On Tue, 1 Jun 2004 10:11:14 +0100, "Richard"
wrote:

|
|"Wes Stewart" wrote in message
.. .
| On Mon, 31 May 2004 15:59:31 +0100, "Richard"
| wrote:
|
| |I'm just about getting around to making my 5 or 6 element marine band
|yagi,
| |covering 156 - 162 Mhz RECEIVE ONLY. Do I have to employ a folded dipole
|
| No.
|
| |or
| |can I use just a simple hertz dipole as the "driven" element? I'm trying
|to
| |get away from using a folded dipole if I can. Any wideband designs out
|there
| |that use a simple hertz dipole that I can scale? I'm looking, but if you
| |have some links aready. TIA.
|
| All dimensions in inches
|
| X Y Z X Y Z Dia
|
| 0.000 -18.553 0.000 0.000 18.553 0.000 0.187
| 9.275 -18.368 0.000 9.275 18.368 0.000 0.312
| 12.250 -17.077 0.000 12.250 17.077 0.000 0.187
| 23.800 -16.672 0.000 23.800 16.672 0.000 0.187
| 34.213 -16.672 0.000 34.213 16.672 0.000 0.187
| 47.211 -16.229 0.000 47.211 16.229 0.000 0.187
|
|
|Sorry I'm a bit uncertain of what you wrote.
|
|Obviously an efficient way to provide data for yagi's, but how do you read
|it?
|
|Is there a webpage associated with the design, so I can look at the general
|construction?
|
|Rich.

This is the format for almost any of the common antenna modeling
programs. See for example:

Free:

http://www.qsl.net/wb6tpu/swindex.html (look for files beginning
4nec2)

Almost free:

http://www.qsl.net/ac6la/multinec.html

Free demo:

www.eznec.com

etc.

With a fixed width font, the columns would line up and each line would
represent one element, starting from the rear (reflector end) of the
boom. The data are the X,Y,Z coordinates of the ends of the elements
and the diameter of that element. Z is this case is zero since the
model assumes free space. It also assumes no effects from the
supporting boom; the usual compensations must be made for this,
depending on the element mounting method. See:

http://www.antennspecialisten.se/en/ham/tech/bc.html

I assumed that you had some experience with Yagi design and
construction and were simply looking for a design that gave both the
required bandwidth and a fifty-ohm feedpoint; something this design
delivers.

I can easily be mistaken, but I don't believe the NBS Yagi designs were
ever promoted as being optimal in any way. Their purpose, I believe, was
to provide a set of easy-to-duplicate designs whose gains were well
documented (although I understand an error was made in measuring the
originals) and which anyone could construct as reference antennas with
known gain.

Roy Lewallen, W7EL

Ian White, G3SEK wrote:

The only special feature of the NBS designs (other than the prestigious
mailing address) was a self-imposed restriction to equal spacings
between directors. That turned out to be an evolutionary dead-end, and
you won't find it in any modern optimized yagi design.

The design proposed by Wes doesn't have that feature - fortunately - and
is actually one of the OWA (Optimized Wideband Array) family. One of the
special features of that family of yagis is a very close spacing between
the driven element and the first director, which raises the feedpoint
impedance to 50 ohms and allows a simple split-dipole driven element.

Even this feature is not original to the OWA family. I don't know when
(or indeed how often) it might have been invented, but it was first
systematically used by DL6WU some 25 years ago. With additional help
from computer analysis and optimization, the OWA family continued the
development.

Modern Y-U design allows for a higher feedpoint SWR. Matching loss is
offset by
higher gain and better consistency of F/B. If no matching, then a shaped,
folded dipole is used. Typically a F/B better than 22 dB (sim) is
acheived with
a 10% bandwidth.

We're now in the excellent position of having several alternative ways
to design a yagi to meet each user's specific combination of
requirements. That approach is certainly one of them.

Roy Lewallen wrote:
I can easily be mistaken, but I don't believe the NBS Yagi designs were
ever promoted as being optimal in any way. Their purpose, I believe,
was to provide a set of easy-to-duplicate designs whose gains were well
documented (although I understand an error was made in measuring the
originals) and which anyone could construct as reference antennas with
known gain.


They were in fact described as "optimum designs" in the original paper,
but it's clear that claim was made subject to a number of reservations.
Most of the hype came from antenna manufacturers who subsequently picked
up the designs from the published reports, and misused the "NBS" name to
boost the credibility of their own products.

With hindsight, the NBS designs were really quite good, and their
forward gain is still competitive with more modern designs of the same
boom length. But modern yagi designs are generally better, because there
have been a further 35-40 years of development and optimization. In
particular, the last 10 years have benefited from the availability of
computer modeling techniques - you can now do more good work in a few
evenings than the NBS program achieved in as many years.

