[TowerTalk] Lightning Protection Question

[jimlux]

Jim Brown wrote:
> On Sat, 11 Sep 2010 21:28:55 -0400, Gene Smar wrote:
> 
>> You overlooked my caveat in my original e-mail:  I have a remote 
>>> switch
>>> in the NEMA box at the base of my tower.  Had I installed the Polyphasers 
>>> at
>>> the shack end, only one coax run (between the shack and switch) would have
>>> been protected; the switch and antennaa coax runs into it would have been
>>> vulnerable to lightning damage.  I installed the Polyphasers in the box on
>>> the tower so as to provide protection for the switch and all three coax
>>> lines going into the switch.
> 
> Gene,
> 
> I didn't misread or overlook it at all. The function of an arrestor is to 
> protect sensitive (and expensive) electronics from a strike. ALL IT DOES IS 
> SHORT THE COAX. That remote switch is neither sensitive nor expensive. 
> Although, as K8RI noted, it's easy to do that with the switch, the coax isn't 
> connected to the shack, so it doesn't matter. So I'm sticking with my (and 
> K1TTT's) advice. Arrestors go as close as practical to the sensitive 
> electronics that they protect. Anywhere else is a waste of money.
> 



Let's look at two scenarios with respect to tower mounted protectors:

1) Direct hit - protector fries (or at least breakdown device requires 
replacement) or relay fries.  Relays and protectors cost about the same 
to replace.
2) Nearby stroke inducing a transient - protector protects without 
dying, relay shorted transient to ground without failing.

Now.. one *can* design a relay box to be more robust.  THis might a good 
application for a traditional spark gap type "lightning arrestor" 
overvoltage protection. A simple gap with a breakdown voltage of say, 
3-5kV would protect most relays, and be able to conduct the full 
lightning current without damage.  The gap would be about 1mm or so, and 
would have a drop of a few hundred volts when the tens of kA going 
through it.

The challenge in design is making sure all the "lightning current 
carrying" paths are AWG12 or AWG10 equivalent so they don't melt, and 
that the gap is setup so that the heat dissipated doesn't cook something 
(thermally).