Best way to model a lightning strike

[Phaedrus]

What's the best way to model a lightning strike? For instance in a simulation or calculation. Is it best modeled as a capacitor? Voltage source? Current source? Transient voltage waveform?

In the context of simulating the effects of lightning strikes on off-line power supplies and equipment.

[NiHaoMike]

A pulsed high current source. Pretty much all you have to do is give it a path around the devices to be protected.

[IanB]

I guess you probably want to model the effect you are defending against?

I believe a lightning strike is complex, in both the time and frequency domain. There is a strong DC element with a large amount of RF superimposed.

In industrial settings like broadcast antennas, they defend against lightning strikes by equi-potential bonding. By keeping a low resistance path between all parts of the apparatus everything rises and falls in potential at the same time and excess potential gradients cannot arise within the system and produce fault currents. So I imagine you could simulate this situation by suddenly stepping one end of the system up in voltage by a large amount?

In other cases maybe RF filtering is important, so you could simulate the injection of a wide band RF signal?

[Rufus]

What's the best way to model a lightning strike? For instance in a simulation or calculation. Is it best modeled as a capacitor? Voltage source? Current source? Transient voltage waveform?

This http://www.hvtechnologies.com/LinkClick.aspx?fileticket=paGvyaJoRgM%3D&tabid=190

Gives you the waveforms and levels (for different equipment categories) used to test aircraft equipment against indirect lightning effects.

The transients are applied with respect to earth directly to 'pins' or to a whole cable bundle. The dc transients are applied with both polarities.

[mikeselectricstuff]

Speaking of which....

[Phaedrus]

Thanks!

I mainly want to compare different SMPS topologies, with and without PFC, with and without MOV and ZVS diodes, to see which combinations have the best survival rate. Best case being the PSU and attached electronics surviving; acceptable being the PSU dead, but attached equipment surviving; worst being both power supply and attached equipment fried. With a low impedance path to ground as well of course, but there's still some energy delivered to the PSU.

[NiHaoMike]

HP used to put a MOV right after the fuse, but moved it in front of the fuse in later revisions. Apparently, they were having too many returns due to the MOV blowing the fuse.