Using AutoEZ I took a shot at modeling the 80m DX Special as described
by AI1H back in 1989. See Frank Witt, AI1H, "The Coaxial Resonator
Match and the Broadband Dipole", QST, April 1989, and "The Coaxial
Resonator Match", ARRL Antenna Compendium Volume 2, 1989.
The model is not an exact duplicate of the physical antenna as built
by AI1H. That would be tough to do. Instead, the model uses NEC
transmission line sections. It is equivalent to Fig 9C in the QST
article, page extract shown here (click to enlarge):
https://s8.postimg.cc/y9fc6pcph/image.png
Just about everything is controlled via variables. Here's a screen
grab of the AutoEZ variables tab:
https://s8.postimg.cc/dqkg138wl/image.png
The wires of the antenna can be configured as a flat dipole, as an
inverted vee, or as a dipole with a catenary sag in the center. As a
flat dipole at a height of 60 ft over average ground the SWR curve and
Smith chart look like this:
https://s8.postimg.cc/j1zclyflh/image.png
https://s8.postimg.cc/awhanu1n9/image.png
The good news is that the SWR curve has the expected "W" shape. The
better news is that the SWR minimums are lower than what Witt showed
in his QST article. The bad news is that the dimensions that the
AutoEZ optimizer came up with do not match what Witt states nor what
is shown by K6TU here:
http://1vc.typepad.com/ethergeist/2011/01/a-broadband-contest-antenna-for-80m.html
At the moment I can't explain the discrepancy. Using "fat" vs "thin"
wires for the actual antenna makes very little difference. I also
tried various combinations of center height and included angle between
the sides. So there must be some additional difference between model
and actual.
Note that when building the actual antenna it is not necessary to use
the outer surface of the RG-213 shield as part of the antenna
structure itself. You can use #14 wire for the entire dipole. At the
feedpoint, which can be centered or offset, connect the open stub in
parallel. Then from the feedpoint drop a short length of RG-213, what
I'll call a feedpoint extension, down to a second parallel junction
where the shorted stub is attached. From there continue the normal
RG-213 run to the station. All of the RG-213 sections will hang
vertically and can be taped together. (Well, I think that's correct.
I haven't built it.)
Later, Herb wrote:
>>> There is no way to cover both CW and at least the low-end phone portion of
>>> 80/75 meters. Unlike dipole that can be broadbanded, I have yet to find a
>>> way a suitable way to broadband a Marconi "T" vertical.
So I created an 80m "T" model, vertical height 50 ft and each arm ~9
ft, with an assumed ground loss of 5 ohms in the radial system. The
same kind of AI1H "coaxial resonator" was added at the bottom. Here's
the result, not bad:
https://s8.postimg.cc/coa9itdb9/image.png
AutoEZ format models for the 80m DX Special dipole and Marconi T
configurations here:
https://ac6la.com/adhoc/80m_DX_Special.zip
If you don't have AutoEZ but you do have Microsoft Excel (not a clone)
you are more than welcome to use the free demo version of AutoEZ to
view the models. With the demo you won't be able to run the optimizer
but you can view the trig formulas used for the various wire
coordinates and you can also see how the transmission lines are
configured.
Then Rick asked:
>>> It should be possible to broadband your vertical using lumped element
>>> circuits at the base. ... Dan: can AUTO-EZ do this optimization?
Yes, as above. Or, if you meant in a more general sense, yes to that
as well. The EZNEC "L Network" and "Transformer" insertion objects
make it possible to model, and hence optimize, discrete components as
well as transmission lines. For more on that see the first three
sections on this page:
https://ac6la.com/aetltu.html
>>> Somewhere in the past, QST had an article about this in "technical
>>> correspondence" that related to dipoles.
Perhaps this: Alan Bloom, N1AL, "Once More with the 80-Meter Broadband
Dipole," Technical Correspondence, QST, Jun 1985, p 42.
Dan, AC6LA
https://ac6la.com
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|
