[BC] AM transmitter lightning sensitivity issue

[Gregory Muir]

To all,

I'm still heading in the direction of looking for obvious transient induced 
problems as opposed to slow static buildup at this time.  I appreciate the 
recommendations regarding static dissipation devices on the towers but there 
are two factors at play here at present.  The first involves the fact that 
the transmitter only responds with reducing power or shutting off to fast 
events, not slow static buildup on the towers.  We have never witnessed the 
transmitter changing power levels with simple static (as in non-changing) 
electrostatic fields (how do I know when the area is under the infleuence of 
an electrostatic charge?  See the last two paragraphs) .  That is a good 
indicator that the static drain hardware in the TU shacks are doing their 
job.

On the other hand, transient - and I stress transient - events are the real 
killer here.  The normal situation places the actual lightining discharge 
points miles away from the site.  Usually storms within the range of 15 
miles distant will start the problem.  That's where the mystery begins.

The second factor involves dealing with a corporate structure that does not 
really honor static dissipation devices mostly from a financial standpoint. 
But there is also a certain mindset at work here, too.  I am quite familiar 
with the products and their uses and do have an open mind about them. 
However there are some very negative naysayers out there (I came across one 
at http://lightning-protection-institute.com/fact-fallacy.htm ).  I do know 
that various systems have been implemented at high strike regions such as 
Cape Canaveral and know that they have been touted as being effective in 
static dissipation.  But I have never read any writeups that involve what 
they could possibly do about distant storm effects if any.  Looking at the 
problem from both ends (both site and storm) I can only draw a theory that 
we are looking at either rapid cloud charge equalization or EMP.  But the 
actual solution lies in finding out how these events are being handled (or 
not being handled) by the transmission system.  That will involve some 
rigorous investigation.

I'm not disregarding the suggestions that all of you have offered.  I have 
noted them.  Being from an analytical background, my mind is working on how 
to implement a simple instrumentation system to see what is happening when 
these events occur.  Power lines have alread been address using line 
analyzers.  It became fairly obvious that storms miles distant had little or 
no effect on line transients.  The same applied to ground differentials.  So 
my attention has turned to the RF side of life.  I know it is always 
difficult to look for fast transient events on an active transmission line 
so am leaning towards disconnecting the meter from the base current metering 
transformer in the TU and connecting its output to a scope to see if 
anything can be observed.  Obviously I have little inclination to be sitting 
in a TU hut during an approaching storm (I have other interests in life, 
too) so will look at a possible analog FO link out to a much safer place. I 
suppose I could use the remote base current metering wiring but am wary of 
the frequency response of this setup given the considerable distance to the 
transmitter building and the cable capacitive effects resulting from that. 
Optical cable is cheap and doesn't degrade fast pulses.

As for later suggestons and comments, the Johnny balls do not have paint on 
them.  They have seen use over the 60 years they have been in place but each 
subsequent tower painter left them alone.  The comment on placing 100k 
resistors across the guy insulators is good.  I noticed that Kintronics now 
offers a commercially built version of a tower guy insulator static drain 
choke.  Probably not as cheap as a resistor (!).  The comment regarding 
installation of ferrites around the transmission line being somewhat useless 
if the line is running through metallic conduit is understood.  But I still 
don't understand why electricians insist on running ground wires through 
metallic conduts that extend from power line transient protection equipment 
(again - !!!).  I've educated a few on this point and had them change the 
conduit to plastic.

The center TU hut (day active) had sustained a fire back in the later 90's 
and was rebuilt.  As part of the effort, a new Kintronics TU was factory 
installed with appropriate grounding.  Inspection of the grounding around 
the new building revealed a fairly comprehensive ground system which appears 
to be of recent vintage.  From that I derived a feeling that it was probably 
of good construction.  The tower has four 4-inch straps extending from the 
base insulator base down the concrete pier to the ground system.  The feed 
from the tower does go through a one-turn 1-inch copper tube loop prior to 
entry into the TU hut with a static drain choke connected immediately inside 
of the bowl insulator.  From there the line extends to the Kintronics TU 
where it meets with a secondary spark gap prior to entering the tuning 
elements.  The entire inside of the TU hut contains copper screening which 
is bonded to 4-inch copper strap running the periphery of the room and is 
connected to multiple copper straps that exit the space on all 4 sides of 
the structure.  All cables leaving the building are bonded to the metal TU 
chassis prior to exiting.  Caps in the RF path in the TUs and phasor?  Yes.

In reference to a recent comment:

"After you make a new installation, while
the transmitter is still in warranty, you should
deliberately create arcs at the two arc gaps
with the transmitter modulated and running
full power. You can use a long screwdriver
to create the arcs. Both should self-extinguish
and, ideally, the transmitter should stay on-the
air. In any event, no damage to anything should
occur."

I happened to have a tower climber on site who decided to "test" the 
transmitter while it was on the air by tapping his Crescent wrench across 
the Johnny balls.  The transmitter immediately shut down without first going 
through the power reduction cycles it normally executes.  This was the first 
event that started me to thinking about how there could be a possibility 
that the transmitter VSWR circuitry had been designed too sensitive for an 
average field environment.

I do have the Nautel site documents in possession.  A handy reference. 
Other excellent references are the Motorola R56 standard for communications 
sites (use this and government contract monitors will not ask questions 
about your work) although it leans more towards land mobile, cellular and 
whatever applications.  Also very good are the Verizon plant practice series 
(fundamentally the old AT&T/Bell System series) that covers site 
installations (if you want to have a Cadillac site design).  But again it 
addresses mostly communications sites.  There are also a few military 
standards that copy the commercial versions (and vice versa).

I once had a non-protein (dogless) form of storm prediction.  Read on...

In my old vehicle I had a permanent-mount VHF whip antenna that was 
installed on the top of the cab.  The associated radio had been removed but 
the cable connector was still hanging under the dash where it was somewhat 
visible.  This particular make of antenna (Antenna Specialists, I believe) 
was a 5/8 wave of the base loaded variety but did not have the classical 
loading coil that shunted the antenna element to ground in a DC path.  When 
in rain containing highly charged raindrops or under non-precipitative 
charged clouds, a tiny, thready, pulsing arc would occur from the center 
contact of the connector over to the grounded shell as I drove.

Aside from being a very interesting conversation piece for riders in my 
vehicle, I also found it to be a good gauge of the magnitude of the charge 
in the clouds above me and a good predictor of an impending lightning 
strike.  With arc frequencies in the range of 0.1 to 1 pps, I knew that we 
had a safe charge potential with little chance of a lightning strike.  But 
if the arc frequency exceeded 1 pps and was observed to slowly increase in 
frequency to as high as 10-20 pps, I was guaranteed that a strike would 
occur somewhere in the surrounding vicinity.  It never lied.  And I'm 
considering installing a comparable setup on my new vehicle except might 
connect the coax center lead to a neon bulb or similar device (fluorescent 
tubes are rather bulky).  But the original arrangement was less distracting 
becasue I could audibly hear the "snap" when things started to happen.  I 
was never taken by surprise by a nearby lightning strike again.

Whew!  I've said enough.  Other duties call...

Greg