
sun,