>- Find a means of simulating the fault condition in the lab. This
>will give enable you to test that you have actually improved your
>circuit, and will tell you what the weak point in your design is. I
>once did this by winding a coil of wire near the circuit under test,
>adding a transformer primary in series to boost the inductance and
>flashing the leads across a battery untill I could the reliably make
>the circuit glitch at a given distance.

I have seen a small light box containing a fluorescent lamp with a normal
inductive ballast used for this purpose. It was very effective in producing
spikes on the mains supply when rapidly switched off and on.

Further to my previous mail, I would be looking at having a reasonably large
capacitor at each stage down the power supply chain that you mention,
possibly with diodes so if the input side really does die with a big
negative spike then the capacitor cannot discharge back into the source.
While you are at it do remember to put a reverse diode across the 7805
input/output to protect it. They go very belly up if there is no voltage on
the input, but a big capacitor still charged on the output.

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu