1/4 WAVE Radial Lengths
Paul Cort-Wright
sufficent when a vertical aerial is elevated above ground. Can anyone
help me as to regards their length.
Most literature states a 1/4 wavelength ( say 63ft for 80 metres)
but surely the length will vary the depending on the height from the
ground.
To get maximum efficency I imagine that the radial should be a
1/4 wavelength when in position, BUT how can it be adjusted to correct
length? Is it possible to couple a GDO to it, or will adjustments in
it`s length cause a change in the swr of the aerial at the feed point?
Am I being Thick or I is it really complex!!!
ANY HELPFUL COMMENTS WOULD BE APPRECIATED
MANY THANKS 73 PAUL G3SEM
--
Paul Cort-Wright
Paul.Cor...@nc.bbc.co.uk
**** Auntie doesn't necessarily agree with my opinions ****
WB3U
Yes it is complex, both electrically and machanically. The electrical
complexity is due to impedance and VSWR changes that result from
the prioximity to ground, as well as the number and length of radials,
and their angle against the vertical element. The mechanical complexity
is realized when attempting to tune the system.
Theoretically, the radials should each be the same length as the vertical
element, but ground proximity and the angle between the radials and
vertical element will affect the impedance, even at resonance.
I doubt the GDO technique will be reliable when you're standing in the
middle of the radiating elements, so I would just cut the radials and the
vertical element to theoretical length, then trim the vertical radiator for
the best match. If you're really worried, you could trim the vertical
element and the radials simultaneously, a little at a time, to maintain
equal lengths. This is the mechanical complexity I mentioned.
Hope this helps. 73,
Jack WB3U
orin
interesting discussion on this topic.
Orin KJ7HQ.
Bob Bruhns
ground* are supposed to be about 5% longer than 1/4 wavelength.
Published directional patterns show maximum radiation is at an
angle somewhat above horizontal, indicating that the radials are
not equivalent to an infinite ground plane. Therefore, as far as
I can tell, the function of radials on a greatly elevated
antennas is simply to provide a virtual ground for the feedline.
I believe that the reason for extending the radials 5% beyond 1/4
wavelength is that a zero exists for RF voltage at the
center/junction when the radials are about that long.
In an elevated feed antenna only a tiny fraction of a
wavelength above ground, capacitive effects would make the
radials appear longer, so that they should be shorter than 1/4
wave plus 5%. Exactly how much shorter would depend on complex
factors such as ground conductivity and dielectric constant, and
this would vary depending on recent rainfall, etc. I think the
goal close to the ground would be to minimize the voltage
difference between the center/junction of the radials and ground
just below it. To measure this, the center/junction of the
radials could be connected to a ground rod directly below. The
radials would then be adjusted in length to minimize RF between
the feedpoint and ground. Since ground conditions will vary, a
compromise adjustment will certainly result, but I'll bet it
won't matter much.
I saw a letter in the American Radio Relay League publication
QST, March, 1993, page 72, "More on elevated radials", which
suggests that on 3.8 MHz, the length of the radials should be
about 1/4 wavelength plus the height of the radials. This is
pretty crude reckoning so you can see that this is a complex
subject and best values will be determined experimentally. If
you are so fortunate as to have access to NEC, you may be able to
model such an antenna over various grounds and get a feel for the
variations.
DON'T USE MININEC FOR THIS. Mininec does not model near-
ground correctly; one of its simplifications is to assume perfect
ground when calculating element currents, only factoring ground
characteristics when figuring far-field values. This is adequate
for antennas 1/4 wave or more in height, but not for antennas
very close to the ground.
Don A. Tiffin
>
> Whilst I appreciate all the details about 4 radials being
>sufficent when a vertical aerial is elevated above ground. Can anyone
>help me as to regards their length.
>
> Most literature states a 1/4 wavelength ( say 63ft for 80
metres)
> but surely the length will vary the depending on the height from the
>ground.
>
> To get maximum efficency I imagine that the radial should be a
>1/4 wavelength when in position, BUT how can it be adjusted to correct
>length? Is it possible to couple a GDO to it, or will adjustments in
>it`s length cause a change in the swr of the aerial at the feed point?
>
> Am I being Thick or I is it really complex!!!
>
>
> ANY HELPFUL COMMENTS WOULD BE APPRECIATED
>
> MANY THANKS 73 PAUL G3SEM
>
>
>
My experience indicates that 1/4 wave length is fine in most cases. The
angle of the radials with respect to the verticle antenna will change the
SWR.
