On Mon, 10 Jul 2006 22:01:31 +0100, "David" nospam@nospam wrote:

Do the radials reflect the wave then?

Hi Dave,

This has already been explained.

The reflecting element on a Yagi manages to reflect most of the wave.

In fact, this is not true at all. Being a reflector is the name put
to the chosen effect, the cause of the effect is re-radiation, not
reflection. Reflection is merely the sum of phases that presents this
effect. One must design, by choice, the "reflector" to exhibit a
phase delay in its size, and in its distance from where the
"reflection" is perceived. Anywhere else and there is no reflection
whatever.

Trying to abstract the action of a yagi reflector to a radial has its
virtues, but explaining reflection is not one of them.

I notice that RF experts cannot agree on whether the radials reflect the
wave or not.

Dr. Johnson (about 1780) offers a comment to errors of perception:
"Johnson having argued for some time with a pertinacious
gentleman; his opponent, who had talked in a very puzzling manner,
happened to say,
'I don't understand you, Sir:'
upon which Johnson observed,
'Sir, I have found you an argument;
but I am not obliged to find you an understanding.'

Antenna theory is often about wires and metallic items reflecting waves, and
the phase of the reflected wave. The phase of the reflected wave can be
constructive or destructive, affecting the impedance of the antenna. If an
antenna is mounted too close to the ground, the reflected wave cancels out
the emitted wave.

Um, yes. This catch-all combines many topics, but does not lead to a
single conclusion however. For instance, your last sentence lacks too
many qualifications to be true, and experience shows it is most
certainly false.

Because a ground plane reflects the wave, the impedance of an antenna can
vary with height.

Poor reading. A ground plane can completely absorb the wave, and the
impedance of an antenna can vary with height. A ground plane can
completely reflect the wave, and the impedance of an antenna can vary
with height.

It seems that the common factor is change of impedance, not the
quality of reflection. Unfortunately I will disprove this utterly,
below.

The problem with your last statement is it separates "ground plane"
from ground (height), and then recombines it to a separate conclusion
(reflection) elsewhere in this posting. The radials of an antenna's
"ground plane" is not the same ground plane of reflection so often
employed in analogies.

Parastic elements on a Yagi have a mutual impedance to each other. Would you
regard the radials on a quarterwave vertical as having a mutual impedance?

The radials increase the conductivity below the radiating element,

A paradox:
As all elements of an antenna radiate, and radials are part of an
antenna, and radiating; then the more interesting question is how can
radials be below the radiating element?

decreasing ground losses. The radials are regarded as a finite or imperfect
ground plane.

Much of these clippings from sources are disconnected from context;
generally suitable only for a limited perspective; and often only for
introduction to a fuller and more complete treatment. This is quite
evident in their failure to support general conclusions.

"Antenna Theory" by Professor Constantine Balanis. Second Edition p 165. A
ground plane formed by a perfect conductor completely reflects the wave. If
the ground is finite i.e. not as conductive, it still reflects the wave but
not as well. The conductivity determines the quality of the reflection.

Not to dismiss Costa out of hand, but an extremely poor conductor,
sal****er, reflects nearly 100% of the RF radiated towards it. Turning
it into a sea of copper would hardly improve the quality of
reflection. Again, poor context reduces information to garbage.

Let's consider this enigma further.

I have a standard quarterwave AM antenna that has 48 radials arranged
around it above an average ground that is in fact better than I can
find here in Seattle. At resonance it exhibits a characteristic Z of
Impedance = 36.15 - J 0.008024 ohms
its peak radiation angle is well above the horizon at 22°
if I were satisfied with 3dB poorer performance, I could depress that
angle to 7°.

Now, to test this ground plane (the earth ground that is, because
nothing in the radial plane is going to change here), I put this
antenna over poor ground (much like I would find in my back yard) and
we find at resonance it exhibits a characteristic Z of
Impedance = 33.62 - J 1.774 ohms
Hmm, nothing to cause the Dept. of Homeland Security to bump the white
house up from yellow alert.
The launch angle climbed to 26°, all of four degrees. If I were
satisfied with 3dB poorer performance, I could depress that angle to
9°.

Now, I put this antenna over good ground (much like I would find
floating in Puget Sound, just viewed out the window here) and we find
at resonance it exhibits a characteristic Z of
Impedance = 36.88 + J 7.924 ohms
Hmm, hardly different from average ground.
However, the launch angle dropped to 7° and I can drop that to 1° for
a 1dB (not 3dB) hit. Clearly the "ground plane" out beyond the
"radial plane" has made a substantial difference to the launch
performance (the combination of all phases at a distance) and for all
practical purposes absolutely no difference to its local operation at
the drive point.

So, with all respects to Costa, but your interpretation of his
clipping leaves much to be desired. Clearly seawater is a miserable
conductor compared to copper, and yet you would have to do a yeoman's
chore of work to build a radial field over average earth to compare
equally.

73's
Richard Clark, KB7QHC

p.s.
and just to be perverse, let's raise this out of Puget Sound, out into
outer space, well out of the way of the "ground plane" and discover,
yes, we lose 3dB gain. So this further compounds the folly of
platitudes that the ground plane, in regard to radials now, improves
launch performance.