US 3604010 A
Description (OCR text may contain errors)
United States Patent lnventors Leonard Schwartz Scarsdale: Robert K. Bevan, Plennantvllle, both of, N.Y. Appl. No. 795,172 Filed Jan. 30, 1969 Patented Sept. 7, 1971 Assignee Singer-General Precision, Inc.
ANTENNA ARRAY SYSTEM FOR GENERATING SHAPED BEAMS FOR GUIDANCE DURING AllRCRAFI LANDING 10 Claims, 26 Drawing Figs.
US. Cl 343/756, 343/768, 343/771, 343/786, 343/840 int. Cl H0111; 13/10 Field oil Search 343/767,
 References Cited UNITED STATES PATENTS 3,017,630 1/1962 Begovich et al. 343/771 3,213,454 10/1965 Ringenbach..... 343/771 3,430,247 2/1969 Wong 343/853 Primary Examiner-Eli Lieberman Att0meySal A. Giarratana ABSTRACT: This specification discloses an antenna system for guiding aircraft during landing. The antenna system comprises slotted waveguides for generating radiobeams of the desired shape to provide wide coverage and eliminate false courses. The spacing between the slots and the coupling of the slots to the waveguides are individually selected to synthesize the desired beam shapes,
PATENTED SEP 7 i971 SHEET 1 OF T Z G F FIG.I.
mw Oa SINE I o ANGLE OFF BORESIGHT DEGREES .mov
0 lb ANGLE OFF BORESIGHT (DEGREES) INVENTORS LEONARD SCHWARTZ 8 ROBERT K. BEVAN ATTORNEY CONDUCTANC E RELATIVE POWER (DB) PATENIEDsEP mu 3,604,010
SHEET t UF 7 a; 18) I I 2 .150
E 140- D- LU D 130- E 0 .no- .l5 .l6 .1? .na .I9 2021 22 16017018090200.2020 P0sT DISPLACEMENT POST ms uxcemsm (INCHES) (INCHES) FIG. l4.
INVENTORS 40 LEONARD SCHWARTZ 8:
ROBERT K. BEVAN o|020304050e070a090 ANGLE OFF BORESIGHT (DEGREES) BY w/ yg ATTOR NEY PATENIEUSEP 7m 3,604,010
SHEET 7 [IF 7 F l G 23..
.018 .Ol6 111.014 5.012 .o|o 2-222 6.004
o IllII -I.Q49||||||11||1|||| l 2 3 4 5 s 024 68l0l2l4l6l8202224262830 SLOT ANGLE DEGREES SLOT ANGLE DEGREES FIG F I626.
' I I66 I67 I68 7 "1 7 R|GHT LEFT UPPER LOWER LOCALIZER LOCALIZER 5195 "6 l4 ARRAY ARRAY ARRAY ARRAY g L] F1 4 'SWITCH 4" I63 I 9 6! SIGNAL SOURCE- 161 mvsn'mnss LEONARD SCHWARTZ 8 ROBERT K. BEVAN ATTORN EY ANTENNA ARRAY SYSTEM FOR GENERATING SHAPED BEAMS FOR GUIDANCE DURING AERCRAFT LANDING BACKGROUND OF THE INVENTION This invention relates to an improved system of producing radiobeams for guiding aircraft and, more particularly, to an improved set of antennas for generating localizer and glide slope beams to provide precise aircraft guidance during landmg.
In aircraft landing guidance systems, shaped radiobeams are transmitted to the aircraft to guide the aircraft both vertically and horizontally. The horizontal guidance portion of the system is referred to as a localizer, and the vertical guidance portion of the system is referred to as the glide slope. The aircraft receives the transmitted radiobeams and converts these signals into indications to the pilot to fly right or fly left and to fly up or fly down depending on whether the aircraft is to the left or right or below or above the desired course of the aircraft. The systems of the prior art have problems of generating false courses for the aircraft and not providing indicating signals when the aircraft is at wide angles from the desired course.
The system of the present invention comprises a novel antenna arrangement, which greatly reduces the problem of false course generation and, at the same time, achieves a wide angle of coverage so that aircraft at wide angles from the desired course receive the correct signal.
SUMMARY OF THE INVENTION In accordance with the present invention, a separate antenna system is provided for the localizer system and for the glide slope system. The localizer antenna comprises two waveguides in a common parabolic reflector each being used to generate a shaped radiobeam. One radiobeam peaks on the right side of the desired course and the other radiobeam peaks on the left side of the course and the aircraft by detecting the relative strength of the two beams receives an indication of whether or not to fly left or fly right to correct its course. Near the correct course, the size of the signal linearly varies with the angle of deviation from the desired course. The right and left radiobeams are time shared and have different amplitude modulations thereon so that the two beams can be distinguished. The beams are provided with a desired shape by a unique synthesis technique which results in the slots having a varied spacing and a varied coupling with the waveguide. The glide slope system also generates two radiobeams above and below the desired course and the aircraft in a similar manner determines by detecting the difference in magnitude in the two received signals whether or not it is above or below the desired course. These beams like the localizer beams are shaped in a particular manner by a unique antenna design involving selected slot spacings and couplings in the waveguides generating the glide slope beams.
Accordingly, an object of the present invention is to provide an improved system for generating radiobeams for guiding the landing of aircraft.
Another object of the present invention is to provide an improved antenna array for generating shaped radiobeams.
A further object of the present invention is to provide an improved antenna for use in guidance systems for aircraft landmg.
A still further object of the present invention is to reduce the problem of false courses and increase the coverage in guidance systems for aircraft landing.
Further objects and advantages of the invention will become readily apparent as the following detailed description of the invention unfolds and when taken in conjunction with the drawings briefly described below.
