On 2/22/2015 4:09 PM, wrote:
John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

I modeled this in free space starting with the elements offset by 0.01
lambda in the Z axis and the length adjusted for resonance.

I then stepped the offset up to 0.1 lambda.

As the offset increased, the gain stayed the same to within a fraction
of a dB but the pattern rotated off the Y axis.

The only effect of resizing the elements for the 0.1 lambda section was
to bring the reactance back to zero; nothing else changed by any
significant amount

This might be usefull if one had only E/W supports and wanted to skew the
pattern from N/S.

Great, Jim. Thanks for that info. That pretty much nails it.

On 2/21/2015 11:35 AM, John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at
all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

Hi, John,

I finally got a chance to give this the consideration it deserves.

A lot is going to depend on how far offset the two halves are displaced.
If they're less than 0.1 or 0.2 wavelengths, you aren't going to notice
the difference. Too many other things will affect the operation -
trees, wire fences, etc.

What will affect it the most will be the feeding of the antenna. My
suggestion would be to feed it from the middle of the difference, i.e.

________
|
|
* *
|
|________

Where the '*'s are the feed point. Both halves are 'L' shaped (at the
feedpoint, anyway), with one going up and one going down. The vertical
pieces are of equal length, as are the horizontal pieces. This will
keep the antenna pretty close to being balanced (as much as any dipole
in real life, anyway).

Also, best would be to run the twinlead perpendicular to the feed -
straight towards or straight away from you in the above picture.

The further you have to displace the two halves, the less it will act
like a dipole. However, what would be more of a concern to me is the
capacitive effects of the ground when the distance of each half is at a
significantly different, yet close to the ground. This could also
unbalance the antenna, throwing off the pattern and the impedance.
Checking each half at the feedpoint with an antenna analyzer (treating
them as a 1/4 wave antenna). But unless the antenna is close to the
ground (in wavelengths) and the two are significantly separated
vertically in wavelengths, I doubt it will be a problem.

--
==================
Remove the "x" from my email address
Jerry, AI0K

==================

On 2/23/2015 9:49 PM, Jerry Stuckle wrote:
On 2/21/2015 11:35 AM, John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at
all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

Hi, John,

I finally got a chance to give this the consideration it deserves.

A lot is going to depend on how far offset the two halves are displaced.
If they're less than 0.1 or 0.2 wavelengths, you aren't going to notice
the difference. Too many other things will affect the operation -
trees, wire fences, etc.

What will affect it the most will be the feeding of the antenna. My
suggestion would be to feed it from the middle of the difference, i.e.

________
|
|
* *
|
|________

Where the '*'s are the feed point. Both halves are 'L' shaped (at the
feedpoint, anyway), with one going up and one going down. The vertical
pieces are of equal length, as are the horizontal pieces. This will
keep the antenna pretty close to being balanced (as much as any dipole
in real life, anyway).

Also, best would be to run the twinlead perpendicular to the feed -
straight towards or straight away from you in the above picture.

The further you have to displace the two halves, the less it will act
like a dipole. However, what would be more of a concern to me is the
capacitive effects of the ground when the distance of each half is at a
significantly different, yet close to the ground. This could also
unbalance the antenna, throwing off the pattern and the impedance.
Checking each half at the feedpoint with an antenna analyzer (treating
them as a 1/4 wave antenna). But unless the antenna is close to the
ground (in wavelengths) and the two are significantly separated
vertically in wavelengths, I doubt it will be a problem.


Good suggestions, Jerry. In fact a modification of your drawing above
could be:

________
*

*________

....which looks like a solution.

Thanks.

On 2/23/2015 11:57 PM, John S wrote:
On 2/23/2015 9:49 PM, Jerry Stuckle wrote:
On 2/21/2015 11:35 AM, John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at
all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

Hi, John,

I finally got a chance to give this the consideration it deserves.

A lot is going to depend on how far offset the two halves are displaced.
If they're less than 0.1 or 0.2 wavelengths, you aren't going to notice
the difference. Too many other things will affect the operation -
trees, wire fences, etc.

What will affect it the most will be the feeding of the antenna. My
suggestion would be to feed it from the middle of the difference, i.e.

________
|
|
* *
|
|________

Where the '*'s are the feed point. Both halves are 'L' shaped (at the
feedpoint, anyway), with one going up and one going down. The vertical
pieces are of equal length, as are the horizontal pieces. This will
keep the antenna pretty close to being balanced (as much as any dipole
in real life, anyway).

Also, best would be to run the twinlead perpendicular to the feed -
straight towards or straight away from you in the above picture.

The further you have to displace the two halves, the less it will act
like a dipole. However, what would be more of a concern to me is the
capacitive effects of the ground when the distance of each half is at a
significantly different, yet close to the ground. This could also
unbalance the antenna, throwing off the pattern and the impedance.
Checking each half at the feedpoint with an antenna analyzer (treating
them as a 1/4 wave antenna). But unless the antenna is close to the
ground (in wavelengths) and the two are significantly separated
vertically in wavelengths, I doubt it will be a problem.


Good suggestions, Jerry. In fact a modification of your drawing above
could be:

________
*

*________

...which looks like a solution.

Thanks.

John,

Yes, if your displacement is less than the width of the balanced line,
that would work great. And if it's that close, you're not going to
notice any difference between it and a dipole without the displacement.

