Mark
Let me only respond to the technical things that you are mistaken on
Radiators do have parts that are inefficient which you
apparently do not accept.
Radiation is created by current. If current was uniform over a radiator
length then the length of the radiator is reduced from 1/2 wave to
wavelength over pi.
This is because voltage becomes more dominant than current at the ends of a
radiator.
If you divide the current curve into uniform radiator length
it should become clear to you that the area under the current curve per unit
length diminishes as the curve moves to zero. This is fundermental but if
you still have problems with this concept by all means continue a technical
dialogue.

Loss less feed systems.
This term is used quite a lot in academia. One can relate it to such things
as household circuits where the radiation is so small it is not considered a
factor in calculations.
.. A 'loss less' feed system in say an antenna would comprise of something
short with respect to wave length and would be voltage dominated so that
radiation is minimised by the low value of current.

Regarding efficiency of magnetic loops.
It is clear in this case that we are dealing with a radiatior that is not
only one tenth of a wavelength but also has an impedance dominated by
resistive losses which means that the efficiency will be extremely low and
possibly only a tenth of what you surmised.
There are ways to ensure that low impedance
problems can be overcome, we see similar problems overcome
in very high gain yagi's which tend to have low impedances as efficiency
increases. This problem can be readily overcome in many cases by adding a
second reflector where its proximity to the driven element
reverses the decline in impedance.by adding a coupling effect.
If I have forgotton something technical that you brought up please let me
know.

Ah yes, the yagi syndrome.
Yagi gain is based on boom length assuming other requirements are met. In
the amateaur world boom length is not really a problem for half of the bands
but it is a problem in that boom length and gain have a limit in scope
as well usuitable for many bands. So I would expect that future enginners
will move away from just yagi's and explore methods where direct coupling of
radiators will occur to remove problems of fractional wavelength portions
spacings as one sees with the yagi aproach. and explore other areas, where
turning radius becomes prominent rather than boom length..
But only the future will tell.,which is the subject of this particular
thread.
Art

"Mark Keith" wrote in message
om...
(Richard Harrison) wrote in message
...


Terman has a comment on page 906 of his 1955 edition regarding
"Close-spaced Arrays-Super-gain Antennas. A review of the behavior of
broadside and end-fire arrays make it appear that in order to achieve
high gain it is necessary that the antenna system be distributed over a
considerable space. However, the antennas of Figs. 23-35 and 23-39
obtain enhanced directivity by employing antennas that are closely
spaced. Moreover, it can be shown that an end-fire (like a Yagi) type of
array that is short compared with a wavelength can theoretically achieve
any desired directive gain provided enough radiators are employed and
they are suitably phased. Such antennas which give great gain using
small over-all dimensions are referred to as super-gain antennas."

Read on. There is a fly in the ointment. Terman says:

" A characteristic of all close-spaced arrays is that as the ratio of
size to antenna gain is reduced, the radiation resistance also goes
down; this is illustrated by Fig. 23-36. The result is a practical limit
to the amount of gain that can be achieved in compact antenna systems,
since as the radiation resistance goes down the fraction of the total
power dissipated in the antenna loss resistance goes up. The Yagi
antenna of Fig.23-39 andf the corner reflector represent about the best
that can be achieved----."

This is the fly I refer to when he keeps talks about "lossless
matching" for small antennas or arrays..

So, Art may be on to something to some extent.

Not anything really new though. There is no free lunch. Many have
tried to find it, but it's almost always spoiled by the time they
do...:/ I've modeled close spaced arrays that had loads of gain, but
to feed them efficiently in the real world is not going to be easy.
I'm not sure what the most efficient fed "very small" antenna is.
Maybe a magloop? Dunno...But even a magloop's efficiency will be lucky
to be over 70%?? or so. Not exactly what I'd call a lossless feed. MK