Ian Jackson wrote:
In message ,
writes





The theoretical gain of a GP with horizontal radials, radials drooping
45 degrees and and drooping 85 degrees is 1.42, 2.22, and 3.67 dbi.

I would have thought that the 'ultimate' would be when the droop IS 90
degrees (ie essentially a sleeve dipole).

With a droop of 90 degrees, is the gain slightly more than 3.67?

You can't physically have a 90 degree droop. The radials would have to
extend horizontally for some distance, then drop to 90 degrees.

That is a different antenna.

How come that you can have a 1/4 wave radiator groundplane type of
antenna with a gain that is more than a halfwave dipole (2.15 dBi) -even
if it is more-or-less a sleeve dipole?

When the radial droop approaches 90 degrees it really isn't a GP antenna
anymore, it is something else.


--
Jim Pennino

Ian Jackson wrote:
In message ,
writes
Ian Jackson wrote:
In message ,
writes



The ideal radial length for ANY ground plane antenna is slightly longer
than 1/4 wavelength, no matter for what frequencey.

Why is this? I would have thought that a 1/4 wave would be best, as it
offers the lowest impedance.

First you have to define what "best" means.

Yebbut ........
You've just said "the ideal radial length for ANY ground plane antenna
is slightly longer than 1/4 wavelength, no matter for what frequency". I
assumed that "ideal" = "best".
.
All antennas are a trade off for impedance, bandwidth, gain and in most
cases physical ability to build the structure.

Changing the radial length will have a small effect on impdedance and resonant
point but changing the radial angle will have a bigger effect on impedance
and a very small effect on resonant point.

True - but what's the angle of the radials got to do with their length?

I would suggest downloading the demo version of EZNEC and modeling a GP to
see what small changes in various parameters do.

I had presumed you had already do this (or something similar) in order
to say that slightly longer than a 1/4 wavelength was ideal. However, I
have always assumed that the steeper the angle of the radials, the more
the groundplane becomes like a vertical halfwave dipole - and the lower
becomes the angle of radiation.


OK, let's look at some numbers and see what is actually happening.

First, design a GP for 28.3 Mhz, 1/2 inch 6061 aluminum tubing with all
elements the same length and look at the element length, impedance and gain
in free space.

Then change the radial droop to 30 degrees and 45 degrees, retune for 28.3
and look at the numbers again.

All lengths are free space wavelengths of the driven element.

droop impedance gain length SWR

0 deg 23.6 Ohms 1.34 dBi .247884 lambda 2.12
30 deg 41.6 Ohms 1.83 dBi .238687 lambda 1.2
45 deg 49.1 Ohms 2.2 dBi .234493 lambda 1.02

OK, now repeat with the radials 5% longer than the driven element.

droop impedance gain length SWR

0 deg 23.3 Ohms 1.29 dBi .245373 lambda 2.15
30 deg 41.3 Ohms 1.81 dBi .236106 lambda 1.18
45 deg 50.4 Ohms 2.19 dBi .232007 lambda 1.0011

From the above the best SWR occurs with radial 5% longer than the driven
element and the droop at 45 degrees.

This is also the point of maximum 50 Ohm bandwidth. I will leave it as an
execise for the reader to get the demo EZNEC and view the bandwidth graphs.

In all cases the elevation angle of maximum radiation is 0 degrees.

Now let's come down from free space and put the longer radial version on a
typical single story house roof mounted on a pole.

The house peaks around here are about 13 feet and 10 foot TV masts are cheap,
so let's mount the antenna at 23 feet, which is .662 lambda at 28.3 Mhz, and
see what happens.

Note than because we are now over real ground vertical lobes are formed.

Again I will leave it as an exercise for the reader to get the demo EZNEC
and view the graphs.

droop impedance max gain length SWR

0 deg 22.6 Ohms 2.48 dBi @ 35 deg .245373 lambda 2.12
30 deg 43.4 Ohms 2.24 dBi @ 40 deg .236269 lambda 1.15
45 deg 51.1 Ohms 1.94 dBi @ 45 deg .231667 lambda 1.022

It should be noted that there is a large second lobe:

0 deg 1.09 dBi @ 12.5 deg
30 deg 1.37 dBi @ 12.5 deg
45 deg 1.66 dBi @ 12.5 deg

So which antenna is "best" in the real world?

I would go for 5% longer radials drooping at 45 degress.





--
Jim Pennino

El 02-04-14 22:32, escribió:
wrote:
El 02-04-14 20:25, Ian Jackson escribió:
In ,
writes
Ian wrote:
In ,

writes



The ideal radial length for ANY ground plane antenna is slightly
longer
than 1/4 wavelength, no matter for what frequencey.

