"If you have one piece of gear with a 10 foot cable to the lightning
impulse. And another piece of gear on the bench connected with a 100
foot cable to the lightning impulse, and then you interconnect the two
with a short jumper, you can see that there might be a problem."
And THAT is the crux of the problem when bonding equipment for lightning
protection. NOTE: I said BONDING not grounding! Consider this... the 10'
cable is connected from a rod/shack/power 'ground' to the radio and the 100'
cable is a coax that is not connected to the rod/shack/power 'ground' but
goes right from the antenna/tower to the radio.
First a little math to show the magnitude of the problem. .. An average
lightning stroke is about 30ka. Lets say it strikes the tower/antenna. So,
down the tower the current goes. When it gets to the bottom it sees a
ground system, lets say its good and fat and looks like 5 ohms to the
current, while going to the shack the cables are above ground and skinnier
than the ground so they may look like 100 ohms, this 5 vs 100 ohms splits
the current so most of it goes into the ground. But even 30ka into 5 ohms
is 150kv at the base of the tower. First assumption is that none of the
insulation between the shield and center conductor is good enough to
withstand the voltage at the base of the tower. So at the base of the tower
you now have 150kv on the shield and center conductor of the coax. note
that this is really the best case, it is likely the connection from the coax
to the antenna flashed over already and the center conductor is at an even
higher voltage but lets go with the 150kv for now.
Next some more math and rough assumptions... we now have a 150kv step at the
tower and it has 2 paths to the shack, one on the coax the other through the
ground. Yes, there will be a voltage step propagating along the ground even
if you don't connect the tower ground to the shack ground. In experiments
I have done the speed of propagation along a shallow buried wire is about
.5c or about 150m/usec (450ft/usec). Lets also assume that propagation
along a conductor above ground was about .9c or about 270m/usec(800ft/usec).
These are both rough figures that depend on the soil, depth of the buried
wire, height of the overhead wire, etc, but they'll do for now. Now in that
100' run to the shack it takes the ground current about .2usec, while the
overhead current takes about .1usec... roughly. So now you have 150kv
arriving at the radio .1usec before the voltage on the ground rises up... we
will ignore the fact that the current in/on the ground is spreading out and
thus the voltage there rises less than it would directly at the tower, that
only makes it worse after the initial .1usec event.
So, what happens now. Well, the current on the shield gets connected to the
radio case(and operator if present) and continues down that 10' wire to the
rod/shack/power ground. However the current on the center conductor gets
inside the case and lets just say that most radios are not protected against
150kv for even fractions of a micro-second so the radio power supply becomes
the best path to ground and out comes the magic smoke. Yes, there are
possible mitigating factors, for instance if the radio is connected to a
circuit that has a protector at the rod/shack/power ground once the current
gets to it from the case of the radio it will equalize the voltage between
the ground and the power conductors which will after .04usec (roughly a 10'
round trip from the radio to protector and back to the radio) equalize the
voltage from the power supply to the rest of the radio, but by then the
damage may already be done.
What is the fix? Well, if everything is bonded to the same point at the
rod/shack/power ground with protectors all the wires get their voltages
equalized so there is no flashovers inside the equipment being protected.
Just note, all it takes is one unprotected conductor and that becomes a
target for a flashover from the 'ground' to whatever it is connected to.
You can do similar analysis for a loop where there is a gap at one point,
but that gets much uglier... just think about the early experiments with
spark transmitters and how a spark would be induced in a receiving loop
across a small gap... but scale up the currents and size of the loops by a
lot.
David Robbins K1TTT
e-mail: mailto:k1ttt@arrl.net
web: http://wiki.k1ttt.net
AR-Cluster node: 145.69MHz or telnet://k1ttt.net:7373
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|