On 1/21/22 9:16 AM, Steve Maki wrote:
Here, maybe this will be easier to read...
<big snip>
3) The On ground system on VERY POOR SOIL is only 1 db better than the
FCP on Average soil. This one was a bit more of a surprise
(intuitively) but also explains why some of the confusion when
comparing performance from station to station.
4) When comparing the 1/4 wave elevated radials directly to on ground
radials for the SAME ground type the. The on ground radials were 1 to
1.5db better than the elevated radials.
There's a fairly simple explanation for this. The epsilon (dielectric
constant) determines how much of the current in the wire on the ground
propagates into the soil, vs the air. (It also affects the resonant
length). The resistivity just determines how deep it goes before it's
turned into heat. low resistivity has deeper penetration. There is some
point (with fairly high conductivity, I'd try starting around 1000 mS/m,
if you're modeling) where it might start to change (i.e. it's being more
of a "conductor" than a "lossy medium). And of course, everyone's
favorite saltwater marsh with epsilon = 80 and sigma = 1 S/m. But that's
not soil in the conventional meaning of the word. (for broadcasters,
it's Nirvana)
The reflection coefficient is a complex number so both e' and e'' factor
into the magnitude. It's just that for usual soil, the e' term dominates.
5) Those of you who have done extensive modeling know that depending
on the length of the horizontal section in the inverted L that there
is some signal degradation ( pattern distortion) with the "lowest"
gain in the direction the L points. and that the longer the horizontal
section the greater the degradation in that direction. The better the
soil type however the less pattern distortion for the same geometry.
6) Ground conductivity is BIG factor and we are talking ground in the
far field no just under the antenna.
Yes, and NEC4 only "sort of" models that well. You can set up two
regions with different soil properties. But in the far field, it's doing
a fairly simple reflection coefficient calculation.
Disclaimer and related topics:
1) This is what the NEC 4 models showed, take that with a grain of
salt. According to some articles I have read in the last two years,
the implication of those that those of us in heavily wooded settings
should take down our 160M wire and concentrate on 10m. I have worked
156 countries on 160m in the last 2.5 years from a heavily wooded
location with a modest 60' high wire T with four elevated radials over
very poor soil (.0012 on average) here in central Florida where half
the time we sit and listen to guys in New England work EU like they
were locals (which to a greater extent they are). Probably would be
closer to 200 worked if it weren't for so few expeditions due to
C-192) The above analysis was done at a take off angle of 15degrees
for the vertical portion of the signal. My experience is that the
horizontal portion of the Inverted L's of modest proportions doesn't
provide much in the way of radiation not even stateside anyway.
3) If your really interested in your native ground conditions google
N6LF, Rudy's work on "OWL probes"
4) A couple of side trips related to elevated radial.
Read N6LF's work on elevated radials VERY CLOSELY before jumping to
conclusions: For reasonable radial lengths (.2 to .4 wavelengths) more
than 4 elevated radials doesn't buy you much if anything. More radials
can help with more even distribution of radial currents which is often
more about pattern distortion than anything else.
If you really concerned about pattern distortion go to a "T" rather
than an inverted L for starters or non resonant radials (see the last
paragraph)
Symmetry is usually good. The T is probably one of the oldest antennas
around (you see it in pictures of early radio stations on buildings and
ships) It's basically a tophat capacitive load on a short (in lambda)
wire, so you get more current in the wire, and it's more uniform.
Parallel wires for the T is also effective: increases C, decreases IR
losses, and can still be done with two supports.
Probably the overall cleanest scheme would be an I - parallel wires on
the top, vertical wire, parallel wires on the bottom. Those folks in the
1920s with their spark gap transmitters didn't have all the theory, but
they DID spend a lot of effort on experimentation.
Really, the big issue with this sort of antenna is that it's vertically
polarized. But not everyone can put up a 160 m Horizontal Pol antenna
at 80m height.
Elevating the radials from 8' to 20' buys you a whopping .2db
improvement ( hardly worth the effort IMO).
Yeah.. Most of the field is close to the wire.
Find a copy of K5IU (sk) article on non-resonant radials if you're
really interested/concerned about the uniformity of radial currents
with fewer (4 or less) in elevated radials. I use non resonant radials
on both 80 and 160 (90' long) and the radial current varies less than
2% radial to radial. I have a copy of the paper somewhere ..contact me
off list if you have really searched and can't find it ...
manuals@artekmanuals.comDaveNR1DX
_______________________________________________
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|