On 10/10/2014 6:49 PM, wrote:
David Platt wrote:
* It is a wire dipole (keep the wire AWG to practical values, please).
* The dipole's length is variable.

I forgot to add to the second * item, that we could use .5 wavelengths,
.25 wavelengths, .1 wavelengths and .05 wavelengths for comparison purposes.

In many of the antenna discussions I've read, the term "dipole" is
often reserved for antennas which are of approximately half-wavelength
resonance at the frequency in question.

That is the amateur radio convention, but not so for a physicist.

When a physicist is talking about 1/2 wave dipoles it is either stated
explicitly or obvious from the context.

"Doublet" is used for a center-fed dipole-like antenna of other
lengths... for example, a 40-meter dipole which is actually being used
on 15 meters would be referred to as a doublet. In this sort of
terminology, one could consider a "dipole" to be a particular special
case of a "doublet".

Again, this is amateur radio convention. I don't think I have ever seen
the term doublet in a "science" publication.

Isn't this a amateur radio group? I did a google search and while
dipole comes up a lot more often that simply indicates they are more
often used.

When I did a search for definitions of doublet I found only a few that
refer to a short dipole as a doublet, mostly it is a synonym for dipole.


Of course, there's the other school of thought which calls *all* of
these antennas "dipoles" (e.g. a "short dipole" or a "5/4-wave
center-fed dipole".

Yep, like you find in physics books, which BTW will always be center
fed unless otherwise specified.

All this language stuff becomes important if you concider the reading
audience can consist of anyone from someone with a passing interest
and little education to a Sheldon Cooper theoretical physicist.

One term I have seen used often is Hertzian dipole for an electrically
short dipole.

--

Rick

On Fri, 10 Oct 2014 20:08:51 -0000, wrote:
(...)

This group is the only one that I read where topic drift ocurrs almost
immediately, often on the first or second followup posting. In other
groups, it usually takes 5 to 7 messages in a thread to produce the
same result. The problem is that everyone (including me) tends to
post about things that they are familiar with, know something about,
or are passionate about the topic (even if they know nothing about
it). If the initial question or request for comments does not fit
these criteria, many people will simply warp the discussion until it
does. For example, I know something about the mecahanics of antennas,
something about using NEC models, and have a fair collection of
anecdotes and photos from my days in the 2way radio biz. However,
this thread is allegedly about antenna theory, where I am seriously
lacking. Since I don't want to twist the topic away from theory and
into reality, I haven't said anything worth reading (including this
rant).

On the contrary, I think it is quite a worthwile effort especially if
you summarize and publish the results for the benefit of all the arm
wavers saying things like "short antennas are poor radiators".

Good idea. I'm curious, but not curious enough to dig through several
hundred off topic comments to excavate relevent information.

Then you can discuss real data instead of arguing about what some naif
pulled out of his ass.

"Data is free. Information must be fought for"
(Former statistics instructor in college about 1968).

Speaking of dipole antennas, I did this study a while back:
http://802.11junk.com/jeffl/antennas/vertical-dipole/index.html
Animated version:
http://802.11junk.com/jeffl/antennas/vertical-dipole/slides/animated-v-dipole.html
It's a 1/2 wave dipole at various heights above a real ground. Any
semblance to textbook dipole pattern is long gone.

I also did a study of monopoles of various lengths above a ground.
There are a few that are less than 1/4 wave long which should help
with some short antenna phenomenon.
http://802.11junk.com/jeffl/antennas/Monopole/index.html
Length Gain
wl dBi
0.050 4.75
0.125 4.85
0.250 5.19
0.500 6.96
0.625 8.01
Notice that the gain doesn't really drop very much when the monopole
is shorter than 1/4 wavelength long. A 1/2 wave dipole exhibits a
similar lack of gain loss for short antennas. So, why are short
antennas generally frowned upon? Lots of reasons but the big one are
losses in the matching networks. the 0.050 wavelength antenna looks
like about 700 ohms impedance. The 0.125 antenna is about


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On 10/10/2014 3:08 PM, wrote:
John S wrote:
On 10/10/2014 1:50 PM, wrote:
John S wrote:
The thread about a new philosophy was about 246 posts long including the
drift into mechanisms of photons and quarks.

