This is the summary as promised. I did as little editing as possible to
keep the content flowing. Pardon me please if this gets a little long.
There are some good numbers to work with here, and some general impressions.
I sent the following;
I have been following a few threads on elevated radials, and I'm getting a
little worried that I may have to switch to littering my south pasture with
the standard 120 quarter wave wires for 160 mtrs.
I saw a posting stating that 4 elevated radials (10 ft high I'll assume),
would be 5-6 db down from the 120 radials on the ground.
What are your thoughts? My modeling program won't handle wires that close
to the ground, it just treats them as perfect ground.
I experimented back in '86, and found the following: keep in mind the
antennas were far from optimum due to space restrictions at the house in
town. All antennas were in place at the same time and could be compared on
the air.
Ant 1 160 dipole, slightly bent, center at 35 feet, ends at 20 feet.
Ant 2 Butternut 40/80 verticalwith the 160 kit, on the ground, using 10
rolls of chicken wire 125 feet long, 4ft wide.
Ant 3 160 mtr inverted-L, base 12ft above ground, (30ft vertical section,
90ft horizontal), 3 quarter wave elevated radials 10 to 12 feet high.
I did alot of on the air testing, which produced very consistent results. I
sounded great on the inverted-L, weaker on the dipole, and very weak on the
Butternut.
The usual s meter readerings, using S9 as a benchmark for the L, resulted in
the dipole being a S3-4, and the Butternut a S1-2. This was consistent
throughout contacts in North America.
The Butternut had a bad heating problem apparently due to inefficiency. When
you put the spurs to it, it would change resonance, leading me to believe
the coil was sagging.
Here at the ranch I have a 135ft tower that is shunt fed, with 4 elevated
radials( the feed point is at 10ft). I have about two weeks before the
ground freezes, and I need to quit scratchin' my noggin and make up my mind
on the ground system.
I'll save the responses and do a summary.
73
Jay WC7M in Wyoming
I received the following;
---------------
W8JIT Tom
There is a big problem comparing two vastly different antennas, or even two
identical antennas if the location is different. Since I made the direct
comparisons, let me tell you how they were done.
In a clear level area of my property, over virgin soil (no chemicals or
buried wires) about 250 feet from the nearest antenna, I installed the
system. I suspended a rope between two tall non-metalic supports about 150
feet apart and had a pully hanging from the rope in the middle. That was my
vertical's support, and it allowed me to raise and lower the same antenna
enough to add elevated radials eight feet high.
I cut and tuned a dipole eight feet off the ground to the test frequency (75
meters), and used that wire length for all radials. I used a pair of large
common mode choke to be sure the feedline was decoupled, and a large L
network (using a vacuum variable and a large coil with a Q of over 200) for
impedance matching at the base of the vertical to be sure the impeance was
always 50 ohms in the feedline.
First I added four radials, and measured FS at just over one mile and
recorded that value. Then I went back, lowered the radials and antenna, and
pegged the bare wire to the dirt with little hooks.
I repeated this with 8, 16, and 60 radials. Here's what I measured. the
percentage and dB reading is referenced to the highest FS I measured, NOT to
a theoretically perfect system:
.>8 foot high counterpoise system.
>4 wires 37 percent -4.3 dB 20.1 mV, 8 wires 58 percent -2.38 dB 25 mV, 16
>wires 86 percent
>-.63 dB 28.7 mV , 60 wires 96 percent -.18 dB 32.5 mV
>
>ground mounted radials
>4 wires 28 percent -5.5 dB 17.5 mV, 8 wires 53 percent -2.73 dB 24 mV, 16
>wires 74 percent
>-1.3 dB 28.4 mV, 60 wires reference 100 percent 0 dB 33 mV.
As a second point, FS was improved almost 6 dB by adding conventional radials
to the six wire elevated radial system at an AM broadcast system.
It's important to remember there there has been no direct verification of NEC
for elevated radials, and it is just a model. It may be right in some cases,
all cases or no cases at all. At this point we just don't know.
It's important to remember we absolutely know 60 1/4 wl radials will be
within a dB or so of perfect, so the chance is up to you. The worse off you
could be is only six dB if you take a chance and it doesn't work as models
show.
73 Tom
--------------
NW6N Bob
Jay at my past QTH in Calif. I used only 4 elevated radials. I do believe
they were very effective. In the past three years I have won many
plaques/certs for zone 3 or pac divisions! Empirical yes, I would not
even consider radials on the ground myself.......
