RadioBanter - View Single Post

May 19th 06, 01:01 AM posted to rec.radio.amateur.antenna

Richard Clark wrote:

On 18 May 2006 08:42:06 -0700, wrote:

I would amend Cecil's statement to:
'We hope antennas are linear systems.'

A nonlinear antenna is a terrible beast indeed. It will mix every
signal it receives with every other signal, creating a nasty mush of
signals. On transmit it is not so bad, maybe generating harmonics and
a wider signal than we'd like.

We've all heard stories of a poor antenna connection causing problems,
or a nearby raingutter joint causing TV inteference. These are
nonlinear antennas.

Hi Glenn,

Cecil is being deliberately obtuse to the matter of linearity. In the
game of describing an antenna as a transmission line, the
non-linearity is compellingly obvious.

If you start with standard twin line, its characteristic Z is
dominated by geometry and a ratio of wire diameter to wire separation.
Pull that twin line apart to construct a V or a dipole, and the
geometry necessarily forces a non-linearity into the picture.

Reggie also has considerable difficulty with this concept too as he
prefers to switch to earth as the main arbiter of transmission line
dynamics. Both seem to abandon the generator's view of a 50 to 70 Ohm
load to replace it with their 600 Ohm concepts so as to artificially
impose their need to see a linear load.

These 600 Ohm concepts are achieved only if the operator strains to
fail most spectacularly. Few debates are won this way, but arguments
successfully persist for hundreds of postings. Even then, these
concepts do not answer the initial non-linearity that inhabits the
system. Both the 50 to 70 to 600 Ohm prognostications are artifacts
of a measurement at the terminal of the non-linear device.

The proof lies along the line, and this returns us to the underlying
concept and argument about the distribution of current along the
length of the quarter wave dipole's arms. This is stated in terms of
the Cosine function. However, as with a deliberate failure forcing an
erroneous general solution, the Cosine distribution is only found in
the extreme (or the fevered dream). A 1mm wire strung 36 meters in
outer space is certainly thin by engineering conventions, but it
doesn't qualify as the current distribution misses the mark of
Cosinality by 5 or 6% (the distribution of a poor fit demonstrates the
non-linearity).

Given Cecil's penchant for abstracting considerable error to general
proofs of his crystalline logic, this may not seem much. However, the
precision above is not outside of achievement, and it does demonstrate
with simplicity that linearity does not reside in the
transmission-line-as-antenna. Pulling this antenna down to earth to
allow the boys their investment in the 600 Ohm concepts does nothing
to recover linearity - if anything, it worsens it (albeit, by very
slim margins). Clearly, the dominant factor in the linearity of the
dipole's characteristic Z is with its own wire. This has been long
reported in the literature (Schelkunoff).

Further, to anticipate this does not demonstrate any spurious
emissions - this is only due to your (not yours, Glenn, your in the
sense of the general reader, and our boys with their loss of
investment) inability to resolve them.

This class of non-linearity falls under the heading of "scattering"
and in these most mundane of applications would barely present
spurious products higher than 70dB below excitation, and only several
parts-per-million from the center frequency (called Stokes shift).
When Cecil comes to the table armed with slop on the order of ±59%
allowable error, such products are swamped in stupidity.

73's
Richard Clark, KB7QHC

Thanks a lot for this one, Richard.

And of course, Cecil has already ignored the important bits and state4d
that 6% is no big deal. Typical.

As I said earlier, the responses to Cecil are where the gold resides.

I had not heard of the Stokes shift, nor the scattering you mentioned.
I have some looking up and reading to do. Which, of course, Cecil does
not, since it's not a Xerox moment.

tom
K0TAR