Don Tiffin
WA5FQV
Austin, Texas
Steve Murdock
Why do the radials have to be any particular length? I think I see why
they have to be at least a quarter wave, but why not randomly longer?
>-----------------------------------------------------------------------
Steve Murdock, N0ZYE || Like the Zen Monk said to the hot dog vendor:
mur...@graceland.edu || make me One with everything
-----------------------------------------------------------------------
D.C. Henderson
: Whilst I appreciate all the details about 4 radials being
: sufficent when a vertical aerial is elevated above ground. Can anyone
: help me as to regards their length.
: Most literature states a 1/4 wavelength ( say 63ft for 80 metres)
: but surely the length will vary the depending on the height from the
: ground.
: To get maximum efficency I imagine that the radial should be a
: 1/4 wavelength when in position, BUT how can it be adjusted to correct
: length? Is it possible to couple a GDO to it, or will adjustments in
: it`s length cause a change in the swr of the aerial at the feed point
Paul
I "tune" my elevated ground radials in pairs like low dipoles by hooking
up a coax and then pruning for lowest VSWR. My experience is that they
end up shorter than the standard 1/4 wave formula probably because of
ground capacity loading.
I have modeled the gound plane antennas using NEC/WIRES. it predicts less
than .5 db improvement using the tuned radials over the "standard 1/4
wave" calculation. In any case the "system" is only truly resonanat at
one frequency anyway and since few of us stay exactly on that frequency
one could conclude why all the bother...close is close enough!! Your
house, the surrounding treees , power lines etc etc probably have as much
an effect if not more and that effect is unknowable.
I do have one recommendation try and get your radials up at least 7 to 8
feet. Not only will it improve the effeciency of the system but it will
avoid the embarssing sitution where you (or the XYL) run into them in the
dark while chasing the neighbors cat with a butcher knife. Not a pretty
picture!
Dave
N0Dh/7
:
: --
Muenzler, Kevin
Paul Cort-Wright (to...@gw.nc.bbc.co.uk) wrote:
: Whilst I appreciate all the details about 4 radials being
: sufficent when a vertical aerial is elevated above ground. Can anyone
: help me as to regards their length.
: Most literature states a 1/4 wavelength ( say 63ft for 80 metres)
: but surely the length will vary the depending on the height from the
: ground.
: To get maximum efficency I imagine that the radial should be a
: 1/4 wavelength when in position, BUT how can it be adjusted to correct
: length? Is it possible to couple a GDO to it, or will adjustments in
: it`s length cause a change in the swr of the aerial at the feed point
The general rule of thumb is that the radials are 5% longer than
1/4 wavelength. You then must tune the length of the radiator
depending on the angle between the radiator and the radials.
If the radials are at 90 degrees to the radiator then the radiator
must be shortened. The impedance of a 1/4 wave ground plane
antenna is around 38 ohms. You can either shorten the vertical
radiator or you can increase the angle from 90 degrees to about
135 degrees. At this length the impedance is around 50 ohms.
But back to your original question, 1/4 wavelength + 5%.
They can be as long as you want them to be but they need to be
at least 1/4 wave +5% to be really effective.
73
Kevin
Legal stuff:
The above opinions are my own and not necessarily those of the staff,
faculty, administration, or lab animals (woof!) of The University of
Texas Health Science Center at San Antonio or anyone else who is not
me. Your mileage may vary. Batteries not included. The perceived
length of this message may not be its actual length.
**********************************************************************
Kevin R. Muenzler, WB5RUE The University of Texas Health
muen...@uthscsa.edu Science Center at San Antonio,
Department of Computing Resources
----------------------------------------------------------------------
The difference between genius and |
stupidity is that genius has its |
limits. The only difference | I am Voltohm of Borg!
between stupidity and intelligence| Resistance is E/I!
is that intelligence may be | Power is EI!
artificial, stupidity is always | You will be attenuated!
genuine. |
**********************************************************************
W8JI Tom
earth. The radial is coupled to the lossy earth and has a very low Q. I
would use as many radials as possible, because radials serve two
functions: 1.) Something for the antenna to "push against". 2.) They
prevent the antenna from "seeing" the lossy earth and re-enforce the
signal by "reflecting" energy that would normally be absorbed by the lossy
earth.
