[Top] [All Lists]

Re: [TowerTalk] Grounding

To: <>
Subject: Re: [TowerTalk] Grounding
From: "Dale Sanders" <>
Date: Sun, 18 Nov 2012 12:51:01 -0600
List-post: <">>
This is the very reason that many of us that are hams and work with this
stuff on a day to day basis make very few comments.  It's those of you that
deal with the equations and book stuff that just don't understand what goes
on in the "real world"  While you may have never seen a conductor melted,
again, have you ever been in a cell site when the storm struck, or worked on
a broadcast facility after it got struck?

As for the cadwelding comments, you need to do your homework.  Burndy has
crimps that have been tested and equaled or exceeded the same connection
that was cadwelded.

If you don't work with this stuff on a daily basis, you might want to listen
to those of us that have been doing it for over 35 years.  We know a little
about what we're talking about.....and equations are fine but mother nature
has never taken a course in quantum physics.

-----Original Message-----
From: TowerTalk [] On Behalf Of K8RI
Sent: Saturday, November 17, 2012 8:33 PM
Subject: Re: [TowerTalk] Grounding

On 11/17/2012 10:51 AM, Jim Lux wrote:
> On 11/17/12 4:40 AM, K1TTT wrote:
>>> Dale - WD4IFR
>>> In any lightning strike, even a 2" solid piece of copper takes on 
>>> the characteristics of a light bulb filament.

I've never seen a conductor melt that has take a lightning strike here.

>> It is simple statements like this that make the rest of what you say 
>> questionable.  The fact is that MOST lightning strokes hitting even 
>> small conductors will not significantly heat them.  To get any 
>> significant warmth in 2" copper is probably even beyond even the 
>> largest measured mega strokes.
I'm not sure I'd go quite that far as I've seen holes blown right through
the wings of large aircraft from mega strokes.

> I agree... take a look at the Preece or Onderdonk equations for fusing 
> current.  Onderdonk takes into account the pulse length.

Many lightning strikes are a repetition of strokes that add up to
substantial time (relative to the length of a single stroke). Yet I have
never lost a coax tied to an antenna that has taken a hit and that includes
8X,  RG-6  CNT-240, LMR-400, Davis BuryFlexTM) and LMR-600.  I have had
insulators fail in N-Type connectors from over voltage.

I have to admit I do not like CN-240 and would never recommend it over 8X.
It has a tiny, copper plated, steel center conductor that is easily broken
with cable flexing.  The shield is foil with braid, but the braid consists
of very find wires that are easily broken while making connections.

The current is a summation of a rather complex wave form with relatively
steep rise and fall times. Most strikes with thousands of peak amps probably
have less than a hundred amps average for less than one second. 
(I'd guess at less than a 1/4 second for the event)  If you are calculating
fuse time, how can you not take into account duration which would be a
summation of the current wave form for the duration of the event.  IE you
need to know how many jouls the fuse can stand for a specific interval.

>>> BTW -- You do not have to use exothermic welds exclusively for your 
>>> ground connections.  One of the companies that make crimp on 
>>> grounding connectors is "Burndy".  These clamps are designed to get 
>>> away from Cadwelding but still require the use of a 12 ton crimper 
>>> to bond the

Big difference between those crimp connections and clamps.

>> I have never cadwelded a ground connection, most commercial 
>> electricians I

All of the ground rod connections in the antenna systems here are cadwelded.
The idea is to keep the "voltage" that eventually gets to the equipment
relatively low while attempting to keep all of the connections going into
the equipment at close to the same potential.
There can be thousands of volts difference from one end of the house to the
other, but hopefully everything in any one room including computers and rigs
are very close together.  The rigs can be bonded to each other, but the
network, cable (broad band) telephone and USB/serial connections from
computers to rigs needs to be kept within just a few volts which is a bit
more challenging.

