Reflect Array 2

Antennas and Microwaves

Entry Page

Contents

Page 2 of 3

Reflect Array

Design and Implementation

Reflect arrays, like Yagi antennas, come into the class of parasitically coupled arrays. In these, energy is coupled radiatively between elements, rather than being distributed in a guided mode though a feed network.

However, while Yagi type antennas couple energy from one element to another in a ‘daisy chain’ fashion, reflect arrays are as their name suggests, are much closer to reflector antennas in operation. In essence, a spherical wave front emanating from a small feed antenna is converted to a plane wave front, by an array of resonant elements instead of a shaped conducting surface. Elements longer than optimum for resonance tend to retard the phase of reflected signals, those shorter than optimum will advance the phase.

Just thinking about the geometry of a feed antenna illuminating a flat sheet reflector, we can visualise signals taking longer to travel to the edges of the reflector than to a point immediately below. As a result, signals at the edge are phase delayed with respect to those at the centre. By carefully choosing the size and spacing of the elements (generally longer and wider in the centre of the sheet, and shorter and closer at the edge) the phase error can be corrected.

In practice the exact sizes and spacing were determined by modelling the antenna in 4Nec2 and using the standard ‘hill climber’ optimizer to maximise forward gain. To reduce the number of independent variables the resonators were grouped in rings e.g.

circumference of elements in ring 1, radius of ring 1, ...etc

My first practical implementation of the antenna was really just a confidence check in the modelling approach. The choice of 2366Mhz for my prototype was governed mainly by the dimensions of raw materials I had around at the time, rather than any burning desire for an antenna at that frequency.

Helix feed point detail

Tuning the helix feed is relatively straightforward. As a rule of thumb, add one turn to the helix compared to the uniformly pitched NEC model, produced by the GH card. Keep this 1st turn close to the ground plane, gradually increasing the pitch to the design value (11.5 Deg). Next, add a lumped capacitance a short distance from the feed-through, here a small piece of brass sheet was used.

Helices are naturally wide band, an octave is possible. This VSWR response was achieved after only a few minutes of ‘tweaking’.