Why is this? I would have thought that a 1/4 wave would be best, as it
offers the lowest impedance.

First you have to define what "best" means.

Yebbut ........
You've just said "the ideal radial length for ANY ground plane antenna
is slightly longer than 1/4 wavelength, no matter for what frequency".
I assumed that "ideal" = "best".
.
All antennas are a trade off for impedance, bandwidth, gain and in most
cases physical ability to build the structure.

Changing the radial length will have a small effect on impdedance
and resonant
point but changing the radial angle will have a bigger effect on
impedance
and a very small effect on resonant point.

True - but what's the angle of the radials got to do with their length?

I would suggest downloading the demo version of EZNEC and modeling a
GP to
see what small changes in various parameters do.

I had presumed you had already do this (or something similar) in order
to say that slightly longer than a 1/4 wavelength was ideal. However,
I have always assumed that the steeper the angle of the radials, the
more the groundplane becomes like a vertical halfwave dipole - and the
lower becomes the angle of radiation.

You are right, very steep radials become the lower half of a half wave
dipole as the currents do not cancel eachother and contribute to the
field of the quarter wave monopole. The "ultimate" version is the
sleeve dipole.

Not really.

When they are in the horizontal plane, the contribution to the total
radiation pattern is very small, and the contribution from the radials
is even zero for the vertically polarized component at zero elevation.

The theoretical gain of a GP with horizontal radials, radials drooping
45 degrees and and drooping 85 degrees is 1.42, 2.22, and 3.67 dbi.

You may check your simulations, as in free space you will not exceed
the half wave dipole gain with near vertical radials (for the quarter
wave version).

A quarter wave monopole with near vertical radials has same current
distribution as a vertical half wave dipole (use sum of current in all
radials). Of course provided that you don't have significant common
mode current in the mast or coaxial cable, as this may increase or
decrease the free space gain.

When extending both sloping radials and radiator you can get more
gain, but you get significant increase in common mode current as the
radial ground no longer act as a floating ground point, and the input
impedance has a reactive part.



The "somewhat longer then 1/4 wavelength" I also noticed with radials
connected to a coaxial braid to form a narrow band common mode choke.
the choking effect (common mode insertion loss) is better when they
are somewhat longer then 0.25lambda (depending in thickness).

The effect of sloping angle on zero elevation gain is small, and you
get hardly measurable more gain when they are almost vertical. Sloping
radials have some other advantage: less birds.

Changing the angle of the radials has little to no effect on elevation gain
unless the radial ends are a very tiny fraction of a wavelength above ground.

I can't match this statement with your earlier gain figures, or did I
misunderstand you.

Elevation radiation angle is almost totally determined by the antenna
height above ground.


Agree, but now the number of variables increases as you need to take
into account nearby ground conductivity and far away ground conductivity.



--
Wim
PA3DJS
Please remove abc first in case of PM

El 03-04-14 1:04, escribió:
Ian wrote:
In ,
writes





The theoretical gain of a GP with horizontal radials, radials drooping
45 degrees and and drooping 85 degrees is 1.42, 2.22, and 3.67 dbi.

I would have thought that the 'ultimate' would be when the droop IS 90
degrees (ie essentially a sleeve dipole).

With a droop of 90 degrees, is the gain slightly more than 3.67?

You can't physically have a 90 degree droop. The radials would have to
extend horizontally for some distance, then drop to 90 degrees.

This is the same as saying, you can't have a 90 degree radiator, as
due to wind it will bend. You know that going horizontally a few inch
and then 90 degrees down doesn't make big difference compared to 85
degrees sloping. You only may experience some length difference to get
lowest common mode current in the mast or feeder. Both option will not
give you more gain compared to a half wave dipole (free space).

That is a different antenna.




How come that you can have a 1/4 wave radiator groundplane type of
antenna with a gain that is more than a halfwave dipole (2.15 dBi) -even
if it is more-or-less a sleeve dipole?

When the radial droop approaches 90 degrees it really isn't a GP antenna
anymore, it is something else.

Is this because of electrical operation (I doubt), or naming convention?




--
Wim
PA3DJS
Please remove abc first in case of PM

Wimpie wrote:
El 03-04-14 1:04, escribió:
Ian wrote:
In ,
writes





The theoretical gain of a GP with horizontal radials, radials drooping
45 degrees and and drooping 85 degrees is 1.42, 2.22, and 3.67 dbi.

I would have thought that the 'ultimate' would be when the droop IS 90
degrees (ie essentially a sleeve dipole).

With a droop of 90 degrees, is the gain slightly more than 3.67?

