Planar antenna array

[15] 3,701,162

[45] Oct. 24,1972

United States Patent

Seaton

<td> [54]</td><td colspan="2"> PLANAR ANTENNA ARRAY</td><td> [56]</td><td> References Cited</td>

<td> 172] [73]</td><td> Inventor: Assignee:</td><td> Arthur F. Seaton, Palos Verdes Estates, Calif. Hughes Aircraft Company, Culver City, Calif.</td><td> 2,557,951 2,967,301</td><td> UNITED STATES PATENTS 6/1951 DeRosa et al..........343/771 X 1/1961 Rearwin....................343/771</td>

<td> [22] [21]</td><td> Filed: Appl. No.:</td><td> March 24,1964 354,341</td><td colspan="2"> Primary Examiner—Benjamin A. Borchelt Assistant Examiner—Richard E. Berger Attorney—James K. Haskell and Noel B. Hammond</td>

<td></td><td></td><td></td><td> [57]</td><td> ABSTRACT</td>

<td> [52] [51] [58]</td><td colspan="2"> U.S. Cl..............................343/771, 343/100 PE Int. Cl..............................................HOlq 13/10 Field of Search............................343/771, 100.3</td><td colspan="2"> A circularly polarized planar array of slot antenna elements provide a pencil-beam pattern substantially perpendicular to the plane of the array.</td>

Claims, 4 Drawing Figures

<img file="US3701162A_D0001.tif" />

PATENTED0CT24 1972

3.701,162

SHEET 1 OF 2

<img file="US3701162A_D0002.tif" />

<img file="US3701162A_D0003.tif" />

-e***M*&lt;

<img file="US3701162A_D0004.tif" />

+4&amp;EA/7

PATENTED OCT 2 4 1972

3.701.162

SHEET 2 OF 2

<img file="US3701162A_D0005.tif" />

/yy£//ra^·

111

3,701,162

PL ANAR ANTENNA ARRAY

The invention described herein was made in the performance of work under a NASA contract and is subject of the provisions of section 305 of the National Aeronautics and Space Act of 1958, Public Law 85568 (72 Stat. 435; 42 U.S.C. 2457).

Planar arrays are used in applications requiring high gain and high aperture efficiency. However, conventional planar arrays are ordinarily linearly polarized. Where circular polarization is a requirement, a parabolic antenna, rather than a planar array of slot elements, is usually employed.

However, parabolic antennas have a lower gain and lower aperture efficiency than a planar array and, in addition, are larger and heavier. Hence, attempts have been made to adapt planar arrays to circular polarization.

The so-called “leaky pipe” waveguide planar array, although usually linearly polarized, can be made circularly polarized but it does not provide a “broadside” beam, that is, a beam perpendicular to the plane ofthe array. The mode of operation is essentially “end-fire,” that is, with the beam substantially parallel to the plane of the array (sometimes called the zero order beam). The leaky-pipe array cannot be made to operate in the broadside mode, and tilts the beam at least 40° off broadside. Aperture efficiency is low because the effective aperture is reduced by the cosine of the beam-tilt angle.

The annular slot waveguide planar array also can be made to provide circular polarization but it suffers from the necessity of spacing the slots a full guide wavelength apart. This results in the generation of secondary maxima, that is, in addition to the broadside lobe, the array generates other large lobes at large angles off broadside. Consequently, the annular slot antenna has low efficiency.

Circularly polarized traveling wave linear arrays providing a broadside beam have been constructed employing nonresonant crossed slots on the broad wall of a waveguide operating in the TE₁₀ mode. These linear arrays may be combined to form planar arrays but such an array also has an extremely low aperture efficiency because, again, a large slot spacing is required. This large slot spacing is necessitated by the fact that crossed slots can be used only on one side of the centerline of the waveguide for a given sense of circular polarization (i.e., right-hand or left-hand circular polarization). Crossed slots on the other side of the centerline radiate a circularly polarized wave of the opposite sense. Broadside operation of the array thus requires that the slots be spaced a full guide wavelength apart. This spacing, which is approximately equal to 1.4 times a free-space wavelength, allows the generation of secondary maxima which seriously detract from the principal maximum, or main lobe.

Accordingly, it is an object of the present invention to provide a circularly polarized planar antenna array having a narrow beam pattern substantially perpendicular to the plane of the array and having substantially no secondary lobes.

Another object of the invention is the provision of a circularly polarized planar antenna array having a high efficiency.

