Page 78 - Antennas for Base Stations in Wireless Communications

Base Station Antennas for Mobile Radio Systems 51

Antenna constructions employing coaxial cables for the feed network
generally use cables with solder-flooded copper wire braid. Cables of this
type are easy to bend; they stay in place once formed; and they have
much lower PIM than cables with conventional wire braid. Some cable-
fed array designs make use of printed-circuit power dividers; others
are all-cable designs, using cables of different characteristic impedance,
sometimes connected in parallel, to form impedance transforming sec-
tions. The electrical design of feed networks is discussed below.
In general, the construction methods for the radiating elements and
the feed system should be chosen together; particular care must be
taken with the interface between them in terms of mechanical integrity,
impedance matching, the generation of PIMs, and cost. This interface is
often between copper or copper alloys (in the feed network) and alumi-
num alloy (for the radiating elements). A direct joint between these has
a high galvanic contact potential and will rapidly corrode in a warm,
humid atmosphere. The availability of reliable low-cost tin plating on
aluminum has provided one popular solution, allowing coaxial cables to
be soft-soldered directly to aluminum alloy dipoles or patches.
For azimuth beamwidths narrower than about 55°, a tier is often
constructed using two radiators. As their lateral spacing is increased to
reduce the beamwidth, the level of azimuth sidelobes rises. The achieve-
ment of low sidelobes with a 30° beamwidth is difficult and solutions
exist in which three elements are used, with their lateral spacing and
relative currents chosen to provide a clean azimuth beam.

2.3.2 Array Design
Having chosen the type of radiating elements and feedlines to be used,
the next task is the design of the array.

2.3.3 Dimensioning the Array
The gain of a BS antenna is typically many times larger than can be
provided by a single radiating element, and it is necessary to array a
number of elements to achieve the required gain and pattern charac-
teristics. If the array were uniform and lossless, its directivity would
depend only on the azimuth pattern of the elements, the electrical
length of the array, and the number of elements used to fill it. In order
to minimize cost, we wish to use the minimum necessary number of
elements, so we must understand how the vertical spacing between the
elements affects array performance. Note that, at this stage, we refer
to directivity because we have not yet investigated the inevitable losses
within the array that will reduce its gain.
A typical relationship between array directivity and element spacing
is shown in Figure 2.9, which relates to a uniform broadside array of