ADMINISTRACIÓN GUBERNAMENTAL DE INGRESOS PÚBLICOS

For practicality, and for easier handling of the antenna, we limit the footprint of the an- tenna to 1m2. There is likely to be a size limitation for the antennas in the SKA low frequency array but we are not yet 100% sure about this. One could argue that at the site there is unlimited space available for large antennas, but large antennas could provide problems in terms of increasing costs, manufacturing and maintenance. The lower fre- quency (long) dipoles are bent in order to fit into this constraint.

lating the two cases. We will simulate the folded and unfolded antennas and discuss the related results.

We start the folding process by considering one zig-zag arm as shown in Figure 4.7 (Step

Figure 4.7: Step by step miniaturisation process. Step 1: original height H. Step 2: folded height

H/2. Step 3: twice-folded height H/4. Step 4: thrice-folded and final height of H/8. Redrawn from Sharma [5].

1). This structure’s height in the y-axis can be defined by H. We first fold the part that is underneath the x-axis by 90◦to lie in the xz-plane as shown in (Step 2). Now the height of the structure is H/2 and thus half of its original height. Next we fold the longer elements in on themselves as shown in (Step 3). The structure is now H/4 high. We can repeat this one last time in order to get to (Step 4) with a structure height of H/8. For our structure we omit (Step 1) to accommodate 4 elements.

One would expect the characteristics to change significantly as folding the elements re- duces the effective electrical length of each element. By folding only the longest dipoles, one expects the high frequency part of the antennas characteristics to stay the same, but that the low frequency part would be affected.

We now compare two of the same antennas with each other. Both are two arm zig-zag antennas with the same design parameters. The one is unfolded (straight) while the other is folded twice, up to (Step 3) in Figure 4.7. The unfolded structure is 2.3 m wide, while the folded structure is 1 m wide. Figure 4.8 shows the two structures.

Here we are merely focussing on the difference as a result on the folding of the arms. In Figure 4.8 the bottom parts of the two zig-zag antennas are 1 m apart. The width of the unfolded structure is 2.3 m while the width of the folded structure is 1 m. At a half-wavelength of 2.3 m, the resonant frequency is about 65 MHz and at 1 m it is about

150 MHz. Thus by shortening the electrical length through bending the elements one ex- pects the lower cut-off frequency to move upwards from 65 MHz to somewhere between 65 and 150 MHz. One determines this point through simulation and that is exactly what will be done next.

Figure 4.8:Two zig-zag antennas to be compared with each other.

The two antennas of Figure 4.8 are simulated in FEKO and the results are shown in the following figures.

Figure 4.9 shows the directivity comparison between the two configurations. Here we clearly see that the directivity only changes a little (< 1 dB) over the entire band width. Thus one can assume that the directivity remains unaltered.

Figure 4.10 shows the reflection coefficient for the two configurations. Here the reflec- tions are exactly the same from about 200 MHz and upward. This was expected to re- main the same as the longer elements do not contribute at the higher frequencies. Below this point both configurations still radiate. The folded configuration crosses the -10 dB point at 102 MHz while the unfolded configuration crosses it at 76 MHz. This means that we have lost some bandwidth unfortunately but this was expected. This shows that by reducing the structures width by more than halving it only reduces the structures band- width at the low frequencies by 26 MHz.

This also suggests that the following characteristics, the polarisation purity and the E- and H-plane beamwidths, we will look at next should follow a similar trend as the re- flection coefficient does. One expects the high and mid frequencies to remain unaltered while the lower frequencies are expected to be different.

Figure 4.9:Comparison of the directivity between the folded and unfolded antenna.

Here one sees the co-polarisation is very similar to the directivity and the cross-polari- sation is much less. The cross-polarisation of the two configurations is very similar over the frequency band except at the lower part. Here the folded cross-polarisation worsens quickly at about 100 MHz and downward.

Next we take a look at the E- and H-plane beamwidth. Figure 4.12 shows the beamwidth at the low frequencies. The unfolded E-plane beamwidth is 73◦ while folding it changes it to 81◦. The unfolded H-plane is 136◦ while folding it changes it to 182◦.

Figure 4.13 shows the beamwidth at the centre frequencies. Here one sees the E- and H- plane beamwidths are similar for the folded and unfolded cases. With the E-plane being 80◦ and 66◦ for the unfolded and folded cases respectively and the H-plane being 109◦ and 98◦for the unfolded and folded cases respectively.

Figure 4.14 shows the beamwidth at the high frequencies. Here one sees the E- and H-plane beamwidths are identical for the folded and unfolded cases. With the E-plane being 72◦ for both unfolded and folded cases and the H-plane being 104◦ for both the unfolded and folded cases.

Figure 4.10: Comparison of the reflection coefficient (S11) between the folded and unfolded an-

tenna with Z0=225Ω.

This shows that by folding the longer elements the antenna’s performance remains most- ly unchanged. Only some bandwidth at the lower frequencies are lost. This is the price paid for in terms of performance for reducing the width of the antenna.