I wish to start this thread by discussing the merits of dipoles with
respect to length. I really don't want to start another thread of
flames. Please help me do this.

Let's confine the discussion to...

* It is a wire dipole (keep the wire AWG to practical values, please).
* The dipole's length is variable.
* It may or may not have wire resistance (state your value).
* The source is placed at the center of the antenna.
* There is no transmission line.
* There are no matching devices.
* This is a theoretical discussion but may include practical parameters.

EZNEC or NEC, provides answers to these questions, I think. I will do
some modeling and I'd like to compare my results to other results.

It does.

Let's kick them around.

There is nothing to kick around unless you want to argue about whether
the various NEC implementations provide valid results.

What NEC will tell you is that as the .5 wave dipole gets shorter and
shorter, the resistive part of the antenna impedance decreases and the
capacitive part increases and the patern and maximum gain changes very
slightly.

As the ratio of resistive input impedance to the wire resistance get
smaller, the I^2R losses increase.

But don't let that stop you from doing it.

People often learn much better from actual doing then they do from lectures.

BTW, you have to pay some attention to segmentation and as you get smaller
and smaller the segmentation limits just how accurate the simulation is.

Ok, I guess I had hoped to have a good conversation concerning the subject.

Jim has convinced me that it is not a worthwhile effort.

Cheers.


On the contrary, I think it is quite a worthwile effort especially if
you summarize and publish the results for the benefit of all the arm
wavers saying things like "short antennas are poor radiators".

Then you can discuss real data instead of arguing about what some naif
pulled out of his ass.

On 10/10/2014 3:08 PM, wrote:
John S wrote:
On 10/10/2014 1:50 PM, wrote:
John S wrote:
The thread about a new philosophy was about 246 posts long including the
drift into mechanisms of photons and quarks.

I wish to start this thread by discussing the merits of dipoles with
respect to length. I really don't want to start another thread of
flames. Please help me do this.

Let's confine the discussion to...

* It is a wire dipole (keep the wire AWG to practical values, please).
* The dipole's length is variable.
* It may or may not have wire resistance (state your value).
* The source is placed at the center of the antenna.
* There is no transmission line.
* There are no matching devices.
* This is a theoretical discussion but may include practical parameters.

EZNEC or NEC, provides answers to these questions, I think. I will do
some modeling and I'd like to compare my results to other results.

It does.

Let's kick them around.

There is nothing to kick around unless you want to argue about whether
the various NEC implementations provide valid results.

What NEC will tell you is that as the .5 wave dipole gets shorter and
shorter, the resistive part of the antenna impedance decreases and the
capacitive part increases and the patern and maximum gain changes very
slightly.

As the ratio of resistive input impedance to the wire resistance get
smaller, the I^2R losses increase.

But don't let that stop you from doing it.

People often learn much better from actual doing then they do from lectures.

BTW, you have to pay some attention to segmentation and as you get smaller
and smaller the segmentation limits just how accurate the simulation is.

Ok, I guess I had hoped to have a good conversation concerning the subject.

Jim has convinced me that it is not a worthwhile effort.

Cheers.


On the contrary, I think it is quite a worthwile effort especially if
you summarize and publish the results for the benefit of all the arm
wavers saying things like "short antennas are poor radiators".

Then you can discuss real data instead of arguing about what some naif
pulled out of his ass.

Apologies for the blank post. I hit the wrong button.

OK, so lets analyze my results:

Conditions are free space, wire is #14 gauge but may have zero ohms
where noted. The antenna is a dipole with the source connected at the
center, F=7MHz. I'm using EZNEC with a source of 1 watt. Antenna
resonance plays no part in this. # segments = 99 unless otherwise noted.

Lambda Wire Rin Gavg(dbi) Gmax(dbi) Efficiency

0.5 zero 80 0 2.16 100%
0.5 #14 73.6 -.09 2.08 98%

0.25 zero 13.2 0 1.85 100%
0.25 #14 13.7 -.17 1.69 96%

0.125 zero 3 0 1.78 100%
0.125 #14 3.25 -.33 1.45 93%

0.05 zero .464 0 1.76 100%
0.05 #14 .556 -.78 0.98 83%

Rin is the terminal resistance only. Gave is the average gain integrated
over the pattern, Gmax is the highest gain detected.