73, NW6N Bob
-------------
N6ND Rick
Hi Jay, I know there's been a lot of nay sayers lately about elevated radials
but I'll tell you what my experiences have been. For about 10 years I've
shunt
fed a 100' tower on 160 with about 60 radials on the ground. I've got a
second
100' tower that is 130' from the first and in-line to JA. The second tower
is
very close to my driveway and a road so there was no way to lay out radials on
the ground for it. As an experiment I put 3 elevated radials on the second
tower and shunt-fed it from the 10' point. In comparing the two towers on the
air (mostly on receive) I couldn't tell any difference between them on signals
that were coming in roughly broadside to the two towers. There is some
coupling between the towers as I do see differences on signals that
originate in line with the towers. After some soul searching, I removed
the ground mounted radials from the first tower and installed 3 elevated
radials on it and re-did the shunt fed. I'm now phasing both towers with
elevated radials and it's working pretty good. I see 15 to 20 dB of F/B on
signals in line with
the towers and on xmit they seem to be better than most antennas in the
Southern Cal area.
120 radials on the ground may be the optimum installation but I'm not
convinced
that all the modeling programs are incorrect in the evaluation of elevated
radials. Give the elevated radials a try this winter and if you think your
signal is down compared to the other WY stations, try the ground radials next
year. Please keep the reflector informed on what you find out.
73 Rick
Rick Craig, N6ND
------------
W8WE John
go wid the elevated radials,,, my understanding is that 4 of them should do the
job....gl 73 john W8WEJ
-----------
KB5UL Charles
I have an 85 feet high wire, with two top load wires ("T") that droop a bit.
The counterpoise is 4 wires about 20 feet high. They are 1/4 wavelength
plus the height (20'). The base of the vertical is at ground level as are
the radials. The radials slope up at about a 75 degree angle from the
ground until reaching 20 feet. I am using a transmission line transformer
to match the c23 ohms to the 50 ohm coax.
That said--it works *very* well, and the horses don't know it is there!
Above ground radials work if your system is set up right. According to
ON4UN the best height for the radials would be 34 feet, but I can't hack
that. Even supported by 10 foot PVC waste-and-drain pipe dropped over a
steel fencepost would be a good bet. Have fun!
73, Charles - KB5UL - New Mexico
------------
N5IA Milt
I used a gamma matched grounded 1/4 wave vertical for about four
years. No radials, but a decent driven ground system. Worked good and
was no slouch.
Two years ago I installed eight 1/4 wave length radials at 10-12
ft. above ground. Some improvements; maybe 1-2 dB at the most, but still
headed in the right direction.
Last year I studied all the reference material during the summer
months and came to the conclusion that the most efficient transmitting
radiator that I could install on my 125' x 190' lot was a full sized
ground plane.
Starting during the Thanksgiving weekend last year I stripped off
all the low radials, constructed an insulated section for the Rohn 25
tower, and installed the insulated section at 1/8+ wave length above
ground.
There were a minimum of three reasons for doing this.
1. I wanted the ground plane as high above ground as possible.
2. I have a large diameter 65' high mulberry tree that is just 30'
from the tower location. I wanted everything clear of it.
3. The 2nd guy attachment level, a logical place to break open the
existing tower, was at 67' above ground.
I then extended the tower to 197' total height. This is 130', a
1/4 radiator, above the insulated base and just 3' short of the
requirement for lighting and painting or strobing by the FAA.
There are 6 guy levels and the guy cables are insulated from the
tower and broken up at 57' intervals.
There are four radials constructed in this manner. Each radial
is composed of four (4) # 12 copperweld wires spaced one foot apart in a
square configuration to create a caged radial. Each radial extends out
1/4 wave length to the top of a 55' high wooden pole at the corner of my
lot. As the property is not square the radials form an "X" and mot a
cross.
The radials are insulated from the tower and ground at the tower
end, and also insulated at the pole end. At each end the four wires
forming the cage are common connected. At the tower end all four caged
radials are connected together and to the shield of the feedline. The
coax center conductor feeds the vertical element directly.
The feedpoint impedance is 32 Ohms. I use Belden 9913 coax cable
with 20 8" diameter coils as a choke at the feedpoint.
The antenna performs superbly. In comparisons with two other
"classic" vertical (1/4 wave verticals with 100+ radials) equipped
stations here in the southwest, this antenna performs as well as the
others in most situations and better in some.
