A New Sum-Channel Radiating Element for a Patch-Monopole Monopulse Feed

Abstract

This article describes a new design for the sum-channel, circularly-polarized, radiating patch element of a multimode patch-monopole monopulse feed. Such feeds are suitable for the prime-focus illumination of a deep symmetric reflector. A relatively simple and compact feed design allows for a compact single-mirror antenna with monopulse tracking. The feed is about a wavelength in diameter, also making it suitable for illuminating smaller antennas, for example, in LEO-satellite ground (client) stations. The design applies a slotted circular patch as a circularly-polarized sum-channel element. The design variables are optimized mainly for the sum-channel aperture illumination efficiency in an S-band satellite ground station. After a few cut-and-try iterations, the final feed prototype was produced and measured. In addition to an additional degree of freedom in the geometry and potentially easier fabrication, the new sum-channel radiating-element shape allows for a slightly better monopulse channel isolation, and a higher radiation-pattern symmetry compared to the previous design.

1. Introduction

The patch-monopole monopulse feed [1] is a relatively simple and compact multimode feed design with a circular cavity that is approximately a wavelength in diameter. This type of monopulse feed is suitable as a prime-focus illumination of a rotationally-symmetric (deep) reflector, a single-reflector paraboloidal (parabolic) antenna [2,3]. It enables inexpensive monopulse antenna designs for stations that benefit from rapid and very accurate antenna pointing, such as LEO (low Earth orbit) satellite ground stations [4,5,6], as illustrated in Figure 1, ground stations of moving aerial or high-altitude platforms, long-distance links (for example, Lunar communications [7]), and similar.

The compactness of the design results in minimal antenna-aperture shading and makes the feed suitable also for small(est) prime-focus paraboloidal antennas, only a few wavelengths in size. The simplicity of the design results in a very manageable fabrication complexity and low production cost. The compact design makes the feed also suitable for mobile radars, aerial–platform communications, and similar applications where compact monopulse antennas are sought.

The main radiating element is a circularly-polarized patch exciting the TE₁₁ mode (circular-waveguide equivalent) as the sum channel. The second radiating element is a linearly-polarized monopole exciting the TM₀₁ mode (circular-waveguide equivalent) for the difference channel. The patch element in a relatively shallow cavity radiates like an open-ended (short) circular waveguide [8,9,10]. Two-dimensional monopulse operation requires a circularly-polarized incident wave, in contrast to the conventional monopulse design with an antenna array and a comparator network [11,12,13,14,15]. The tracking information is obtained by quadrature detection with the sum signal as the reference.

This article briefly describes the research, development, and prototype measurements of a patch-monopole monopulse feed design with a novel, sum-channel, radiating patch element. The evolved design employs, as a circularly-polarized, sum-channel radiating element, a slotted circular patch [16,17,18] instead of an elliptical patch [18,19] or similar [20,21] from the previous design [1]. The slot allows for an additional separation of the sum-channel radiating element from the rest of the feed. The design variables are optimized mainly for good radiation properties, including the sum-channel aperture illumination efficiency in a typical satellite-receiving antenna (spillover illumination is minimized). The slot is designed so that the two fundamental modes of the patch are phase-shifted in quadrature for circular polarization. A technical drawing of the proposed design is shown in Figure 2.

As in the previous design, the sum-channel patch is left-handed-circularly polarized (LHCP) and it is optimized for a symmetric paraboloidal antenna with a focal-distance-to-diameter ratio ${\displaystyle \frac{\mathrm{F}}{\mathrm{D}}}$ of 0.4. The center frequency of operation is 2.21 GHz. Compared to the previous design, the new patch geometry provides an additional degree of freedom while the circular patch outline is somewhat simplified, allowing for a potentially easier fabrication. The new design shows a slightly improved monopulse channel isolation, and a high radiation-pattern symmetry.