With only four wires close to the earth the radials can not effectively
shield the earth from the antenna, you can lose up to 6 dB of useful
signal from that alone. The other loss would be the I sqrd R loss in the
ground system. Some IEEE papers I have show that at 1/4 wave above earth
you need 8 radials (actual field stngth measurements - NOT NEC model) to
approximate 120 ground mounted radials, and at 1/8 wave height they needed
16 radials!
D.C. Henderson
: With only four wires close to the earth the radials can not effectively
: shield the earth from the antenna, you can lose up to 6 dB of useful
: signal from that alone. The other loss would be the I sqrd R loss in the
: ground system. Some IEEE papers I have show that at 1/4 wave above earth
: you need 8 radials (actual field stngth measurements - NOT NEC model) to
: approximate 120 ground mounted radials, and at 1/8 wave height they needed
: 16 radials!
Tom
How about giving the IEEE paper refernces if you have them, would like to
read emas I'm sure others would, for that matter I bet they are availble
though FTP somehwre????/
DAve
N0DH/7
W8JI Tom
Oops, I think they were from proceedings of IRE. I found my notes from
over ten years ago and will locate copies to confirm info before giving
data. I got involved in this after Arch Doty had a CQ article about
elevated radials. I was working with Joe Dobelsy in Cleveland Ohio. Joe
was a pioneer at Smith Engineering (of AM broadcast research fame) in
Cleveland. He found the papers for me in his archives. Can anyone fax me
the recent papers? It's not that I don't trust NEC, it's just that
everything modeled should have confirming measurements. It's very
difficult to model the earth near an antenna.
Jay Kesterson K0GU
: With only four wires close to the earth the radials can not effectively
: shield the earth from the antenna, you can lose up to 6 dB of useful
: signal from that alone. The other loss would be the I sqrd R loss in the
: ground system. Some IEEE papers I have show that at 1/4 wave above earth
: you need 8 radials (actual field stngth measurements - NOT NEC model) to
: approximate 120 ground mounted radials, and at 1/8 wave height they needed
: 16 radials!
I know two people who get the IEEE papers. What they got out of the papers
was that a single 1/4 wave vertical with four radials had about the same
impedance and efficiency as a ground mount with 120 radials. Can you give
more details on your paragraph above?
I'll throw this out even though you don't seem to be a NEC fan. I modeled
a 1/4 wave vertical at 1.830 MHz, elevated 10 feet, over average ground
with four radials. Then added four more radials and ran it again. The gain
increased by .07 db. I think there is no doubt more radials will help, but
how much?
I find NEC to be quite accurate in real life designs. Have the IEEE papers
(or anyone else) stated differently?
73, Jay K0GU ja...@fc.hp.com
W8JI Tom
claim that four elevated radials are enough has never appeared with
supporting field strength measurements. While NEC is a very good program,
earth effects are the most difficult thing to account for. Shouldn't
someone somewhere measure this in the real world before we throw away the
research from the past? It can't be that difficult to measure a four wire
elevated radial system and add more wires while watching the field
strength a few wavelengths away. We don't need a pattern plot, just a FS
measurement!!
Forrest Gehrke
WT> Here's more food for discussion. The info I keep seeing supporting
WT> the claim that four elevated radials are enough has never appeared
WT> with supporting field strength measurements. While NEC is a very
WT> good program,
WT> research from the past? It can't be that difficult to measure a four
WT> wire elevated radial system and add more wires while watching the
WT> field strength a few wavelengths away. We don't need a pattern plot,
WT> just a FS measurement!!
Also, provided it is a single element (with nothing nearby coupling)
measurements of the resonant self-impedance will be an indicator of
the amount of ground loss.
A quarter-wave length element should be near 36.5 ohms resistive at
resonance for least ground loss.
* RM 1.3 02583 * Xeroxes - persian photocopy king.
Jay Kesterson K0GU
> Here's more food for discussion. The info I keep seeing supporting the
> claim that four elevated radials are enough has never appeared with
> supporting field strength measurements. While NEC is a very good program,
> earth effects are the most difficult thing to account for. Shouldn't
> someone somewhere measure this in the real world before we throw away the
> research from the past? It can't be that difficult to measure a four wire
> elevated radial system and add more wires while watching the field
> strength a few wavelengths away. We don't need a pattern plot, just a FS
> measurement!!
A friend of mine is trying to come up with the IEEE papers. If I get the
info on them I will pass it along.
A couple of comments he passed on to me:
He was familiar with some of the work of Arch Doty. He was under the
impression that Doty was using a elevated ground plane (not resonate,
lots of wires) as opposed to (tuned) elevated radials. Perhaps I (we)
have taken this sentence out of context and it doesn't have anything
to do with the discussion? Or maybe we are comparing apples and oranges?
