Topband: superdirectivity

[Tom Rauch]

> i am slowly taking on board tom's assertion that
> ultimately the s/n on topband is determined by the
> directivity of the antenna. it is a largely sound concept
> which will probably never find its way into the textbooks
> on antenna theory because it has relevance only to the
> dx community.

It's not my assertion, it is a well-established fact echoed by many. 
It also applies to any receiving system, no matter what the 
frequency unless the system is limited by INTERNAL noise.

For example, a common myth exists that an 8 dBd gain antenna 
on 10 meters improves S/N ratio by eight dBd on receive. That isn't 
so, just like it isn't true on 160. 

It is always the ratio of response to desired signal compared to the 
response to noise source directions (and polarization) that causes 
the receiving improvement, as long as the system limits on 
external noise.

When receiving consideration of gain (even at a certain TO angle) 
is, for the most part, a total waste of time.

> receive antennas. i am looking for a broadband antenna
> with a beamwidth better than a 4 wavelength beverage

A four wavelength Beverage is easy to beat because it is too long 
for 99% of receiving cases. Not only does the long antenna 
become ineffective because of ground characteristics (phase shift 
and loss), a profound amplitude and phase shift occurs in as little 
as two or three wavelengths of space on 160 meters when receiving 
propagated signals.

It is impossible to reliably and directly combine sky-wave signals 
received at points separated by as little as 3 wavelengths on 160 
meters, because the phase (and level) of sky-wave signals at those 
points varies from second to second. 

While the entire length of the physically-large antenna contributes 
to noise power, the signal (being from a single coherent source) is 
not often added. As a matter of fact, signals at distant points of the 
antenna are just as likely to subtract as add when the antenna is 
made physically large.

I suspect the main reason a few people (not very many) report an 
improvement with really long antennas is because the antenna self-
terminates from losses. It also could be an emotional improvement 
(we all "feel better" with larger antennas) rather than a true result.

> and a little less demanding on real estate, especially something
> portable which can be constructed on the beach.  the array
> of pennants (i am more inclined towards flags which can use
> their support structure as part of the antenna) has considerable
> merit, a broadband cardioid must be a great building block for any
> superdirective array, but i can't see why an array of verticals can't
> be equally effective. the search continues ..

There is almost nothing really new to be discovered in antennas, 
just slightly different ways of doing the same thing.

In the early 70's, when I started playing with loops, W1BB sent me 
information on a company in Canada manufacturing phased arrays 
of terminated loops (today we would call the very same antennas 
K9AY loops, or Flag antennas). 

The arrays were not "superdirective", because of system internal-
noise limitations. We can not construct a true superdirective array 
with four elements ( a single Flag is really two short phased-
verticals with a slightly radiating feedline between the elements) 
spaced at large spacings. Because the "feedline" (the wire 
between the vertical sections) radiates, Flags and other large 
terminated loops would be very poor choices for use in a 
superdirective array.

Superdirective arrays must have multiple elements with close 
spacing. Since they have highly destructive wave interference in all 
directions, even in the desired direction, the last thing we want is 
"unclean" cell patterns caused by radiation from wires that really 
just act like feedlines.

I have slight mutual impedance related errors in an array of short 
resistor-loaded verticals with three element 1/8th wl endfire 
spacing, and 340 foot broadside spacing. (This antenna is the 
equivalent of the one Earl is waiting for someone to try, except it 
uses verticals instead of Pennants...hi hi.)

Those errors would become profound with smaller element 
spacings necessary in superdirective arrays, and the array would 
be far too complex.

73, Tom W8JI
W8JI@contesting.com 


--
FAQ on WWW:               http://www.contesting.com/FAQ/topband
Submissions:              topband@contesting.com
Administrative requests:  topband-REQUEST@contesting.com
Problems:                 owner-topband@contesting.com