David Robbins wrote:

SNIP


lightning doesn't go on for milliseconds,

SNIP.

More specific information. The continuing portion of a lightning stroke,
values up to 600 amperes, can go on for almost 300 milliseconds.
Reference, WS-118-41129, Paragraph 3.9.XX [XX= I forgot the sub
paragraph], Lightning [USAF system specification for the WS-118 Missile].

There is a USAF model, released in 1982, that encompasses 90% of all USA
lightning strokes. It is as follows:

Peak stroke; 0 to 100,000 amperes in 1 microsecond.
Peak decay; 100,000 amperes to 25,000 amperes in 25 microseconds
First tail; 25,000 amperes decaying to 600 amperes in 1 millisecond
Continuing current; 600 amperes constant for 300 milliseconds.

There may be up to 6 continuing strikes with amplitudes at 1/4 to 1/2 of
the above.

50 micro-seconds is a relatively
long stroke. 30kA can go through a 12ga copper wire with no damage for
10-20 microseconds. in most cases there will actually be very little
voltage between wires of a coax or twin lead just because their insulation
will break down or the feedpoint of the antenna will arc over

SNIP: Arcs can sustain 100s of volts in the dynamics of the arc [plasma
or carbonized material].

.... both are
naturally occuring spark gaps that actually work very well to protect
equipment from direct strikes. assuming of course the tower and feedline
have good grounds. where people have problems is they don't ground the
shield of the coax to a single point ground along with the power lines, so
they get differential voltages between grounds that has no place to go but
through the equipment.

SNIP: Great advice!

properly grounded installations with relatively
small arresters to limit voltage on the center conductor of the coax
relative to the shield are very effective. for tube type receivers a simple
spark gap is adequate, for transistorized stuff you may need lower voltage
protection and should probably get something commercially made for the job.

SNIP: for a high power solid state station, 1500 watts, the matched RMS
voltage is 274 volts, the maximum peak to peak is 274*2.828 = 774 volts p-p.

Any surge device must accommodate the high RMS voltage and yet the
receiver/transceiver front end must tolerate 774 volts p-p without damage.

My solution, disconnect the antennas, radio power lines, etc.

Deacon Dave, W1MCE