You can't physically have a 90 degree droop. The radials would have to
extend horizontally for some distance, then drop to 90 degrees.

This is the same as saying, you can't have a 90 degree radiator, as
due to wind it will bend. You know that going horizontally a few inch
and then 90 degrees down doesn't make big difference compared to 85
degrees sloping. You only may experience some length difference to get
lowest common mode current in the mast or feeder. Both option will not
give you more gain compared to a half wave dipole (free space).

If you do this you do not have a ground plane antenna; you have an asymmetric
dipole with one skinny element and one fat element.


That is a different antenna.




How come that you can have a 1/4 wave radiator groundplane type of
antenna with a gain that is more than a halfwave dipole (2.15 dBi) -even
if it is more-or-less a sleeve dipole?

When the radial droop approaches 90 degrees it really isn't a GP antenna
anymore, it is something else.

Is this because of electrical operation (I doubt), or naming convention?

Actually both.

BTW, in retrospect I don't think that 3.67 dbi number for 85 degrees is
correct.

The limiting gain should be that of a vertical dipole as you pointed out.

I think the problem is that most analysis programs have issues with very
close wires and very small angles.

This can be seen by analyzing a fan dipole and decreasing the angle between
the elements. Eventually the results stop making any sense.


--
Jim Pennino

Wimpie wrote:
El 02-04-14 22:32, escribió:
wrote:
El 02-04-14 20:25, Ian Jackson escribió:
In ,
writes
Ian wrote:
In ,

writes



The ideal radial length for ANY ground plane antenna is slightly
longer
than 1/4 wavelength, no matter for what frequencey.

Why is this? I would have thought that a 1/4 wave would be best, as it
offers the lowest impedance.

First you have to define what "best" means.

Yebbut ........
You've just said "the ideal radial length for ANY ground plane antenna
is slightly longer than 1/4 wavelength, no matter for what frequency".
I assumed that "ideal" = "best".
.
All antennas are a trade off for impedance, bandwidth, gain and in most
cases physical ability to build the structure.

Changing the radial length will have a small effect on impdedance
and resonant
point but changing the radial angle will have a bigger effect on
impedance
and a very small effect on resonant point.

True - but what's the angle of the radials got to do with their length?

I would suggest downloading the demo version of EZNEC and modeling a
GP to
see what small changes in various parameters do.

I had presumed you had already do this (or something similar) in order
to say that slightly longer than a 1/4 wavelength was ideal. However,
I have always assumed that the steeper the angle of the radials, the
more the groundplane becomes like a vertical halfwave dipole - and the
lower becomes the angle of radiation.

You are right, very steep radials become the lower half of a half wave
dipole as the currents do not cancel eachother and contribute to the
field of the quarter wave monopole. The "ultimate" version is the
sleeve dipole.

Not really.

When they are in the horizontal plane, the contribution to the total
radiation pattern is very small, and the contribution from the radials
is even zero for the vertically polarized component at zero elevation.

The theoretical gain of a GP with horizontal radials, radials drooping
45 degrees and and drooping 85 degrees is 1.42, 2.22, and 3.67 dbi.

You may check your simulations, as in free space you will not exceed
the half wave dipole gain with near vertical radials (for the quarter
wave version).

Addressed in another post.

A quarter wave monopole with near vertical radials has same current
distribution as a vertical half wave dipole (use sum of current in all
radials). Of course provided that you don't have significant common
mode current in the mast or coaxial cable, as this may increase or
decrease the free space gain.

In the simulation there is no mast or cable.

When extending both sloping radials and radiator you can get more
gain, but you get significant increase in common mode current as the
radial ground no longer act as a floating ground point, and the input
impedance has a reactive part.

There is always a frequency where the reactive part is zero.

See my long post comparing configurations.

As the simulations have no cable, you can not see any common mode current
effects.

The "somewhat longer then 1/4 wavelength" I also noticed with radials
connected to a coaxial braid to form a narrow band common mode choke.
the choking effect (common mode insertion loss) is better when they
are somewhat longer then 0.25lambda (depending in thickness).

The effect of sloping angle on zero elevation gain is small, and you
get hardly measurable more gain when they are almost vertical. Sloping
radials have some other advantage: less birds.

Changing the angle of the radials has little to no effect on elevation gain
unless the radial ends are a very tiny fraction of a wavelength above ground.

I can't match this statement with your earlier gain figures, or did I
misunderstand you.

See my long post comparing configurations.

Elevation radiation angle is almost totally determined by the antenna
height above ground.


Agree, but now the number of variables increases as you need to take
into account nearby ground conductivity and far away ground conductivity.

Only if you want to simulate some particular and specific place.

