At 05:15 PM 7/8/2006, Michael Tope wrote:
>With all this talk about lightning, I was thinking it might be
>prudent to throw in a few heavy guage bare copper conductors
>into the radial system that I am planning to install for my
>short 160 meter vertical:
>Most of the radials will be light guage copper (perhaps
>insulated depending on price), but I don't suppose it would
>hurt to throw in some big conductors for lightning protection.
>So the questions is, what size should I use #4, #2, ???
>stranded or solid (presumably solid is best), and most
>importantly what is my best bet in terms of price (electrical
>wholesaler, surplus, grand theft (just kidding), etc).
Check the current scrap price for copper, (there's a bunch of places in San
Fernando that buy scrap metal) and that will give you a sort of "floor"
The code requires a certain size if you're doing concrete encased, but I
suspect that this isn't really an application contemplated by the
code. You're sort of designing for yourself.
The maximum current will be at the place where the wire enters the soil,
and it will gradually decrease as it moves down the wire. I suppose you
want, as a thinner size bound, to keep the wire from melting from the
current. For that, you could use one of the standard fusing current equations:
The equations are usually for free air... given the thermal conductivity of
soil compared to air, you can probably take a bit more current, so the free
air number is a conservative estimate. I suspect you will find that the
diameter is surprisingly small, even for 20 kA kinds of current.
Then, the next thing you'll want to decide is what's the maximum "step
potential" gradient you want across the soil.. i.e. how many volts/meter is
acceptable. That will determine the effective resistivity of the soil/wire
combination. You can also use this to determine what the ground resistance
to "infinite distance" is, which would essentially tell you the peak
voltage at the base of the antenna (= Istroke/Rinfinity). For instance,
AWG 10 wire is 1 ohm/1000ft, so at 20kA, you'd see about 20V/foot along the
There's also the inductance issue... more small wires in parallel is
probably better than few thick wires, because you're getting closer to
approximating a flat sheet. Skin effect will make all the current flow in
the outer surface of a thick wire, so except from a thermal mass issue, big
wire is economically bad: cost goes as the square of diameter (mass),
impulse impedance goes as diameter.
The optimum use of copper, ignoring your valuable labor, is probably some
sort of brushy forking arrangement with few thick conductors at the base
connecting to many smaller conductors as you move out, etc. A sort of
dendritic pattern, if you will. [This would be an excellent exercise in a
calculus class, for an analytic solution, or a software class, for an
algorithmic search method]
>TowerTalk mailing list
TowerTalk mailing list