For us inner city people who would like to operate the 160 meter cw contest,
a 160 meter dipole is out of the question. Even an 80 meter wire is not
possible.

Modeling a 14 gauge copper 68 foot dipole on 2 Mhz at 35 feet over real
ground shows a feedpoint impedance of 7-j2700 ohms. This translates to a
3.11 dB total loss in a 50 foot long low-loss open wire feedline. If the
2700 ohm reactance can be eliminated, the total line loss becomes a more
much acceptable .25 dB. This can be accomplished with a stub but that is
good at only one frequency.

If the antenna copper loss is excluded in the simulation, Z is 2.8-j2700
ohms and the feedline loss is still less than 1 dB after reactance
cancellation.

Use two parallel feedlines. Clip a 60 foot line onto the dipole. Connect a
30 microhenry high Q variable or roller coil at the station end and adjust
it for +j2700 ohms at the antenna end. This acts as a variable stub. Voila!
the reactance is cancelled out.

Notwithstanding the lowered radiating efficiency of a short dipole or the
match loss introduced by the antenna tuner, at least the line loss is
minimized to an acceptable level.

There are a couple of interesting things this analysis reveals.

The first is that wire loss is reducing the antenna efficiency to 2.8/7
= 40%, or a loss of just about 4 dB. So making it out of a larger
diameter wire, like the shield of a piece of coax (with the coax intact
so that the shield stays full size), or a number of parallel wires,
would be a useful improvement if physically practical.

Another thing is the importance of the loading inductor or stub Q. If
the overall Q is, say, 100, then the inductor/stub will introduce 27
ohms of additional loss resistance. This would reduce the antenna gain
(due to efficiency reduction) by nearly another 7 dB (assuming an
initial 7 ohm feedpoint impedance). If the overall Q is 400, it'll
introduce 6.75 ohms of resistance, for about 3 dB extra loss. So it pays
to make the stub/inductor Q as high as possible. Some antenna
manufacturers would have us believe that stubs are lossless, or at least
low loss. They're not -- it's not hard at all to make an inductor whose
Q is considerably better than a stub. Open wire stub loss can be
determined by modeling as wires, with and without wire loss. Of course,
if the stub is made from twinlead or window line, its loss will increase
considerably when wet.

Roy Lewallen, W7EL

Alfred Lorona wrote:
For us inner city people who would like to operate the 160 meter cw contest,
a 160 meter dipole is out of the question. Even an 80 meter wire is not
possible.

Modeling a 14 gauge copper 68 foot dipole on 2 Mhz at 35 feet over real
ground shows a feedpoint impedance of 7-j2700 ohms. This translates to a
3.11 dB total loss in a 50 foot long low-loss open wire feedline. If the
2700 ohm reactance can be eliminated, the total line loss becomes a more
much acceptable .25 dB. This can be accomplished with a stub but that is
good at only one frequency.

If the antenna copper loss is excluded in the simulation, Z is 2.8-j2700
ohms and the feedline loss is still less than 1 dB after reactance
cancellation.

Use two parallel feedlines. Clip a 60 foot line onto the dipole. Connect a
30 microhenry high Q variable or roller coil at the station end and adjust
it for +j2700 ohms at the antenna end. This acts as a variable stub. Voila!
the reactance is cancelled out.

Notwithstanding the lowered radiating efficiency of a short dipole or the
match loss introduced by the antenna tuner, at least the line loss is
minimized to an acceptable level.

"W5DXP" wrote in message
...

Could a parallel reactance change the series impedance to 50-jX
and then +jX series reactance be added?
--
73, Cecil http://www.qsl.net/w5dxp

Kind of what I was thinking, but I thought of putting some reactance in
SERIES to make the real part of the PARALLEL impedance around 50 ohms. Then
tune with more parallel reactance. There might be a problem with the Q
getting too high, though.

Tam/WB2TT

By golly, you've discovered the L network!

Roy Lewallen, W7EL

W5DXP wrote:
Tom Bruhns wrote:

Some comments in addition to Roy's...

First, if you're cancelling out the reactance by _paralleling_ +j2700
ohms, then the net impedance will be very high, NOT 7 ohms. My
calculator says (7-j2700)||(0+j2700) is (1041428+j2700). You'll be
wanting to put that +j2700 in series to see 7 ohms net. It's likely
to upset any balance you were trying to maintain...


Could a parallel reactance change the series impedance to 50-jX
and then +jX series reactance be added?