2. Materials and Methods

The main course of the research consisted of a study and review of the literature study, then the design synthesis, which was followed by a few prototype iterations to optimize the design. The initial patch dimensions can be obtained from resonance-frequency approximations [22]. Any general-purpose antenna-simulation software, employing, for example, the method of moments, suffices for simulating and optimizing such a feed. At the current stage of research, the feeding circuit consists of coaxial lines below the cavity floor, which complicates electromagnetic modeling of the whole structure. We only simulated the cavity part of the antenna for coarse-tuning of radiation properties, and the experimental work was mainly focused on prototype fabrication and measurements. The optimized physical and electrical dimensions (at 2.21 GHz) of the proposed monopulse feed design are shown in

Table 1. Physical and electrical dimensions (at 2.21 GHz) of the proposed monopulse feed design.

Unit	d	h	e	a	f	u	v	m
mm	96.0	11.0	3.5	69.2	11.0	7.0	18.5	27.5
$\lambda$	0.71	0.08	0.03	0.51	0.08	0.05	0.14	0.20

and correspond to the marking in the technical drawing in Figure 2.

The idea for prototype fabrication is the same as that of the design itself: simplicity and straightforwardness. Prototypes were fabricated manually, by using only some standard tools such as a ruler and compass, metal cutters, a drill, a soldering iron with approximately 50 watts of thermal power, standard tin and flux, etc. The used materials and components were also standard. The cavity and the patch element are both made from 0.4 mm thick brass sheet. The patch element is held up by the center conductors (and/or dielectric) of both sum-channel feedpoints, made from flanged SMA (SubMiniature version A) connectors. The monopole is easily made from a standard SMA cable connector with a flanged mount and a suitable coaxial cable (such as RG-405) with an outer cladding stripped to make a monopole. A fabricated prototype is shown in Figure 3.

The sum-channel patch requires a balanced feed for a symmetric, nonsquinted radiation pattern and least coupling to the difference channel. The prototype uses a simple balun made out of a parallel tee connection of coaxial cables that are ${\displaystyle \frac{1}{2}}$ $\lambda$ and 1 $\lambda$ in length. The patch is best fed at a point where the impedance is around 100 $\Omega$ [23]. A parallel connection of the two arms results in an impedance of around 50 $\Omega$. The monopole feedline is made for the best impedance matching of the difference channel. The length differences in the sum and difference channels manifest as an additional constant phase shift of the resulting monopulse ratio, which needs to be spatially aligned with the antenna (calibrated). The final feed prototype during radiation-pattern measurements, including the feeding circuit, is shown in Figure 4.

The radiation efficiency is expected to be high since the prototype is made out of good-conductivity metals and low-loss dielectrics: air in the cavity, PTFE (polytetrafluoroethylene, also known as Teflon) or PTFE-like in coaxial lines, etc. Simulations show a sum-channel radiation efficiency above 99% for an ideally-fed antenna and above 95%, including a simple coaxial network balun. The experimental verification [24] in a less-than-ideal reverberation chamber revealed a sum-channel radiation efficiency above 80%.

3. Results

3.1. Scattering Parameters

The scattering parameters were measured together with a simple feeding circuit: a coaxial sum-channel balun and a difference-channel monopole impedance-matching stub. The feeding circuit was only coarsely optimized with the conventional cut-and-try method, so only the measurements are provided in this subsection. The simulations indicated that the return loss is not a limiting factor of the design with a good feeding circuit, and confirmed that the patch impedance remains predictable [23]. The return loss frequency characteristic is shown in Figure 5. It indicates a nonideal sum-channel balun for matching the target frequency of 2.21 GHz. The crosstalk between the sum and difference channels of the measured prototype is around −35 dB in the band, and is also shown in Figure 6.

3.2. Radiation Parameters

The total sum-channel directivity of the proposed design at the center frequency of 2.21 GHz is simulated to be 8.9 dBi. Indoor laboratory measurements in a semi-anechoic space confirmed the simulation results within 0.4 dB of the difference in both radiation-pattern cross-sections (cuts), even with the nonideal balun. The radiation-pattern cuts of the sum channel are shown in Figure 7 and those of the difference channel are shown in Figure 8. The horizontal cut plane (mark H) is parallel to the line between the patch feed points, and the vertical cut plane (mark V) is perpendicular to them.

