Broadside and End-Fire Arrays: How Phasing Shapes an Antenna Pattern

Put two identical antennas near each other, feed both from the same transmitter, and something interesting happens: the pattern changes. Depending on how far apart they are and how their currents are timed relative to each other, the pair can send its signal out to the sides, along the line between the antennas, or into one direction only. That is the principle behind every phased array that hams use, from a pair of vertical wires on 40 meters to the stack of dipoles on a broadcast station’s tower. This guide explains the two basic families, the broadside array and the end-fire array, shows how spacing and phase decide which one you get and what a real station can expect from them.

Two Antennas, Two Variables

When two antennas radiate together, their fields add in space. In some directions the waves arrive in step and reinforce each other; in others they arrive out of step and cancel. Two things decide where each case happens: the distance between the antennas, measured in wavelengths, and the phase difference between the currents that feed them. Everything else follows from those two numbers and from the pattern of a single element. The Wikipedia article on antenna arrays explains that small antennas around one wavelength in size, such as quarter-wave monopoles and half-wave dipoles, do not have much directivity, and that combining several of them is the way to create a directional high-gain antenna.

The same article gives the two definitions that matter here. Arrays in which the elements are fed in phase are broadside arrays, and the main lobe is emitted perpendicular to the plane of the elements. An end-fire array is a linear array in which the direction of radiation is along the line of the antennas, and the elements must be fed with a phase difference equal to the separation of adjacent antennas. The rest of this guide shows what those statements mean in practice.

The Broadside Array

In a broadside array the elements are fed in phase. Imagine two vertical wires side by side, one half wavelength apart, both fed with the same current at the same time. For a listener standing exactly halfway between them, on the line perpendicular to the line joining the wires, the two signals arrive together and add. For a listener standing along the line of the wires, the signals have traveled different distances, half a wavelength apart, and they cancel. The result is a figure-eight pattern with two lobes pointing to the sides and deep nulls along the line of the elements.

The gain depends on the spacing. For an ideal pair of point sources at half-wavelength spacing, the calculated directivity is about 3 dB more than for one source. At a quarter wavelength the gain falls to under 1 dB, because the pair does not have enough separation to narrow the beam much. Stacking more elements in the same line narrows the beam further, which is how a curtain array of many dipoles produces a sharp beam. The Wikipedia article on curtain arrays describes the large shortwave broadcast antennas that consist of many horizontal dipoles suspended in a vertical plane, in front of a reflecting screen of wires, which emit a beam at a shallow angle just above the horizon. Vertical stacks of Yagis, as in our article on stacked Yagis, use the same principle to gain a few decibels in the vertical plane.

The End-Fire Array

An end-fire array points its main lobe along the line joining the elements. To make the energy from both elements arrive together in the forward direction, the element at the rear must be fed earlier in time, which is the same as saying that the front element lags in phase by the amount that the wave needs to travel between them. For a spacing of a quarter wavelength, the phase difference is 90 degrees. In the opposite direction the signals then cancel, because the extra path and the phase difference add up to half a cycle. The pattern is a heart shape, the cardioid, with a single main lobe forward and a deep null to the rear.

For an ideal pair, the directivity of a quarter-wave end-fire pair is also about 3 dB over a single element, but it comes with a very high front-to-back ratio in theory. That combination, a modest gain and a strong null behind, is the reason why end-fire pairs are used on the low bands to reject noise and interference from a certain direction, and why directional receiving antennas use them. The Yagi is a parasitic end-fire array: according to the Wikipedia article, it has a single driven element with several parasitic directors in front and a reflector behind, and the phase of the currents in the parasitic elements is determined by their lengths and spacings rather than by a feed line.

Spacing and Phase at a Glance

The gain figures in the table were calculated for two ideal point sources. Real antennas add their own patterns: two vertical monopoles, for example, have an omnidirectional pattern in the horizontal plane, so the array pattern is the pattern that you see in the horizontal plane, while the vertical pattern is that of the single vertical. The mutual coupling between nearby elements also changes the feed point impedances and the current distribution, which is why a real design needs measurement and adjustment, not just a calculation.

Creating the Phase Difference

The phase difference is produced by the feed system. The simplest method uses a length of coaxial cable as a delay. The phase shift of a cable is proportional to its electrical length, and its electrical length is its physical length divided by the velocity factor, expressed in wavelengths. A quarter-wave section of cable gives 90 degrees. Because coaxial cable has a velocity factor of typically about 0.66 or 0.8 depending on the dielectric, the physical length is shorter than a free-space quarter wavelength, a point we explained in our guide to velocity factor. Always use the real velocity factor of your cable, or, better, measure it.

