Re: Topband: Modeling W8JI and Hi-Z 8 Circles

[Dan Maguire via Topband]

Based on some user feedback, the 8-circle models mentioned earlier have been 
modified with an additional variable to control the azimuth direction of the 
main lobe.  That lets you change the direction just by changing (or doing a 
sweep on) a single variable.  So you can easily:  1) see how the azimuth 
patterns overlap as the array is steered in 45° steps, and 2) compare patterns 
against other models which may be fixed in a particular direction.

For case 1, here's how the Hi-Z type (8 elements in use) 8-circle array 
patterns would overlap as you turn the knob on the control box.  This is for a 
200 ft diameter array with ±106° phasing, 20° TOA for the patterns.  Compass 
rose angles are shown in parenthesis.

http://ac6la.com/adhoc/8circle7.png

For case 2, this is how a W8JI type (aka BroadSide/End-Fire or BSEF, 4 
elements in use) 8-circle array with 0.604 wl broadside spacing and 125° 
phasing compares against a 4-square receiving array with 0.125 wl element 
spacing and Crossfire feeding.  Both arrays use the same basic element, the top 
hat model rxvrhat.ez from Tom's page.  At a 20° TOA the 4-square has a gain of 
-22.43 dBi compared to -10.95 dBi for the 8-circle.  (But the RDF is only about 
1 dB lower for a diameter of 94 ft compared to 348 ft for the 8-circle.)  In 
order to get an accurate comparison of the pattern *shapes* on a common polar 
chart, the *gain* of the 4-square has been normalized to that of the 8-circle 
by reducing the value of the swamping resistors.  Original gain below on left, 
normalized gain below on right. 

http://ac6la.com/adhoc/8circle8.png

The 4-square can be fed either as Crossfire or BroadSide/End-Fire using exactly 
the same phasing lines, a neat idea picked up from IV3PRK.  In the patterns 
below, the swamping resistors are back at the normal values (to give 75+j0 at 
the feedpoint for a standalone element).  With Crossfire feeding the gain at 
20° TOA is down 7.45 dB compared to BSEF feeding but the RDF is 2.1 dB better, 
below left.  Note that a design criteria for IV3PRK was good rejection at 
backside 45°, below right.

http://ac6la.com/adhoc/8circle9.png

For more details on this dual-feed 4-square see the Receiving Antennas 
portion of the IV3PRK site and follow the links on the left side of the page.

http://www.iv3prk.it/rx-antennas.htm 

Finally, thank you Jim Brown for mentioning the 2014 Contest University W3LPL 
Receiving Antennas presentation.  Lots of good info there.  Here's the direct 
link.

http://www.contestuniversity.com/attachments/W3LPL_Receiving_Antennas_2014.pptx

These are the revised 8-square models with the additional variable to change 
the pattern direction.

http://ac6la.com/adhoc/W8JI_8_Circle__TopHat.weq
http://ac6la.com/adhoc/W8JI_8_Circle__AL-24.weq
http://ac6la.com/adhoc/Hi-Z_8_Circle__TopHat.weq
http://ac6la.com/adhoc/Hi-Z_8_Circle__AL-24.weq

And here's the 4-squareRX model with a variable that switches between Crossfire 
feed and BSEF feed.

http://ac6la.com/adhoc/4SquareRX__TopHat.weq

Dan, AC6LA
_
Topband Reflector Archives - http://www.contesting.com/_topband

Topband: Modeling W8JI and Hi-Z 8 Circles

[Dan Maguire via Topband]

The recent thread on 8 circle arrays prompted me to create a few different 
AutoEZ models, mostly because I was curious about the relationship between 
array size, element phasing, and number of active elements (4 with W8JI, 8 with 
Hi-Z).

