Towertalk
[Top] [All Lists]

[TowerTalk] "75-Meter Dipole Stretcher"

To: <towertalk@contesting.com>
Subject: [TowerTalk] "75-Meter Dipole Stretcher"
From: "Paul Christensen" <w9ac@arrl.net>
Date: Sun, 15 Oct 2017 22:24:30 -0400
List-post: <mailto:towertalk@contesting.com>
The November, 2017 issue of QST has an article titled: "The 75-Meter Dipole
Stretcher."  As one might guess, it's the author's fix for the age-old
problem of attaining reasonably low SWR from a dipole at both ends of the
75/80m bands when using coaxial transmission line.  The author's solution
applies a relay-switched series inductance that can be inserted into a
coaxial transmission line at any multiple of a 1/2 wavelength.  Recall that
the feed point impedance of an antenna repeats every 1/2-wave multiple down
a transmission line, but is precise only on lossless lines.  The author is
using RG-8X coax in his installation.   

What caught my attention was this statement in the article's second
paragraph:  "The solution is to add an inductor at a multiple of a
half-wavelength-long feed line, and its *effect will be the same* as a coil
at the dipole feed point."  

Really?  The same as if the inductor was at the feed point and there's no
loss?  I decided to apply some antenna and transmission line modeling tools
to find the answer.  The author gives no specifics on where he sets his two
frequency points, but it seemed reasonable to use 3.9 MHz and 3.6 MHz.
Since 3.6 MHz is electrically short (and capacitively reactive) when a 3.9
MHz dipole is resonant, the added inductive reactance will cancel the
antenna's capacitive reactance.  

The dipole's complex impedance was calculated using #14 copper wire up
1/4-wave above the ground, using moderate ground conductivity.  Ground
conductivity over soil for a dipole 1/4-wave above ground won't matter
significantly in the calculation, but it's an input field and I decided to
populate it.      

I started with NEC 4.2 and adjusted the flat-top dipole length at 3.9 MHz to
reach close to zero reactance which results in antenna resonance.  The final
values are a dipole length of 121.5 ft. and that yields a complex impedance
of 86.3+j0.  At 3.6 MHz, the dipole's complex Z is 65.1-j113 and the
transmission line length between the antenna and inductor is adjusted to
reach the first electrical half-wave (per the author).  That result is 105
ft. (due to RG-8X velocity factor).   

Efficiency results:  At 3.9 MHz, 83W from a 100W source is dissipated by the
author's dipole antenna.  At 3.6 MHz, that's reduced to 61 watts.  So, when
using RG-8X, it's hardly the same as placing a coil at the feed point where
the inclusion is much closer to lossless.  The combination of high SWR
(about 7:1 at 3.6 MHz) and line loss reduces power dissipated at the antenna
by 27% or 1.4 dB.  

True, we're not going to see a 1.4 dB difference in real world operating.
My point:  the article states that placement of the inductor down the
transmission line is a lossless remedy.  It's definitely not.

Finally, I wanted to see loss figures when using Belden 8213 (RG-11) which
has a characteristic Z of 75 ohms.  At 3.9 MHz, the antenna dissipates 94W
of power and at 3.6 MHz, it's still a very respectable 86W.  The author can
achieve better efficiency by changing his RG-8X with 113 ft. of RG-11 type
(or equivalent).  Also, total system loss will increase by the degree to
which additional transmission line is used in front of the switched inductor
to reach the transmitter.  However, that loss will be smaller than it is on
the half-wave section between the antenna and switched inductor.  

Conclusion:  (1) if using this system to cover all of 75/80m, know that
network placement away from any antenna (i.e., L/C components) can, and will
affect system loss when high SWR is present on a lossy transmission line;
and (2) as proven above, use a good 70-75 ohm coax with this antenna in a
flat-top configuration to reduce system loss.  Even better, I would use
600-ohm open feeders and a tuner to reduce system losses even further.
However, that's not a solution for everybody. 

Paul, W9AC

_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

<Prev in Thread] Current Thread [Next in Thread>