BRIEF DESCRIPTION OF THE DRAWINGS FIGS. 1 and 2 are graphs illustrating the shapes of radiobeams which can be generated by a particular radiating source;
FIG. 3 shows two curves illustrating the shape of two ideal beams which are desired to be generated by the localizer portion of the system of the present invention;
FIG. 4 is a graph illustrating how a desired beam shape is synthesized by matching the beam at selected angles;
FIG. 5 is a graph illustrating the pattern radiated by a dumbell slot;
FIG. 6 is a graph illustrating the ideal array factor pattern for one of the localizer antenna arrays;
FIGS. 7 and 8 are graphs illustrating the amplitude and phase distribution of the electric field along the antenna array to approximately synthesize the desired beam pattern for the localizer system;
FIG. 9 illustrates the slotted side of a waveguide used for generating one of the localizer beams;
FIG. 10 is a sectional view taken through FIG. 9 along the lines 10-10;
FIG. I l is a graph illustrating how the locations of the slots in the waveguide shown in FIGS. 9 and 10 are determined;
FIGS. 12 and 13 are graphs which are used to obtain the desired coupling of some of the slots for the waveguide in FIGS. 9 and 10;
FIG. 14 is a perspective view of the localizer antenna system of the present invention;
FIG. 15 is a graph illustrating the shapes of the beams produced by the localizer system of the present invention;
FIG. 16 is a graph illustrating the ideal desired shape the upper glide slope beam to be produced by the glide slope antenna system;
FIG. 17 shows a front view portion of the waveguide array for producing one of the glide slope beams;
FIG. 18 shows a side view of the waveguide array portion shown in FIG. 17;
FIG. 19 is a graph illustrating the pattern radiated by the open end of a waveguide;
FIG. 20 is a graph illustrating the ideal array factor patterns for the glide slope system;
FIG. 21 is a graph illustrating amplitude and phase distribution along the antenna array to approximately synthesize the desired upper glide beam pattern;
FIGS. 22, 23 and 24 are graphs from which the slots in the upper glide slope antenna are designed;
FIG. 25 is a perspective view of a portion of the complete assembly of one of the glide slope arrays; and,
FIG. 26 is a block diagram illustrating the time shared operation of the system.
DESCRIPTION OF THE PREFERRED EMBODIMENTS When a radio signal is radiated by an aperture in the form of a continuous line source of a finite dimension, the strength of the signal will vary with the direction of radiation in a manner depicted by FIG. 1, if phase and amplitude of the electric field is constant from one end of the aperture to the other. In FIG. 1, the strength of the radiated signal is plotted against the sine of the angle of radiation in a plane containing the line of the radiating aperture. In FIG. 1, zero angle is in a direction perpendicular to the line of the radiation aperture. It will be noted in FIG. 1 that the null points 10 of the radiated signal are equally spaced except for the two null points on opposite sides of the main radiated beam, which is represented by that portion of the graph in FIG. 1 designated by the reference number 11. The spacing between these two null points on opposite sides of the main beam 11 is twice that between the remaining null points. This relationship between the spacing between the null points is obtained because the signal strength is plotted against the sine of the angle of radiation rather than the angle of radiation. The width of the spaces between the null points is inversely proportional to the length of the aperture. Ac-
cordingly, the beam width of the main beam 11 is inversely related to the length of the aperture.
If the phase of the electric field at the aperture varies linearly from one end of the aperture to the other while maintaining the amplitude of the electric field constant, the entire pattern will be displaced with respect to angle of radiation so that a pattern such as that illustrated in FIG. 2 is produced. In FIG. 2, like that of FIG. 1, the strength of the radiated signal is plotted against the sine of the angle of radiation. If the total phasevariation from one end of the aperture to the other amounts to 2N7r, then the pattern will be displaced two N spaces, one space being defined as the distance between two null points. If the phase variation from one end of the aperture to the other is selected so that it is an integral multiple of 211', then the main beam of the displaced pattern will occur at a null point of the pattern shown in FIG. 1. if several such displaced patterns are superimposed, whereby each displaced pattern is produced by means of a phase variation from one end of the aperture to the other, which phase variation is a multiple of 21r, then the main beam of each pattern will occur at a null point of all the other patterns. This result obtains because the null points are equally spaced when the strength of the signal is plotted against the sine of the radiating angle and because the superimposed patterns will be displaced from the pattern shown in FIG. 1 by an integral number of such spaces as a result of the phase variation to produce each pattern being selected to be a multiple of 211-. If electric fields with phase variations to produce the patterns with different displacements are superimposed on the aperture, then the radiated patterns will be superimposed. By selecting the phase variation of each superimposed electric field to be multiples of 211-, the main beam of each superimposed pattern will occur at a null of all the other patterns. The amplitude of each superimposed pattern will correspond to the amplitude of the corresponding superimposed electric field.
To approximately synthesize a desired pattern of radiation, superimposed patterns of radiation such as those shown in FIGS. 1 and 2 are selected so that the main beam of each of the superimposed patterns matches the desired pattern at each of a plurality of points. The superimposed patterns are selected so that the main beam of each superimposed pattern occurs at the null point in every other superimposed pattern, the desired pattern can be matched exactly at these points by synthesis. The antenna system of the present invention uses this method of synthesizing desired radiation patterns.
The antennas of the present invention radiate shaped patterns to guide an aircraft along a particular course. Two antenna systems are used, one of which guides the aircraft in azimuth and the other of which guides the aircraft in elevation. The antenna system which guides the aircraft in azimuth is referred to as the localizer system, and the antenna which guides the aircraft in elevation is referred to as the glide slope system. The localizer system comprises two antenna arrays, each of which radiates a shaped beam, which beams ideally are of the radiated signal is plotted against the angle of radiation in a given plane. The curve 13 depicts the desired shape of a radiobeam to be radiated on the right side of the desired course, which is referred to as the boresight, and the curve 15 depicts the desired shape of a radiobeam to be radiated on the left side of the boresight. The beams corresponding to the curves 13 and 15 will be referred to as the beams 13 and 15 respectively. These two beams will be produced alternately in a time-shared relationship and will be distinguished by receiving equipment on the aircraft to be guided by means of amplitude modulation on the two radiated beams. Receiving equipment on the aircraft to be guided receives the two beams and compares the relative strengths of the two beams. If the aircraft is directly on course, the two beams will be received in equal strength. If the aircraft is to the right of the boresight, the, beam corresponding to the curve 15 will be received in greater strength than the beam corresponding to curve 13. The receiving equipment on the aircraft compares the relative strengths of the two received beams and provides an indication from these relative strengths. If the beam 13 is received in greater strength than that of the beam is, meaning that the aircraft is to the right of the boresight, the receiving equipment will produce a fly left" signal indication indicating to the pilot to fly his aircraft to the left to bring the aircraft back on course. Similarly, if the receiving equipment detects the beam 15 in greater strength than the beam 13, the receiving equipment will produce a fly right signal to indicate to the pilot to fly his aircraft to the right to bring it back on course. Because of the shape of the two beams represented by the curves 13 and 15, the difference between the strengths of the two beams received by the aircraft when the aircraft is near the desired course or boresight will vary linearly with the angle that the aircraft is from the boresight. Accordingly, the receiving equipment can produce an output signal which has a magnitude varying linearly with the angle with which the aircraft is off the desired course or boresight. The angle through which this linear indication is provided is referred to as the linear course width". At angles greater than the linear course width, it is desired for the fly right or fly left signals produced by the receiving equipment to be maximum or full scale.