For the vast majority of problems you just use average ground values.

EZNEC does allow you to change the ground values if you want to see what
happens in someplace like a desert with poor ground.


--
Jim Pennino

wrote:
Wimpie wrote:

snip

A quarter wave monopole with near vertical radials has same current
distribution as a vertical half wave dipole (use sum of current in all
radials). Of course provided that you don't have significant common
mode current in the mast or coaxial cable, as this may increase or
decrease the free space gain.

In the simulation there is no mast or cable.

This got me to wondering...

I took the model from the longer comparison post, i.e. radials at 45 deg
and 5% longer than the driven element, and looked at the currents.

The sum of the currents in the 4 radials is exactly equal to the current
in the driven element.

Then I added a 10 foot conductive mast.

This had a very small effect on the resonant frequency, SWR, and pattern,
but now there was a significant difference between the current into the
driven element and the radial/mast structure.

This says to me that you DO want to provide current choking at the feedpoint
for a GP as even if the mast is non-conductive, the outer coax shield is
conductive.


--
Jim Pennino

El 03-04-14 18:41, escribió:
wrote:
El 02-04-14 22:32, escribió:
wrote:
El 02-04-14 20:25, Ian Jackson escribió:
In ,
writes
Ian wrote:
In ,

writes



The ideal radial length for ANY ground plane antenna is slightly
longer
than 1/4 wavelength, no matter for what frequencey.

Why is this? I would have thought that a 1/4 wave would be best, as it
offers the lowest impedance.

First you have to define what "best" means.

Yebbut ........
You've just said "the ideal radial length for ANY ground plane antenna
is slightly longer than 1/4 wavelength, no matter for what frequency".
I assumed that "ideal" = "best".
.
All antennas are a trade off for impedance, bandwidth, gain and in most
cases physical ability to build the structure.

Changing the radial length will have a small effect on impdedance
and resonant
point but changing the radial angle will have a bigger effect on
impedance
and a very small effect on resonant point.

True - but what's the angle of the radials got to do with their length?

I would suggest downloading the demo version of EZNEC and modeling a
GP to
see what small changes in various parameters do.

I had presumed you had already do this (or something similar) in order
to say that slightly longer than a 1/4 wavelength was ideal. However,
I have always assumed that the steeper the angle of the radials, the
more the groundplane becomes like a vertical halfwave dipole - and the
lower becomes the angle of radiation.

You are right, very steep radials become the lower half of a half wave
dipole as the currents do not cancel eachother and contribute to the
field of the quarter wave monopole. The "ultimate" version is the
sleeve dipole.

Not really.

When they are in the horizontal plane, the contribution to the total
radiation pattern is very small, and the contribution from the radials
is even zero for the vertically polarized component at zero elevation.

The theoretical gain of a GP with horizontal radials, radials drooping
45 degrees and and drooping 85 degrees is 1.42, 2.22, and 3.67 dbi.

You may check your simulations, as in free space you will not exceed
the half wave dipole gain with near vertical radials (for the quarter
wave version).

Addressed in another post.


My results (IE3D, now Mentor Graphics Hyperlynx):
Quarter wave radiator over 4 quarter wave radials, no sloping:
impedance at resonance 23 Ohms, Gain at zero elevation: 1.52 dBi

0.625 wave radiator over 4 quarter wave radials, no sloping:
Gain at zero elevation: 1.52 dBi, 2.29 dBi at 20 degr elevation.

0.5 wave radiator over 4 quarter wave radials, no sloping:
Gain at zero elevation: 2.05 dBi.

Quarter wave radiator over 4 quarter wave radials, 45 degrees sloping:
Impedance at resonance 54 Ohms, gain at zero elevation: 1.97 dBi

Quarter wave radiator over 4 quarter wave radials, 85 degrees sloping:
Impedance at resonance 74 Ohms, gain at zero elevation: 2.14 dBi

All in free space, without a mast.

Adding a mast, especially for the sloping case can give large
deviation depending on the CM impedance as seen from the floating
ground. I did simulations and current measurements for my own mast,
but the results cannot be applied to other configurations.

As I stated before, the difference between the configurations is
hardly measurable. Nice to see that the over-rated 5/8 lambda antenna
doens't perform better then the quarter wave antenna (at low elevation
angle).

Though the design is more demanding, I prefer the half wave option as
you can use less, sloping, shorter radials without running into common
mode mast current problems.



--
Wim
PA3DJS
Please remove abc first in case of PM

Wimpie wrote:

snip

My results (IE3D, now Mentor Graphics Hyperlynx):
Quarter wave radiator over 4 quarter wave radials, no sloping:
impedance at resonance 23 Ohms, Gain at zero elevation: 1.52 dBi

0.625 wave radiator over 4 quarter wave radials, no sloping:
Gain at zero elevation: 1.52 dBi, 2.29 dBi at 20 degr elevation.

