[Antennaware] 4 Sq, verses 3 El endfire

Terry Conboy n6ry at arrl.net
Thu Sep 11 00:38:44 EDT 2008

You can use the same delay lines and phase angles with the 4 Square, 
but you have to make the length of the diagonal equal to the 80m 
spacing between elements 1 and 3 in the linear endfire array.  (So 
the spacing between adjacent elements is 1.414 * 40m.)  This gives 
-30 dB back lobes, but the azimuth pattern is quite different, with 
two fairly large low-angle sidelobes (-12 dB at +/- 114 degrees at 
0.53 and 1.06 MHz).  Even though this looks like a poorer pattern 
than the 3 element inline array, the 4 square has a narrower main 
lobe and an RDF that is 1 dB higher, because of deep nulls at +/- 78 
degrees that aren't present with the 3 element array.

The back lobes of both the 3 element and the 4 square grow with 
increasing frequency until they are about -22 dB at 2.65 MHz.  The 3 
element main lobe gets quite broad in azimuth (157 degrees) and is 
down about 0.5 dB on boresight.  With the 4 square, the main lobe is 
just 73 degrees wide, however those pesky sidelobes are only -5 
dB.  The RDF of the 4 square is about 2 dB higher at 2.65 MHz.

Life is all about compromise, eh?

73, Terry N6RY

At 04:39 AM 2008-09-10, Andrew Ikin wrote:
>Many thanks Chris and Ken for your responce to mail below:
>I trust that a brief description of the array will answer your questions.
>The proposed antennas will be receive only 1-2m active Vertical ( 
>FET/Bipolar buffer amp) hence there will be no element mutual 
>interaction. The centre element is run in anti-phase compared to the 
>outer elements, the amplitude is the sum of the outer elements also. 
>All 3 elements are combined using zero delay for Element 
>1;  0.73x40m (95ns) for the centre (2) Element and 0.73x80m (190ns) 
>for the Element 3.
>The delaylines provide near optimum phasing for reduced rear 
>vertical lobe from 0.53MHz to 2.65MHz. Going above 2.65MHz or 
>0.35Lambda spacing the Gain and F/B starts to reduce.
>The array was modelled at 0.53,1.06,1.59, 2.12 and 2.65MHz using the 
>degrees phasing calculated from the delaylines i.e. the correct 
>phasing tracks the frequency range.
>Eg. for 0.53MHz the phasing is 0, -162, -324 degrees and for 1.06MHz 
>the phasing 0, -143, -286 degrees. The delaylines were determined 
>from a similar model for a 3 Loop phased array.
>I have used a similar phasing scheme for the MW 2 Element Phased 
>loop array and a Phased Vertical array.
>"Hello Guys
>Having struggled to model an 3 element endfire Active Vertical array 
>using the same Phasing lines from 0.53MHz to 2.12MHz. I find that I 
>can achieve a very good elevation plot with very low rear lobes 
>(-30dB). Even the plot at 2.65MHz only shows 2 small (-22dB) rear 
>lobes!!!  Also this model is very tolerant of several dB amplitude 
>difference on the outer elements providing that the centre element 
>is the sum of the 2 outer ones.
>I have tried to do the same with the 4 Square, but there is always a 
>large rear lobe on the elevation plot 10-15dB higher than the 3 
>element endfire, no matter how I adjust the phase. Am I doing 
>something wrong or is there a fundamental problem with the 4 Square?!
>Both arrays have 40m Element spacing.
>Andrew G8LUG (not active)"

More information about the Antennaware mailing list