Lee,
We probably will just have to disagree about this.
From my viewpoint, the behavior isn't too much different than a big yagi
stack or other antennas we are used to.
The size of the array generally sets the directivity limits. We can add more
elements that are closer-in than optimum, and that can certainly help if the
size is smaller than optimum, but the trade is gain or pattern cleanliness
and sharpness for size.
The forward two elements and back two elements are too close to contribute
broadside pattern, which is what provides the clean pattern absent major
side lobes in the full size 8 circle. As a matter of fact, adding them in
destroys some of the broadside directivity.
If, however, we make the array so small that it loses broadside pattern
multiplication, then we can see an increase in directivity through the small
endfire length increase.
A .327wl radius array gives about .25 wl endfire spacing in the primary
cells (the center elements), and is not improved in pattern quality by
adding the forward and rearward cells. The two forward pairs and rearward
pairs are not only too close to have broadside pattern contribution, they
are closer endfire. They are about 75% of the endfire spacing in the central
quad, and nearly 40% of the broadside width. They certainly can contribute
endfire, but they actually remove broadside directivity in the process!
In an optimum size array the amplitude ratio from the primary quad has to be
4:1 or 5:1 or more to prevent some pretty significant pattern null area
deterioration when the additional 4 elements are added, because they
deteriorate broadside pattern multiplication faster than they contribute
endfire gain (at ~.187 spacing when the primary endfire cell has .25 wl
spacing).
If the array is made so small that there is little broadside contribution
from array width, then the addition of the four will improve things. There
isn't any broadside pattern to hurt. That isn't the same as a broad general
statement that using more of the elements allows the array to be made
smaller, unless we want to compromise pattern to have the same directivity.
I go through similar things with Yagi arrays. All of my Hygain 5 element
Yagis have been changed to four elements, and my KLM six elements have
become 5's. :) It isn't so much they work better, they just work different
in a way that is a better compromise for pattern, bandwidth, complexity, and
gain.
Everything is a compromise. If the target is maximum directivity and a clean
pattern (more like a flashlight), the array has to be large. It can never
be the same if small, or we all be running multi-element short boom antennas
in close-spaced stacks.
I do agree, however, if space is so limited the array can't use broadside
multiplication (which is the same as stacking gain in a Yagi array) then all
active elements with more elements is better.
73 Tom
----- Original Message -----
From: "Lee K7TJR" <k7tjr@msn.com>
To: "'Tom W8JI'" <w8ji@w8ji.com>; "'Bob Tabke'" <bob@onehorsecreek.com>;
<topband@contesting.com>
Sent: Monday, December 15, 2014 11:09 PM
Subject: Re: Topband: 8 circle: DXE vs Hi-Z
The primary difference between DXE and Hi-Z 8 circle arrays is the fact
that
Hi-Z uses ALL 8 verticals actively at the same time where DXE uses only 4
at
a direction.
Using all 8 verticals allows the use of a smaller diameter and a
performance edge on Directivity over the larger 4 active array.
All 8 element arrays do NOT work the same way.
Lee K7TJR
-----Original Message-----
From: Topband [mailto:topband-bounces@contesting.com] On Behalf Of Tom
W8JI
Sent: Monday, December 15, 2014 7:24 PM
To: Bob Tabke; topband@contesting.com
Subject: Re: Topband: 8 circle: DXE vs Hi-Z
- DXE wants a 320' diameter and Hi-Z wants 200' for optimum performance.
It's hard to tell what DXE performance is because it does not disclose
RDF, beam width or F/B. And neither vendor supplies EZNEC files so I
can see the effect of varying the layout. So I'm not sure how to
decide what array size is best for me. It would be wonderful if
someone has a model for these two systems.
Bob,
The ideal spacing of arrays like this is entirely dependent on the
frequency
range and goal you have for pattern or directivity. It is NOT dependent on
the design or manufacturer, there are no magical space saving tricks.
The circle diameter determines both endfire and broadside spacing, and
spacing determines the beamwidth. Something in the 330-350 foot range
across
the element pairs is near optimum for 160 directivity. You can use it down
to spacings where the element-to-element spacing is about 35-40 feet on
160,
but it might as well be a four element vertical or some other array at
that
spacing. You can narrow the 160 pattern by going larger than 350 feet,
but
the array can develop unwanted lobes. If element-to-element goes over 135
feet or so, you will start to have F/R issues.
This is the way every single eight element circle will work.
The primary difference between the DXE and Hi-Z is construction quality,
and
that determines cost. The DXE unit is a metal case that serves as a
groundplane for the connectors, and a much better PC layout. How much that
translates into better performance depends on how "pure" the rest of the
installation is. If the installation is sloppy or compromised, that will
set
the limit more than construction.
The DXE is a nearly direct copy of what I use here, which is a very clean
layout with PC mounted connectors and a super good groundplane between
connectors to prevent ground loops that affect performance and minimize
chances of lightning damage.
One reason I especially worry about connector grounding is my eight
verticals are spread in around a 350ft circle, and each has several 70 ft
buried radials. The large physical size of a system like this sets the
system up for large common mode currents in storms, it is actually a
yearly
event here to melt the shields off at least one cable with a nearby
lighting
hit (within a few thousand feet) because of ground loop currents, and yet
I
almost never have box troubles.
I use a 20ft vertical with a small loading coil and series load resistor
in
my elements, and a three wire hat. Mine is single band 160 (although I'm
very slowly working on a 16-element circle for 160-80). People who operate
here just love the 8 circle.
I can send you an EZNEC file that would roughly approximate the array.
73 Tom
_________________
Topband Reflector Archives - http://www.contesting.com/_topband
_________________
Topband Reflector Archives - http://www.contesting.com/_topband
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4794 / Virus Database: 4235/8744 - Release Date: 12/15/14
_________________
Topband Reflector Archives - http://www.contesting.com/_topband
|