He also said there have been actual measurements taken and that several
AM broadcast stations are now using verticals with a small number of
tuned elevated radials.
Hopefully I can get the details before too long.
73, Jay K0GU ja...@fc.hp.com
W8JI Tom
any field strength measurements, and you are correct..it was insulated
radials vs bare wire. I hope I find the old info I had again soon. Someone
from Ga Tech told me that Smith Electronics did a study a few years ago
and he recalled it taking 60 radials at ten feet to equal 120 on ground,
or something similar. This I can swallow because the current wouln't be so
concentrated in 60 wires, and the coupling to earth and near field losses
would be minimized. If we can't find anything I can take some data here
but it will be over very poor soil. It would take only one afternoon to do
this (on 40 meters). I can start with four and measure FS change in one
fixed direction several wavelengths away as I add wires. Wonder why no
Hams have even done crude test??
W8JI Tom
I intended to say better.
I don't believe any Hams have made FS measurements. Doty made thousands of
current measurements with a very crude instrument but never even one
ground
wave S meter test!
The problem I have with the NEC models is they have the most problem with
near field coupling, and soil is a very strange and difficult thing to
model.
Friends of mine like John (Jack) Belrose disagree but the arguement alway
hinges on how good NEC is at modeling antennas close to the earth.
The biggest reason Ham's use the elevated systems are because they sortof
follow anything published, while commercial people have to prove it (with
proof of performance tests). I think popular opinion is also somewhat
confusing when it deals with what a radial (or ground system) actually
does.
Everyone seems to think it "couples" power to the dirt or collects return
currents from the earth. It actually does the exact opposite, it shields
or
isolates the radiation field from the earth! It also gives the other
terminal of a marconi something to "push against".
That is the crux of my disagreement or questioning. The current density
near
the center and the electric field intensity at the ends of the elevated
radials are so much higher when the current can only divide between four
wires, considerable displacement currents must occur near the radials.
Also
anything within a few dozen to a few hundred feet is coupled rather
tightly
to the radials by the strong electric and magnetic fields. Only conductors
well removed or symmetrically placed around the four radials will not be
coupled into the system.
This could be fodder for a great long debate. Perhaps it would get Maxwell
off the hook for a while!!!
Gunnar Hedby FOA 72
The problem isn't that simple I guess. To see what happens when adding
more radials to an omnidirectional antenna, you need to measure at all
elevations between 0 and 90 degrees. What happens when adding the radials
is that the antenna pattern changes. Since the total amount of energy is
constant, the gain you obtain at some elevation you loose at some other.
So, the difficult thing is not to measure, but to do it right and to know
what your doing.
/Gunnar (SM5EQW)
-- Life is a journey of learning --
--- What you don't know today, remains to be discovered tomorrow ---
===============================================================
= =
= Gunnar Hedby =
= National Defense Research Establishment (NDRE) =
= PO Box 1165 Phone: +46 13 318136 =
= S-581 83 Linkoping Fax: +46 13 128372 =
= SWEDEN gun...@lin.foa.se =
= =
===============================================================
Forrest Gehrke
WT> I've been checking and checking. The problem with Doty is he never
WT> made any field strength measurements, and you are correct..it was
WT> insulated radials vs bare wire. I hope I find the old info I had
WT> again soon. Someone from Ga Tech told me that Smith Electronics did
WT> a study a few years ago and he recalled it taking 60 radials at ten
WT> feet to equal 120 on ground, or something similar. This I can
WT> swallow because the current wouln't be so concentrated in 60 wires,
WT> and the coupling to earth and near field losses would be minimized.
WT> If we can't find anything I can take some data here but it will be
WT> over very poor soil. It would take only one afternoon to do this (on
WT> 40 meters). I can start with four and measure FS change in one fixed
WT> direction several wavelengths away as I add wires. Wonder why no
WT> Hams have even done crude test??
Tom,
You have to keep in mind when you make these measurements that as
you add radials you are affecting the vertical lobe pattern. I
expect that as more radials are added two effects will be seen:
1. The vertical lobe will be lowered. If the field
strength measurement is made at a high enough elevation, the
signal will drop. On the other hand one taken a low elevation
will find the f.s. raised. What is being proved?
2. As more radials are added, radiator resistance loss will
decrease. This will raise the radiated power and f.s. increased.
Again, what is being proved?
