Howdy, all.
<<<Introduction:>>>
Peter's post here follows a direct inquiry some weeks ago that I was
unable to fully answer before now.
Wire mats connected to ground radials have been around a long time and
instances reported improving sparse on/in ground radial systems. So an
intuited extension to a mat on the ground beneath an FCP is quite
understandable. It's also a reach for many people to see the FCP as
anything but another kind of radial, just folded up to make it smaller
and get away using only one.
I get that. To be truthful, ten or twelve years ago, **I** would have
a really hard time believing the stuff on k2av.com. There was too much
deeply embedded contrary "traditional" understanding back then. W0UCE
and I had an occasional tough ride disentangling from the "wisdom of
the elders" getting to the understandings of the current day.
I still think that the FCP should have been invented some time in the
late 1950's. If my elmer had invented the FCP, I'd have had no trouble
absorbing it and weaving it into my plans, and you would have heard of
it long ago from some "revered" source. But sorry, you got me, and I
know I don't have the personal appeal of Mr. Rogers or Mr. Science to
ease a gritty informational transition with a fetching personality.
All that makes the question worth a careful, respectful answer,
however much a PITA the work to generate the associated confirming
math. The math work highlights pesky details that make a huge
difference in what a "ground screen" or "mat" does or does not do,
ground radials vs. FCP.
Digging deep into the question necessitated a pile of NEC 4.2 model
runs, some with more than 18,000 segments. Even with my 3 GHz PC,
model runs using high accuracy ground and the NEC 4 double precision
engine sometimes took 15 or 20 minutes to complete and post up
figures.
At some point, running the usual necessary sanity checks on results
exposed artifact issues with very small segments and close wire
spacing. I was starting to wonder if NEC 4.2 could actually do the
problem correctly.
Fortunately I had a flash of memory about the NEC 4.2 upgrade which
introduced "extended accuracy" ground, something about artifact fixes
in the new method. I reran in extended accuracy, and results started
clearing sanity checks.
So *everything*, hours and hours, had to be run all over again in
extended accuracy. The extended accuracy ground method can better than
double the program run time. So this has not been a simple or quick
question to analyze carefully. I've settled on 48 wires, notched 1/10
inch into the ground, 66 feet long, spaced one inch, directly
underneath the FCP. The 30 minutes plus for solution with mat is at
the limit of workable program run times in extended accuracy. I start
runs with the mat on the model and get up and go do something else.
<<<Short Answer:>>> The ground screen has no effect beneath an FCP.
I ran NEC 4.2 with and without a ground mat, over extremely poor
ground (.001,3). *With* the mat on ground the program computed average
3D gain at -6.25 dB. *Without* mat on ground returned -6.27 dB. Two
hundredths of a dB difference in the results is probably correctly
dismissed as "noise". If it was precise it would be the difference
between 355.7 and 354.1 watts left from 1500, not discernable on
analog power meters. Using "ghastly ground" (.0005,1), returned -7.63
vs -7.66 dB, or 258.9 vs. 257.1 watts.
With either set of ground constants, the NEC 4.2 "with" and "without"
plot patterns were identical. The differences in the overlaid plot
values were less than the width of a pixel. Expanding the overlaid
traces to a foot across on the monitor did not show any divergence.
Ghastly ground (.0005,1) is what I call the ground description,
deliberately chosen, that causes the most loss in NEC antenna
calculations. It is used to super-emphasize ground sensitivities, and
make sure solutions work and are optimized over the oh-so-common urban
and suburban poor, very poor and extremely poor ground. This yields
solutions that possibly somewhat uncritical with "average" ground, are
centered in the narrow optimum solution ranges often found with poor,
very poor and extremely poor ground.
K0RF (and a co-conspirator whose call I can't find in any of my email)
ran an interesting test at 160 MHz over an MFJ VHF RF analyzer. They
built two inverted L antennas from SO239 chassis connectors and #12
bare wire. One was over four quarterwave radials. The other was over
an FCP. The feedline was coax adapters from the MFJ to the SO239.
The MFJ stood upright. Each individual antenna's feed characteristics
were observed while a large copper sheet was pushed underneath the
MFJ. The meter indications on the L over 4 radials displayed a large
variation between "with" and "without" the copper sheet. On the L over
FCP, the readings did *not* change between copper and no copper. This
demonstrated the FCP's deliberate design to minimize net counterpoise
fields at ground.
<<<Long Answer:>>>
As far back as 1937, Brown, Lewis and Epstein (BL&E) examined effects
of ground screens in their seminal study of MF antennas and ground
radial fields. The BL&E study measured antenna system performance at a
mile with/without a mat in various radial configurations. The ground
screen results are included in their published study in Proceedings of
the IEEE, Vol 25, No 6, June 1937, p 782.