First thing I did was educate myself.  For W8JI type arrays:

http://w8ji.com/small_vertical_arrays.htm
http://n3ujj.com/manuals/8%20Circle%20Vertical%20Array%20for%20Low%20Band%20Receiving.pdf
http://www.dxengineering.com/parts/dxe-rca8b-sys-3p
http://static.dxengineering.com/global/images/instructions/dxe-rca8b-sys-3p-rev3.pdf
Low-Band DXing, 4th ed, ON4UN, Chapter 7, Sections 1-21 and 1-30

And for Hi-Z type arrays:

http://www.k7tjr.com/need.htm
http://www.kkn.net/dayton2014/HiZ_DAYTON_2014_7n2.pdf
http://www.hizantennas.com/HiZ8-16080_users_guide.pdf
http://www.hizantennas.com/Hiz_8_16080_manual.pdf
http://www.dxengineering.com/parts/hiz-8a-lv2-160-2

Then I created models where the array size (diameter) can be controlled via a 
single variable, along with another variable to control the phasing.  For 
example, here's the AutoEZ Variables sheet tab for a W8JI type array.  The Hi-Z 
is similar except that array size is specified in feet (or meters) rather than 
wavelengths.

http://ac6la.com/adhoc/8circle1.png

Since everything is controlled by variables you can run variable sweeps 
changing one or more parameters.  Here's the W8JI array with the spacing (B) 
held constant while the phase delay (P) is swept from 80 to 140 degrees.  For 
each test case AutoEZ will automatically calculate the RDF (last column).

http://ac6la.com/adhoc/8circle2.png

When the calculations finish you can step through the 2D patterns.  Here are 
the elevation and azimuth (at 20° TOA) patterns for 0.604 wl broadside spacing 
and 125 degree phasing, as shown by B and P in the lower right corner.

http://ac6la.com/adhoc/8circle3.png

You can run similar sweeps changing the array size while holding the phase 
delay constant, or hold both size and phase constant and do a frequency sweep, 
or use the Generate Test Cases button to create any combination.  For 
example, the setup below would vary broadside spacing B from 0.50 to 0.70 
wavelengths; for each B the phase delay P would be varied from 115 to 135 
degrees; all at a constant frequency of 1.85 MHz.  That would be 25 test cases. 
 You can run thousands if you like.

http://ac6la.com/adhoc/8circle4.png

You can also show 3D patterns.  Here's an example of a Hi-Z array, diameter 200 
ft with ±106 degree phasing, along with the 2D elevation pattern.

http://ac6la.com/adhoc/8circle5.png

And here's how the RDF for a 200 ft Hi-Z array varies as the phase is swept 
from ±100 to ±112 degrees.

http://ac6la.com/adhoc/8circle6.png  

For both of the array types, I created one model using W8JI-style top hat 
loaded vertical elements (per the sample model on Tom's site) and a second 
model using simple aluminum tube elements (per the four-section, 23.25 ft, Hi-Z 
AL-24).  Here are the models.  Save to your computer then use the AutoEZ Open 
Model File button.

http://ac6la.com/adhoc/W8JI_8_Circle__TopHat.weq
http://ac6la.com/adhoc/W8JI_8_Circle__AL-24.weq
http://ac6la.com/adhoc/Hi-Z_8_Circle__TopHat.weq
http://ac6la.com/adhoc/Hi-Z_8_Circle__AL-24.weq

Please note that these models have not been reviewed or approved by the authors 
of the references cited above.  Any mistakes or misinterpretations are strictly 
mine.  And the models may or may not be an accurate representation of any given 
commercial package.

In all the models, a single variable (X) controls the segmentation.  You can 
reduce that to speed up the calculations.  You can also run a sweep on X to 
do a convergence test for model accuracy.

For comparison with the 8 circle arrays, here's the W8WWV Benchmark Beverage 
model.  With this one you can sweep the length and/or other parameters.

http://ac6la.com/adhoc/W8WWV_Beverage.weq

Greg's Beverage page for reference.

http://www.seed-solutions.com/gregordy/Amateur%20Radio/Experimentation/Beverage.htm

Calculated results for this Beverage are shown in section Calculate 3D Data 
here.

http://ac6la.com/aeuse3d.html

===

AutoEZ is an antenna modeling program which uses Microsoft Excel in combination 
with EZNEC.

http://ac6la.com/autoez.html

Current AutoEZ users, see here for recent maintenance updates and instructions 
on how to get the latest release.

http://ac6la.com/aechanges.html

Dan, AC6LA
_
Topband Reflector Archives - http://www.contesting.com/_topband