As a result of this development, you can now have a better combination
of features - for example, cleaner patterns, wider bandwidth and/or
greater tolerance to dimensional errors, and easier impedance matching -
and keep the good forward gain as well.

Many modern designs have been developed as complete 'families', with
simple design rules that let you add or remove elements (changing the
boom length accordingly) to create new yagis, each of which will be
close to optimum for its boom length. The NBS yagis don't have that
'family' feature - each one is an individual design, and attempts to
adapt them have generally not been successful. This lack of adaptability
is a direct consequence of the original decision to use the same spacing
between all elements; that's why I described it as an "evolutionary
dead-end".

It costs exactly the same to build a good, modern yagi as it does to
build an older, inferior design - the only differences are in *where*
you apply the hacksaw and the drill. Therefore there's not much point in
building an inferior design... which is what the NBS yagis have now
become.

For many years, the greatest value of the NBS yagis was that they had
accurate gain measurements, so they could be used as reliable benchmarks
in antenna gain shootouts. (The known error in the gain measurements
applies to the 2-element yagi only.) Even that use has now been
overtaken by computer modeling.

In summary, the NBS yagis deserve respect for their major contribution
to the art and science of yagi design, but they are now mainly of
historical interest.

For examples of modern yagis, and tips on construction, see the 'VHF/UHF
Long Yagi Workshop' on my website.


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

"Wes Stewart" wrote in message
...
On Mon, 31 May 2004 15:59:31 +0100, "Richard"
wrote:

|I'm just about getting around to making my 5 or 6 element marine band
yagi,
|covering 156 - 162 Mhz RECEIVE ONLY. Do I have to employ a folded dipole

No.

|or
|can I use just a simple hertz dipole as the "driven" element? I'm trying
to
|get away from using a folded dipole if I can. Any wideband designs out
there
|that use a simple hertz dipole that I can scale? I'm looking, but if you
|have some links aready. TIA.

All dimensions in inches

X Y Z X Y Z Dia

0.000 -18.553 0.000 0.000 18.553 0.000 0.187
9.275 -18.368 0.000 9.275 18.368 0.000 0.312
12.250 -17.077 0.000 12.250 17.077 0.000 0.187
23.800 -16.672 0.000 23.800 16.672 0.000 0.187
34.213 -16.672 0.000 34.213 16.672 0.000 0.187
47.211 -16.229 0.000 47.211 16.229 0.000 0.187

IOW:
X Y Z Dia

REF End 1 0.000 -18.553 0.000 0.187
End2 0.000 18.553 0.000


DE End1 9.275 -18.368 0.000 0.312
End2 9.275 18.368 0.000


DIR1 End1 12.250 -17.077 0.000 0.187
End2 12.250 17.077 0.000


DIR2 End1 23.800 -16.672 0.000 0.187
End2 23.800 16.672 0.000


DIR4 End1 34.213 -16.672 0.000 0.187
End2 34.213 16.672 0.000


DIR5 End1 47.211 -16.229 0.000 0.187
End2 47.211 16.229 0.000

I'll see how I get along with the 4Nec2 program.

What is the DE,just a regular hertz dipole? Is DE specified as to what it is
(hertz, folded) somewhere in 4nec2?

In the Geometry Edit I see "Radius" for wire diameter I suppose. Do I
calculate the radius from the dia given in the table above? TIA.

Ian,

Thanks very much for the correction and additional information. Like a
lot of other folks, I learn a lot from this newsgroup.

Roy Lewallen, W7EL

On Wed, 2 Jun 2004 18:44:28 +0100, "Richard"
wrote:

|IOW:
yes

|I'll see how I get along with the 4Nec2 program.
|
|What is the DE,just a regular hertz dipole? Is DE specified as to what it is
|(hertz, folded) somewhere in 4nec2?

No, it's not specified as such. Modeling programs assume straight
"wires" so to model a FD you have to "construct" one from straight
wires. This is not trivial however, so don't try it until you
understand the pitfalls of "closely-spaced wires."

I strongly recommend that you download Roy's EZNEC demo, if for no
other reason than to get the excellent "Help" section.


|In the Geometry Edit I see "Radius" for wire diameter I suppose. Do I
|calculate the radius from the dia given in the table above? TIA.

Yes.


Also, when you get to building the real antenna, you have the problem
of how to feed the DE. The design in question offers a nice 50 ohm
impedance, but just opening the center and feeding directly with coax
is going to be problematic. In the Arizona desert where I live I've
gotten away with pigtailing the coax and using ferrite sleeve baluns.
Since this is a marine band antenna, I assume there's water around
someplace and that may be unsatisfactory for you.

What did you have planned?