> like it was with the beverage connected. I then terminated both ends of
the
> beverage against elevated "T's and presto the pattern was 23db off the
sides
> and agreed with the computer model instead of bottoming out 15-16 db as it
> did before. In all of these case I kept a duplicate of the original
antenna
> in service in four directions so I could listen on the band and see if I
Hi John and all,
First I want to be clear I really respect John and all the work he does. The
curiosity in me just makes me analyze things. Maybe we can think about this
and come up with a better test.
There are two issues. The first we agree totally on, so I'll completely bury
the first issue.
The first issue is the idea sloping ends causes the Beverage's vertical ends
to electrically vanish. The Pennant, despite having one end sloped and one
end vertical, has nearly identical vertical sensitivity from both ends. That
fact alone blows that theory right out of the water. There can be no debate
at all the sloped wire behaves almost exactly like a vertical of the same
height, or the Pennant would have nearly zero F/B ratio. It's clear-cut
logical proof a sloped wire acts like about the same overall vertical drop
(or the Pennant could have no null).
As for the change in directivity being attributed to the vertical ends when
the counterpoise is Tee'd, I can't speak to that. I've never done a full
pattern test when that is done. All I've done is compared sloping feeds next
to vertical feeds, and found no difference at all. It not only makes no
logical sense, it made no detectable difference in performance. I can say
one thing that is very true, we have to be very careful when looking at
nulls! That is the worse thing imaginable to test.
Here are the conflicts I see. None disagree with your results, but question
how much the vertical section means:
1.) Over perfect ground where the Beverage's gain is almost entirely from
the vertical ends, the peak gain is -22dBi. The deepest side null is
about -44dBi. The average gain is -29dBi.
2.) Over lossy ground the Beverage's peak gain is -15dBi, the deepest null
is about the same at -48dBi, the highest sidelobe -23.5dBi, and the average
gain is -25.5dBi.
3.) The radiation of a "T" over lossy ground contains significant higher
angle radiation. It is a myth that a T section does not radiate even in a
**perfect** world with perfect balance. A T (same height as the Beverage
earlier) termination over lossy ground shows a 5.4dB null in what otherwise
should be an omni pattern. The average gain of that T is -36.7dBi over
perfect ground, and -33dBi over real ground.
Number 3 tells me a horizontal termination wire (I modeled ONE end) has a
large effect, distorting the pattern 5.5dB. It obviously radiates nearly as
much as the vertical wire or the pattern would not have that significant
distortion. This fits with everything I know about and have measured with
antennas.
Number 1 and 2 tell me the peak and minimum sidelobe gain doesn't change
that much between a perfect Beverage and a perfectly DEAD horizontal section
over perfect ground.
Number 1 and 2 maximum sidelobe response compared to #3 average gain tells
me the vertical ends appear to be down at least ten dB from the peaks
sidelobe response. Even if the sidelobe response was all due to the ends (it
isn't) the vertical ends appear to be attenuated ten dB more meaning they
can't affect the average gain of the nulls significantly.
So here is my question. If we know the T wire radiates almost the same
amount as the vertical wire and we know the sidelobe nulls and peaks don't
appear to change much between a perfect and real ground Beverage, how do we
know the pattern change isn't happenstance because the test method added a
null? After all, the horizontal section added its significant radiation to
the mix.
None of this disagrees with the results. It just demonstrates when we are
measuring nulls and modifying an antenna, we have to be careful stating the
causes. The conclusion would be valid if the T ground didn't contribute
significant radiation to the system compared to the vertical ends.
Unfortunately that only happens when the T is replaced with multiple
elevated radials or placed over perfect earth.
No hurry, but let's think about all this.
73 Tom
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|