Unless I have done something wrong, I see that a dipole that is .05
wavelengths long is within 20% of being as efficient as a half-wave
dipole. Even including wire resistance.

I invite discussion in any case.

John S wrote:

snip

OK, so lets analyze my results:

Conditions are free space, wire is #14 gauge but may have zero ohms
where noted. The antenna is a dipole with the source connected at the
center, F=7MHz. I'm using EZNEC with a source of 1 watt. Antenna
resonance plays no part in this. # segments = 99 unless otherwise noted.

Lambda Wire Rin Gavg(dbi) Gmax(dbi) Efficiency

0.5 zero 80 0 2.16 100%
0.5 #14 73.6 -.09 2.08 98%

0.25 zero 13.2 0 1.85 100%
0.25 #14 13.7 -.17 1.69 96%

0.125 zero 3 0 1.78 100%
0.125 #14 3.25 -.33 1.45 93%

0.05 zero .464 0 1.76 100%
0.05 #14 .556 -.78 0.98 83%

Rin is the terminal resistance only. Gave is the average gain integrated
over the pattern, Gmax is the highest gain detected.

Unless I have done something wrong, I see that a dipole that is .05
wavelengths long is within 20% of being as efficient as a half-wave
dipole. Even including wire resistance.

I invite discussion in any case.

The diameter of #14 solid wire is 0.0641"; how about a line for #8, which
is 0.1285"?



--
Jim Pennino

Jeff Liebermann wrote:

snip

Speaking of dipole antennas, I did this study a while back:
http://802.11junk.com/jeffl/antennas/vertical-dipole/index.html
Animated version:
http://802.11junk.com/jeffl/antennas/vertical-dipole/slides/animated-v-dipole.html
It's a 1/2 wave dipole at various heights above a real ground. Any
semblance to textbook dipole pattern is long gone.

Yep, ground has a huge effect on some types of antennas.

An instructive slide show would be the vertical pattern of a horizontal
1/2 dipole at .1, .2, ... .5 wavelengths over ground.

Another one would be a 3 element beam at those heights.

I also did a study of monopoles of various lengths above a ground.
There are a few that are less than 1/4 wave long which should help
with some short antenna phenomenon.
http://802.11junk.com/jeffl/antennas/Monopole/index.html
Length Gain
wl dBi
0.050 4.75
0.125 4.85
0.250 5.19
0.500 6.96
0.625 8.01
Notice that the gain doesn't really drop very much when the monopole
is shorter than 1/4 wavelength long. A 1/2 wave dipole exhibits a
similar lack of gain loss for short antennas. So, why are short
antennas generally frowned upon? Lots of reasons but the big one are
losses in the matching networks. the 0.050 wavelength antenna looks
like about 700 ohms impedance. The 0.125 antenna is about

Nope, the vertical does the same thing when shortened from 1/4 as a
dipole shortened from 1/2 wave.



--
Jim Pennino

On Sat, 11 Oct 2014 11:47:22 -0500, John S
wrote:

Unless I have done something wrong, I see that a dipole that is .05
wavelengths long is within 20% of being as efficient as a half-wave
dipole. Even including wire resistance.

Sounds about right except that it doesn't include any losses
introduced by the necessary matching network and real ground losses at
HF frequencies. Expanding my table to include radiation efficiency:

http://802.11junk.com/jeffl/antennas/Monopole/index.html
Length Gain Radiation
wl dBi Efficiency
0.050 4.75 99.09%
0.125 4.85 99.66%
0.250 5.19 99.93%
0.500 6.96 99.97%
0.625 8.01 99.93%
In other words, there's nothing inherent in the length of the radiator
that would affect radiation efficiency. If there is a drop in
radiation efficiency, then it's mostly due to ground losses, material
losses I2R, and matching losses).

I invite discussion in any case.

