Richard (Rick) Karlquist wrote:
"Diffraction is reciprocal, regardless of asymmetries in the terrain."
HFTA would seem to suggest otherwise.
I generated four arbitrary terrain profile files and fed them into HFTA
last evening. Each of the four terrain profiles had a peak 660 feet
high (I said it was arbitrary) 5,000 feet distant from the antenna. The
peak for the first profile was broad and smooth on both the near and far
side. The peak for the second profile was sharp and steep on both near
and far side. The peak for the third profile was sharp on the near side
and smoothly broad on the far side, and the peak for the fourth profile
was smoothly broad on the near side and steep on the far side. I
assumed a yagi antenna on 14 MHz (8 elements to get more gain
visibility) at 70 feet above ground.
HFTA shows markedly different takeoff angle profiles for the four
terrains. In general, HFTA says a peak with a steep near side slope and
a sharp peak will diffract the signal lower than a more rounded near
side slope and a broad peak. A peak that is symmetrically sharp appears
to bend the signal the most, although not much more than if only the
near side is steep. A peak that is broad on the near side and steep on
the far side is almost identical to a peak that is broad on both sides.
For the heights, distances, and terrain shapes I arbitrarily chose,
either of the terrain profiles with a steep near side slope gave at
least ten db stronger signals than either of the terrain profiles with a
broad near side slope at all takeoff angles of six degrees or less. At
higher angles, the plots tended to be similar with lots of crossing back
and forth among them.
I played around a bit with different antenna heights and the decibel
difference between the plots varied somewhat, but the general
relationships held. I haven't tried to change the distance or height of
the peak to see what combinations might have the most effect ... I
simply picked some numbers for a first pass comparison.
Unless I messed up (probable), the two relevant plots from HFTA should
(might) be available by clicking on the links below:
http://www.mediamax.com/ab7e/Hosted/Diffraction1.jpg
http://www.mediamax.com/ab7e/Hosted/Diffraction2.jpg
The links for the four terrain files are:
http://www.mediamax.com/ab7e/Hosted/rnd_near-shp_far.PRO
http://www.mediamax.com/ab7e/Hosted/round_symmetric.PRO
http://www.mediamax.com/ab7e/Hosted/sharp_symmetric.PRO
http://www.mediamax.com/ab7e/Hosted/shp_near-rnd_far.PRO
If those files aren't accessible, someone please let me know and I'll
try to fix the links. I can also send them as attachments directly to
anyone who asks.
I'd appreciate any comments on this quick and maybe questionable
analysis, but it seems to suggest that signals approaching a
non-symmetrically shaped terrain feature from different directions could
skew differently, and therefore that a signal going one way along a
specific path might have a different strength than a signal going the
other direction along that exact same path. It has occurred to me that
a signal approaching a terrain feature from slightly above the horizon
might behave differently than a signal approaching that same feature
from below, but since HFTA won't handle negative takeoff angles I can't
really check that out, and at six degrees or less I wouldn't think the
difference would be large anyway.
Simple diffraction aside, real life terrain contains more than one
feature that would make full reciprocity even less likely, in my
opinion. According to the algorithms built into HFTA, the net energy
leaving at any particular angle is the result of multiple combinations
of reflections and refractions, refractions of reflections, reflections
of refractions, and so on. Intuitively, it seems impossible to me that
a ray arriving from a distance at that same angle can somehow split
itself into those same components in reverse. As a minimum, some of the
reflecting surfaces available to the outgoing ray might be totally
shadowed to the return signal by secondary terrain features.
But again, if I'm wrong here I'd appreciate someone correcting me with
enough explanation that I can understand it.
73,
Dave AB7E
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|