I ask the question about what is being proved because there
has to be a comparison. Compared to what? Answering that
question requires similar measurements of a ground mounted
vertical measuring f.s. with varying number of radials on
the ground as one possibility.
Re the Doty articles. I never saw the CQ article but I did
see the QST article of some years ago. My criticism of it
was that his data showed he had a double ground; one at
the point of measurement and another at the generator.
(The article didn't state that, his current measurements
at the radiating element showed it).
Since this ground loop existed, his measurements of ground
currents under the counterpoise were flawed; consequently his
conclusions in this article had to be questioned.
--k2bt
* RM 1.3 02583 * "All these *wires*! Why do they call it `wireless'!?"
Forrest Gehrke
WT> Also anything within a few dozen to a few hundred feet is coupled
WT> rather tightly to the radials by the strong electric and magnetic
WT> fields. Only conductors well removed or symmetrically placed around
WT> the four radials will not be coupled into the system.
There is this to consider about raised radials also:
A top hat having a peripheral wire connecting the ends of the
spokes allows a much smaller diameter than if there is no
peripheral wire. This indicates that a significant electric
field exists out at the edge.
I believe this also applies to an elevated radial system.
--k2bt
* RM 1.3 02583 * OS/2 is a bowl of cherries... Windows is the pits!
W8JI Tom
>more radials to an omnidirectional antenna, you need to measure at all
>elevations between 0 and 90 degrees. What happens when adding the
radials
>constant, the gain you obtain at some elevation you loose at some other.
>So, the difficult thing is not to measure, but to do it right and to know
>what your doing.
>/Gunnar (SM5EQW)
The problem is simple. All verticals shorter than 1/2 wl try to radiate
max field along the ground. Unless the ground radials are really long
(several wl) you won't change the Brewster angle and the overall elevation
pattern any meaningful amount. Typical practical radial systems primarily
effect the overall efficiency of the system, and that can certainly be
measured at any angle in the main lobe.
The gain you realize with a better ground system when using a vertical is
due to a reduction in loss, not in a shifting or concentrating of the
pattern. You don't loose anything at any angle, it simply "gets bigger all
over".
Of course I assume the system is reasonably symmetrical, so radiation from
the radials cancels in the far field. Example: If one elevated radial is
used, this wouldn't be a good test. But then who would use such a lousy
system!
Mike, KN6IS
> The problem isn't that simple I guess. To see what happens when adding
> more radials to an omnidirectional antenna, you need to measure at all
> elevations between 0 and 90 degrees. What happens when adding the radials
> is that the antenna pattern changes. Since the total amount of energy is
> constant, the gain you obtain at some elevation you loose at some other.
> So, the difficult thing is not to measure, but to do it right and to know
> what your doing.
>
> /Gunnar (SM5EQW)
>
> -- Life is a journey of learning --
>
> --- What you don't know today, remains to be discovered tomorrow ---
>
> ===============================================================
> = =
> = Gunnar Hedby =
> = National Defense Research Establishment (NDRE) =
> = PO Box 1165 Phone: +46 13 318136 =
> = S-581 83 Linkoping Fax: +46 13 128372 =
> = SWEDEN gun...@lin.foa.se =
> = =
> ===============================================================
What do you think of this idea, (since I do not have a anocaic chamber)??
I have been experimenting with a weather ballone on 160 meters at 500' AGL
using a vertical. What if I take my ballone and attach a 10 or 2 meter
(not sure what equipment I can get) antenna to it and with a reciver on
the ground attached to the ballone antenna measure the FS say at least 5
wl away from the omnidirectional source antenna up to 90 deg. If I used a
10 meter setup that would be about 170 feet away from the source on say,
29.000 MHz. I have 5 acres of flat desert so no problem with trees or any
other obstruction. With some help from my friends I could make
measurements while changing the radial configuration. Just a thought!!
Any other ideas??
Mike, KN6IS
--
Mike, KN6IS
Forrest Gehrke
WT> The problem is simple. All verticals shorter than 1/2 wl try to
WT> radiate max field along the ground. Unless the ground radials are
WT> really long (several wl) you won't change the Brewster angle and the
WT> overall elevation pattern any meaningful amount. Typical practical
WT> radial systems primarily effect the overall efficiency of the
WT> system, and that can certainly be measured at any angle in the main
WT> lobe.
Now we have to know what you mean by "meaningful", Tom.
Would you consider bringing down the main lobe 10 degrees
meaningful? This can be done with 1/4 w.l. radials by
going from 60 to 120 radials. This is what I measured
when I did it.
