[Top] [All Lists]

Re: [TowerTalk] Grounding Coax at Tower

To: "'Jim Lux'" <>
Subject: Re: [TowerTalk] Grounding Coax at Tower
From: "David Robbins K1TTT" <>
Date: Thu, 28 Feb 2008 22:13:50 +0000
List-post: <">>
> > For those interested in such things the estimated surge impedance
> (similar
> > to coax characteristic impedance) for this tower works out to 240ohms,
> most
> > of it is pure inductance of course.  To get an idea of the total
> possible
> > voltage, if there were no ground that would make a voltage of about 7MV
> at
> > the top of the tower, with the ground connected that is reduced to 1.7MV
> at
> > .5usec into the stroke, well before the peak current is reached.  With
> the
> > insulators flashed over that peak voltage is only 800kv, this is because
> > lots of the current is drained away from the tower on the wires.
> 1.7 MV at the top of 120 feet. Is the voltage profile reasonably uniform
> down the tower (e.g. at 12 feet it's 170kV, etc.) or is it distinctly
> nonlinear?

For normal size towers it is pretty linear just because there is not enough
time for it to not be, very tall towers would likely be more non-linear.
The more interesting effect is the shape of the pulse.  Due to the
inductance and the current reflection from the low ground impedance the peak
voltage occurs before the max current in the stroke.  The rising edge causes
the highest voltages, but the time the peak current occurs most flashovers
have occurred and the reflected wave from the bottom of the tower has
reduced the voltage all the way back to the top.  In fact in this case the
peak voltage at the top of the tower happens at .5 usec.  Lower down the
voltage is more closely related to the current into the ground which looks
like the stroke waveform so at 10' the peak voltage is about 1.5usec.

> On tesla coils, where I (and others) have done lots of modeling and
> measurements, you get a funny distribution.  It's not linear because the
> C from a section of the secondary to the ground plane is larger at the
> bottom than at the top. Sort of like a tapered transmission line.
> That's for the fundamental excitation. There's also other effects from
> fast transients (when you get a spark, the rise time is tens of
> nanoseconds).  The folks killing their garage door openers and other
> consumer electronics with tesla coils kill them from the induced voltage
> from the field from the sparks hitting the ground.
> Even worse are small Marx banks.  2 or 3 nanosecond rise times aren't
> unusual, and a 300kV spark a couple feet long from low inductance
> capacitors.... that's a solid state equipment zapper.

big marx banks are even more fun.  At the epri lab in lenox they have a nice
one.  it can do something like 5MV when set up for max voltage, and we were
able to get about 30ka into single ground rods for full scale testing of
grounding with 280kj pulses.  If you blow up the first picture in the pdf
below it's the skinny tower just to the right of the big round tank looking
building right in the middle.  About 100' tall with an elevator up the
middle of it to service and configure it.

David Robbins K1TTT
AR-Cluster node: 145.69MHz or telnet://


TowerTalk mailing list

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