My only complete contest so far with it was the CQ 160 SSB test
in February. According to the claimed high scores for the contest listed
in the August CQ mag, I made the top 50 in the world. That is all
classes of stations, High Power, Multi Op, etc., and I operate Low Power
(100 Watts with an old Yaesu FT-980).
My goal was to break into the world top 10 Low Power and it would
appear that I was able to do that. And that is from the land locked
southwestern New Mexico. Have to wait until the December issue of CQ
arrives to find out the final full listings.
My intent in this letter is not to brag, but to give you some
information on what I have been able to accomplish here by myself on a
small town lot. There has been one individual on the reflector who has
really been down playing the elevated radials because no one has done any
real scientific antenna field range measurements.
My studying leads me to believe, and now somewhat proven in
practice, that elevated radials at a low height are good but not nearly
as good as a full radial field. But if a true ground plane is created
such that the radials are reasonably removed from earth proximity, then
they are just as efficient as the classic 120 radial ground mounted
antennas.
I hope this is of some use to you in making a decision for your
antenna system. Good luck, good DX, es 73 de
---Milt Jensen @ ARS N5IA---
-------------------
Arlen
Some References: QST Aug 88 P35, feedback Oct 88, p44
QST Mar 93, p72
QST June 94, p 45
QST Dec 94, P84
I have been using an L that is 60 vert and 70 ft horizontal for a couple of
years now. I wrote it up in the last listed reference. Mine works quite well
considering that it is only 60 (was 51) feet vertical. How do you measure
effectiveness? I guess with a field strength meter and a well defined
experiment where all of the parameters are known and controlled.
The main detractor for elevated radials has been W8JI, Tom, who says that
his experience shows several db difference between elevated and in-ground
radials. I do not remember all of the details. A number of others have
commented that elevated radials seem to work well for them. Not much data,
just gut feelings from some pretty savvy operators.
One thing I found was that the elevated radial "system", in my case, has a
very high Q. I can only QSY about 20kc before I get up to 1.8 or so. I am
thinking about some top loading changes that might buy me a little more width.
With in ground radials, you have something that is pretty permanent. You
don't just pick it up and move it if you change your mind on things (as I do
frequently).
I have the opportunity to run an experimentby taking FS measurements on the
existing system and then laying say, 60 radials. Lotta work. Furthermore, I
do not have a stable field strength meter with which I could take accurate,
stable measurements over a period of time.
In ground radials will no doubt give you a greater bandwidth due to the
ground resistance being in series with the feeding impedance. If you use
elevated radials with you system and it is series fed (insulated at base) be
prepared for a lower than normal impedance. Most elevated 1/4 W antennas
will turn out to be around 28 ohms, not 36. I do not know why.
Considering the time contstraints, I would be very tempted to go with the
elevated radials.
I have a continuing interest in what you do.
INK N4OO Sopchoppy (north) FL
--------------------------
N7CL Eric
There seems to be a lot of confusion over the issue of elevated radials and
vertical antenna ground system losses in general. To compound the problem,
none of the most popular antenna modelling programs handle any ground
system other than an infinite perfectly conducting plane very well.
My experience during testing of antennas for military use suggests the
following:
1. There are a number of components to what we routinely group together as
"ground losses". I think the understanding of what is going on may be
enhanced when we start talking about these loss components separately.
I would like to apologize in advance for a somewhat oversimplified
but nonetheless long winded treatment of these topics.
First, there is what I like to call the "connection" loss. This is the
effective resistance of the earth terminal connection of the antenna
system to the flow of whatever RF current the antenna system is
attempting to pump into and out of the earth on the wire that connects
it to earth. This is most easily visualized when the ground system
consists of a single ground rod at the base of a 1/4 wave (or shorter)
antenna.
Second, there is the loss due to the interaction of the near-field
energy storage fields of the antenna with nearby lossy ground,
vegetation and structures. This loss component behaves slightly
differently depending on wheter the antenna in question does most of
its near field storage in the magnetic field or in the electric field.
For our usual amateur discussion of short linear verticals and
horizontal dipoles, the near field storage is predominantly electric.
Lets refer to this as "near-field" loss.
Third, there is the RF radiation far-field interaction with the
(somewhat less nearby) ground around the antenna. This is the beam
forming or elevation pattern affecting interaction of the RF field with
the surrounding ground. This interaction is very difficult to
completely describe since the nature of it changes with the distance
from (and therefore the angle to) the antenna's phase center and the
plane of the "RF earth". The frequency involved is another variable
factor here. As someone earlier pointed out, the "skin depth" of the
earth is significant at 1.8 MHz. But as the grazing angle is decreased
to approach and exceed the critical angle, very little penetration
occurs (this is not to say that loss is completely eliminated).
