[geeks] Whee! Lightning strikes, AGAIN!

[der Mouse]

> I would guess that the surge didn't come from the switch and only
> passed thru it using a common path shared by the used ports.  Without
> looking inside I would think it was a ground path(s).  All the signal
> leads would pass [straight] into the chip and it would be harder for
> the surge to keep moving to ports on other chips.

I _think_ twisted-pair uses transformer isolation between the cable run
and the hardware; if so, this means there's no DC path between any of
the wires and the rest of the machine until you exceed the insulation
breakdown voltage.

One possibility is that the lightning-induced surge was enough to do
just that.  There's a _lot_ of power riding a lightning strike, so it
doesn't take much of it coupling onto wires to do damage.

Another is that the differential power induced was enough to blow the
transformer winding, in which case the port will be functionally dead
even if nothing else is damaged.  (Twisted pair should make this
difficult; the point of twisted-pair is to induce the same effect on
each wire of the pair.  But there usually are small parts that aren't
twisted, such as etch runs between jacks and circuitry, and lightning
is funny stuff - I wouldn't rule it out without checking, unless I'm
totally wrong about the isolation transformers.)

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse at rodents-montreal.org
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B