In the system of the present invention, each of the beams 13 and 15 are produced by a separate antenna array, which is designed to produce the desired beam shape by a beam synthesis technique employing the principles described with reference to FIGS. 1 and 2 above. To approximately synthesize the beam 13, the length of the radiating aperture must first be selected. In order to achieve a linear course width of pulse or minus 3.5", the aperture is selected to be 6 wavelengths long. As pointed out above, the width of the main beam of each of the superimposed patterns is inversely related to the length of the radiating aperture. The width of the superimposed beams determines the steepness of the curves l3 and 15 and the steepness of the curves 13 and 15 determines the linear course width. The point at which the two curves 13 and 15 cross each other on the boresight is defined as the crossover point. The level of the crossover point relative to the beam maximums can be readily varied by shifting the directions of each of the patterns after they have been synthesized in a manner to be described below.
Since the width of the spaces between null points in the superimposed patterns is inversely related to the length of the radiating aperture, the number of match points which occur at the null points depends upon the length of the aperture and is equal to twice the length of the radiating aperture in wavelengths plus one. Thus, the 6 wavelength aperture will have 13 match points. One of the match points is at angle zero and six match points are located at angles on each side of angle zero. These match points occur at angle 0 such that sin 6=N.\/L in which L is the waveguide length, A is the wavelength of the radiated signal and N is any integer such that NAIL is less than or equal to 1. Accordingly, the match points on each side of angle zero will occur at angles of 9.6", 19.9", 30, 41.9", 56.6 and Since the desired pattern as shown in FIG. 4 is a minimum on the left side of angle zero, all the match points on the left side of the angle zero require no matching of superimposed patterns. In addition, a match point at angle zero would result in an extremely peaked amplitude distribution of the electric field along the array, which is difficult to obtain. Accordingly, the match point at angle zero is not used. This leaves six match points at the angles listed above. Thus, in FIG. 4, the desired pattern represented by the curve 13 is matched at the six match points by the main beams of six superimposed narrow beam patterns. The main beams of these superimposed narrow beam patterns are represented by the curves designated 21-26, which plot the strength of the main beams of these superimposed patterns as a function of the angle. The main beams of the superimposed narrow beam patterns corresponding to the curves 21-26 for convenience shall be referred to as the beams 21-26 respectively. As shown in FIG. 4, the beam 21 is produced at an angle of 9.6". This beam would be produced by a linear phase variation in the electric field from one side of the aperture to the other of 211'.
The peak of the beam 21 matches the curve 13 at the maximum point of curve 13. The beam 22, pointed at an angle of 19.9", will be produced by a linear phase variation from one end of the aperture to the other of 47?. The peak of this beam matches the curve 13 at the angle l9.9, 4 decibels below the beam maximum. Similarly, the beams 23-26, pointing at angles of 30, 41.9, 56.6 and 90 are produced by linear phase variations on the aperture of 611-, 811', 1011' and l2n-respectively The peaks of these beams match the curve 13 at these angles down 8 decibels 12.3 decibels 16.6 decibels and 30 decibels from the maximum of the curve 113. By superimposing electric fields on the aperture with phase distribution to produce the beams pointing at the angles as shown in FIG. 4 and with amplitudes so that the peaks of the beams match the curve 13, the curve 113 could be approximately synthesized. The curve 13 would be exactly matched at the match points because all of the superimposed beams except for the matching beam are at null at the match points. The distribution of the amplitude and phase of the total electric field along the aperture resulting from superimposing the electric fields to produce the matching beams as described above is given by the following equation:
n and x is the distance from the center of the aperture. Equation (1) gives the total resulting electric field distribution in real and imaginary numbers, the real number representing one component of the filed and the imaginery number representing a component 90 displaced in phase from the component represented by the real number. In order to achieve the amplitude and phase distribution along the aperture represented by equation (1), it is necessary that the amplitude and the phase be independently adjusted along the array. A practical microwave antenna array which possesses the characteristic of permitting independent adjustment of amplitude and phase is the travelling wave slotted waveguide array. The phase of the electric field radiated from a slot in the waveguide is determined by its position along the array while the amplitude of the electric field can be adjusted independently by adjusting the slot coupling. Since the slotted waveguide is a discrete rather than a continuous line source, the pattern obtained will only approach the synthesized function which would be obtained by the superimposing of narrow beam patterns containing the beams 21-26. Moreover, in determining the amplitude of the electric field along the array, it is necessary to take into account the element factor, which is the reduction in the amplitude of the radiated signal due to the fact that the radiated is by slots rather than by continuous line source. In the antenna array for the localizer beam, the slots in the waveguide are horizontal dumbbell slots, which are used so that the radiated beam will be vertically polarized. The experimental radiation pattern of a dumbbell slot in a plane containing the length of the slot is illustrated in FIG. 5. The overall pattern of the slotted waveguide antenna will be the product of the linear array pattern multiplied by the element factor. Since we are dealing with decibels, this multiplication can be taken into account by simple addition. In other words, in order to produce each of the narrow beams 21-26 as shown in FIG. 41, by the slotted waveguide array, the element factor plus the array factor must add together in decibels to produce each of the beams 21-26. From FIG. 5, the element factor for each of the match points can be determined. As shown in FIG. 5, the pattern produced by a dumbbell slot is at a maximum at The pattern is reduced a negligible amount at the match point of 96 and, accordingly, the element factor at 9.6 in decibels is zero. At the match point of 19.9, the curve in FIG. shows the radiated pattern to be down 0.75 decibels. Similarly, the element factor at each of the other match points 30, 4l.9, 566 and 90 is determined from the curve in FIG. 5 to be 1.8 decibels, 2.5 decibels, 6.6 decibels and decibels respectively. The desired array factor can be determined by subtracting in decibels the element factor at each of the match points from the total desired pattern at each of the match points shown in FIG. 41, thus, the array factor at 9.6 is down 0 decibels from the beam maximum. The array factor at 199 is equal to 4 decibels down from the maximum minus the element factor of 0.75 or, in other words, 3.25 decibels down from the maximum. The array factors at each of the remaining match points of 30, 41.9, 56.6and are determined in a similar manner to be 6.2 decibels, 9.8 decibels and 10 decibels respectively down from the maximum. Table I shows the total pattern, the element factor, and the array factor determined for each of the match points.