The other methods use lumped components or hybrid couplers. A hybrid coupler divides the power equally between two outputs with a fixed 90 degree phase difference, which is exactly what a quarter-wave end-fire pair needs. Switching the connections of the coupler reverses the direction of the lobe, so a pair of verticals can be switched electrically to point either way without moving anything. Larger systems use a controller to apply several phases and powers to four or more elements, as our article on the four-square vertical array explains. These systems can choose between four directions with the push of a button.

A Worked Example for 40 Meters

To see how the numbers work, take two vertical wires for 40 meters at 7.1 MHz. The free-space wavelength is about 42 meters, so a half wavelength is about 21 meters and a quarter wavelength about 10.5 meters. For a broadside pair you place the verticals 21 meters apart and feed them in phase with equal lengths of cable from a common junction; the pattern is a figure-eight at right angles to the line joining them, so if you align the line east to west, the lobes point north and south. For an end-fire pair you place them 10.5 meters apart and delay the feed of the front element by 90 degrees; the single lobe points along the line toward the delayed element, and the null points the other way.

Notice the trade-off. The broadside pair needs twice the space and gives two lobes, so it covers two opposite directions but it cannot suppress an unwanted station in one of them. The end-fire pair is more compact and gives a single direction, but its gain is only about the same, and the null is sharp only if the phasing is accurate. For a small garden, the end-fire pair is often the better choice, and for a large field with a need to cover two directions, the broadside pair is simpler to build because it needs no delay line. Whichever you choose, measure the SWR and the pattern in use, and compare reports from stations in the different directions.

What a Real Array Needs

• Identical elements. The elements should be the same, with the same height, radials and surroundings. An unequal pair produces an unequal pattern.
• Accurate spacing. The spacing of a quarter or half wavelength is calculated for the band, and the real distance in a garden has to be measured on the ground. An error of ten percent changes the nulls noticeably.
• Equal currents. A feed system that divides the power equally is needed, and the currents should be checked with a current probe or by observing the pattern on the air.
• A good ground system. For verticals, each element needs its own radial system, as our guide to vertical antenna radials shows. The radial fields may overlap but must not be shared in a way that upsets the currents.
• Patience in tuning. Tune each element alone, then connect the phasing system, then adjust. Expect a few rounds of measurement and change.

Where Each Type Is Used

A broadside pair is attractive when you want a bidirectional pattern with some gain and you have the space. On 40 meters, two verticals half a wavelength apart, which is about 20 meters, give a figure-eight that can cover two directions, for example Europe and North America from a location in between. An end-fire pair of verticals suits the low bands, where you want a unidirectional pattern and a null toward a noise source or a strong local station. A pair of horizontal wires, a stack of dipoles or a collinear array on the VHF bands makes use of the same principles at a smaller size.

When you read the specification of a commercial antenna or the description of a station in a contest report, the words broadside and end-fire tell you how the antenna is built. A stack of two Yagis one above the other is a broadside array of directional elements. The Yagi itself is an end-fire array. A rotary beam and a phased vertical pair are therefore not rivals but members of the same family, and our guide to antenna radiation patterns helps to interpret the plots that come with them.

The Bottom Line

A broadside array has its elements fed in phase and radiates perpendicular to the line of the elements, while an end-fire array has a phase difference that matches the spacing and radiates along the line. With two ideal elements, half-wavelength spacing in phase gives a figure-eight with about 3 dB of gain, and quarter-wavelength spacing with 90 degrees of phase gives a cardioid with about the same gain and a deep null behind. Real arrays need identical elements, accurate spacing, equal currents and good grounds, and the phase is made with delay lines or hybrid couplers. Understanding the two families makes any antenna description easier to read, whether it is a pair of verticals, a stack of Yagis or a curtain array.

Frequently Asked Questions

What is a broadside array?

An array whose elements are fed in phase, so that the main lobe is perpendicular to the plane of the elements.

What is an end-fire array?

A linear array whose radiation goes along the line of the elements. The feed phase difference equals the separation between adjacent elements, in electrical degrees.

Is a Yagi a broadside or an end-fire array?

An end-fire array: it has a driven element, a reflector behind it and directors in front, all along one line.

How much gain does a two-element array give?

For ideal point sources, about 3 dB at the classic combinations of half-wave spacing in phase or quarter-wave spacing with 90 degrees. Real antennas deviate because of coupling and ground.

How do I make the 90 degree phase shift?

With a quarter-wave section of coaxial cable, shortened by its velocity factor, or with a hybrid coupler that gives a fixed 90 degree difference.

Why would I choose an end-fire pair for the low bands?

Because it gives a single forward lobe and a deep null to the rear, which helps against noise and interference from the opposite direction.