[time-nuts] GPS antenna and lightning

kevin-usenet at horizon.com kevin-usenet at horizon.com
Sun Oct 4 08:57:22 UTC 2009


Although not exactly trivial, it is actually not only possible but
surprisingly easy to design a setup that can indeed withstand repeated
direct lightning strikes without damage.

Large AM transmitters deal with this all the time.  Polyphaser
sells a lightning strike *counter* so you can schedule maintenance
on your lightning protection:
http://www.comm-omni.com/polyweb/lsc.htm

What you have to remember is to use electrical Aikido.  There is
nothing you can do that will possibly STOP a lightning strike, so don't
even try.  Your goal is to DIVERT the strike away from passing through
your equipment.

Some people really seem to have a hard time abandoning the mental image
of "stopping" lightning, so please forgive me as I hammer the point home
with several different examples...


It's not unlike dealing with explosives.  A *typical* lightning strike
dumps 500 MJ of energy, or 0.12 tons of TNT equivalent.  Wrapping that in
steel is called a "fragmentation grenade" and just creates more damage.
Explosives storage facilities don't try to contain that; rather, they
have strong walls facing directions to be protected and weak walls (and
roofs) in other directions so the blast can escape before damaging the
strong walls.

Consider a lightning strike to be the closest thing to an ideal current
source you are going to encounter.  You *cannot* stop it with series
impedance alone, no matter how high; you have to provide it with a very low
shunt impedance.

Likewise, remember that it's not the volts that kills you, it's
the amps.  It doesn't matter if your entire equipment bench bounces
1 MV, as long as all potential *differences* are small emough that
no damaging currents flow.

The basic layout is the "moat and drawbridge" illustrated at
http://www.sigcon.com/Pubs/news/7_02.htm

Lightning will take the lowest-inductance path(s) to ground.  Your goal
is to make sure that all paths through your equipment pass over a SINGLE
drawbridge, which is tied together with surge-diverting devices such as
spark gaps, gas discharge tubes, MOVs, transzorbs, etc.

Separate the stages with a bit of series impedance like unsaturatable
air-core coils.

In a typical AM transmitter shack, the drawbridge will take the form of
a big well-grounded steel plate on one wall sized to handle kilo-amp
currents without damage.  All wires, WITHOUT EXCEPTION, entering or
leaving the shack pass through feedthroughs in that steel plate.

Any damaging current would have to pass over the drawbridge, loop through
your equipment, and back to the exact same drawbridge to ground.
This is a dead-end path that current is not going to flow through.

It's not hard to make the high-frequency impedance of that loop something
like 1M times larger than the direct path through the steel grounding
plate.

Then, of a large but not extraordinary bolt of 100 kA, only 100 mA goes
through your equipment.  Suddenly, it's a lot less threatening.


But the secret is making sure that *every wire* to your protected
equipment has a low-impedance path to *every other* wire.  It's not a
matter of protecting them individually, because the protection does not
STOP lighting current.  You have to tie them all to the SAME ground point.

For example, if you have equipment plugged into two different surge
suppressors, you can have lightning pass in one and decide to send half
of its energy out through the other via your equiment.  On the way,
it lets out all the magic smoke.  :-(

You have to consider the impedance between each possible pair of wires.
Where is the shunt path, and why is its impedance many orders of magnitude
lower than the path through the protected equipment?

Just one little wire that's not tied into the system provides a path
that will let damaging currents come in through any other wire, no matter
how well "protected" they are.



More information about the time-nuts mailing list