Buck wrote:
On Fri, 21 Mar 2008 15:50:45 -0700, Jim Lux
wrote:



Since the roof of a house is fairly large, a more conventional random
length loop tuned with something like a LDG or SGC autotuner at the
feedpoint might be a good approach. Icom IC-AH4 could probably also work.


That is certainly an idea I hadn't thought of. It sounds cheaper and
easier to build. The same concept could be used with raising and
lowering the antenna, but instead of a 'loop', it could be a wire.

Thought: let's assume one puts up a loop on the roof. The ground side
is a single wire running over the peak of the house roof, folds up at
90 degrees for 3-6 feet, back across the roof 6 foot above the peak
and back down to the hot lead of the autotuner. Would that not
circumvent the need for the capacitor at the far end? and would the
loop have to be trimmed to a particular size to operate the various
bands, or could the tuner actually operate with most any size loop.

Probably would work ok. The other approach would be a more conventional
dipole looking thing, with the tuner in the middle. To a first order,
they'll work the same. A long skinny (much wavelength wide) loop
looks a lot like a folded dipole. So, how well it works will probably be
determined more by what impedance happens to appear at the feedline, and
how well the tuner matches it. If the tuner's at the end (either of the
skinny loop or a long random wire) then the impedance tends to be high
(close to resonance). If it's in the middle, it's low.

In your situation, I might be tempted to try hanging a vertical wire
down the side of the house, hooked to one side of the tuner, and the
other horizontalish wire 6 ft off the roof. RFI/EMI might be an issue..

You could also do something like run a couple wires on the roof at right
angles to the ridge line, hook those to one terminal of the tuner, then
hook your wire running along the ridge to the other terminal.


I've fooled around a lot with various "throw the wire on the roof"
situations (out of laziness, more than anything), and, in general, it
seems to work better (no quantititive results) when it looks more like a
fan dipole (feedpoint in the middle-ish) than a loop (feedpoint on the
edge).