[BC] More sparks

[clive@citiria.com]

>First, consider the shunt fed tower as an autotransformer.

*** Eh? It is not an autotransformer. Picture a narrow vertical triangle. The
right side is vertical. That side is earthed at the base. The left (slant) side,
at the base, is the driven part. The lines go all the way to the top of the
antenna. It is a folded dipole with top loading. I don't know what kind of
antenna you might be referring to.

                  /|   (Top loading radials not shown)
                 / |
                /  |
               /   |
              /    |
     50 ohm Feed   |
 ==================!======= Ground ===

>Second, remember that lightning is an ALTERNATING current.

*** No it isn't; it is a wideband pulse that contains frequencies well up into
the GHz region, but most energy is concentrated into the DC-17KHz region, thus
that's why 'growlers' can be heard at VLF. See appropriate books.

>Third, consider that the autotransformer "primary" is quite
>sympathetic to ground strikes in the near vicinity.

*** No autotransformer exists in the circuit (above) which can be viewed as one
half of a folded dipole with the other half being the ground image. The typically
low radiation resistance of a short vertical is multiplied by the folding to give
50 ohms j0 in a similar way that a standard dipole has Z=75 and folded, Z=300.
Therefore implying that the radiation resistance of the unfolded mast would be
about 12.5 ohms. Of course, the top radials change the characteristics quite a bit.

>The saving grace is most shunt feeds are quite inductive at
>the drive point, hence they typically have a large series
>capacitor bringing the feedpoint reactance to null.

*** The feed point is 50 ohms +-j0 when properly adjusted.

>However, it is vitally important to have a good spark gap,
>preferably of the Jacob's Ladder variety at the feed point
>end of the feed wire to prevent development of a pulse
>sufficient to break through the series capacitor dielectric.

*** Ah HA! Bad practice. Series capacitors are an absolute no-no.
Every RF designer and antenna guy I ever met knew that.
THAT is why you're getting lightning problems!

>In the case of a direct heavy strike on the tower, you may 
>find the arc gap is overwhelmed and the capacitor arced.

*** That's why it is bad practice.

>Thus begins the "interesting" project. Fortunately, the odds
>of a strike of this magnitude are low. :)
>
>Skirt fed monopole/unipole radiators generally do not exhibit
>quite the same behavior as the older slant wire feeds, and are

>best understood using a somewhat different concept. Owing to
>installation variations, these can be somewhat unpredictable 
>WRT lightning, however the hazards for the ATU are similar to 
>those of a series fed owing to the impedance of the supporting 
>structure below the attach point of the drape wires.

*** NO ATU. The attach point is the top of the mast. The support structure is the
mast itself.
Please provide calculations that support your point of view . . . 
I have not analysed the actual voltage present at the top of the earthed mast
when a strike occurs because I don't have the necessary information (series
resistance and structure self-inductance, just for starters) to allow that.
I do have photos of some of my installs of this type and I don't recall that we
installed any arc horns at all, but I will check. 
- Clive

>The bottom line is that if you can transmit RF, you are also
>a lightning "receiving" station. 

*** Not necessarily. I can transmit RF using a magnetic loop antenna if I care
to, and that wouldn't be prone to lightning! And neither would be a CFA, if you
believe in them (I am a sceptic but await the forthcoming tests near the Isle of
Man with considerable interest.)

Best wishes
Clive

Now regarding the next one:

>> Well, let's see if I can put some figures to that assertion.
>
> And you did quite well, with one criticle error.....
>
> And it is....
>
>> The mast end is earthed,
>
> It may appear grounded at visual observation of the obvious, but it is
> some number of degrees, and some turns ratio above ground, and
> therefore your voltage calculation is actually in series with, and additive
> to, the lightning voltage at that point on the tower, stepped up by the
> transformer effects of transmission line theory !

*** The base of the tower sits directly on the terminating plate for all the
ground radials. If that isn't earth, tell me what is, please. But I think I know
what you mean. You mean the antenna is not a perfect ground because of its
self-inductance.
My calculations relate strictly to the assertion made by Dana that the magnetic
field induced by the lightning current racing down to ground, induces a high
voltage in the slant wires. I calculated 800 V and that voltage is between the
feed point and the top of the tower. 

> ONE end of the tower is grounded, but that's not even ( electricly ) close to where
> the shunt tap is located.

*** Correct. There is some missing information, as I said above, relating to the
voltage at the top of the tower being faced with a self-inductance (the tower) in
its path to ground. A good question is whether this voltage is additive to or
subtractive from the induced voltage, because I did leave out one item from my
Faraday equation: the sign of the induced voltage is negative. (- dW/dT) volts.

But I am mainly a transmitter person; I need to talk to some dedicated antenna
engineers.