The patch has a bandpass frequency characteristic. The axial ratio is 0.1 dB at the center frequency, where the circular polarization is optimal, and the 3 dB bandwidth of the axial ratio is 32 MHz. The total directivity and the LHCP gain frequency characteristics are shown in Figure 9. As explained in Section 2, the radiation efficiency seems to be in the 80% to 90% range, resulting in a center-frequency LHCP realized gain value of around 8 dBi. The peak aperture illumination efficiency is calculated to be 75% at an ${\displaystyle \frac{\mathrm{F}}{\mathrm{D}}}$ ratio of 0.4. The difference radiation pattern, shown in Figure 8, exhibits a strong dip, suitable for tracking operation. A quick test of the monopulse operation of the feed with a transmitting circularly-polarized antenna and a vector voltmeter at the receiving side showed that the feed is working as expected.

The rotational symmetry of the sum-channel pattern at the reflector’s edge is close to 0 dB according to the simulation results. The measurement results agree well, considering a semi-anechoic-only measurement space. The difference in both points of a cut intersecting the reflector’s edge is 0.4 dB for both cuts. The values between both cuts differ by 1.3 dB, and, on average, the measurements correspond well to the simulation results at the reflector’s edge. Backside radiation results are not accurate due to suboptimal measurement space and the effect of the mounting structure.

4. Discussion

The radiation-pattern measurements correspond well to measurements at about the same solid angle as the illuminated reflector is located, so the taper efficiency is easily calculable. The spillover-efficiency calculation is more problematic, since spillover radiation measurements and simulation results are only rough approximations. The simulation model is too ideal and the measured weak backside-radiated power is sensitive to a nonideal measurement space, while also affected by the mechanical mounting structure, cabling, and a measurement probe (as seen in Figure 4). The same applies to determining the depth of the difference-channel minimum. The sum-channel radiation pattern tapers off by about 9 dB at the edge of the reflector. Including the edge attenuation due to the reflector shape (3 dB for ${\displaystyle \frac{\mathrm{F}}{\mathrm{D}}}$ of 0.4), the edge attenuation amounts to about 12 dB. There is still confidence in the results around the reflector’s edge, where the spillover radiation is the strongest. The weaker side and back radiation affect the spillover-efficiency approximation less.

Compared to the elliptic-patch design [1], the diameter of the feed is slightly reduced, while the new design still increases the directivity. A feed of this size shades only 2% of the aperture area of a reflector that is only 5 $\lambda$ in diameter. More than the new patch shape, the narrowband design is responsible for the increased directivity, and also a shallower cavity, to a lesser extent. The feed design is optimized for a relatively narrowband sum channel to see the achievable efficiencies of such a design. The crosstalk between the channels of −35 dB is suitable for tracking operation. Since the tracking control loop is very slow compared to the sum-channel data rate, the difference channel is much less demanding in terms of bandwidth and the required signal-to-noise ratio.

For future work, a wideband optimization at a minimal and acceptable performance loss in the middle of the band should be tested with the design to find its practical limits. The feeding circuit elements could be integrated directly into the feed structure for an even more compact, repeatable, and integral solution. It should also be possible to scale the design to at least the X band, make a dual-polarization version of the design, or develop a simple radome for such a feed. Finally, it is impractical to wholly measure the feed in an indoor test space, and the final antenna efficiency, gain, and noise temperature can only be predicted. In the future, we will also focus on field-testing the feed, illuminating a real ground-station antenna.

5. Conclusions

The main contribution of this article is the introduction of a new sum-channel radiating element with a comparable shape complexity to the patch-monopole monopulse feed design. The new sum-channel patch shape allows for conductive separation of the patch to the other parts of the feed.

Both sum and difference channel radiation patterns are as expected in the design phase. The sum-channel center-frequency LHCP gain of the feed is around 8 dBi. The feed exhibits a somewhat narrowband sum-channel in the 30 MHz range as it is primarily optimized to test achievable efficiencies of the design. The sum-channel aperture illumination efficiency is 75% at a target ${\displaystyle \frac{\mathrm{F}}{\mathrm{D}}}$ ratio of 0.4. We were unable to measure radiation efficiency very well, but the design with low-loss dielectrics should not be problematic or limiting in this regard.