2. Short (1/4 wave or less) base fed vertical antennas require significant
RF current to flow in the ground return terminal of the feedpoint.
If the base is at ground level and no metallic ground "screen" (to mean
either radial system or actual mesh screen covering a large enough
area) is provided, then the losses are dominated by the first two kinds
("connection" and "near-field") for obvious reasons. The third kind
(radiation field losses) are also present but are swamped in magnitude
by the first two. I am assuming a single ground rod attachment for
ground return currents here.
If we add just two slightly elevated resonant 1/4 wave radials (one at
0 and the other at 180 degrees), we can reduce the "connection" loss
component to a very small value. The RF ground return current can be
made to flow almost exclusively in the resonant radials with very
little loss. Radiation fields from these radial wires are very small
due to nearly complete cancellation in the far field. However, we are
left with a very significant amount of "near-field" interaction loss,
and the radiation field interaction loss.
How significant is the "near-field" component? Usually between 4 and 6
dB depending on the exact nature of the local earth and surroundings.
The farther we raise the base (and the two radials) from the earth, the
more we can reduce the effect of the "near-field" losses. How far must
we elevate the antenna to eliminate the "near-field" losses? Our work
(mostly between 9 and 18 MHz) showed diminishing returns setting in
around 3/8 wavelength above earth surface and loss of measurability
somewhere just beyond 1/2 wavelength. On top band, even the 3/8
wavelength number translates into a very diffcult support structure.
Imagine a 1/4 wave vertical with its BASE at almost 200 feet!
If instead of raising the structure, we begin adding radials to
"screen" the "near-fields" from "seeing" the underlying lossy earth, we
can also reduce the effect of the "near-field" losses.
Did someone ask how many radials does it take to eliminate the loss? I
thought so. Well, in a nearly ideal flat, large enough, open field
without any vegetation we found that in terms of length (assuming ideal
screen density), returns again diminished in the 3/8 to 1/2 wavelength
range (for a full size 1/4 wavelength radiator). In terms of screen
density (with various lengths of radial), diminishing returns began
when the distance between the open ends of the radials was less than
0.03 wavelength. Loss of measurability occurred at around half of that
or about 0.015 wavelength. Note that there is nothing resonant in this
ground screen. It can be replaced (or large areas of its central zone
can be) with square welded intersection mesh. Using the mesh has no
measurable effect so long as the comparison is between mesh with 0.015
wavelength or less openings and an identical (size and shape) radial
screen with no more than 0.015 wavelength spacing between the radial
ends.
So, does this answer the question "Do elevated radials work?" Yes. The
answer is that they are effective at reducing the "connection" loss. And
if they are "elevated" far enough (along with the base of the antenna),
they work as well as a full density on-the-ground screen in terms of
radiation effeciency (ignoring changes in the shape of the resulting
elevation pattern).
So it is probably the case that both the guy who says "I added 4 elevated
radials to my vertical antenna and the performance improved greatly" and
the guy who says "I and others have evaluated 4 elevated radials against a
full ground screen and the elevated radials loose by 4 - 6 dB" are correct.
Interestingly, our work also showed that physically short (less than 1/8
wavelength) top loaded (hat + inductor) verticals required _less_ ground
screen in terms of radial length to get to diminishing returns. Required
screen density was the same. Unfortunately, as the radiator gets shorter,
the losses from the ratio of radiation resistance to ohmic losses in the
radiator and loading and matching system component losses overcome the
benefit of needing a smaller ground screen area.
Anyone notice that I didn't talk about radiation field interaction losses
very much? There are a number of reasons for that. First, I'm not
confident that I can quantify it very well. Second, there is very little
that the average ham (or on top band, even the rich, obsessed, landed gonzo
contester) can really do about it. This is because the distances from the
base of the antenna which are important for DX (low angle beam forming)
range from about 0.75 to 3 wavelengths. The exact boundaries are somewhat
dependent on the height of the antenna's effective phase center above the
plane of the surrounding "RF earth". Those numbers are approximately
correct for a ground mounted 1/4 wave vertical. So the "perfect" RF
radiation field ground plane requires enough 1580 foot long radials to keep
the tip to tip distance to 0.015 wavelength. That is 630 1580 foot long
radials! Even Uncle Sugar declined to do that.