NEC: Power Efficiency vs. Radiation Efficiency
L. B. Cebik, W4RNL
http://www.antennex.com/w4rnl/col0504/amod75.html
Lots of examples of how "efficiency" calculations work, and how
various common antenna configurations affect the results. (I need to
re-read the article as there's plenty about this which I don't
understand very well). Test cases 5 and 6 are short monopoles, which
should have something to do with short dipoles. From the bottom of
the page:
"Unlike the vertical monopole, the horizontal dipole shows much
more regular changes of radiation efficiency with changes of
soil type, ranging from 80.01% over very good soil to 65.93%
over very poor soil."

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Jeff Liebermann wrote:
On Sat, 11 Oct 2014 11:47:22 -0500, John S
wrote:

Unless I have done something wrong, I see that a dipole that is .05
wavelengths long is within 20% of being as efficient as a half-wave
dipole. Even including wire resistance.

Sounds about right except that it doesn't include any losses
introduced by the necessary matching network and real ground losses at
HF frequencies. Expanding my table to include radiation efficiency:

There is no feed because it is the ANTENNA that is being analyzed, not
an antenna SYSTEM.

And while I don't know if the simulation included it, NEC can include
the ground losses for the ANTENNA.

http://802.11junk.com/jeffl/antennas/Monopole/index.html
Length Gain Radiation
wl dBi Efficiency
0.050 4.75 99.09%
0.125 4.85 99.66%
0.250 5.19 99.93%
0.500 6.96 99.97%
0.625 8.01 99.93%
In other words, there's nothing inherent in the length of the radiator
that would affect radiation efficiency. If there is a drop in
radiation efficiency, then it's mostly due to ground losses, material
losses I2R, and matching losses).

Ground and I^2R losses of the antenna are shown by NEC.

Matching losses are NOT part of the antenna.

I invite discussion in any case.

NEC: Power Efficiency vs. Radiation Efficiency
L. B. Cebik, W4RNL
http://www.antennex.com/w4rnl/col0504/amod75.html
Lots of examples of how "efficiency" calculations work, and how
various common antenna configurations affect the results. (I need to
re-read the article as there's plenty about this which I don't
understand very well). Test cases 5 and 6 are short monopoles, which
should have something to do with short dipoles. From the bottom of
the page:
"Unlike the vertical monopole, the horizontal dipole shows much
more regular changes of radiation efficiency with changes of
soil type, ranging from 80.01% over very good soil to 65.93%
over very poor soil."

Yep, and once the issue of size versus efficieny is put to rest, it would
not be a bad idea to look at the real effects of ground, both in terms
of height in wavelengths and soil quality.



--
Jim Pennino

wrote in :

Yep, and once the issue of size versus efficieny is put to rest, it would
not be a bad idea to look at the real effects of ground, both in terms
of height in wavelengths and soil quality.


Thankyou. This is the bit I am most weak on (though I did not grasp all the
previous detail, the ground loss was an issue not lost on me, and is the
one I need most to solve in my next efforts). What is this NEC program, and
where can I find it? Sorry, but I have to ask, or Google will likely flood me
with Nippon Electric Company details.

Lostgallifreyan wrote:
wrote in :

Yep, and once the issue of size versus efficieny is put to rest, it would
not be a bad idea to look at the real effects of ground, both in terms
of height in wavelengths and soil quality.


Thankyou. This is the bit I am most weak on (though I did not grasp all the
previous detail, the ground loss was an issue not lost on me, and is the
one I need most to solve in my next efforts). What is this NEC program, and
where can I find it? Sorry, but I have to ask, or Google will likely flood me
with Nippon Electric Company details.

Look at this for an overview of NEC.

http://en.wikipedia.org/wiki/Numeric...magnetics_Code

At the bottom under External links you will find both free and commercial
implementations.

EZNEC by W7EL is popular among hams and has a free demo version that is
fully functional but limited in how complex a model you can generate.

NEC itself just crunches and produces numbers, but there are several
versions, including EZNEC, which have graphical interfaces to make
it easier to build the model and view the results.

I've used EZNEC+ for years.



--
Jim Pennino