3 Feed pattern measurements

The antenna test range of the Integration and Tests Laboratory (a satellite dedicated facility) at the National Institute for Space Research - INPE - was used over a period of 3 weeks in order to obtain full beam patterns of the GEM backfire helical feed antennas at 408MHz and 1465MHz. For the measurements, a vertically polarized transmitter was located on a tower 25 m above the ground and 80 m in front of an anecoic chamber. The antenna under test sat on a plate attached to the head of the fiber glass support arm of a platform with 3 degrees of freedom (slide: horizontal motion along the axis between transmitting and receiving antennas; roll: rotation of the head support plate about the slide axis; azimuth: horizontal scanning motion).

During the measurements, the upper section of the support arm was surrounded with Eccosorb in order to avoid undesired strayed signal from the obstruction behind the head support plate. Furthermore, since a backfire helix radiates in the direction of its ground plane, PVC extensions were customized to position the helix upside down on the head plate and to direct the feed cable toward its connector at the ground plane. Preliminary tests were conducted at different positions along the slide axis to match the phase center of the feed antenna with the rotation axis of the support arm. The backlobe structures of the feeds were also obtained by adjusting their ground planes onto the PVC extensions attached to the head plate.

The beam patterns were obtained by measuring the power response of the antennas with polar angle $\theta$ while the platform rotated through $360\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ in azimuth. The measurements were taken at $1.6\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ intervals at 408MHz and every $0.2\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ at 1465MHz. The full spatial response was generated by repeating the azimuth scans for a sequence of equally-spaced roll angles. Although a $180\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ range in roll angle would have sufficed to cover all space directions, the helical antenna is capable of shifting the phase of the received signal as it turns around its main beam axis (Kraus 1988). Roll angle test measurements with the 408MHz helix were consistent with this prediction and, in this case, the entire $360\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ range in roll angle was covered at $4.8\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ steps. No significant phase shifting was noticed with the 1465MHz helix, for which $10\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ roll angle steps were used.

As required by a polar angle resolution of $1\ifmmode^\circ\else\hbox{$^\circ$ }\fi$ in the diffraction model we apply in the next section, the measured responses were regridded and interpolated to accomodate a $1^\circ\times 1^\circ$ spatial resolution. Diagrams of the resulting power patterns $P_{\rm n}(\theta ,\phi )$ down to the 20-dB level are displayed in Figs. 7a,b. Their mean response averaged over $\phi$ produces the pattern profiles $P_{\rm n}(\theta)$ shown in Fig. 8. The radiometric characterization of these backfire helices is further illustrated in Fig. 9, showing the antenna solid angle as a function of the polar angle $\theta$, and Table 1 gives the directivity, D, main beam efficiency, $\epsilon_{\rm M}$, and the beam solid angle fraction, $\epsilon_{\rm h}$, intercepted by the co-rotating ground shield (halo) of the GEM parabolic reflector. The 10-dB points attain 93.8% and 62.5% of the total dish illumination at 408MHz and 1465MHz, respectively.

Figure: Polar diagrams of the 408MHz and 1465MHz feed patterns using the mean backfire response in the $\phi$-plane. Vertical reference lines delimit the sidelobe structure within the field of view of the halo and the width of the assumed main beams

Figure 9: Antenna solid angle growth with polar angle for the beam patterns in Fig. 8 and for the full 3-d measured response of the backfire feeds. The vertical lines are as in Fig. 8

   

Table 1: Measured radiometric properties of backfire helical feeds

	$P_{\rm n}(\theta)$			$P_{\rm n}(\theta ,\phi )$
	408MHz	1465MHz		408MHz	1465MHz
D	5.32	8.19		6.92	13.56
$\epsilon_{\rm M}$	0.87	0.71		0.87	0.71
$\epsilon_{\rm h}\times 10^2$	5.90	9.36		5.89	9.32

Experimental reports on monofilar axial-mode helical antennas have seldom focused the backfire type. End-fire helices of equivalent design characteristics, for example, do not depend critically on frequency over the range studied here (see Paper I); whereas Table 1 clearly favours a frequency dependence for the backfire mode. A few authors have also attempted to describe the radiometric properties of the backfire helix from analytical, numerical and experimental points of view (Sexson 1965; Johnson & Cotton 1984; Nakano et al. 1988). No definite consensus has yet emerged from these studies, since the mechanical design of the helices under investigation was substantially different for each author. Our backfire feeds, which follow the design considerations of typical Kraus coils (Kraus 1988), show a substantial narrowing of the beamwidth with increasing frequency which disagrees with the predictions of earlier studies (Sexson 1965; Nakano et al. 1988).