And an impedance in the hundreds of Ohms.

0.5 wave radiator over 4 quarter wave radials, no sloping:
Gain at zero elevation: 2.05 dBi.

And an impedance of about 1,000 Ohms.

Quarter wave radiator over 4 quarter wave radials, 45 degrees sloping:
Impedance at resonance 54 Ohms, gain at zero elevation: 1.97 dBi

Quarter wave radiator over 4 quarter wave radials, 85 degrees sloping:
Impedance at resonance 74 Ohms, gain at zero elevation: 2.14 dBi

All in free space, without a mast.

Again, in free space the maximum is ALWAYS at zero elevation.

Adding a mast, especially for the sloping case can give large
deviation depending on the CM impedance as seen from the floating
ground. I did simulations and current measurements for my own mast,
but the results cannot be applied to other configurations.

As I stated before, the difference between the configurations is
hardly measurable. Nice to see that the over-rated 5/8 lambda antenna
doens't perform better then the quarter wave antenna (at low elevation
angle).

I wouldn't call an impedance range of 20 Ohms to 1000 Ohms "hardly
measurable". In real life you have to feed the thing.

Though the design is more demanding, I prefer the half wave option as
you can use less, sloping, shorter radials without running into common
mode mast current problems.

And requires some sort of feed arrangement to transform 1,000 Ohms into
50 Ohms.

In my opinion, dealing with the added complexity of impedance matching,
which is almost always narrow banded, is not worth a dB or two of gain.

I think I will stick with 5% longer radials at 45 deg and some ferrite at
the feed point.



--
Jim Pennino

El 03-04-14 20:54, escribió:
wrote:

snip

My results (IE3D, now Mentor Graphics Hyperlynx):
Quarter wave radiator over 4 quarter wave radials, no sloping:
impedance at resonance 23 Ohms, Gain at zero elevation: 1.52 dBi

0.625 wave radiator over 4 quarter wave radials, no sloping:
Gain at zero elevation: 1.52 dBi, 2.29 dBi at 20 degr elevation.

And an impedance in the hundreds of Ohms.

From my memory it was closer to 50 Ohms, but for the gain figures
this isn't important.


0.5 wave radiator over 4 quarter wave radials, no sloping:
Gain at zero elevation: 2.05 dBi.

And an impedance of about 1,000 Ohms.

Depends strongly on thickness/lambda ratio, therefore I didnt mention
the value, and it isn't important for the gain.

Quarter wave radiator over 4 quarter wave radials, 45 degrees sloping:
Impedance at resonance 54 Ohms, gain at zero elevation: 1.97 dBi

Quarter wave radiator over 4 quarter wave radials, 85 degrees sloping:
Impedance at resonance 74 Ohms, gain at zero elevation: 2.14 dBi

All in free space, without a mast.

Again, in free space the maximum is ALWAYS at zero elevation.

Except for the 5/8 lambda, as I mentioned.


Adding a mast, especially for the sloping case can give large
deviation depending on the CM impedance as seen from the floating
ground. I did simulations and current measurements for my own mast,
but the results cannot be applied to other configurations.

As I stated before, the difference between the configurations is
hardly measurable. Nice to see that the over-rated 5/8 lambda antenna
doens't perform better then the quarter wave antenna (at low elevation
angle).

I wouldn't call an impedance range of 20 Ohms to 1000 Ohms "hardly
measurable". In real life you have to feed the thing.

We were discussing gain....


Though the design is more demanding, I prefer the half wave option as
you can use less, sloping, shorter radials without running into common
mode mast current problems.

And requires some sort of feed arrangement to transform 1,000 Ohms into
50 Ohms.

In my opinion, dealing with the added complexity of impedance matching,
which is almost always narrow banded, is not worth a dB or two of gain.

I think I will stick with 5% longer radials at 45 deg and some ferrite at
the feed point.


What someone will use, depends on many factors (not only electrical
ones).

I have some experience with HV, and I know how to design these
circuits without loosing useful bandwidth, so I prefer the half wave.
I like racing and wood pigeons, but not on my antennas! In residential
areas over here, generally people don't like to see lots of aluminum
in the air.

So for my situation it is not the less then 0.5 dB addditional gain
over a classic 4 radial GP, but just the common mode issue together
with visible apearance (I don't want 4 radials). Note that we have
maximum 400WPEP in PA-land, that also makes it easier.



--
Wim
PA3DJS
Please remove abc first in case of PM