A big thank you to Frank, VO1HP, St. Johns, Newfoundland, whose
successful version of the L + FCP + short tower is the genesis for
this method.
FCP refers to the 5/16 wave single wire folded counterpoise for low
bands, described at http://www.w0uce.net/K2AVantennas.html At some
point an illustrated version of this post will be available on that
web page.
Caveat:
-----------------------------------------------------
I don't have Frank's setup, nor does anyone close by (all using
trees), so I haven't been able to do personal hands-on independent
validation. But the NEC4 ** modeling supports both Frank's experience
in Newfoundland and the "suck out" degradation I have experienced with
vertical wires close to resonant/near-resonant towers. Since we've
had so many inquiries about doing L+FCP off a tower, we decided to
release this information anyway and let people give it a try. I
expect we will gain more experience rather quickly. We will keep
readers posted on developments. ***
Super short version:
----------------------------------------------------
This is a simple solution of limited application range for reducing
lossy interaction between a 160 meter inverted L and a short tower
supporting the bend. The method does not involve tower detuning, or
complex anti-resonance schemes.
Up top, support the bend in the L 1-3 feet from the tower. Locate the
FCP feed and isolation transformer box 13-15 feet horizontally from
the tower. Pull the vertical run straight from the isolation
transformer box up to the bend. The wire up to the bend will be at an
angle to the tower, 10 or more degrees off vertical.
===> There are some nasty devils in the details, ways to lose all
your gains. Read everything carefully. <===
Short version:
----------------------------------------------------
Some installing an FCP under an "L" must support the L's bend with
their short tower because it's the available high point on a small
lot. This tells how to do that without losing gains from FCP to losses
where the tower goes into the ground. It requires the tower plus top
load being resonant well above the 160 meter band. Performance
*improves* with *poorer* RF connection between the tower and ground.
The method consists of :
a) how to run the "L" and locate the FCP in relation to the tower.
b) removal of existing tower base radials, ===> BUT lightning
ground rods and connections remain. <===
c) breaking up destructive resonances in other close antenna + cable
connections by using or adding balun(s) and common mode choke(s) WITH
FULL RATINGS ON 160 METERS.
If your tower is bristling with large antennas up top this almost
surely won't work for you. If it's taller than 65-70 feet and has an
HF beam, this solution most likely won't work for you. This solution
is specific to a single wire antenna, commonly called an "inverted L",
run vertically, then bent to a horizontal direction with the bend
supported by a tower. The far end of the "L" away from the tower can
be very miscellaneous, even a partial inverted U, without affecting
the solution.
Long version:
------------------------------------------------------
Frank has a small lot in St. Johns, Newfoundland, with a layer of dirt
on the big rock that is the island. The bend in his "L" is supported
by a 55 foot crank-up tower topped with a small tribander. Short
miscellaneous radials were not working with his "L". So he converted
to an FCP. He kept the tower supported bend, but the FCP had to be
located away from the tower due to yard circumstances. Frank then put
up surprising summer numbers at EU RBN sites, and worked
(unprecedented for him) new top band countries in summer season.
That REALLY got my attention. Low band vertical wires closely
supported by a tower have always have possibility for severe
interaction loss. One can model a driven vertical wire next to a
tower where the tower's induced current exceeds the driven current on
the wire. For all practical purposes the vertical wire was just
another way to load the tower. Modeling a correspondent's actual
tower/tribander/L/FCP dimensions always came out middling to poor to
awful. On small lots the tower was just one more way to push power
into a lossy medium. Until VO1HP.
Frank gave me the exact layout for antennas and back yard and some
data about his local geology. I modeled Frank's setup and monkeyed
with it a while. I finally figured out why it worked and that led to
a method for shorter tower/L/FCP users that minimizes the usual demons
of tower-supported vertical wires. Frank's setup matched the method
in all but one detail. His FCP itself is per the web page.
Keeping the bend close to the tower up top, and slanting the wire away
from the tower going down reduces current in the tower, in turn
reduces tower induced ground loss. The method doesn't eliminate all
tower base loss, but reduces it significantly. A two to one tower
current reduction is a four to one reduction in power loss, three to
one current reduction a nine to one loss reduction.
I have done many NEC4 runs on variations of the VO1HP solution,
looking for defeating sensitivities, including modeling wires and coax
attached to the tower and nearby. There is a common, particularly
destructive sensitivity that requires treatment: A 160 or 80 or 40
meter inverted vee or dipole, fed with coax, either on the tower or
nearby. Nearby vertical coax runs for other supported antennas may
also need treatment. See the text in 8) and 9) below.
------------------
The Essentials of the VO1HP small tower L+FCP method, as optimized.
------------------
1) This solution is for a short to medium tower, supporting a small to
medium triband beam just above the top of the tower with mast and boom
not insulated from the tower, a very common situation for 1/4 wave
L+FCP+tower inquiries. For a tower 60' and less with a small or
medium tribander up top, the solution seems equivalent to a tree
supported 1/4 w L/FCP of the same size, shape and ground
characteristics.
The solution works for towers up to about 70 feet with *small*
tribanders up top. With a lot of top aluminum at 65 or 70, or by 80
feet regardless of top load, the performance has already fallen off
the cliff into steeply increasing loss and the method can't be used.
The problem is the tower plus top-loading getting closer and closer to
resonance and increasing tower current.
2) Support and insulate the L's bend 1-3 feet from the tower. This
allows a rigid support which maximizes the height of the bend. Because
neither end of the wire moves, this keeps wind stress from trees from
affecting the wire at the isolation transformer box, or changing the
spacing between wire and tower, and allows pulling the wire straight
from the bend to the transformer box, required for best results.
