[Top] [All Lists]

[Towertalk] single point ground

To: <>
Subject: [Towertalk] single point ground
From: (Tom Rauch)
Date: Wed, 24 Jul 2002 05:35:42 -0400
Hi Jim and all,

> First, you need to recognize that the 2nd floor windows are too 
> far from ground to be of value.  What you do there is cosmetic,
> and weather seal only.  

I agree as far as grounding to earth, but disagree as far as 
lightning protection. I'm doing something in a similar situation for 
a friend of mine right now.

One of the main goals in adding lightning protection is to clamp 
everything to the same potential so everything moves together. 
Grounding is second in importance to having everything entering the 
room tied together.

I would install a room ground to the same common point as the cables 
at the window. I'd have a plate inside the room where ALL of my power 
lines, telco, and equipment ground. I'd have one plug leave that 
common point to an outlet (or a 240 plug also, if you needed a 
separate circuit) and run all the equipment to a nice bypassed outlet 

I'd use the widest flashing possible to a ground, and route the 
coaxial cables down to a ground before bringing them up to the room.
Then there is one or two plugs to yank during non-use, and minimal 
worry if you forget.
> Second, in addition to providing a ground panel at ground level
> for antennas and control cables,  on the downstream (station) 
> side of them, I would route all cables through a piece of iron
> pipe about 4' long.  Ground it, as well, to the single point.

Why iron and why only 4'?? You might as well use any metallic 
conduit, or better yet just directly bury the cable. I think the idea 
of using iron pipe must come from some misplaced notion that ferreous 
metals add inductance to the path, but the effect is negligible much 
above dc. Since lightning has the bulk of energy above a few dozen 
kHz and contains no dc, the effect of that pipe is minimal.

There was a myth that even made it into the handbooks that fine steel 
wool would act as an RF choke or balun. In fact, it lowers the 
impedance of the cable shield passing through it by increasing the 
effective diameter.

As an example, when you drive your big hulking steel and iron car 
over the imbedded road coils operating an oscillator at a few kHz 
that sense for traffic lights, the oscillator frequency INCREASES. 

> It will act as a choke for transients which get across the
> ground panel, forcing arcs to ground at the SPG.

>From experience making direct RF impedance measurements I'm positive 
any change in series impedance of the system would be minimal. You'd 
be MUCH  better off to simply coil the coax in several large turns in 
air or on a PVC form than to stick it in a pipe.

The pipe does decrease shunt impedance, and if long enough will help 
move current out to the surface of the pipe due to skin effect, but 
that has nothing to do with it being iron. It can be any conductive 
material, and copper or a low resistance would actually be better.

Look at transformer and choke designs. Without thinly sliced 
insulated  laminations a soft-iron core has no ability to increase 
inductor or conduct magnetic flux even at 60Hz. At higher 
frequencies, you have to powder and the insulate each particle of the 
core to have it be effective.

With lightning, you'd need a high impedance from a few Hz up to many 
hundreds of  MHz, with the impedance peak in the dozens or hundreds 
of kHz range.

Someone either has a wild imagination or total lack of RF experience 
if they are telling people a short iron pipe would act like a 
significant impedance to a time-varying current like lightning.

Even a few turns of wire (say ten feet of overall length) in a large 
air-core loop would have significantly more series impedance than 
five feet through thick iron pipe.      
> I am assuming that what you do at the SPG provides an ESD gap,
> so excess voltage is dumped to ground across all coaxes and
> rotor/control cables.  ICE does this.  

All I do is use shielded control cables, and ground the shields. You 
could do the same by burying the cables or running them through 
any metallic  conduit (even diamagnetic materials) that is best 
grounded at BOTH ends, or at least the house end. The protected 
length has to be long.

I have diode clamps on dc control lines using 1N5408 diodes and 
electrolytic dump capacitors for minor surges. I have MOV's on my 
power lines and telco lines at the common point shack ground, and RCS-
8V's on antenna leads.

There isn't a SINGLE lightning bypass on any of my coaxial cables, 
including a repeater system that runs all the time and is in a 
shed under my 300 foot tower. Lightning hits have melted the shields 
of cables back in my receiving arrays and vaporized antenna 
wires, but I've yet to lose a single piece of equipment in the house 
of any type despite leaving everything connected.

Grounding at the towers and using proper common grounds with power 
and telco lines is the key. That is what makes or breaks the 

If it were my system, I'd bring the cables past the telco and power 
line entrance and ground them there, and then come up to the room 
entrance. I'd have everything in the room grounded or bypassed to the 
common point room ground, as well as all the feeders and control 
cables bundled with shields grounded at the service entrance of the 

73, Tom W8JI 

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