[TowerTalk] lightning tolerance

[Jim Lux]

I wonder if some of the time and effort put into building elaborate 
grounding systems for towers and such might not be better invested in 
making the actual signal paths more lightning tolerant.

For instance, it makes no difference whether the tower rises 1kV, 10kV 
or 30kV relative to the shack: they're all enough to cause potential 
problems.  So, is it worth it to put in more ground system than is 
necessary to make sure the tower isn't physically damaged?

For control lines, particularly for slow speed things like rotor 
controllers, I think it wouldn't be too hard to build bullet proof 
galvanic isolation or transient suppression.  Enough L in series and C 
in parallel, along with appropriate over voltage transient suppression.


it's the broadband RF signals that will be the challenge. but what is 
the real magnitude of it? You could probably help a lot (particularly on 
higher bands) with a DC block between equipment and feedline; but since 
160m isn't that far from where the energy peak is for lightning, I don't 
know that this is the silver bullet.  A DC block (and a DC ground for 
the center conductor, on the antenna side of the interface) would sure 
solve the electrostatic charging thing, though.

I think the "common mode" problem can be fairly easily solved with 
galvanic isolation techniques (transformers, capacitors, etc.), it's the 
differential mode that is going to be challenging: e.g. the voltage 
between center conductor and shield.

A kilowatt into 50 ohms is a bit more than 220 Volts, and you've already 
got to stand that.  If you put transient suppression that clamps at 750 
or 1000 Volts, you've already limited some of the problem.

The receiver front end is a challenge. The traditional approach is to 
put some sort of low voltage clamp (back to back zeners, for instance) 
across the input that limits the voltage to less than what the first 
mixer can take.

Or, to put a narrow band filter.  Maybe the long term solution to HF rig 
lightning protection is to put in a good filter that passes the HF band 
of interest, but is a short at other frequencies (including the 
lightning transient).  For "one frequency at a time" operation, that 
filter could be quite high Q, and acts as a preselector and could be 
tunable.

But more folks are doing waterfall and pan displays, so I think a 
"practical" solution would need to pass an entire ham band.