The Radiation Pattern of the Antennas


Joachim Köppen Strasbourg 2013


The antenna is perhaps the most important element in a receiving system, because its properties strongly influence the overall performance: The more strongly the antenna's sensitivity is restricted to a narrow beam towards the source or the other station, the better is its capability to receive only the desired signal and to reject any signals coming from other directions: hence the better is the signal-to-noise ratio.

The design of good antennas requires accurate measurement of the antenna pattern, to verify that the theoretical calculations do indeed yield the desired properties. Since the proximity of any material - such as conductors like metal structures or the iron in concrete buildings, or the absorptive character of stone - can strongly influence the antenna properties, measurements usually require to be done on antenna ranges where the device under test is placed as far away as possible from any such structures. However, the antenna in a real environment will perform differently from such an ideal situation.

This is an early attempt to measure the horizontal pattern of the 144 MHz antenna by observing the signal strength of an amateur radio FM repeater station in Switzerland:

In his Individual Project in 2013, Tingwei Guo measured the radiation patterns of the ground station's antennas and compared them to the simulations based on an accurate modeling of the geometry of the antennas. This is how he did it:

Simulation

First, all the dimensions of the antennas were measured with a ruler and band measure. This means all the lengths, diameters, and relative positions of all the elements were measured. Here is a schematic drawing of the horizontally polarized 432 MHz stack:

These data were entered in the software MMANA-GAL written by Macoto Mori, JE3HHT, a Japanese Radio Amateur and his friends. This program allowed to compute the theoretical patterns for the vertically polarized 144 MHz stack:
which can also be shown in 3-D version:
and the 432 MHz stack

Measurement

The next step was to measure the patterns:

Comparison with Simulation

Comparing the measured and simulated patterns of the horizontal plane showed that the primary lobes in both antennas were in excellent agreement:

However, there were significant differences in the side lobes, although there was a certain similarity, such as the positions of the nulls. In the 144 MHz pattern there was also a strong feature, as indicated by the red arrow, which had no symmetric counterpart on the other side. The 432 MHz pattern also exhibited an asymmetry ... This could be traced to reflections of the radio waves from the generator by a tall vertical lightning rod in about 3 to 4 meters distance from the Ground Station antenna!

The patterns in the vertical plane also showed excellent agreement for the main lobe, but strongly differing side lobes, both at 144 MHz:

at at 432 MHz:
As there are quite a few metallic structures, like the railing and the floor of the roof, as well as the platform on which the antenna is mounted, it is very likely that this environment affects the side lobes.

The characteristics of the antennas

Instead of specifying the detailed patterns, it is more convenient for operations to look at the properties of the main lobe:

HPBWHPBWFront/Back ratioFront/Back ratio
antennaH planeV planesimulatedmeasured
144 MHz H-pol.462220dB--
144 MHz V-pol.562221dB10dB
432 MHz H-pol.261417dB--
432 MHz V-pol.281415dB12dB

Comparison with Theory

For a first estimate, we may use the simple formula between the antenna gain and the HPBW valid for parabolic dish antennas:

HPBW = 180 / sqrt(10^(gain_dBi/10))
For the above example on 144 MHz, we know the gain of a single antenna by 12.2 dBi which gives 44, close enough to our measurement.

Why do we compare the horizontal HPBW with that of a single antenna? Because the two single 144 MHz antennas are stacked vertically: The resulting horizontal pattern is equal to the of the single antenna, while the vertical pattern is narrower (about one half: 22)!


| Top of the Page | Back to the MainPage | to my HomePage |


last update: Apr. 2013 J.Köppen