The thimble summary of BL&E on ground screens: With a dense radial
field, 113 radials, there was no change in field strength at one mile,
measured *with* and *without* a screen, relative field intensity 1.0.
With a sparse in-ground field of 15 radials, relative field strength
at a mile with screen was 0.785, without was 0.555, a measured 3 dB
difference. Current day successful ground screens have their ancestral
roots in the BL&E experiments. The BL&E study was run at 3.0 MHz, with
a 9 foot square copper screen. This comparative result can be
generated roughly the same in NEC 4.2.
HOWEVER, the circumstance of the BL&E mat was NOT raised
radials/counterpoise above an at-ground mat 8 or 10 feet below. The
*at-ground* BL&E mat was connected to the *at-ground* radial system.
The mat was RF-conducting as *part of* the counterpoise, making the
combined counterpoise dense in the area right around the ground level
vertical feed, the area of highest field values and highest potential
loss.
With an FCP there is quite a difference in the ground screen situation
vs. that tested in BL&E:
1)
It is easy enough in NEC 4.2 to show that most of the at-ground field
intensity from an L/FCP is *not* from the FCP. More than 95% of the
ground field beneath the FCP is from the aerial wire(s). The X, Y and
Z axis electrical fields in a NEC generated near field table can tell
you what is producing field intensity. Deliberately, in the models I
ran, the horizontal wire is in the Y axis, the FCP is at right angles
in the X axis, and the vertical wire is in the Z axis. Thus the X, Y
and Z columns in the near field table tell you which wire(s) is
responsible for the field intensity at the point(s) where the table
was generated.
The FCP has fields from its three major wire extents arriving at
ground. The fold design produces such a reduction in the vector
addition of the three FCP fields at any point on ground, that the FCP
*net* field value at ground is far less than that of the aerial wire.
Without the mat underneath, running 1500 watts, at 1/10 inch below ground:
The FCP, 10 feet overhead, is responsible for 0.62 volts per meter
(V/m) field strength.
The horizontal wire 80.5 feet overhead is responsible for 2.43 V/m.
The vertical wire starting at 10 feet up and rising, is responsible
for 23.1 V/m !
The ratio of FCP to vertical wire fields at ground remains roughly the
same anywhere in a 35 foot radius around the point directly below the
vertical wire.
So if one wants to talk about the effect of a ground screen located
below an L over FCP, the ground screen's effect on the antenna will be
characterized by the screen's effect on fields ***from the aerial
wire, not the FCP.***
2)
Connecting a ground mat to a ground counterpoise means that the
currents in the mat are *driven*, amplitude and phase are controlled
by the RF feed voltage and phase, and will be in a fixed relationship
with amplitude and current in the aerial wire(s). The mat currents
will have the same useful phase and amplitude relationship with the
aerial wire currents as the radials themselves.
The ground mat underneath an elevated L/FCP will have current
amplitude and phase determined by the aerial wire, the physical length
of wires in the mat, the spacing of wires in the mat, the velocity
factor of wire in the specific dirt, resistivity and dielectric loss
of the dirt itself, the instant moisture content, etc.
In effect the mat is a parasitic element on the ground beneath the
L/FCP. It cannot change parameters in the L/FCP unless the L/FCP
induces strong current in the mat wires, and the current in mat wires
in turn produces significant fields at the FCP which in turn induce
modifying currents in the L/FCP. This is the same process as parasitic
elements in a yagi.
With 5 amps current maximum in the FCP from 1500 watts at the
feedpoint, current in the mat wires ten feet below is in single digit
milliamperes. This occurs because the mat wires electrically are quite
less than a halfwave, so harder to induce, and the mat wires "adopt"
or couple some of the resistivity of the ground around them, making
the wires appear as long skinny resistors. Thus the feedback is
reduced to such a low level that the FCP for all purposes is
independent of the mat.
73, Guy K2AV
On Mon, May 14, 2018 at 10:46 AM, Peter Bertini
<radioconnection@gmail.com> wrote:
> Has anyone with poor soil considered or tried adding a few
> hundred square feet of cheap galvanized wire mesh on the ground beneath
> the FCP to reduce ground losses? I suspect
> improving the near field ground losses would help slightly,
> but was curious if anyone actually tried it?
>
> Pete
> _________________
> Topband Reflector Archives - http://www.contesting.com/_topband
_________________
Topband Reflector Archives - http://www.contesting.com/_topband
|