The last column in table 1, labeled C merely amounts to a conversion of the decibel values of the array factor into relative electric field magnitudes taking the electric field magnitude at the match point at 96 to be one. If the array factor values are plotted as a function of angle, the curve shown in FIG. 6 results. This curve represents the desired array factor pattern. The problem resolves itself into constructing a slotted waveguide array in which the phase of the slots and the slot couplings are adjusted to produce the array factor values given in table I or, stated another way, to match the array factor pattern shown in FIG. 6 at the match points. To determine the desired phases and couplings for the slots, the values of C, in table 1 are substituted in equation (1) and the resulting amplitude and phase distribution of the electric field along the array is plotted. The resulting amplitude distribution of the electric field along the array is shown in FIG. 7 and the resulting phase distribution of the field along the array is shown in FIG. 8. These functions, as illustrated, have considerable amount of ripple, which must be averaged out in a practical embodiment as illustrated by the dashed lines in FIGS. 7 and 8. In FIGS. 7 and 8, the array length zero represents the middie of the waveguide. It will be noted that both the amplitude and phase distribution along the array is symmetrical about the center.
In order to provide the array factor pattern represented by the match point values shown in table I, the amplitude and the phase of the slots spaced along the waveguide are made to match the average curves shown in FIGS. 7 and 8. First the slot spacing of the slots along the waveguide are selected so as to conform to points on the curve shown in FIG. 8. Then the coupling of these slots is adjusted so as to provide electric field amplitudes to match points on the amplitude distribution curve shown in FIG. 7. Since the phase distribution of FIG. 8 and the amplitude distribution of FIG. 7 are symmetrical about the middle of the waveguide, the slot spacing and electric field amplitudes need only to be determined for one side as they will be identical on the other side of the waveguide.
In FIG. 9, which illustrates the slotted side of the waveguide, the reference number 35 designates the waveguide and the dumbbell slots in the waveguide, which are 13 in number, are designated by the reference numbers 41-53 respectively. As shown in FIG. 10, the waveguide is rectangular in section, the wider walls being referred to for convenience as the front wall and backwall respectively and the narrower walls being referred to as the sidewalls. The dumbbell slots are cut into the front wall. The inside dimension of the front and backwalls is selected to be 0.622 inches and the inside dimension of the sidewalls is selected to be 0.31 1 inches for the signal frequency of 15.5 gigaI-Iertz. The waveguide is fed from the right end as viewed in FIG. 9 and matched impedance load is provided at the other or left end of the waveguide.
As shown in FIG. 9, the dumbbell slots extend in the direction of the waveguide and are provided with difi'erent horizontal spacings along the waveguide in order to provide points on a predetermined phase distribution curve. The couplings of the dumbbell slots with the waveguide are adjusted to provide electric field amplitudes to match points on the amplitude distribution curve shown in FIG. 7. The coupling of a dumbbell slot with the waveguide can be adjusted by changing its distance from the centerline 55 of the waveguide. A dumbbell slot on the centerline 55 would have no coupling with the waveguide and the greater distance the slot is moved from the centerline 55, the greater will be the coupling of the slot with the waveguide. The coupling of the slots 46, 47 and 48, which are located on the centerline 55, is provided in a different manner other than by spacing them from the centerline as will be described below. The phase of the electric field radiated from each of the slots above the centerline 55 would be opposite in phase from that which would be radiated from a slot similarly positioned below the centerline 55. The slots 46, 47 and 48 are coupled in a manner to the waveguide so that the slots 46 and 48 act as if they were below the centerline and the slot 47 acts as if it were above the centerline. The reason for the different manner of coupling the slots 46, 47 and 48 to the waveguide will be explained below.
In FIG. 11, the average phase distribution curve of FIG. 8 has been replotted as a function of A, from the middle of the waveguide to one end. In FIG. 11, this curve is designated by the reference number 54. The symbol A is the wavelength within the waveguide and is equal to 1.26A, the wavelength in space. In FIG. 11, the curve designated by the reference number 57, represents the phase of the electric field that would be transmitted by a slot positioned above the centerline 55 as a function of the distance of the slot from the middle of the array, which is represented by the dividing line 60. The curve 59 represents the phase of the electric field that would be radiated by a slot positioned below the centerline 55 as a function of the distance of the slot from the dividing line 60. Note that the curves 57 and 59 indicate that for any given position along the array, a slot above the centerline 55 will radiate a signal 180 out of phase from that below the line. The points 71 through 77, at which the curves 57 and 59 cross the phase distribution curve 54, indicate the positions where slots may be located to match points on the phase distribution curve 54, the points 71 through 74 where the curve 57 crosses the phase distribution curve 54 indicating distances of the slots from the dividing line 60 for slots above the centerline and the points 75 through 77 at which the curve 59 crosses the distribution curve 54 indicating the distances of the slots positioned below the centerline from the dividing line 60. The point 71, where the curve 57 intersects the distribution curve 54, indicates that a slot above the centerline 55 at the middle of the array will match the phase distribution curve. The point 72, where the curve 57 next crosses the distribution curve 54, indicates that a slot above the centerline should be spaced from the dividing line 60 by an amount equal to 0.727%. The remaining points 73 and 74, where the curve 57 crosses the phase distribution curve 54, indicate that slots above the centerline should be spaced from the dividing line 60 by amount equal to 1.59M and 2.45%. Similarly, tlepoints 75, 76 and 77 where the curve 59 crosses the phase di.-.ribution cure 71 indicates that the slots positioned below the centerline should be spaced from the dividing line 60 by amounts equalto 0.326)\,,, 1.16) and 2.02)\,,.
Because the phase distribution curve 54 is symmetrical about its midpoint as illustrated in FIG. 7, the slots will have the same spacing from the dividing line 60 on both sides of the dividing line 60. Accordingly, the determination of the spacing from the dividing line 60 on one side as is done in FIG. 11
will provide a determination for the spacing of the slots from the dividing line on the other side. After the spacing of the slots from the dividing line 60 is determined, the amplitude of the electric field to be produced at each of the slots is then determined from the curve shown in FIG. 7. The coupling of the slots with the waveguide is chosen so as to provide electric field amplitudes to match the curve shown in FIG. 7. In this manner, the slotted waveguide is made to match points on the amplitude and phase distribution curves shown in FIG. 7 and 8.
