On Nov 15, 6:47*am, Richard Fry wrote:
On Nov 15, 12:23*am, Art Unwin wrote:

*What one gains from this aproach is that any radiator of any shape,
*size or elevation can provide figures in the order of 100% as long as
the radiator is a multiple of a wavelength where it is *resonant at
exact and repeatable measurements.

"Figures in the order or 100%" of what?

All radiators of all sizes and shapes will radiate on the order of
100% of all the r-f energy that can be coupled into them through their
input terminals, whether or not those conductor sizes/shapes are
naturally resonant at the applied frequency.

But the fact remains that natural resonance does not occur in
electrically small radiators -- while their radiation resistance is
very small, and their feedpoint is very reactive. *These realities
make it very difficult to supply r-f power to such a radiator without
relatively high losses.

As a consequence, the efficiency of the transmitter SYSTEM
(transmitter + radiator + matching network, + r-f ground loss in the
case of monopoles) can be very low.

To illustrate, the link below leads to a calculation of the
performance of a 3-meter monopole system on 1500 kHz. *Due to the low
radiation resistance and system losses, and even though the short
monopole itself is nearly 100% efficient at radiating the power across
its feedpoint, that radiator receives only about 0.37% of the power
available from the transmitter. *So the system efficiency is very
poor.

Such an electrically short radiator (no matter what its shape) is not
very useful compared to a naturally resonant 1/4-wave monopole or 1/2-
wave dipole -- both of which can radiate nearly

The use of the term "nearly" does not imply total accuracy.
To use Maxwell's equations for accuracy one cannot introduce metrics
that are not absolute.
1/4 or 1/2 wave radiators cannot supplant the "period" of a wave form
and thus introduce inaccuracies. The use of different algarithums in
programing accentuate or minimise the effect of these inaccuracies
thus providing different results. Same goes for close spaced wires
where the use of "near" accurate capacitances by avoidance of all
other proximety effects again take away from the accuracy of Maxwell's
equations. An accurate measurement of resonance of a mesh as I have
shown on my web page need not be dissed because of the presence of a
computer program.


100% of the available
power.

The calculations in the link below were made using standard equations,
in a spreadsheet format to make it easy to follow and confirm.
Properly constructed/used NEC models will verify the spreadsheet
calculation, and the statements about the dipoles mentioned above.

There is no cause to distrust NEC when it is properly understood and
properly used.

http://i62.photobucket.com/albums/h8...5on1500kHz.gif

RF