The bottom line is that any individual amateur has to be limited to what he
_can_ do at his QTH. To that end, my recommendation for anyone driving
a ground mounted short vertical antenna is to use as many on the surface
radials as he can reasonably achieve. But no more than are required to
limit the tip to tip distance to 0.03 wavelengths or a bit less. Note that
the less space you have for radial length, the smaller the number of
radials you need to get into diminishing returns. And, yes, DX can be
worked with numbers of radials ranging from zero through "enough".
However, with "enough" you will be louder.
73, Eric N7CL
-------------------
ON4UN John
did you read my recent article in CQ Contest magazine? You must make a clear
distinction between radiation efficiency and reflection effciency. As far as
radiation effciency is concerned, even a single radial can get you 100%
efficiency.
However, buried, classic-style raduials also improve the reflection
effciency of the ground near the antenna. With ground reflection you can add
6 dB or northing at all! When you have a bad ground to start with, the
buried (or laid on top of) radials will certainly give you an almost handful
of dB's of impriovement in reflection effciency.
My 80 m vertical system (4 square) has a single elevated radial on each
vertical (17 ft elevation), and 250 buried radials.
PS. The simple modeling programs based on Mininec all take the radiation
efficiency as 100%, and only model for reflection effciency. You need Nec to
be able to model with radials close to ground.
73
john, ON4UN
Is your feed point really elevated? That means that the part of the tower
below the feed point must be electrically isolated from the part above. If
you have 4 elevated radials, and a short 10 ft connection to the (poor?)
ground, then a large part of the return currents will flow through the
ground. This is what you must prevent. Elevetaed means elevated and isolate!
----------------
W8JIT Tom
>My experience during testing of antennas for military use suggests the
>following:
> First, there is what I like to call the "connection" loss. This is the
> effective resistance of the earth terminal connection of the antenna
> system to the flow of whatever RF current the antenna system is
> attempting to pump into and out of the earth on the wire that connects
> it to earth.
Right on. I like to view this current as the current created because the
antenna needs something for the other terminal to "PUSH" against. The sum of
currents flowing in ALL the wires running up vertically must be equal to the
flow of current into the ground system, or else the feedline will radiate!
> Second, there is the loss due to the interaction of the near-field
> energy storage fields of the antenna with nearby lossy ground,
> vegetation and structures. This loss component behaves slightly
> differently depending on wheter the antenna in question does most of
> its near field storage in the magnetic field or in the electric field.
Here's where most stuff we are exposed to falls apart. It is impossible to
have a time varying magnetic field without an accompanying electric field,
and vice versa. All three fields exist very near the antenna.
There's no such thing as a Magnetic radiator and noise is electromagnetic,
JUST like a signal. Even a small "Magnetic" loop antenna is greatly affected
by ground loss near the antenna. Trying to sort out these fields and blame
earth loss on one or the other gets us into all sorts of confusion.
> As someone earlier pointed out, the "skin depth" of the
> earth is significant at 1.8 MHz.
Up to 50 feet or so, depending on soil.
> If we add just two slightly elevated resonant 1/4 wave radials (one at
> 0 and the other at 180 degrees), we can reduce the "connection" loss
> component to a very small value.
Measure the impedance of a dipole at ten feet. The impedance of one radial
will be roughly HALF the impedance of the dipole. The antenna's impedance can
be (and usually is) less than the impedance of the single radial. Most of the
energy in the radial excites charges (current) in the earth, and is consumed
as heat. Some of it radiates mostly straight up, like a dipole at that
height.
> How significant is the "near-field" component? Usually between 4 and 6
> dB depending on the exact nature of the local earth and surroundings.
Exactly what I measured here, and what someone else measured at an AM BCB
station. Amazing coincidence isn't it, the same amount at three locations
with vertical, the same error in Beverage current tests, and in the test of a
low dipole in Thailand?
>So it is probably the case that both the guy who says "I added 4 elevated
>radials to my vertical antenna and the performance improved greatly" and
>the guy who says "I and others have evaluated 4 elevated radials against a
>full ground screen and the elevated radials loose by 4 - 6 dB" are correct.
It's pretty difficult to measure 6 dB by over the air tests. QSB alone can
mask the change, plus the antennas would have to be over identical soil. In
an A-B test here with a FSM, that's exactly the change I measured. About 5 dB
going from 4 elevated radials to 60 1/4 wl radials.
73 Tom
--------------------
73
Jay WC7M in Gillette, Wyoming
On the 1AB Ranch in Campbell County
--
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Sponsored by Akorn Access, Inc & KM9P
|