Towertalk
[Top] [All Lists]

Re: [TowerTalk] Lightning suppression through coax loops

To: towertalk@contesting.com
Subject: Re: [TowerTalk] Lightning suppression through coax loops
From: "Roger (K8RI)" <K8RI-on-TowerTalk@tm.net>
Date: Fri, 28 May 2010 02:37:02 -0400
List-post: <towertalk@contesting.com">mailto:towertalk@contesting.com>

On 5/28/2010 1:39 AM, jimlux wrote:
> Roger (K8RI) wrote:
>    
>> On 5/27/2010 9:27 PM, Gary Schafer wrote:
>>      
>>> Actually, Polyphaser recommended against putting loops in the coax line. The
>>> thought was that a loop would greatly increase the field and act as an
>>> antenna that would  increase the chances of induced current into other
>>> conductors.
>>>
>>>        
>>    The problem with loops, or I should say the potential problem with
>> loops is lightning does not like to travel in circles.Give it a corner
>> or curve and it's likely to get off the conductor and go some where else.
>>      
>
>
> Uhhh... I'd kind of like to see a physics explanation of not like to
> travel in circles or bends and getting off and going somewhere else.
>
>    
Well, lightning likes to take the shortest path although that may not be 
a straight line
Any conductor, straight, bent, circular, or crooked has inductance.
The rise time of a pulse creates a reverse EMF. The steeper the rise 
time the greater the reverse EMF.
A discontinuity in the coax such as a bend or coil increases the 
inductance at that point.
In the case of a coil and a rapid rise time the shortest path may be 
from turn to turn, to another cable, or to the tower rather than to 
continue through the coil.
I've seen lightning strike the top of a tower and get off half or 3/4 of 
the way down without following the guy wires. This was explained to me 
that the rise time created enough reverse EMF to make it easier for the 
lightning to get off the tower and jump to ground rather than proceed 
straight ahead even though the tower provided a substantial conductor.
> A loop has inductance. There will be some amount of voltage drop across
> the inductance.  But a 1 turn loop doesn't have a heck of a lot of
> inductance (I think 6" diameter is like 0.5 uH)... about the same as a
> half a meter of straight wire.
>
> Lightning has a rise time of about 1 microsecond, and peaks at say, 50kA
> (higher than most strokes, but makes the math easy).. so the voltage
> drop is  Ldi/dt is 0.5E-6*50E3/1E-6... about 25kV... that's big, but not
> huge.. not going to leap feet, in any case.  Maybe an inch or so.
> Certainly would punch through the jacket on most coax and arc to any
> touching metal.  (hence the electrical code requirements to space
> conductors that might carry lightning away from other ones)
>
>    
Yet I've seen lightning bolts in both photos and in person, that jumped 
many feet, or rather 50 to 60 yards when they had only 50 to 100 foot of 
tower between them and ground.


Having worked in industry I've seen electricity do some strange things. 
Two memorable ones were an electrician prying the cover off an overhead 
buss to add another switch.  That buss was composed of 3 1" copper 
rods.  When he pried off the cover the screwdriver slipped and contacted 
one of the phases (3 conductors - 3 phase) At that point the whole 
overhead buss with the three conductors and steel enclosure started 
burning like a fuse. Man, what an arc. Noise, light, fire, and motion.  
The current in those conductors was causing them to jump and arch 
between 6" and a foot until some one could make the mad dash to the sub 
and throw the "big switch".  Needless to say the burnt buss was not the 
expensive part.  Another was an engineer forgot three rather small 
grounding lines in a building sub.  Now these things were meant to 
prevent static buildup so the wire was less than #16. Probably 20 or 
22.  When the main was thrown to provide power to the sub it found three 
direct shorts to ground.  It instantly vaporized the three grounding 
lines and actually bent the 1/8" thick steel doors with reinforcing 
strips that were on the sub.

The energy in lightning bolts is many times the stuff we were working 
with, so I can understand the induced and reverse EMF values being 
enough to jump quite a ways. Normally in a transmitter, or a house they 
*can't* jump more than a foot or two without running into another 
conductor.  But lightning is highly variable so the voltages and 
currents vary by at least a magnitude and probably more than several 
magnitudes when anomalous lighting strikes are taken into account.  I 
don't worry about the small ones any more. As I've said many times, the 
tower here has taken at least 17 verified hits in what is now the last 
decade. How many times it has actually been hit is anyone's guess.  I do 
know that I've had no damage while several neighbors have had extensive 
damage, and I saw one pine tree blown apart with enough force to throw a 
section of trunk over 50 feet and stick it over 3' deep into the yard 
which was not muddy.

> I'm pretty sure the injunction against bends and loops in lightning
> conductors is for mechanical reasons.  The force tending to expand the
> loop *is* quite high.   A 50kA pulse rising in a few microseconds will
> easily rip a 1" diameter coil of AWG 10 wire apart (see, e.g., quarter
> shrinkers)
>    
I've seen less bend one inch buss bar.
> You don't want your lightning conductor ripping off the wall or loose
> from the supports.
>
>
> Running a 6" diameter coil with 2 turns and 1" long through the Wheeler
> formula, I get about 1 uH..  Not a particularly effective choke at 1
> MHz... a Z of about 6 ohms.
>    
True, but the one thing about lightning is its inconsistency.

What do you get for a square steel tower, 2' on a side? Yet the 
inductance was enough to cause the bolt to get off the tower and jump 
sideways.

73

Roger (K8RI)
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
>    
_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

<Prev in Thread] Current Thread [Next in Thread>