"Jeff Liebermann" wrote in message
...
On Fri, 4 Sep 2009 00:56:39 -0700 (PDT), wrote:

Sure, you can get fairly close to isotropic with the right
system, but how are you going to do it by tipping a
vertical? The likely results do not fit my idea of isotropic.

I forgot to connect my comments to the original question. Sorry(tm).
You're correct. There's no way to get a good isotropic radiator
pattern with a simple vertical radiator. However, you can still get
fairly close if you make the antenna sufficiently small relative to
the operating wavelength. As the physical antenna size approaches a
point radiator, the pattern starts to look rather spherical.


That doesn't sound right. The directivity gain of an infinitesimal electric
doublet (i.e. a dipole with infinitesimal length) is about 0.4 dB less than
that of a half-wave dipole. Its similar lemniscate-of-rotation radiation
pattern results from the symmetry about its axis. The results for monopoles
derived from these forms of dipole won't be too different. It is rather
obvious that a receiving dipole of any polarisation won't receive much
signal from the end of a transmitting dipole or monopole, however it's
oriented - there's no apparent length over which unopposed current is
flowing so there's no 'moment' in that direction.

One solution to spherical radiation is the Lindenblad array (and variants
that others have chosen to re-name) which presents finite resolved
components of the lengths of some of its dipoles in all directions ... but
the tilt of the elements has nothing at all to do with Art Unwin's 'theory',
it's simply a matter of making sure there's a resolved component in each
direction. Of course, a Lindenblad designed for a near-omni pattern
achieves this in respect of circular polarisation so it would be
ineffiecient in a system where a linearly polarised antenna is used at the
other end of the link.

I was once told a true isotropic radiator would have to be circularly
polarised because it would be so small that it could contain nothing with a
defined axis of symmetry ... that is, the antenna would have the form of an
infinitesimal sphere. The question then is 'which sense of circular
polarisation' ... which undoubtedly has nothing at all to do with Coriolis
force! The real answer is that it doesn't matter because, as you mentioned
(below) such an antenna has an infinitesimally small radiation resistance
and cannot be made to radiate.

Chris



Unfortunately, the gain drops, efficiency drops, and feed point
impedance drops, resulting in a rather inferior antenna.

There's also a question of how close to perfection does the spherical
pattern need to become? Within 0.1dB, 1dB, 3dB, etc???? Offhand, I
would guess anything within a few dB of spherical could be considered
isotropic, as in all the patents I noted.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On Fri, 4 Sep 2009 16:44:48 +0100, "christofire"
wrote:

"Jeff Liebermann" wrote in message
.. .
However, you can still get
fairly close if you make the antenna sufficiently small relative to
the operating wavelength. As the physical antenna size approaches a
point radiator, the pattern starts to look rather spherical.

That doesn't sound right. The directivity gain of an infinitesimal electric
doublet (i.e. a dipole with infinitesimal length) is about 0.4 dB less than
that of a half-wave dipole.

I'll plug a series of shortened dipoles, possibly with loading coils,
into 4NEC2 and see what happens. You may be right.

As I recall, the big holes in the pattern, that are inline with the
elements gets smaller is diameter as the antenna gets electrically
smaller. The rounded circular donut pattern tends to flatten. I
wanna play with the models to be sure. This still begs the question
of how close to spherical does the pattern need to be in order to call
it isotropic? Dunno.

I was once told a true isotropic radiator would have to be circularly
polarised

"Near isotropic circularly polarized antenna"
http://www.google.com/patents?id=saMgAAAAEBAJ
CP satellite antenna used on Intelsat V. I've been looking at the
patent for a while trying to figure out how it works.

Yeah, it should be CP because that would correctly fit the definition
of the field being identical along the sphere, in all possible
measurement antenna orientations. Note that the isotropic simulator I
posted is *NOT* circularly polarized. If you plug the deck into 4NEC2
and instead of looking at the total gain in the 3D window, look at the
vertical and horizontal gains individually, you'll see something
really ummm.... interesting.
http://802.11junk.com/jeffl/antennas/isotropic/isotrop2-vert.jpg
http://802.11junk.com/jeffl/antennas/isotropic/isotrop2-horiz.jpg
Needless to say, that the polarization is not even close to being
uniform over the sphere. (I'll add these to the menu as soon as I can
figure out what the latest JAlbum update broke in my photo
collection).

Drivel: Just got 4NEC2 setup on my new computah (Dell Optiplex 755
E8500 with 4GB). A messy tower and antenna simulation, that took over
an hour on my old PIII/1GHz clunker, now takes about 4 minutes. I'm
happy (for now).




--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558