Yet another object of the present invention is to provide an efficient circularly polarized high-gain antenna which is light in weight and which has a thin, flat configuration.

In accordance with these and other objects of the invention, there is provided a thin rectangular enclosure 5 having one broad surface slotted to form radiating elements and the opposing broad surface corrugated to form a slow-wave structure. The rectangular enclosure forms a wide rectangular waveguide operating in a multimode fashion to effectively produce a plurality of vir^υ tual waveguide linear arrays extending side-by-side and operating in the traveling wave mode. Circular polarization is obtained by using nonresonant crossed slots on one side of the centerline of each virtual ₁₅ waveguide and pairs of separated nonresonant slots on the other side of the centerline. Each pair of separated slots is centered between the crossed slots to obtain the correct sense of rotation and phase. Alternate linear arrays are displaced by half a guide wavelength to ac20 count for the phase reversal of the alternate half cycles of the multimode feed arrangement. The antenna is excited from feed slots in the bottom of the first corrugation trough of each linear array. The feed slots are shunt-type slots in a feed waveguide extending tranver25 sely across the ends of the linear arrays. The feed waveguide operates as a standing wave array and is centerfed by means of a magnetic loop coax-towaveguide transition. Conductive strips serving as mode suppressors extend from the corrugations to the ³⁰ aperture plate along the line of the virtual waveguide walls.

The following specification and the accompanying drawings describe and illustrate an exemplary embodi₃₅ ment of the present invention. Consideration of the ³ specification and the drawings will provide an understanding of the invention, including the novel features and objects thereof. Like reference characters denote like parts throughout the figures of the ₄₀ drawings.

FIG. 1 is a plan view of an embodiment of a planar antenna array in accordance with the invention, with a portion of the aperture plate broken away to show the interior;

FIG. 2 is a perspective view, partly broken away, of a portion of the planar array of FIG. 1;

FIG. 3 is a schematic diagram indicating the current distribution in a portion of the planar array of FIGS. 1 and 2; and

FIG. 4 is a vector diagram indicating the relative exciting voltages for three different instants of time of the slots shown in schematic diagram of FIG. 4.

Circularly polarized arrays can be constructed by using crossed slots on the broadwall of a waveguide at a 55 point of circular polarization of the magnetic fields within the waveguide. However, efficient aperture illumination is difficult to obtain using these slots along since they must be spaced a full guide wavelength apart in order to keep the radiated fields in phase. This ⁶⁰ results in a free space distance between elements of approximately 1.4 multiplied by a free space wavelength. This large spacing is undesirable because it results in a loss of aperture efficiency in the generation Of seconda., ry maxima (also known as second order beams or grating lobes).

In a linearly polarized broadwall shunt slot array it is conventional to place consecutive slots at a half guide

3,7' wavelength spacing but on opposite sides of the centerline, thereby keeping all slots in phase. This technique will not work with crossed slots because only the transverse components of current on the waveguide walls are reversed on the opposite side of the centerline. The longitudinal components are in phase at all points across the guide at any given transverse plane. This situation results in circular polarization of the opposite sense being radiated from crossed slots placed on the opposite side of the guide.

To solve these problems, the present invention employs additional slots in a new arrangement which together with a slow wave structure produce adequate interelement spacings. These additional slots are also of the nonresonant type, and may be considered to be the two orthogonal portions of a crossed slot separated by a spacing of half a guide wavelength. The separated slots are located on the waveguide relative to the position of the normal crossed slot such that at broadside both the normal crossed slot and the separated crossed slot radiate a circularly polarized wave in phase and of the same sense.

In FIG. 1 there is illustrated an exemplary embodiment of a circularly polarized planar array 10 constructed in accordance with the present invention. There is provided a thin, rectangular enclosure 11 made of metal or other conductive material. The rectangular enclosure 11 forms a wide rectangular waveguide suitable for excitation in a multimode manner. In the present example, the enclosure 11 is 38.5 inches square and 1 inch thick. One broad surface 12 of the enclosure 11 is slotted to form radiating elements. The other broad surface 13 of the enclosure 11 is corrugated to form a slow-wave structure.

