Richard Clark wrote:
On Fri, 15 Feb 2008 11:40:50 -0800 (PST), Swoozie Pellegrino
wrote:

What might the impedance of this look like, say, if I
took the last 10 meters of a 100' 50-ohm coax piece
and slice it to split out into a 20-meter sleeve dipole
and hung the dipole part vertically from the balcony?
Can I just screw the other end onto the back of
my transceiver and go? Or....

Sure, if you can live with the mismatch. Better if you can tune it
for one band. Or simply stick with my suggestion and run it into a
tuner which you will probably need anyway. The difference between any
of these will be undetectable at the far end of the QSO.

To reduce the chance of a hot chassis due to common mode currents, you
should use a 1:1 W2DU style BalUn (aka choke).

73's
Richard Clark, KB7QHC

I've found that the trick of folding the braid down over the coax does a
very poor job of decoupling the line in an imitation of a "sleeve" or
"bazooka" dipole. It turns out that you really need a high Z0 for the
decoupling sleeve, and the only practical way I know of to do that is
with a larger diameter pipe, and air insulation between the pipe and the
coax. I've also found that a choke balun made of multiple turns on a
single core, by itself, anyway, doesn't provide adequate impedance. I'd
worry that a W2DU style balun would cause a lot of loss in this
application, but it should be possible to model it reasonably well with
a series of loads and find out. (You'd first need to determine the
actual Z of the type of core used, at the operating frequency.) You
might be able to decouple the line adequately with a very high impedance
and low resistance resonant current balun (common mode choke), perhaps
one made by winding coax on a plastic pop bottle. You'd probably need a
second current balun of some type about 1/4 wavelength down the line. My
experience is that it's a trickier problem than most people think.

Roy Lewallen, W7EL