>> know would probably look at you with complete confusion if you asked 
>> them to do that.  The basic brass clamp sold in every hardware store 
>> is what probably 99% of the ground connections in this country are 
>> connected with is just fine.

I've found those to be dangerous. Electricians and the power company use
them at the entrance/meter. I've never seen one of those connections last a
year at this location.  The soil is quite acetic and usually very moist
about 6 months out of the year. After a year you can dig the dirt away and
simply lift the clamp off the ground rod.  Not what I'd call good for
safety.  OTOH they did require 3 rods and they only used #6 stranded (the
green wire from the breaker panel to the meter and ground)

   After all a discharge that has just jumped a mile to the
>> ground

>> isn't going to be stopped by a few microns of oxide in a connection 
>> between two conductors.  Even lightning rod installers use easily 
>> hand crimped splices and connectors.

I'm far more interested in the thousand volt voltage drop across that tiny

The antenna systems that are above the top of a 100' 45G have taken many 
hits. (17 verified in the first 6 years, none in the last 5 or 6) On 
several occasions the strike has removed all the water proofing from the 
connections along with all the silver plating leaving the connectors 
with a "sand blasted" appearance and no trace of what they were plated 
with.  OTOH those 5 runs were all still connected.  I lost one 
polyphaser with no damage to any equipment.

> the need for exothermic welding or good clamps isn't for lightning.  As
> 'TTT says, a few thousandths of an inch gap or oxide isn't going to make
> any difference.
> The good quality bond is for more mundane electrical safety reasons,
> when the 110V wire shorts to the metal case, you want the circuit
> breaker to trip, or, at least, the voltage on the case relative to your
> bare feet to be limited.

If memory serves, The GFIs depend on very little current flowing through 
the green wire and they are quick. I don't remember which manufacturer 
put on the demo, but the guy used his body for grounding. When he took 
hold of the second wire the GFI tripped. He said he didn't feel 
anything.  I don't have that much faith in man made objects.

In the shop the electrician used a GFI duplex outlet as the first in 
each string, so it serves as the GFI for the entire string without 
requiring an expensive GFI breaker in the panel.

>> Oh, one more while I'm at it:
>>> copper braid that are cadwelded (exothermic) to the tower legs.  From
>> Braid?? Cadwelded??  I always thought this was a no-no.  and who runs
>> braid
>> for a lightning or safety ground?  Don't all codes specify solid
>> conductors?

We use all stranded in this area. The electrical inspector really didn't 
care what I used, but I went with #2 stranded as it was all I could get. 
  Nothing was available larger than #6 or #8 solid and that was at Lowe's.

> You can use stranded in larger sizes. There's a lot of 2/0 stranded used
> for this kind of application, for instance.  (because handling solid 2/0
> is hard work?)

That too was unavailable so we had to use 3/0 Al from the meter to the 
panel. That has a second problem. The Al cold flows so you have to 
retighten the bolts.  Stretching my memory, after 1 year it took over 1 
full turn to get the bolts tight. After another year it took about a 
half turn and the 3rd year added another 1/8th to 1/4 turn.  10 years 
later I've not been able to do any additional tightening.

> I'm too lazy to go get my code book and check, but there IS some
> threshold size for it.
> I wouldn't use welding cable (zillions of tiny strands, so it's real
> flexible).  I had bad luck using welding cable for pulsed power
> applications, although I never spent much time figuring out why.

Probably all those zillions of tiny connections/diodes in there<:-))

>> I wouldn't trust braid for an outdoor application like that anywhere,
>> unless
>> maybe you are talking about braid that uses something  like 10ga strands.
> AWG 2/0 stranded...19 strands of something like 16.

The strands in the #2 I use appear to be about #14.  The stuff is about 
as flexible as a rod.  It's available in various configurations.


Roger (K8RI)

> _______________________________________________
> _______________________________________________
> TowerTalk mailing list


TowerTalk mailing list


TowerTalk mailing list

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