Hi Pete,
I haven't got my copy yet (my fault not theirs), but I just can't
hold out any longer on this point.
To: <towertalk@contesting.com>
>Date: Mon, 27 Jul 1998 18:45:34
>From: Pete Smith <n4zr@contesting.com>
>
>I've been reading the N0AX/K7LXC tribander test booklet with
>great interest since its arrival Saturday. There are a lot of
>caveats that deserve attention.
>
>That being said, I have one big question. A horizontally
>polarized antenna over real (but ideally flat) ground has no
>radiation at zero degrees elevation angle. The report says that
>the two antennas were, "to all intents and purposes," at the
>same elevation. So what was the nature of the received energy
>in the test setup?
There is a very good discussion of this in the latest Microwave
Journal (page 78 July 98 issue). It is written for VHF and up
but it applies equally well to these HF tests. In a month or
two, this may even be available on their website.
Assuming that the antennas were too close together for the test
to have been contaminated by near vertical ionospherically
reflected signals, the vast majority of the signal was probably
due to the "direct wave" (remember line of sight?). This signal
was probably attenuated somewhat (but not completely) by a ground
reflection component. So long as the separation distance wasn't
too far to permit a direct path to exist, the test methodology is
probably perfectly valid for doing direct comparisons of relative
performance.
A good caveat for those wishing to duplicate the tests is "don't
run test measurements when you can hear stations within a few
hundred miles of you being propagated into your test area". If
the ionosphere is no good for high angle signals, your results
are less likely to be contaminated. Actually 3 wavelengths of
separation on the lowest band to be investigated is plenty of
distance to get valid comparison results. Beyond that distance,
the closer, the better.
>It clearly was not the nicely-formed, x-degrees-above horizontal
>lobe we see in the modeling programs. Even the TA trace shows a
>deep null at zero elevation angle.
All of these packages are assuming distances beyond LOS range for
the patterns. And, I don't think any of them are accounting for
the dispersive variation of both the magnitude and phase of the
earth's reflection coefficient as the TOA changes (particularly
at low angles). I could be wrong about that. If so, someone
please let me know (Roy L. etc.).
Even if the packages do correctly acount for the earth reflection
coefficient changes with changing TOA and frequency, there is no
way for them to account for the actual elevation structure
(except possibly TA) and losses of the real path in question. So
that even if the bare smooth earth's reflection coefficient was
-1, the amount of signal that actually got there (to the earth
surface) would not be what was required to produce a perfect
cancellation of the direct wave.
Think of it this way. If the zero signal at zero relative angle
assumptions were true in the real world, almost no amateur
communications could take place at VHF and UHF.
Snip...
>Tentatively, I'd conclude that the relative field strength on
>the very lower edge of the main lobe is about the same for all
>three antennas (relative to the main lobe, that is). This may
>mean that the measurements in the report are a valid
>representation of the relative gains of the antennas tested, at
>least within the precision of the technique.
>From my limited discussions with Ward so far about their
methodology, my belief is that the tests are very probably a
valid indication of relative performance between the antennas
that were evaluated.
>But I sure wish I knew what was really being measured.
>
>73, Pete Smith N4ZR
>n4zr@contesting.com
>
>"That's WEST Virginia. Thanks and 73"
73, Eric N7CL
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|