Towertalk
[Top] [All Lists]

[Towertalk] 468/f

To: <towertalk@contesting.com>
Subject: [Towertalk] 468/f
From: jljarvis@adelphia.net (jljarvis)
Date: Wed, 26 Jun 2002 09:33:07 -0400
Mauri,

You are absolutely right.  I wouldn't DREAM of messing with a 
complex multi-element array without prior modelling.  In fact,
I am in process of doing just that, with a stack of lpda's
right now.

You are also correct that trimming an antenna to make the vswr
read right at the end of a coax transformer isn't optimizing
the antenna.  

However, if I had tried to model the antenna in question, complete
with provision for tower guys, I'd STILL be at it today, 3 days 
after the contest where I needed to use it.  

Moreover, we had a raging debate about vertical vs. horizontal.
So it was constructed as a GP with two radials, which was to be
installed as a sloper.  (I was opposed to that, but it's easy to
fix.)  When the tower crew put a guy wire in the direction of
desired slope, we had no choice but to put it up as an inverted Vee. 
One snip of the pliars to remove one radial...another snip to knock
off the 5% added to the remaining radial, et voila, le Dipol.  

Simple solutions for simple problems; sometimes best.  

In fact, the ARRL has NOT dropped 468/f from the antenna book...but 
they DO have a discussion of the impedance of a dipole in free space,
vs. frequency, which is pretty complete.  

Jim, n2ea


-----Original Message-----
From: i4jmy@iol.it [mailto:i4jmy@iol.it]
Sent: Wednesday, June 26, 2002 6:24 AM
To: jljarvis@adelphia.net
Cc: towertalk@contesting.com
Subject: Re: [Towertalk] 468/f


With a good modeling software and proper skill is nowadays possible to 
realize even very complex (yagi) antennas (like those with open sleeve 
tapered elements) and predict results, even SWR dips within KHz.
With cut and try method it's practically impossible to obtain 
reasonable results or optimize when more than one variable is present, 
not counting that SWR at TX side, typical effort, is often only a 
transformated impedance (by cable lenght) of an antenna that's not 
resonant, not a cure.
 
When a computer model is correct but practical results doesn't appear 
coincident with predicted ones is almost sure that the antenna in 
object is not working because suffering by interactions or detunings 
caused by objects or other antennas, not because the model is incorrect.

A cut and try procedure that would minimize SWR doesn't solve the 
problem at all, likewise wrong is modeling too close from ground, from 
other antennas elements or objects.

Both will not make a poorly installed antenna efficient or properly 
working.

73,
Mauri I4JMY


---------- Initial Header -----------
 
>From    : towertalk-admin@contesting.com
To      : <towertalk@contesting.com>
Cc      : 
Date    : Wed, 26 Jun 2002 04:17:32 -0400
Subject : [Towertalk] 468/f

> 
> >And for how long did 468/f rest resplendent in the books,
> >unquestioned, until Joe Common Ham got his mits on antenna modeling
> >software?"
> 
> When I was a kid, the ARRL handbook offered 468/f as an approximation 
> of the length of a dipole @ 1/2 wave height.  It was stated rather cle
arly 
> that trimming would be required to accomodate site-variables.  
> 
> Last week, I cut an 80m dipole using 468/f as a starting point. Went u
p as
> an inverted Vee.  Took perhaps 10 minutes with a meter to trim off a f
oot 
> from each end, precisely as expected.  (I always add a foot on 80 & 40
,
> just to give me room to trim.  THAT they didn't put in the handbook!) 
  
> 
> Evidence suggests that 468/f is still a workable approximation.
> 
> It would seem that Joe Common Ham may suffer from an over-reliance on 
> model data, or a lack of common sense in applying it.
> 
> Perhaps it's the expectation that you can predict unseen environmental
 
> variables with any precision whatever.  
> 
> Jim Jarvis, N2EA
> Essex, VT
> 
> _______________________________________________
> Towertalk mailing list
> Towertalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
> 

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