When the enclosure 11 is excited, it operates as a plurality of rectangular waveguide linear arrays extending side-by-side. In the present example, none virtual waveguides are formed. No waveguide walls are necessary to separate the virtual waveguides because the electromagnetic fields operate as if there were solid walls between the virtual waveguides. However, due to the lack of mirror image symmetry in the slots on each side of the virtual walls, mode suppressors 14 are employed. The mode suppressors 14 are elongated metal strips extending along the virtual walls between the aperture surface 12 and the crest of the corrugated surface 13, and form partial waveguide walls. The mode suppressors 14 suppress undesirable propagation modes in the virtual waveguides.

The first virtual waveguide is indicated by the bracket 15 extending across the broad dimension thereof at the feed end. On one side of the waveguide centerline, there is a row of crossed-slot elements 16. Each crossed-slot element 16 is formed of two intersecting orthogonal slots 17, 18. The slots 17, 18 are disposed at an angle of 45° with respect to the sides of the enclosure 11 so that they can be elongated without extending past the waveguide edge. The crossed-slot ⁶⁰elements 16 are located at the point in the waveguide where the wall currents are circularly polarized and are spaced approximately one guide wavelength apart. Actually, the spacing is somewhat less than one wavelength, as will be discussed hereinafter.

On the other side of the waveguide centerline, there is a row of separated slot-pair elements 20. Each

01,162 separated slot-pair element 20 is formed of two nonintersecting orthogonal slots 17', 18'. These slots 17', 18' are also oriented at an angle of 45° with respect to the sides of the enclosure 11, and at an angle of 90° with ⁵ respect to each other. Each slot designated 17' of a separated slot-pair element 20 is parallel to a slot designated 17 of a crossed-slot element 16. Similarly, each slot designated 18' of a separated slot-pair element 20 is parallel to a slot designated 18 of a crossed^υ slot element 16. The centers of the two slots 17', 18' of each separated slot-pair element 20 are spaced nominally half a guide wavelength apart, as modified by spacing factors to be discussed. Each separated slot15 P³³ element 20 is centered longitudinally along the guide between two crossed-slot elements 16. With this slot pattern, the crossed-slot elements 16 and the separated slot-pair elements 20 radiate a circularlypolarized wave in-phase and of the same sense.

2o The slot pattern for each of the other waveguides is the same except that the slot pattern is displaced by half a guide wavelength in alternate waveguides to account for the phase reversal of the alternate half cycles of the multimode excitation. The planar array 10 may 25 be considered to be nine linear arrays, each having slotpair elements 16, 20 staggered across the centerline to obtain in-phase, circularly polarized radiation from an interelement spacing of half a guide wavelength.

FIG. 3 schematically shows the currents for the TE₁₀30 mode on the broadwall of a slotted rectangular waveguide representing one of the slotted virtual waveguides of the planar array 10 of FIGS. 1 and 2. By noting the direction of the current lines exciting the crossed slots 17, 18 as the wave progresses down the ³⁵ waveguide, it can be determined that the radiated field is left-hand circularly polarized for a wave coming out of the page. The field is circularly polarized rather than elliptically polarized because the slots 17, 18 are dis₄₀ placed laterally from the centerline to the location where the magnitude of the current remains constant as the wave propagates or, in other words, where the magnitudes of the longitudinal and transverse currents (which are 90° in time-phase) are equal.

The separated slots 17', 18' are disposed on the opposite side of the centerline and centered between the crossed slots 17, 18 to provide the correct sense of rotation and phase with respect to that of the crossed slots 17, 18. The lateral displacement of the separated 50 slots 17', 18' from the centerline is identical to that of the crossed slots 17, 18, because the same criterion holds. FIG. 4 vectorially illustrates the relative exciting voltages in the slots 17, 17', 18,18' at three successive instants of time differing from each other by 45° in ⁵⁵ phase of the wave traveling in the direction of propagation shown in FIG. 3. The summation of the vectors, indicated as E_t in FIG. 4, represents the addition that takes place in the direction of the main beam. As may be seen from FIG. 4, the total electric field propagated in the direction of the main beam from the two separated slots 17', 18' is always equal in amplitude, orientation and phase to the electric field propagated in that direction from the crossed slots 17,18.

