I can't seem to find a chart anywhere to tell me how big my traces need to be to handle a high current (500A) short duration (120us) lightning pulse.
Has anybody come across a chart that covers this?
All the charts I see are for continuous current.
I'm not an expert. But, here are some ideas.
You don't want to allow nearly all of the surge to conduct through your pcb to earth ground.
You will find that it is totally economically unfeasable to protect your system from a direct hit indicated by your 500amp spec. Not to mention your power cord and all wiring to the ground rod at the breaker panel will vaporize even if the pcb could handle it.
Most of what I have learned has been from Broadcasters keeping transmitters on the air in all weather. http://www.bdcast.com/fgal/white_paper/ground_BCEWHP.pdf
Phone companies rely on arc gap and MOV devices in the demarcation
box where the phoneline enters the building to shunt a lightning surge to earth ground. Cable companies
protect their systems by using a coax CATV grounding block connected to a
water pipe or the building ground rod.
I don't think forgoing similar devices is worthwhile.
On your pcb at the IO connector, you can use MOVs to dump any surge that leaks to your pcb to earth ground, not DC ground.
At the connector on your pcb, your breakdown voltage calculations will be constrained by the voltage rating of the MOV, usually around 450volts. Which is even easier to design to knowing the breakdown voltage of soldermask or fiberglass/epoxy surrounding internal trace layers is in the kV range. How thick should the traces be? I can only guess because there are a few variables to deal with. Instantaneous resistance of the MOV during a surge and the actual voltage of a surge. Ohms law for gamblers.
Thanks for the response, Robert.
This is for an aviation application, where we need to protect against surges coming in through our connector.
We have protection circuitry, but I've got to get there from the connector.
I will take a look at the broadcasting link you sent. That should be interesting.
Have you considered taking apart a surge protector outlet strip to see what they used? It isn't avionics or direct hit but it might give you a clue.
All I've ever worked with is stuff sticking out of the ground.
I wish you had a more knowledgeable source than us to ask.