The resulting horizontal slot spacings and electric field amplitudes at the slots for the radiated signal frequency of 15.5 gigal'lertz determined from the curves in FIGS. 7, 8 and 11 as described above, is given below in table 11.
In the column labeled Amplitude in table II, the figures are the relative amplitude of the electric field to the maximum amplitude which is arbitrarily assigned on amplitude of one as is done in the curve of FIG. 7. In the column labeled Spacing the horizontal spacing of each of the slots is given in inches from the adjacent slot nearer to the center slot. The dumbbell slots are selected to be 0.313 inches long with end poles 0.072 inches in diameter in order that the slot resonate at the radiated frequency of 15.5 gigaI-Iertz.
As pointed out above with respect to FIG. 8, the three middle slots 46, 47 and 48 lie on the centerline 55 of the waveguide and are coupled to the waveguide by a method other than by being spaced from this centerline. The slots 46, 47 and 48 are not coupled to the waveguide by being spaced from the centerline 47 because these slots must radiate substantial amplitudes as will be apparent from the curve in FIG. 7. To achieve the necessary amplitude in the radiated signal, the displacement of the slots from the centerline 55 to achieve the necessary coupling would have to be great. Accordingly, these slots would depart materially from the ideal line-radiating source which is being simulated by the waveguide array. Such wide displacement of the slots from the centerline as would be necessary to achieve the required coupling would result in unacceptable side lobes being produced particularly in planes out of the plane perpendicular to the front wall of the waveguide. For this reason, the slots 46, 47 and 48, which would require the greatest spacing from the centerline, are located on the centerline and are coupled to the waveguide in a different manner other than by being spaced from the centerline. The slots 46, 47 and 48 are each coupled to the waveguide by means of cylindrical conducting posts 81 which, as shown in FIG. 9 and in the sectional view of FIG. 10, extend all the way through the waveguide from the front wall to the backwall and are each positioned opposite a corresponding slots. The posts 81 in effect make the slots 46, 47 and 48 electrically off center in the waveguide and in this manner provide coupling of the slots to the waveguide. The amount of coupling that each of the slots 46, 47 and 48 will have with the waveguide will depend upon the position of the corresponding post 811. The nearer the post is to the centerline 55, the greater will be the coupling of the slot to the waveguide and the further the post 81 is from the centerline, the smaller the coupling of the slots will be to the waveguide. In order to prevent the post from causing a large impedance mismatch at the slot, and resulting in a phase shift at the slot, it is necessary to have the post coupled slot resonant in a transverse plane with a capaciu've reactance. As a result, the phase shift at the resonant frequency will be zero. This capacitance is provided by means of a capacitive iris, which is in the form of a plate 83, one for each post 81, within the waveguide extending across the waveguide from the post to the opposite sidewall of the waveguide. The plate 83, as shown in lFlG. E0, is positioned in contact with the backwall of the waveguide opposite the slotted front wall leaving a rectangular opening between the plate 83 and the corresponding slot. The posts 81 and plates 83 are electrically connected to the waveguide walls by soldering. The horizontal spacing of the three of the slots 46, 47 and 48 is increased to provide proper coupling. This can be done without changing the shape of the pattern which is produced if the same increment is added to each of the spaces between adjacent slots. The changing of the spacing between the slots in this manner will not change the general shape of the phase distribution curve which the slots match. The new phase distribution curve will still have a relatively steep linear middle portion and relatively flat linear end portions and it will still be symmetrical about its midpoint.
In order to keep the same pattern shape, new amplitudes for the radiated electric field must be determined from the curve of FIG. 7 for the new slot positions. The effect of adding the same increment to the horizontal spacing between the slots is to rotate the pattern toward perpendicular to the array or, in other words, toward the boresight. The final horizontal spacing between the slots is selected so that the main beam will be positioned at the desired angle to overlap with the left beam to be produced in a similar manner on the other side of the boresight to provide a crossover point at the desired level. The final spacing between the slots is given in table III below.
As will be apparent, the 0.025 inches was added to the horizontal spacing between each of the slots given in table II. Table III also gives the amplitude of the radiated electric field from each of these slots in the column labeled Relative Power Radiated." These values are taken from the curve in H6. '7 in the same manner that the amplitude values in table II are obtained.
As pointed out above, the coupling of the slots 41-45 and d9-53 depends upon their spacing from the centerline 55. The coupling is related to the spacing from the centerline in accordance with the following formula:
rrlt 'lrX in which g is the slot conductance, A is the wavelength within the waveguide, 1\ is the wavelength of the radiated signal in space, a is the inside dimension of the front and backwalls of the waveguide, b is the inside dimension of the sidewalls of the waveguide and X is the slot offset from the centerline 55. The slot conductance g is defined as the ratio of energy transmitted through the slot to the remaining energy which travels down the waveguide from the slot. First the conductance for each slot is determined and then the necessary otfset to achieve this conductance can be determined from equation (2). To determine the conductance necessary for each slot to produce the relative amplitude of radiation of each slot listed in table III, it is necessary to take into consideration the efficiency of the array. If the slots are numbered from the load end of the array, the conductance g for any given nth slot is given by the equation below:
in which 1; is the array efficiency and P,- is the total power radiated by all of the slots. Assuming an efiiciency of 50 percent, the required conductance of each slot to obtain the amplitude distribution of table III can then be determined from equations (3) and (4). The effieiency assumed is selected to a little under the maximum efficiency which can be obtained from the array for the electric field values in table III. After the conductances have been determined, the displacement of the slots 41-45 and 49-53 from the centerline is determined in accordance with the formula in equation (2). The spacing of the slots determined in this manner from the centerline 55 is given in table IV below:
To determine the proper post displacement and iris depth to achieve the required conductance for the center slots 46, 47 and 48, it was first determined how the iris depth varies with the post displacement in order to provide resonance and how the conductance varies with the post displacement when combined with the resonant iris. The results of this experimental data is illustrated in FIG. 12 and 13. FIG. 12 illustrates how the conductance of the slot varies with the post displacement when the post is combined with a resonant iris and FIG. 12 illustrates the variation in the depth of the iris with the post displacement to achieve resonance. To determine the required post displacement and iris depth for the center slots, first the required conductance of these slots is determined as explained above from equation (3) and them from these conductance values, the post displacements are obtained from the curve in FIG. 12. Having the post displacements, the iris depths are obtained from the curve in FIG. 13. The post displacement fro the slot 46 is determined to be 0.192 inch and the depth of the corresponding iris is determined to be 0.133 inch. The post displacement for the slot 47 is determined to be O.l76 inch and the corresponding iris depth is determined to be 0.146 inch. The post displacement for the slot 48 is determined to be 0.208 inch and the corresponding iris depth is determined to be 0.125 inch.