--
==================
Remove the "x" from my email address
Jerry, AI0K

==================

John S wrote:
On 2/22/2015 4:09 PM, wrote:
John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

I modeled this in free space starting with the elements offset by 0.01
lambda in the Z axis and the length adjusted for resonance.

I then stepped the offset up to 0.1 lambda.

As the offset increased, the gain stayed the same to within a fraction
of a dB but the pattern rotated off the Y axis.

The only effect of resizing the elements for the 0.1 lambda section was
to bring the reactance back to zero; nothing else changed by any
significant amount

This might be usefull if one had only E/W supports and wanted to skew the
pattern from N/S.

Great, Jim. Thanks for that info. That pretty much nails it.

Well, not quite as it turns out.

I did it again in free space looking at the 3D pattern.

What happens is that as the separtion increases, the familiar doughnut
shaped 3D pattern start to tilt along the axis of the doughnut.

At 0.01 lambda the tilt is about 2 degrees; at 0.1 lambda about 24 degrees;
at 0.2 lambda 56 degrees.

I then put it over real, average ground with the lower element at .5 lambda.

What happens is difficult to describe in a few words and your really
should download the demo version of EZNEC to see it for yourself.

Essentially what happens is the pattern beging to favor one direction,
the elevation angle of the lobes go up, and multiple lobes start appearing.

The bottom line is that below about 0.02 lambda (about 16 inches at 20M)
you will not see any practical difference in the real world.

Above that an the pattern gets so weird I see no application for it.



--
Jim Pennino

On 2/24/2015 1:08 PM, wrote:
John S wrote:
On 2/22/2015 4:09 PM, wrote:
John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

I modeled this in free space starting with the elements offset by 0.01
lambda in the Z axis and the length adjusted for resonance.

I then stepped the offset up to 0.1 lambda.

As the offset increased, the gain stayed the same to within a fraction
of a dB but the pattern rotated off the Y axis.

The only effect of resizing the elements for the 0.1 lambda section was
to bring the reactance back to zero; nothing else changed by any
significant amount

This might be usefull if one had only E/W supports and wanted to skew the
pattern from N/S.

Great, Jim. Thanks for that info. That pretty much nails it.

Well, not quite as it turns out.

I did it again in free space looking at the 3D pattern.

What happens is that as the separtion increases, the familiar doughnut
shaped 3D pattern start to tilt along the axis of the doughnut.

At 0.01 lambda the tilt is about 2 degrees; at 0.1 lambda about 24 degrees;
at 0.2 lambda 56 degrees.

I then put it over real, average ground with the lower element at .5 lambda.

What happens is difficult to describe in a few words and your really
should download the demo version of EZNEC to see it for yourself.

Essentially what happens is the pattern beging to favor one direction,
the elevation angle of the lobes go up, and multiple lobes start appearing.

The bottom line is that below about 0.02 lambda (about 16 inches at 20M)
you will not see any practical difference in the real world.

Above that an the pattern gets so weird I see no application for it.

I'll do what you suggest. Thanks for your efforts. Much info there.

On 2/24/2015 6:51 AM, Jerry Stuckle wrote:
On 2/23/2015 11:57 PM, John S wrote:
On 2/23/2015 9:49 PM, Jerry Stuckle wrote:
On 2/21/2015 11:35 AM, John S wrote:
Suppose I have a dipole where the two quarter-wave elements are not
exactly aligned. That is, the axis of, say, the left element is
displaced by some amount from the axis of the right element.

Then, what about feeding them from a balanced line?

The questions a

* How will the antenna impedance and current distribution suffer, if at
all?

* What feed system is required? 90 degrees to the axis is preferable,
but that means that the twin lead has one side longer than the other.

Is this interesting? Or, has it been discussed before? What was the
outcome, if any?

Thanks, Guys.

Hi, John,

I finally got a chance to give this the consideration it deserves.

A lot is going to depend on how far offset the two halves are displaced.
If they're less than 0.1 or 0.2 wavelengths, you aren't going to notice
the difference. Too many other things will affect the operation -
trees, wire fences, etc.

What will affect it the most will be the feeding of the antenna. My
suggestion would be to feed it from the middle of the difference, i.e.

________
|
|
* *
|
|________

Where the '*'s are the feed point. Both halves are 'L' shaped (at the
feedpoint, anyway), with one going up and one going down. The vertical
pieces are of equal length, as are the horizontal pieces. This will
keep the antenna pretty close to being balanced (as much as any dipole
in real life, anyway).

Also, best would be to run the twinlead perpendicular to the feed -
straight towards or straight away from you in the above picture.

The further you have to displace the two halves, the less it will act
like a dipole. However, what would be more of a concern to me is the
capacitive effects of the ground when the distance of each half is at a
significantly different, yet close to the ground. This could also
unbalance the antenna, throwing off the pattern and the impedance.
Checking each half at the feedpoint with an antenna analyzer (treating
them as a 1/4 wave antenna). But unless the antenna is close to the
ground (in wavelengths) and the two are significantly separated
vertically in wavelengths, I doubt it will be a problem.


Good suggestions, Jerry. In fact a modification of your drawing above
could be:

________
*

*________

...which looks like a solution.

Thanks.

John,

Yes, if your displacement is less than the width of the balanced line,
that would work great. And if it's that close, you're not going to
notice any difference between it and a dipole without the displacement.

That was my thinking as well. Thanks for that.