3) The FCP feed point and isolation transformer box should be 13-15
horizontal feet from the nearest tower surface. This means you can't
support the FCP feed point on the tower, or directly below the bend.
This results in the wire running down from the bend at 10 degrees or
more off tower vertical. Less separation than 13-15 feet is lossier
*and* more separation than 13-15 is lossier. This is true with the
shorter tower heights we are talking about.
4) The azimuth direction from the tower to the FCP feed can be any
angle relative to the horizontal wire of the L. There is only a very
mild advantage to placing the FCP feed in the quadrant underneath the
horizontal. Practically, it doesn't matter. Let the necessities of
the lot dictate. The FCP feed can be on the other side of the tower
from the horizontal or out to the sides from the horizontal, in
addition to being underneath the horizontal. Support the bend 1-3 feet
to the side of the tower going past to use an FCP feed away from the
horizontal.
5) The distance from the FCP to the tower at any point along the wires
should be no less than the half the distance from the FCP to the
ground. The combination of (2), (3), (4) and (5) should allow a great
deal of flexibility for FCP placement to deal with physical issues on
the small lot.
6) If you have radials attached to the tower base, they should be
removed or disconnected. You want the tower to appear to RF like an
uncooperative resistive short non-resonant conductor. IMPORTANT: Do
not remove connections to grounding rods. These are required for
lightning, and to satisfy building and electrical codes. Since they
are notoriously poor RF connections, removing them gives you no RF
advantage and does not help this method.
7) This method also minimizes induced current in miscellaneous tower
conductors, thus *not* requiring special treatment of tower conductors
*just* for the FCP. Exception: See 8). *Other* reasons to treat
tower conductors still apply, such as grounding, blocking common mode
noise on conductors from the house to keep it from going up the tower
and getting into other antennas, etc, just as they did before.
8) IMPORTANT: If you have a 40, 80 or 160 meter dipole or inverted
vee, and possibly other 40/80/160 antennas, either nearby or with its
feed coax going up the tower, you must treat them or they may defeat
your L+FCP+tower setup.
Both sides of the dipole or inverted vee must be connected to feed
coax via a common mode blocking balun RATED FULL PERFORMANCE ON 160
METERS, regardless of the normal band of use. On 160 meters neither
wire of the dipole or vee must be able to "see" the common mode
looking through the wire-to-coax connecting device. If either wire of
the dipole or vee has an "easy ride" to the common mode on the feed
coax shield on 160, it becomes the equivalent of a massively
top-loaded small tower, and defeats the method. Other 40/80/160 meter
antennas may present the same problem.
If the vee/dipole center is mounted on the tower, the 160 rated balun
at the top will suffice. If the vee's apex is not supported by the
tower, and the coax is not running down the tower, but is nearby, the
induction to the coax from the L's vertical wire is NOT minimized with
the slant wire trick, and a 160 meter rated common mode block is
required at the ground, before any grounding, in addition to the 160
meter rated common mode blocking at the feedpoint.
9) Other nearby coax fed antennas with a vertical coax drop not on the
tower need a 160 meter rated common mode block at the ground, before
any grounding. When the RF coming off the L is looking around in the
near field, you don't want it to see any ad hoc 160 antennas coupling
to the ground. You want the RF to get a bad taste in its mouth if it
decides to use the tower.
10) Metric values: Tower roughly 20 meters or less and FCP feed 4 to
4.5 meters horizontally from tower. Bend in the L supported 0.3 to 1
meter from tower.
-----------------
Other considerations
-----------------
Don't start on a program of changes related to this posting unless you
are within the size limits. A resonant or near-resonant tower/beam
combo supporting an L/FCP can be remarkably lossy. If you don't fit
the limits, there may be a different technique for you later, but we
don't have anything proven. VO1HP on the other hand, has a proven
installation of this method.
Many have tried loading RF to a tower on 160 using whatever short,
sparse, miscellaneous radials they could manage in the cramped
confines of the small lot. If that really worked, they would be happy
with their loaded tower and we wouldn't be having this discussion. If
one is putting up an L/FCP, the radials are not needed. DISCONNECT
THE RADIALS FROM THE TOWER! They are lossy if connected to the tower.
Radials lower the impedance to ground that a non-resonant tower
otherwise presents against induction from the vertical wire,
increasing the induced current, and increasing lost power. You either
do tower radials well, or you do an FCP. But not both. No combos.
A ground rod(s) for lightning may be required by building or
electrical code, or a UFER ground (controversial) may suffice. I will
not weigh in on that. But the lightning ground rods are notoriously
bad at RF. DO NOT REMOVE lightning grounds -- a lot of risk for NO
payoff. DO NOT AVOID INSTALLING LIGHTNING GROUNDS to "help" this
project. You won't help it. Follow your building and electrical
codes.
If you do this, please let me know how it goes.
73, and GL
Guy K2AV
------------------------
** Modeling this circumstance requires the ability to specify
conductors below ground. NEC4 can process underground conductors, but
NEC2 based processes can't. Only users with the NEC4 program itself
or running the NEC4 engine in a shell (e.g. EZNEC Pro), with the
appropriate NEC4 license, can currently model this problem.
*** I am trying to get in touch separately with all those who have
corresponded with me in the past, mentioning both "tower" and "FCP" in
the same email and without header reference to TopBand. I have no way
to know if any of these receive TopBand. Most correspondents have not
posted here on TopBand reflector, that I have been able to tell,
though many read and do not post.
_______________________________________________
Remember the PreStew coming on October 20th. http://www.kkn.net/stew for more
info.
|