The above description has been of one waveguide array for producing the beam 13, which as shown in FIG. 3 is one of the two beams of the localizer antenna system. The other waveguide for producing the beam is designed in the same manner to simulate the beam 15 on the other side of the boresight.
The orientation of the two waveguides of the localizer system is illustrated in FIG. 14. As shown in FIG. 14, the waveguide 35 is positioned alongside of a waveguide 91, which produces the beam 15 on the opposite side of the boresight from the beam produced by the waveguide 35 as shown in FIG. 3. The pattern of slots in the waveguide 91 is a mirror image of the pattern in the waveguide 35 in order for the waveguide 91 to produce the beam 15, which is the mirror image of the beam 13. The post positions and iris depths in the middle three slots of the waveguide 91 will be the same as in the corresponding mirror image slots in the waveguide 35. The waveguides 35 and 91 are mounted in a cylindrical parabolic reflector 93 and the slotted sides or front sides of the waveguides 35 and 91 face the reflector. Radiation absorbing walls 95 and 97 extend out from each end of the parabolic reflector 93 to a point just beyond the waveguides 35 and 91. The waveguides 35 and 91 are oriented so that the loaded end of each waveguide is adjacent the absorbing wall and the waveguides are fed from the ends adjacent each other in the middle of the parabolic reflector. With the waveguides oriented in this manner, the beam 13 is produced by the waveguide 35 will extend out from the reflector on the side of the boresight on which the absorber 97 is placed and the beam 15 produced by the waveguide 91 will extend out from the reflector on the side of the boresight on which the reflector 95 is placed. The waveguides 35 and 91 are fed by means of waveguide feeds 99 and 101, respectively, which waveguide feeds extend along the walls of the waveguides 35 and 91 opposite the slotted walls and along the inside of the absorbing walls 95 and 97 and through the parabolic reflector 93. With the waveguides 35 and 91 positioned in this manner, the absorbing wall 95 will tend to cut off the undesirable side lobes produced by the waveguide 35 on the wrong side of the boresight, which side is the left side in FIG. 3, and the absorbing wall 97 will tend to cut off the undesirable side lobes produced by the waveguide 91 on the right side of the boresight as seen in FIG. 3.
FIG. 15 illustrates the actual patterns produced by the localizer antenna system illustrated in FIG. 14 with the waveguide elements 35 and 91 designed as described above. It will be noted that both of the two beams provide coverage for wide angles and provide substantial clearance over the side lobes except at angles of plus or minus 73. A linear course width of plus or minus 3.5 is provided and a null sensitivity of 1.6 decibels per degree is obtained. The pattern shown in FIG. 15 is in the plane of the antenna arrays perpendicular to the front walls of the waveguides. The same wide coverage and high clearance over side lobes is obtained through plus or minus 24 in elevation.
The synthesis procedure described for the localizer antenna system was also utilized to design the waveguide element producing the upper glide slope beam. Because vertical polarization is also required in the glide slope beams, edge cut slots, which are slots cut into the narrower sidewalls of the waveguide, are used to simulate the line source in the glide slope antenna arrays. For the upper glide slope beam, the beam shape desired in a plane extending vertically from the boresight is illustrated by the curve in FIG. 16, which plots the strength of the ideal beam as a function of the elevation angle above the boresight. The curve is more peaked than the localizer curves because the coverage required is not as great as in the localizer system.
FIGS. 17 and 18 illustrate how the edge cut slots are coupled with the waveguide forming the antenna elements of the glide slope antenna system. FIG. 17 shows a front view of the narrow side of the waveguide in which the slots are cut and FIG. 18 illustrates a view of the wide side of the waveguide with a portion thereof cut away to illustrate the slots. As shown in these figures, the slots which are designated by the reference number 111 are cut at an angle to horizontal in the waveguide, which is designated by the reference number 113 and which is oriented vertically. The slots 111 are 0.034 inch wide. The amount of coupling of the slots with the waveguide will vary with the angle of the slot to horizontal with the coupling increasing with increasing angles. Each slot would radiate a signal which was polarized perpendicular to the line of the slot. Accordingly, to achieve vertical polarization, each slot feeds into an open waveguide element 115. The open waveguide elements 115 have their wide dimension oriented horizontally and their narrow dimension oriented vertically so that out of the open ends thereof they will radiate a vertically polarized electric field. The narrow sidewalls of the elements 115 extend to the back of the waveguide 13 as is shown in FIG. 18. V-shaped conducting plates 1 16 attached to the ends of the guides 115 and extending between adjacent pairs thereof protrude out beyond the open ends of the waveguides 1.15 to reduce coupling between the waveguides 115. Alternatively, absorbing material can be placed in the spaces between the waveguides 1 15 to reduce coupling.
The angle and position of the slots are selected to produce electric fields with amplitudes and phases in a manner similar to that described with reference to the localizer antenna arrays so as to simulate a line source and approximately synthesize the desired radiated beam shape. As in the case of the i0- calizer antenna elements, it is necessary to consider the element factor in the glide slope antenna element. The element factor for the glide slope antenna is the pattern produced by the open waveguide elements 115. The pattern produced by an open waveguide element with a narrow dimension of 0.31 l inch, which is the vertical dimension of the waveguide elements 115, in the plane of the narrow dimension is illustrated in FIG. 19. After taking into account this element factor, the ideal glide slope array factor pattern to be provided by the upper glide slope antenna element is illustrated in FIG. 20. The upper glide slope array factor pattern is designated by the reference number 121. The lower glide slope array factor is represented by the curve designated by the reference number 123.
