Towertalk
[Top] [All Lists]

Re: [TowerTalk] Help in planning a new QTH

To: K4SB <k4sb@bellsouth.net>, towertalk@contesting.com
Subject: Re: [TowerTalk] Help in planning a new QTH
From: Jim Lux <jimlux@earthlink.net>
Date: Wed, 16 Mar 2005 12:10:36 -0800
List-post: <mailto:towertalk@contesting.com>
At 08:59 AM 3/16/2005, K4SB wrote:
>Michael Tope wrote:
> > Steve,
>snipped....
> > to do for a fixed element length yagi. It could very well be that
> > a SteppIR can eek out an extra .5dB of forward gain when
> > you get rid of the bandwidth constraints.
> > 73 de Mike, W4EF................................
>
>I would tend to agree more with Steve and Dave on this matter. There
>is simply no way in the world ( short of an advanced antenna range )
>to determine how an individual element in the SteppIR might be changed
>to get additional gain.


There are two possible strategies:
Do modeling, and optimize for desired performance, running several models 
with small variations to take care of tolerances.  This will prevent you 
from getting into configurations that are excessively "picky".

Adjust it on the fly for observed performance.  I think some folks are 
doing this by hand, but there's no real reason why you couldn't have a 
computer start with a modeled starting point, then make small changes in 
element lengths to increase a desired signal or reduce an undesired signal 
(essentially doing what Excel's solver does).  There are a significant 
number of problems with this strategy (fading being but one), however, over 
the long term, with good data recording, it could probably optimize your 
antenna for your location.

(The strategy would be one like the optometrist does when checking your 
correction.  The computer would offer two choices and switch between them 
and ask you whether you preferred "A" or "B".  It would then log this 
(along with azimuth and any other relevant data).  Over some time span 
(say, months, or years), you could arrive at an optimized 
configuration.  This is really an implementation of a "genetic" algorithm 
in the optimization world.)




>As an example, the element length and spacing on my 5 element
>monobander on 20 * and its twin on 15 ) is so critical I had to use
>millimetres for the tips and element spacing. As it stands right now,
>the antenna models at 50 +/- j0, and shows 15.93 dbi of gain with 55.6
>db F/B. All this modelling was done with EZNEC. (DOS and Windows
>versions) I can vary the spacing of the 3rd director ( or length ) and
>that pattern goes to hell in a hand basket.

Which is an indication that the design is overly sensitive.

null depth (which is what you're really looking at here) is exceedingly 
sensitive to small changes, particularly with nulls more than 30 dB 
down.  A deep null is really the result of all the radiation from all parts 
of the antenna, in that particular direction, exactly cancelling.  It 
doesn't take much of a difference to destroy the null.

  Think of it this way... if you're radiating a kilowatt in the desired 
direction, inadvertently radiating 1 watt in the wrong direction makes your 
sidelobe performance 30dB.

(People who design radar antennas with better than 50 dB sidelobes have 
quite a challenge in front of them... Radiate a kilowatt, and the biggest 
error can be 10 milliwatts..)

On typical amateur antennas (probably ANY HF antenna), a 50 dB null is 
spurious and unrealistic.



>Admittedly, 55.6 db F/B is excessive, but I have always gone with the
>theory that you
>shoot the "juice" in the desired direction at all costs.

Maximum F/B is NOT maximum forward gain.  As you say, on Tx, you want to 
squirt the juice in the desired direction, but changing from a 40 dB F/B to 
a 30 dB F/B is like radiating 1 watt more out of a kilowatt. That's the 
same as your feedline loss changing by 0.004 dB.

I would venture to guess that a very high F/B design will have more 
resistive loss in the elements (almost by definition, it's highly coupled 
among the elements), and will actually radiate less power in the desired 
direction than a lower F/B design.

What you want on TX is max foward gain.  What you want on receive is good 
F/B and good directivity (to suppress interference from the non-desired 
direction).  Since, on HF, most of the noise isn't from the receiver, 
you're willing to give up some antenna efficiency, in exchange for looking 
at a smaller piece of the sky. Not to mention suppressing that guy coming 
in on the reverse path behind you.

I don't know that anyone is doing this with the SteppIR (I don't know that 
the response time is fast enough), but it might be worthwhile switching 
between a low loss, high efficiency mode on Tx, to a higher loss, 
superdirective, narrow beam mode on Rx.



>As an after thought, I firmly believe that antenna height is the
>overall controlling factor. As some of the guys have pointed out, my
>TH-11 at 55' will whip the snot out of that big 20 under 5000 miles.


_______________________________________________

See: http://www.mscomputer.com  for "Self Supporting Towers", "Wireless Weather 
Stations", and lot's more.  Call Toll Free, 1-800-333-9041 with any questions 
and ask for Sherman, W2FLA.

_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

<Prev in Thread] Current Thread [Next in Thread>