Furthermore, results then verified it; I could never work the
Central Asian USSR republics on 75M because I couldn't hear
them (using a 4square aimed in their direction). After raising
the number of radials under each vertical to 120 - 130,
I worked all of them in a few weeks.
I doubt that the reduced radiation resistances involved had
any significant bearing on the array for *receiving* signals
because this was going from unreadability to 3/3 or 4/4 copy.
(Those fellows never ran much power nor good antennas!).
This had to be somewhere between 3 and 6db improvement
in signal and is not accounted for by increased gain due
to a small amount of reduced loss.
--k2bt
* RM 1.3 02583 * If turning it on doesn't help, plug it in.
W8JI Tom
>measurements of the resonant self-impedance will be an indicator of
>the amount of ground loss.
>A quarter-wave length element should be near 36.5 ohms resistive at
>resonance for least ground loss.
Why were you posting replys here????
W8JI Tom
WT> I've been checking and checking. The problem with Doty is he never
WT> made any field strength measurements, and you are correct..it was
WT> insulated radials vs bare wire. I hope I find the old info I had
WT> again soon. Someone from Ga Tech told me that Smith Electronics did
WT> a study a few years ago and he recalled it taking 60 radials at ten
WT> feet to equal 120 on ground, or something similar. This I can
WT> swallow because the current wouln't be so concentrated in 60 wires,
WT> and the coupling to earth and near field losses would be minimized.
WT> If we can't find anything I can take some data here but it will be
WT> over very poor soil. It would take only one afternoon to do this (on
WT> 40 meters). I can start with four and measure FS change in one fixed
WT> direction several wavelengths away as I add wires. Wonder why no
WT> Hams have even done crude test??
>Tom,
>You have to keep in mind when you make these measurements that as
>you add radials you are affecting the vertical lobe pattern. I
>expect that as more radials are added two effects will be seen:
>1. The vertical lobe will be lowered. If the field
>strength measurement is made at a high enough elevation, the
>signal will drop. On the other hand one taken a low elevation
>will find the f.s. raised. What is being proved?
Not true at all unless the radials are very very long in terms of
wavelengths and in proportion to antenna height. It is a misconception
that adding radials "lowers the wave angle" of a vertical.
>2. As more radials are added, radiator resistance loss will
>decrease. This will raise the radiated power and f.s. increased.
>Again, what is being proved?
>I ask the question about what is being proved because there
>has to be a comparison. Compared to what? Answering that
>question requires similar measurements of a ground mounted
>vertical measuring f.s. with varying number of radials on
>the ground as one possibility.
Exactly my proposal and exactly what I did.
>Re the Doty articles. I never saw the CQ article but I did
>see the QST article of some years ago. My criticism of it
>was that his data showed he had a double ground; one at
>the point of measurement and another at the generator.
>(The article didn't state that, his current measurements
>at the radiating element showed it).
That was my conclusion also. The currents in each radial wire did not
equal the current in the radiator. If Doty understood antennas, he would
have caught that error.
>Since this ground loop existed, his measurements of ground
>currents under the counterpoise were flawed; consequently his
>conclusions in this article had to be questioned.
>--k2bt
Sorry the delay, I didn't know you were posting here.
W8JI Tom
Field strength
measurements were made with an old Texscan FSM. I checked the calibration
on it with a good
attenuator pad, so I am sure it is very linear. I used my 80 meter mobile
antenna and drove to a
clear area about 1/2 mile from my location and parked in the same spot.
The antenna tested was a
1/4 wave vertical wire about 200 feet from my other antennas and my house
in an area with no
buried or overhead wires. My soil is clay and rock, probably very poor
conductivity.
I measured the impedance of the elevated ground wires with a GR bridge. To
measure the
resistance of the wire as a counterpoise, I used a 10 foot base loaded
antenna with a very good
coil (Q>300) as my reference. By measuring the change in impedance of the
feedpoint with this
antenna from over a standard ground, I believe a good estimate of ground
resistance can be
made. I accurately know the resistance of this short antenna when worked
against 83 surface
mounted 1/4 wave radials and a small 10 foot square ground screen. The
resistance of the radials
can be assumed to certainly less than two or three ohms of the system
resistance.
While my measurements can be open to some error, they are certainly more
accurate than an "on
the air" impression. Here are my measurements:
Resistance
Single 16 ga wire 8 feet high fed like dipole showed resonance when each
side was 232 / F . That
was 62 - 3/4 feet each side at the test frequency of 3.7 MHz. The
resistance was 60 ohms, the
3:1 SWR bandwidth of 170 kHz up and down was nearly all due to reactance.