The lower glide slope beam must be made narrow to minimize ground reflection. The upper glide slope beam must be similarly steep at the crossover point to provide a linear course width. To provide the upper beam with sufficient steepness at the crossover point, a 28 wavelength array is used. With an array of this length, the array factor pattern shown in FIG. 20 can be matched at 29 points if the match point at angle zero is included. The angles of the match points are obtained from the expression sine I =N/28. N equals any integer from one to 28 and 1 is the angle at a match point. As with the localizer antenna system, the array factor pattern is matched at each of the match points with the main beam of a narrow beam pattern that would be produced with a particular linear phase shift across a line source with a dimension of 28 wavelengths. A match point at angle zero is also included in the synthesis of the array factor pattern 121 in addition to the other fourteen match points at elevation angles greater than zero. No matching of the pattern is provided at elevation angles below zero because the pattern is desired to be at a minimum at these match points. After the upper glide slope array factor pattern has been matched at the match points and the value of the electric field C,, at each of the match points is noted, the amplitude and phase distribution curves are determined from equation l The resulting amplitude and phase distribution curve for the upper glide slope is illustrated in FIG. 21. The curve 131 represents the amplitude distribution plotted from the center of the array to one end with the center of the array being designated position 0 and the curve 133 representing the corresponding phase distribution. As shown in FIG. 21, the curves 131 and 133 have considerable ripple and in a practical embodiment, this ripple must be averaged out resulting in the amplitude and phase distribution curves 135 and 137 respectively represented by dashed lines where they depart from curves 131 and 1133. The slot coupling and slot displacement from the array center in the upper glide slope array are selected to match these average distribution curves. The phase of the signal radiated from a slot at a given position will depend upon the direction in which it is rotated from horizontal. A slot rotated in one direction from horizontal will provide a phase 180 displaced from the phase provided a slot is rotated in the other direction from horizontal. By using slots rotated in both directions from horizontal, the number of points at which phase distribution curve 131 can be matched is doubled. In the resulting waveguide array, the slots will alternate in the direction of rotation from horizontal. It was found that if the slot positions were selected to match the initial linear phase rise exactly rather than following an average value in this initial portion, the ripple in the resulting output pattern would be reduced. For this reason, the curve 137 is shown to follow the initial linear portion of the curve 133 exactly.
As seen in FllG. 21, the averaged amplitude and phase distribution curves 135 an 137 for the upper glide slope antenna array have the same general shape as the curves have for the localizer antenna arrays. The amplitude distribution curve is peaked in the center and is symmetrical about a line drawn through the center of the curve. The phase distribution curve has a relatively steep linear portion in the middle and flatter portions on the end which are approximately linear. This curve like the glide slope phase distribution curve is symmetrical about its center point, only one half of the curve being shown in F1621.
First the slot spacings to match points on the average phase distribution curve 137 were determined. Then the spacing between each pair of slots was reduced by the same amount equal to 0.025 inch. As pointed out above, this changing of the spaces between adjacent slots merely rotates the direction of the the resulting pattern. In this case, with a decrease in the spacing, the rotation is away from normal thus raising the elevation angle of the main beam of the pattern. The spacing between the resulting slots was reduced because otherwise the spacing between the slots near the center of the array would have been too close to half a wavelength of the waveform within the waveguide, which would have resulted in an impedance mismatch at the slots. After the spacing between the slots was computed in this manner, the relative electric field amplitude at each of the slots to match the average amplitude distribution curve 135 was determined. The resulting spacing and electric field amplitudes determined are given in table V below:
In table V, the slots are assigned numbers 1-39, with the slot 20 being in the middle. The slot 1 is at the load end of the waveguide and the slot 39 is at the feed end of the waveguide, the slots in between being numbered consecutively.
As pointed out above, the conductance of each of the slots will vary with the angle of the slot to horizontal increasing conductance being obtained with increasing angle. FIG. 22 is a curve illustrating how the conductance varies with increasing slot angle for the waveguide with the dimensions used in the array of the present invention. This curve was obtained experimentally. In order to obtain increased conductance from a given slot angle, a special waveguide having thin walls with reduced dimensions was used. The inside dimensions were selected to be 0.477 inch by 0.311 inch. The wall thickness of the guide is 0.040 inch with the exception of the slotted wall which has a reduced thickness of 0.034 inch'in order to permit the slot depth to penetrate the sidewalls to obtain resonance. FIG. 23 is an expansion of the curve shown in FIG. 22 illustrating how the conductance of the slot varies with the angle at small angles. This curve was determined from the formula: (5)
g==g sin 0 in which g is the conductance determined from the curve in FIG. 22 at an angle of 15.
In order for the slot to be resonant, the depth of the slot must be properly selected. The proper depth to obtain resonance will vary with the angle of the slot in accordance with the curve shown in FIG. 24. The conductances needed to produce electric fields to match the amplitude distribution average curve 35 are first computed from the equations (3) and (4) above. After the conductances are determined for each of the slots, the slot angles and slot depths are determined from the curve shown in FIGS. 22, 23 and 24. Table VI gives the resulting slot depths and slot angles for each of the slots numbered numerically as in table V.
TABLE VI Slot Slot Slot Slot Slot Angle Depth Slot Angle Depth Number (Deg) (inches) Number (Deg) (inches) FIG. 25 illustrates a portion of the assembly of the upper glide slope antenna array. The waveguide 113 of the upper array is fed from the top of the array so that the pattern will be above the boresight. Two straight reflectors 141 and 143 extend out from the array in the direction in which it is radiating to form an H plane horn. The reflectors 141 and 143 are attached to the sides of the open waveguide elements 115 and bend away from them at the open ends thereof forming an angle of 13 with a plane defined perpendicular to the slotted wall of the waveguide. Accordingly, the reflectors 141 and 143 form an angle of 26 with respect to each other. The reflectors extend out 5 inches from the ends of the waveguide elements 115.
In accordance with the present invention, the above described antenna array for producing the upper glide slope beam may be used in combination with either of two antenna arrays for producing the lower glide slope beam. One lower glide slope antenna array is to be used for applications in which the boresight is at a relatively low elevation angle of 3.75". The other lower glide slope antenna array is used for applications utilizing higher boresight elevation angles. The firstmentioned lower glide slope antenna array is an antenna array designed merely to produce a narrow beam pattern designed in accordance with the principles described in Antenna Engineering Handbook by H. Jasik published by McGraw-I-Iill in 1961 starting on page 2-25. This antenna is utilized in the applications for low boresight elevation angles to reduce the problem of ground reflection. The specific antenna array comprises a waveguide like the upper glide slope antenna array having edge cut slots positioned at varying angles to obtain conductances. However, in order to obtain the narrow beam, the slots are equally spaced in accordance with the principles given in the above-mentioned Handbook.
