Topband: Shunt fed towers and common mode chokes
Tom W8JI
w8ji at w8ji.com
Fri Dec 7 12:21:17 EST 2012
> That's pretty close. It's a 110 foot tower, with a 4 element 40 (linear
> loaded) on top, and a KT-36XA at 75 feet. These two antennas have their
> feedlines isolated from the tower, and the feed methods at the antennas
> are isolated from the booms, so there is no low Z common mode path, and
> neither balun seems to be heating. The two-dipole 80 meter array is at
> 105'. Isolating the switching network could be challenging. It's a large
> metal box containing not only relays, but several big air variable caps
> and inductors. Some kind of polycarbonate or other material ?
I've had similar problems when I installed 160 dipoles on 160 shunt fed
towers, and on my insulated 300 foot tower. Since many people probably do
multiple duty with towers, this is a useful problem to look at.
If it were my problem, and I could not run control voltages through the
coax, I would use an AIR core bifilar choke with one winding being the coax
and the other being the control. I would stay away form cores entirely,
because you need a bifilar winding and the system needs to be reliable and
immune to lightning. It doesn't matter that the control and coaxial feed
cables are dissimilar conductors, they are still in bifilar mode so CM
voltages are equalized. Since the choke is on the feedline side, and not the
tuning side, you can use any type and length of coaxial line that suits
your power needs, like RG400 cable (if you shield it from UV light) and any
type of reasonable size control line (although a shielded line with the
shield parallel connected to the coax shield would certainly be best).
I would absolutely ground the shield to the tower.
I would make a support of fiberglass rod for the box, and add the isolation
choke between the box and the tower.
If you wanted to save a few turns, the coax (box) to tower could have
intentional resonating capacitance. It only needs be across the coax shield,
because mutual coupling will take care of the control line conductors. That
would reduce bandwidth, but even fairly high values of C (50 pF or more)
would still be fairly wide bandwidth on 160. You probably actually want the
choke to be capacitive in reactance, placing resonance below 160 meters.
That would reduce choke (actually it is functioning as a trap) heating.
Worse case heating is when resonance is above 160. Even with 50,000 ohms of
resistance, you are going to have somewhere around 10% of applied power as
heat.
I'd be very careful about the 80M baluns you use, because the one size fits
all rule won't work well when the system is excited on 160. I'd use a higher
Q more reactive type of balun on 80 in this case.
I don't see many other solutions to your problem, other than high-Q resonant
160 meter chokes in each dipole feeder. While that is possible, it is a more
complex engineering problem and likely would restrict bandwidth. The
mechanical issue of suitably isolating the box from the tower is easier than
the electrical issues caused by mounting it to the tower.
Another entirely different concept (that works well with simple dipoles)
would be an isolation transformer for the coax, but it will not solve
control line issues. If your system used an isolation (primary-secondary)
transformer for RF, you would still be stuck with control wire issues for a
floating box (each conductor would require something like an amplifier's RF
plate choke) and the open circuit impedance of the transformer if the box.
Also, open circuit impedance of the transformer is critical in your
particular application.
While I'm sure a tower mounted box could be used, the electrical design is
far more complex and more compromised.
73 Tom
More information about the Topband
mailing list