Here is some further info about lightning protection from many years
of win and lose experience.. In operating seismic telemetry stations
with VHF transmitters it is obvious that the antenna in the tree or tower
is the lightning rod, and the seismometer buried in the ground is the
ground rod (or will try to be). So successful lightning protection is 
essential to equipment survival. In general, we have had more losses of the
VHF radio transmitter than of the telemetry electronics itself. There
were also notable disasters, like when the lightning arced right through
the 1/4" ABS pipe vault from the case of the L4-C to ground. (There is no
rule that lightning has to even LIKE my grounding system and use the 
ground rod properly!).

But for the most part the equipment survives. Unfortunately, one never
knows for sure when the protection works; only popped fuses, transistors,
amplifiers and smoked boards say it didn't.
So the protection is multi-level and inserted in any power or signal
interconnection, including the coaxial line to the antenna and the 
DC power controllers for the solar panels. It only takes about 20 volts 
to burn out the RF output transistor of the radio transmitter which 
runs at 14 volts, so a system based on 18 volt MOVs is needed..

Until the advent of MOVs (metal oxide varistors) transorbs were the
device of choice. These are essentially high power zeners designed to
take a surge, and actually have a dissipation rating in joules. They
are expensive, hard to test (without blowing them), and give no external
sign that they are in fact blown. But they are fast, since they have 
low capacitance.

So MOVs are preferred, since they clearly self destruct if their power
rating is exceeded.. But they are also large capacitors, so they are 
slow. This requires some delay element (usually an inductor) to force 
the surge voltage to rise and fire the protection devices. Inductors 
are also used to isolate a VHF signal from the lower frequency transient.

I generally use several levels of protection, including the 60-volt 
gas discharge devices (available from Newark, Mouser, etc, for under $3.)
These latter may not protect a 12-volt transistor, but will help 
reduce the physical damage when the lower voltage MOVs are overwhelmed,
such that the repair may only involve replacing the transistor.
For capacitance-sensitive applications, like  a 50-ohm antenna feed,
a small 70-volt neon bulb will work. I installed them at the UHF or N
connector of all the VHF transmitters and receivers. They tend to explode
if overloaded.

I have also found that lightning suppression systems should not be
built on a printed circuit board; the lightning just flashes across
the surface, vaporizing the foils. I use high-barrier terminal strips
with point-to-point wiring. Cleanliness is also important also; dried
sweat is a good conductor. Fuses are installed in-line by using the
GJV type inside clear shrink tube, connected with crimp connectors. 
 They are used in both sides of the DC power and signal (both siesmic 
 and FM carrier) lines.These limit the damage when the protection 
 devices conduct the surge to ground.

For testing the lightning protection, we use a small electric fence
transformer. It could spark across an 1/8" air gap. We use a 110v lamp
in series with the AC primary to limit the power.

For the ground rod, standard copperweld rods are used; since my stations
are in generally wet areas (Southern Missouri, the Aleutians, etc), I
have not needed to salt the ground as Mike L. recommends. But location of
the rod is important; it should provide as direct a path from the 
antenna's coaxial feed to the ground at the base of the tower. For a
building, the rod should be installed as close to the probable entry
location of any lightning. Most buildings have the lightning ground
rod installed right under the electrical meter. The electrical panel
ground may also be connected to the cold water pipe.

I there is further interest, I can scan the circuits I use, particularly
those used with the seismic amplifiers and telemetry.

Another note about protecting the AC line: I have had unlimited success
using ferroresonant constant voltage transformers to block transients
on the incoming AC line. These are made by Sola, etc, and are expensive,
costing about $1 per watt of capacity. But they will not pass any power
other than a 60 hz sine wave. They last almost forever, so finds at surplus
stores usually work. After about 20 years the resonating capacitor fails
and the insulation dries out, since they run hot.. They have no "moving
parts", like the relays in the inexpensive switched line regulators, and 
will put out 110 volts from an input of 90 to 140 volts. But, as I said, 
their main virtue is that they only output a 60 hz sine wave. I have
used one at a key telemetry station with 200ft tower since 1972, and
have never had any lightning damage. I have also been at the 'head-end"
building of a cable TV system, where we rent space 300ft up on the tower,
when lightning has hit the tower, and everything survives, including
the local seismometer/telemetry unit.

Regards,
Sean-Thomas


_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)