The planar array 10 may also be considered to be two interlaced linearly polarized planar arrays orthogonal to each other and fed in time quadrature. That is, all of the slots oriented in the direction of slots

3,701,162 and 17' may be viewed as one linear array arranged in rows of slots, with the slots closely spaced more or less end-to-end, and the rows widely spaced. The rows extend at an angle to the sides of the enclosure 11 along lines through the centers of the slots, such as the row of 5 slots indicated by the line in the direction of the arrows 19—19 shown in FIG. 1. Similarly, all of the slots oriented in the direction of slots 18 and 18' may be viewed as rows of slots forming the second linearly polarized array orthogonal to the first. ¹⁰

To adequately suppress secondary maxima (second order beams or lobes) in order to obtain high aperture efficiencies, the interrow spacing has been reduced to less than a free-space wavelength. This is accomplished by means of the corrugated surface 13 which forms a ¹⁵slow-wave structure that provides a 25 percent reduction in guide wavelength. When the array 10 is considered as two interlaced linearly polarized planar arrays, it is apparent that the separation between slots in 20 the same row, such as the row indicated by the arrows 19—19 in FIG. 1, is much less than a free-space wavelength. By use of the corrugated surface 13 as a slow-wave structure, the orthogonal separation between rows is reduced to 0.8 of a free-space 25 wavelength. In this manner, an aperture efficiency of 92 percent was obtained.

To obtain maximum gain, the planar array 10 is uniformly illuminated. Accordingly, because this is a traveling wave array and not a standing wave array, the 30 coupling coefficient of each succeeding slot 17, 17', 18, 18' is made progressively larger as the distance down each virtual waveguide from the feed point becomes larger. The coupling coefficient is increased by increasing the length of the slots which makes them ³⁵more nearly resonant. In this manner, each pair of slots radiates the same amount of power. The last slot is adjusted to couple out as much of the remaining power as possible. To compensate for phase differences thus introduced, the spacing between the slots is gradually decreased as the distance from the feed point becomes larger. The end of the enclosure 11 farthest from the feed end is closed or shorted in the present example, although it may also be left open if desired. ₄₅

To obtain a predetermined space coverage and good impedance match, the beam is tilted 6° back towards the feed end by decreasing the slot spacing an additional amount. The result of this reduction of slot spacing is to reduce the row spacing by some fraction of the 50 percentage of reduction in slot spacing. This reduction in slot spacing is also highly desireable from the standpoint of input impedance. A separation of one guide wavelength between the crossed-slot elements 16 results in a separation of a half guide wavelength 55 between the separated slots 17', 18'. The disadvantage of exactly half a guide wavelength spacing is that reflections from the slots would all add in-phase, thereby producing a high voltage-standing-wave-ratio at the input to the linear arrays. ⁶⁰

The linear arrays are fed from a rectangular feed waveguide 25 (best seen in FIG. 2) operating as a standing wave array and located underneath the radiation virtual waveguides and at one end of the array 10. Coupling from the feed waveguide 25 to the linear arrays is accomplished by nonresonant longitudinal shunt slots 26 located in the broad wall of the feed waveguide and opening into the first trough in the corrugated surface 13. The coupling or feed slots 26 are spaced apart by half a free space wavelength and are disposed on the same side of the centerline of the feed waveguide 25 to provide 180° phase shift between each of the feed slots 26. The feed waveguide 25 is terminated in shortcircuits at each end, one-quarter guide wavelength beyond the last feed slots 26. The reactive component of the admittance is tuned out at each feed slot 26 by an inductive post 27 disposed at each feed slot 26. A coaxial connector 28 couples to the feed waveguide 25 by means of a coupling loop 30 located opposite the central feed slot 26.

For operation at a frequency of 2,295 megacycles per second, the broad dimension of each virtual waveguide is 4.267 inches and the narrow dimension is 0.991 inch. The corrugated surface 13 has a dimension of 0.491 inch from the crests to the troughs, and the corrugations are 0.1 inch in width and are separated by 0.4 inch. By constructing the array 10 of thin aluminum (on the order of from 0.003 — 0.010 inch thick, for example), the total weight of the array 10 is very small, 8½ pounds for example.

An embodiment of an array constructed in accordance with the present invention provided an over all efficiency of 70 percent, aperture illumination efficiency of 92 percent and a gain of 27.0 decibels. The measured ellipticity of the polarization was 1.0 decibel.

Thus, there has been described an efficient, circularly polarized planar antenna array providing a pencilbeam pattern substantially perpendicular to the plane of the array by virtue of employing crossed slots and separated slots in a new arrangement, in conjunction with a slow-wave stnicture.

While only one embodiment of the invention has been shown and described, variations may be made, and it is intended that the foregoing disclosure shall be considered only as illustrative of the principles of the invention and not construed in a limiting sense.