Two crossed half wave wires fed against each other in an X were 22.7 ohms.
There were no
mutual coupling effects noticed in this test, so the wires were not
"talking" to each other.
The same two wires fed like this > < each leg 1/4 wl and still a 90
degree angle were 38.5
ohms. In this case mutual coupling effects were noticed. The mutual
coupling lessened the effect
of the second wire.
Four 1/4 wave wires in 90 deg X with short base loaded vertical in center
added 12 ohms to
base impedance, so the X arrangement should certainly be around 10 to 14
ohms of "push"
resistance.
Eight wires added 7 ohms to the base impedance from over the reference
radial ground.
Field strength measurements follow. Four radials systems were tested, 4,
8, 16, and 60 wires.
The highest reading came from 60 surface radials, so all other systems are
dB down from that
system, or percent of the 60 radial surface 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.
Looks like elevated radials work better if only a few wires are used. At
my location I would need
at least 16 for decent eff with a 1/4 wave vert, and certainly many more
in a phased array or with
a shorter radiator. With 16 wires the difference is only around 1/2 dB
between either system. It
looks like if a large system is used the systems work about the same.
It's amazing how fast a day goes by when your having fun.
W8JI Tom
Any amount that would affect the accuracy of FS measurements enough to
render
efficiency calculations suspect.
>Would you consider bringing down the main lobe 10 degrees
>meaningful? This can be done with 1/4 w.l. radials by
>going from 60 to 120 radials.
I beg to differ with you, adding 1/4 wl radials will NOT appreciably lower
the angle of radiation of a 1/4 wl vertical, that is a wive's tale.
Antennas
form patterns by wave interference, and the ground directly below a
vertical
contributes very little to the shape of the pattern. In order to
substantially affect the wave angle of a ground mounted vertical radiator,
the radials would hve to extend many wavelengths.
Do a wave interference plot of the array using the known field strength of
a
1/4 wave wire at any wave angle and you will see that only a very small
percentage of change occurs. It is the ground a large distance (several
wavelengths) that affects the Brewster angle and the E field pattern of a
vertical. For example, chapter 20 of Jasik's Antenna Engineering Handbook
shows equations for calculating the E field pattern of a vertical radiator
(20-5). The equations include no adjustments for ground losses near the
radiator. They only include current distribution in the radiator, height
of
the radiator, and even the cross sectional derived impedance of the
radiator.
If all of that isn't reason enough, then consider that the measurement
would
be at zero degrees, and any improvement there would only come from an
increase in desirable low angle performance.
>This is what I measured when I did it.
Please explain. How was that wave angle change measured? In order to
confirm
an improvement at one specific angle or a general "lowering" of the wave
angle, it would be necessary to plot the pattern periodically at several
angles from the horizon up, and do this several wavelengths or more from
the
array. Out is no problem , but that's a very long way to go up!
>Furthermore, results then verified it; I could never work the
>Central Asian USSR republics on 75M because I couldn't hear
>them (using a 4square aimed in their direction). After raising
>the number of radials under each vertical to 120 - 130,
>I worked all of them in a few weeks.
Is propagation that stable? How do you know that wasn't simply because the
efficiency of the array increased? A four square array has a low radiation
resistance and depends more heavily on a low loss ground to "push
against".
It also depends on a good ground to minimize unwanted coupling through the
cables (parallel currents). Adding radials affects all those things, but
not
the wave angle (for short systems). I called VS6DO on 160 for weeks, it
wasn't until I mowed my lawn that he finally heard me! I never could
figure
that out. ;-)
>I doubt that the reduced radiation resistances involved had
>any significant bearing on the array for *receiving* signals
>because this was going from unreadability to 3/3 or 4/4 copy.
>(Those fellows never ran much power nor good antennas!).
>This had to be somewhere between 3 and 6db improvement
>in signal and is not accounted for by increased gain due
>to a small amount of reduced loss.
Loss (or efficiency) generally isn't important for receiving, but array
directivity and isolation of the antenna from noise sources are. Aside
from
changes in propagation that constantly occur making over the air minute by
minute comparisons almost meaningless, I would suspect day by day or week
by
week are even less reliable.
