Receiving Four Square Antenna Array 

Four small elements in a square layout, between 1/8- and 1/4-wl side length, can be used to form a four- direction array. Performance will be approximately equal to a 1 wl Beverage. The elements can be similar to those described in small vertical arrays.  

As mentioned earlier, conventional transmitting phasing systems might be easy to obtain, but they are NOT correct when used in receiving arrays with short elements. Since each element has the same impedance in a lossy receiving array, the phasing unit "sees" 75-ohms from each element. When the antenna presents equal impedances at each port, the phasing system should source equal power at each port (which is also equal voltage and equal current)!

Use of a single 180-degree phase-inversion transformer for the center element along with delay lines approximately equal to effective element spacing create a very wide bandwidth phasing system. The phasing system is usable over at least one octave with appropriate element spacing and design! 

Upper and Lower Frequency Limits

Assuming you have broadband active elements of good design, the array will be useful when array side-length ranges from ~1/16wl spacing to 1/3 wl. The phasing system and delay lines are usable at any frequency, but the array spatial delay will provide useful end-fire patterns where array side-length is approximately 1/3-wl or longer. The appearance of grating lobes and dimpling of the front lobe limits upper frequency usefulness.

As frequency is decreased, array sensitivity eventually drops to unusable levels. The effect comes from two effects:

	Elements become electrically shorter, reducing element and array sensitivity.
	Element spacing becomes smaller in electrical degrees, reducing array sensitivity.

Shorter elements, in electrical degrees, mean less sensitivity. Sensitivity obviously must drop with frequency, for a fixed element height. 

Element nulling always fully subtracts with reduced spacing. With any fixed spacing distance, a lower frequency array sensitivity limit appears when "forward" voltage phase vectors become too close to -180 degrees.   

Element and array phasing-sensitivity eventually push the antenna into the receiver system's noise-floor, and this becomes the lower frequency cutoff. Pattern is maintained down to that cut-off, assuming the elements remain matched to the feedlines. 

Even without broadband active elements, multiband or broadband arrays are practical with ONE fixed element group and a single unchanged phasing system. Element resonance and resistance would be the only parameter that would require switching.

Pattern with 70- foot per side spacing, on 1.8 MHz is: 

There is a slight distribution problem with smaller size arrays because of mutual coupling. This can be compensated if 1/4-wl feedlines are used on the lowest band with the phasing coupler shown. RDF (receiving directivity factor) is around 10dB. This is about 3dB higher than the best single Pennant antennas, and on par with a very long Beverage.

If the array is 140 foot per side, the pattern is:

Doubling array size only provides about 1dB more directivity (note that HPBW only changes a few degrees), although signal level increases about 4dB! This array would make an ideal portable antenna for limited-space two-band operation, a SPDT relay at each antenna could be used to switch loading networks for each band.

The phasing system should look like this:

 

 

 

T1 is a 4:1 impedance ratio (2:1 turns ratio) auto transformer.  A suitable 100 kHz to 20 MHz transformer would use 5 to 10  turns of #22 wire (wire size is not critical) in a twisted pair, connected with opposite ends in series. A  Fair Rite Products 2873000202 core is recommended, and available from me if you can not find one. (A turn is one complete pass of the twisted wires through both holes in the binocular core). This transformer steps the 18.75-ohm impedance, produced by paralleling four 75-ohm lines, up to the feedline output impedance of 75-ohms.

T2 is a 1:1 ratio phase inverting transformer. Wire size is not critical as long as the winding fits the core. A good choice is a twisted pair of #22 wires, making five to ten passes through a Fair Rite Products 2873000202 core. This is the proper core!  This is the same transformer used in T1, with a different circuit connection. Again, I will help you with complete tested transformers if you can not locate materials.  

A low attenuation value ~ 75-ohm attenuator is inserted in the line to the front antenna and rear antennas. This attenuator compensates for slightly different losses in transformers and delay lines. Although it absolutely isn't a requirement when using wider spacings (1/4 wl) with proper transformer and element construction and materials, they can be used to "tweak" the array to extremely high F/B ratios. The attenuators become more and more important as the array becomes smaller or less perfectly constructed and tuned. 

DL1 should be slightly shorter electrically than the (around .9 times the) corner-to-corner distance across the array (diagonal distance). This line MUST be the same impedance as the element feedlines.

DL2 and DL3 are exactly half the length of DL1. These lines must be the same impedance as the array's element feedlines. 

NOTE: The delay lines are slightly shorter than element spacing. This elevates the null a small amount above the horizon directly off the rear, forming a cone-shaped null reaching the ground an equal number of degrees either side of the rear. I generally use about 10-20 degrees less than the element physical separation distance of a line running between diagonal corners of the array for delay lines on the highest band.

 If you model this antenna, remember phase is inverted. Effective phase shift from the delay line actually becomes a phase lead that is 180- degrees around a circle from the delay you might expect. For example, a 90- degree delay line actually provides 90- degrees phase advance (or -270 delay) because of the inversion, while a 60- degree delay line provides 120- degrees advance (which is the same as 240-degree delay) .

This "trick" causes phase-shift to track changes in element spacing with frequency, allowing the same delay line to work on multiple bands. The delay always is the proper length, which is slightly shorter than element spacing. It also causes the array to fire in the direction of the common-point element, since the feedline attachment point is the direction of lagging phase.

See the Crossfire phasing article.

Why Are the Lines nearly equal to the spacing for the elements they serve?

For an explanation, see the Cross-fire phasing page on this site! 

The following system is useful for electrically rotating the array:

The antenna ports on the left from top to bottom start at the front of the antenna and from a top view go clockwise around the antenna. Zero voltage is position one at zero degrees, + is 90 degrees CW, - is 180 degrees, and AC is 270 degrees.  

RL2 can be two separate DPDT relays with coils in parallel. I use 12V DIP relays with 220 uF 25v capacitors across relay coils. Any common silicon power rectifier will work for the diodes. If you have high RF levels near the control line, bypass it with a .1uF disc 50v capacitor. I route control signals through the feedline, but you must not have water in the feedline or poor shield connections to do that!

If you can not find parts, I have parts. I'm in the process of ordering commercial PC boards right now, to supply turn-key components kits or systems. I'd be happy to help with anything you need, as these projects move along.