"Dave VanHorn" wrote in message

Though it's entirely done in the HF spectrum, a pretty detailed analysis is
presented he
http://www.cebik.com/58-3.html
A good ground, and cleaner near-field space, is easier to come by at VHF and
UHF, so I would expect results to be somewhat better than what was seen here
even at the high end of the HF spectrum.

It's more decoupling than anything. Although the usual 5/8 GP with 1/4
wave radials is a flawed animal from the git-go. I think the best
article to describe the effect is from a Dr. Reynolds , "I think thats
the name anyway", that wrote an article for AEA about this problem.
They put out a small brochure with the article and some pictures. They
described the problems with most of the common verticals used. IE: 1/2
waves, 5/8's, and collinears. The result of all that led to the
development of the AEA isopole. Probably the best decoupled dual 5/8
collinear ever designed. And thus , the highest performing compared to
less well decoupled competitors. Thats why cushcraft modified their
ringo ranger, and added a decoupling section, and renamed it the ringo
ranger 2. The effects of a lack of decoupling was glaring when
compared to an antenna of the same appx size, using good decoupling.
The RR was also an appx dual 5/8, although slightly perverted in
dimensions...The RR2 is a good antenna. But the isopole will still
usually beat it. The 5/8 GP or other poorly decoupled antenna does
have the rare chance of the feedline currents adding in phase and
creating some gain, but this is like a one in twenty chance...Like
going to Vegas...Doesn't usually work out that way for most people.
Never did for me...I've never had an elevated 5/8 GP on 2m that was
worth a hoot. Not a one...
Only on a car were they ok. But you look at a car...It's large enough
to usually provide a lower 5/8's of sorts, and also there is no
feedline radiation to skew the pattern upwards. The feedline radiation
if any, is shielded by the car body.
HF is a whole different story. On 10m, a 5/8 GP is best over both a
1/4 wave and a 1/2 wave. I've tested this many times in the real
world...
HF is less critical as far as using a real low wave angle, and also
the average angle used , even locally, is probably slightly higher.
But it doesn't apply to VHF or UHF. It's a whole different world
there, and feedline decoupling is by far the most critical part of a
good antenna. Not brute gain numbers. It won't do any good if the gain
is not where you need it. And thats under 5 degrees for local VHF. MK

Mark makes a good point. Those with EZNEC or other modeling program
might try this -- build a model of a ground plane antenna, and do an
elevation plot. I'll describe the process with EZNEC, but you can do the
same with other programs. With EZNEC, start with the example description
VHFGP.EZ, and change the Plot Type to Elevation. Click FF Pat to run a
pattern. In the 2D window, select Save Trace As, enter a name, and click
Save. Then, from the junction of the ground plane and vertical wire,
extend a wire downward for 3/4 wavelength, by adding a wire (in the
Wires Window) with end coordinates 0,0,5 and 0,0,4.25 wavelengths. Give
the new wire 15 or so segments. Change the Units to Inches or
Millimeters, and make the diameter about equal to the outside diameter
of your feedline coax. Do a pattern calculation on the new antenna. In
the 2D window, select Add Trace, and enter the name of the trace you
just saved to show the original pattern for comparison. The distortion
of the pattern is due to current induced on the outside of the feedline
by the antenna -- the radiating feedline is as much a part of the
antenna as the intended radiator. (Those of you enamored with with CFA,
EH, or other small antenna take note!) In fact, feedline radiation can
actually contribute more to the overall field than radiation from the
supposed "antenna". You can see this current in the View Antenna display
after doing a calculation. You'll find that very small changes in the
length of the feedline wire make profound differences in the pattern.
(The phenomenon is similar to what happens when you change the length of
a Yagi element.)

This open ended feedline of course doesn't do a good job of representing
most real feedlines, and the length was chosen to be nearly resonant.
The real feedline is likely to be longer, and take a circuitous path,
perhaps through the mains wiring, eventually ending at ground, and might
or might not resonate. The pattern you'll actually get depends on this
path and length, which are often hard to predict or quantify. But the
program shows that pretty severe pattern distortion is possible for some
feedline lengths, so you might or might not experience it. It also
illustrates that the ground radials do an ineffective job of isolating
the outside of the feedline from the antenna. I believe it's very
possibly a reason people have such varying opinions about ground planes
of various lengths and of J-poles -- and why people seem pretty
uniformly satisfied with the AEA Isopole. It's not spectacular, but it's
consistent, and independent of the feedline.

To further your education, try different ground radial angles and
lengths in the model antenna, and different radiator lengths. Another
fun experiment is to put a "balun" -- a resistive or reactive load of
about 1000 ohms -- at the "coax" feedpoint connection (wire 6, 0% of the
way from end 1), and see what that does to the current and pattern. Try
moving it, and adding a second one at various spacings. Interesting,
isn't it?

The conclusion you should reach is that a description of the vertical
and ground plane wires alone isn't an adequate description of the actual
radiating structure -- the feedline is an integral part of the system
and has to be included if you're to have any idea of how the antenna
will really work.

Hopefully, this exercise will help you understand why, when you ask a
question about how well a "simple" antenna will work, some of us hem and
haw and answer "It depends".

Roy Lewallen, W7EL