It is a fact that improving a ground system to isolate a receiving array
from
effects of unwanted conducted and intercepted signals "pumping the ground
system" with unwanted signals. It is also a fact that the largest
advantage
of a low band receiving array is the NULL in the direction of noise
sources
(directivity), not the gain in the direction of the desired signal.
Terry Burge
>It is a fact that improving a ground system to isolate a receiving array
>from
>effects of unwanted conducted and intercepted signals "pumping the ground
>system" with unwanted signals.
> It is also a fact that the largest
>advantage
>of a low band receiving array is the NULL in the direction of noise
>sources
>(directivity), not the gain in the direction of the desired signal.
>
Now that statement I will disagree with having ran a set of 1/2 wave
slopers on 80 meters during CQWWPHn. One set was sloped towards the
NE and the other to the SW off a 110 foot tower. There was easily
a 5db or more difference on the S-meter between the two that made
a great deal of difference in the received signals from VK,ZL,VS6, etc.
And the noise background had little/nothing to do with it.
Under different cercumstances with a carrier or teletype or something
interfering it might be a different story but that inplies a high amount
noise, not ordinary noise floor static.
Terry
KI7M
W8JI Tom
>>It is a fact that improving a ground system to isolate a receiving array
>>from
>>effects of unwanted conducted and intercepted signals "pumping the
ground
>>system" with unwanted signals.
>HuH???
Remember the coax is a two terminal line. It depends on equal out of phase
currents on the two conductors to prevent radiation or inception of EM
waves. If the ground system at the antenna is less than perfect, parallel
currents on the shield of the transmission line will not be fully
reflected by the ground. Instead they will make the ground system voltage
at the feedpoint wiggle up and down with the signal.
The antenna connects to the other wire of the feedline (the center
conductor) and provides a different voltage reference for the "wiggling"
ground voltage. Parallel currents on the shield cause a voltage
differential between the shield and center conductor any time the ground
is less than perfect. This voltage differential causes an out of phase
equal current signal that propagates back through the feedline to the
receiver. The poorer the ground and the lower the effective antenna
impedance at the ground end, the more pronounced the effect.
So anything that causes noise currents to flow on the outside of the cable
can get right back into the cable at the far end. That is why I put chokes
and grounds along my receiving array feedlines, and on any antenna I use
for receiving! ........ I hope this was more clear. :-)
>> It is also a fact that the largest
>>advantage
>>of a low band receiving array is the NULL in the direction of noise
>>sources
>>(directivity), not the gain in the direction of the desired signal.
>>
>Now that statement I will disagree with having ran a set of 1/2 wave
>slopers on 80 meters during CQWWPHn. One set was sloped towards the
>NE and the other to the SW off a 110 foot tower. There was easily
>a 5db or more difference on the S-meter between the two that made
>a great deal of difference in the received signals from VK,ZL,VS6, etc.
>And the noise background had little/nothing to do with it.
On transmitting gain is important, on receiving gain is unimportant and
directivity is important. Gain is the improvement in one direction
compared to some type of reference antenna. Directivity defines the ratio
of an antenna's response in a favored direction to the response in all the
other directions.
An antenna could have no gain over a dipole, but have 20 dB of directivity
improvement. It would offer a 20 dB improvement on receiving (if it was
favoring the desired signal and the noise was evenly distributed in all
directions) but would obviously offer no improvement on transmitting.
A Beverage is an antenna that operates in this manner. It has loss (over
dipole) in the favored direction, but it also has a few dB of directivity.
So it improves reception but would stink for transmitting. In HF
applications, the ONLY way an antenna improves reception is with
directivity.
>Under different cercumstances with a carrier or teletype or something
>interfering it might be a different story but that inplies a high amount
>noise, not ordinary noise floor static.
Ahh, but even the ordinary noise floor comes from many different
directions. It is an accumulation of thunderstorm, power line, and other
terrestrial noise sources (cosmic and galactic noises do not penetrate the
ionosphere well on 80 meters). Directivity is exactly why the sloper helps
on receiving. It would still help the same amount if the directivity
remained the same and the gain went negative. You only need enough "gain"
to increase the receiver's noise floor a fair amount when you hook up the
antenna, and that can be a "negative gain" or loss (i.e. Beverage, small
loop). Beyond that single restriction, gain is meaningless on receiving.
If you doubt this, set your receiver for maximum gain on a DX station and
add a high power ten dB attenuator to the cable. You will still copy the
DX station with exactly the same S/N ratio even though your S meter is 10
dB lower, but when you transmit you will lose ten dB of signal to noise
ratio at his end.
73 Tom