Table VII given below gives the specific slot angles relative to horizontal for each slot in the lower glide slope array for use with relatively low boresight elevations.
The slots, which as indicated in table VII are 37 in number, are numbered consecutively from the load end of the array. In order to produce the beam below the boresight, the load is positioned at the top of the array and the array is fed from the bottom end. The spacing between the slots, as indicated above, is equal and in the specific embodiment is selected to be 0.566 inch. Each of the slots is cut to a depth of 0.61 inch. The inside dimension of the waveguide is 0.311 inch by 0.477 inch. The waveguide wall thickness is 0.040 inch with the exception of the slotted wall which is 0.034 inch.
When higher boresight elevations are desired, such as from 3.5 to 6.5", a slotted antenna which is designed in the manner similar to that for the upper glide slope beam is used. The pattern shape is synthesized in the same manner as the upper glide slope beam except that the pattern coverage is cut off at 30 to reduce ground reflectionpThe resulting waveguide has 35 slots with angles and spacings as indicated in table VIII below.
In table VIII, the slots are numbered consecutively from the load end of the array. As in the other lower glide slope array, the array is fed from the bottom with the load being placed at the top. The column labeled Slot Angle" gives the angle of the slots relative to horizontal and the column labeled Slot Positiom gives the distance that each slot is located from the center slot, which is slot 18 located at the midpoint of the array. The slot depth of each of the slots in this antenna array is 0.060 inch. The dimensions of the waveguide forming the array are the same as those of the other lower glide slope ar ray. As in the case of the localizer antenna system, the two arrays of the glide slope system radiate energy in a time shared relationship, and the glide slope arrays in turn radiate in time shared relationship with the localizer arrays.
As shown in the block diagram in FIG. 26, the 15.5 gigaI-Iertz signal to be radiated is generated by a signal source 161. The signal is fed to a switch 163, which directs the signal in sequence to the right localizer array 165, the left localizer array 166, the upper glide slope array 167, and the lower glide slope array 108. The signal modulation by which the beams are distinguished is provided by the signal source switching from one modulation to another in synchronism with the operation of the switch 163. In this manner, the necessary time sharing is provided.
The antenna system thus described provides wide coverage in both the localizer and the glide slope systems. False courses are eliminated except at extremely wide angles from the boresight without the use of additional beams. Because the glide slope system is separate from the localizer system, the two systems may be separated thus providing greater flexibility.
The principles of the present invention are not limited to localizer or glide slope systems in which just two-shaped beams are generated. In the glide slope antenna system in particular itrnay be desirable to generate three or more beams in a time shared relationship, one or more of which may be shaped in accordance with the principles of the present invention. These and many other modifications may be made to the above described specific embodiment of the invention without departing from the spirit and scope of the present invention, which is defined in the appended claims.
I. An antenna array for generating a shaped pattern for aircraft guidance during landing comprising an elongate hollow waveguide, a matched impedance load at one end of said waveguide, a wall of said waveguide having slots of substantially uniform size and shape formed therethrough and distributed along the length of said waveguide, the spacing of said slots along the length of said waveguide being arranged symmetrically about the midlength thereof and comprising a plurality of unequal spaces of selective size so that the phase of the electric field radiated from said slots will match a predetermined phase distribution curve along said waveguide, said predetermined phase distribution curve having a linear relative steep middle portion and relatively flat end portions and being symmetrical about a midpoint thereof, and coupling means disposed inside said waveguide, said coupling means including a plurality of coupling portions respectively disposed adjacent selected ones of said slots, said coupling portions being arranged symmetrically about the midlength thereof so that the amplitude of the electric field radiated from said slots matches a predetermined amplitude distribution curve which is peaked in its center and which is symmetrical about a line passing through said center.
2. An antenna array as recited in claim 1 wherein there is provided means to polarize the electric field radiated by each of said slots in the same direction.
3. An antenna array as recited in claim 2 wherein said means to polarize the field radiated from each of said slots comprises an open ended rectangular waveguide element for each slot positioned to receive in one end the energy passing through such slot and to radiate the energy from its opposite open end.
4. An antenna array as recited in claim 3 wherein means are provided to reduce the mutual coupling between the open ends of said waveguide elements.
5. An antenna array as recited in claim 4 wherein said means to reduce mutual coupling comprises V-shaped conducting plates extending between said rectangular waveguide elements and protruding out beyond the open ends of said waveguide elements.
6. An antenna array as recited in claim 3 wherein reflecting plates flare out at angles from the open ends of said waveguide elements to define a horn into which said open ends radiate.
7. An antenna array as recited in claim ll wherein said predetermined amplitude distribution curve has the shape of the dashed line shown in FIG. 7 and wherein said predetermined phase distribution curve has the shape of the dashed line shown in FIG. 8.
8. An antenna array as recited in claim ll wherein said slots include a first set of said slots being coupled to said waveguide in a first manner and a second set of said slots being coupled to said waveguide in a second manner.
9. An antenna array for generating a shaped beam comprising, a waveguide provided with a series of dumbbell slots positioned along the centerline of a broad face thereof said series being composed of first and second sets of slots, said first set of slots being positioned in the middle of said waveguide with respect to the longitudinal dimension thereof, the slots of said first set being located on the centerline of said waveguide and being coupled thereto by a conductive post positioned opposite each slot of said first set perpendicular to the broad face of said waveguide, a conductive plate mounted within said waveguide opposite each slot of said first set to provide capacitive impedance at the slots to resonate with the inductive impedance provided by the corresponding post, said second set of slots being positioned symmetrically on each side of said first set as respects the longitudinal dimension of said waveguide, said second set being coupled to said waveguide by being ofi'set laterally from the centerline thereof.
10. An aircraft guidance antenna system comprising a first waveguide provided with a series of dumbbell slots positioned along the centerline of a broad face thereof, said series being composed of a first set of slots positioned in the middle of said first waveguide with respect to the longitudinal dimension thereof, said first set of slots being located on the centerline of said first waveguide and coupled thereto by a conductive post positioned opposite each slot of said first set perpendicular to the broad face of said first waveguide, a conductive late mounted wlthln said first waveguide opposite each slot 0 said first set to provide a capacitive impedance at the slots to resonate with the inductive impedance provided by the corresponding post, a second set of slots positioned symmetrically on each side of said first set as respects the longitudinal dimension of said first waveguide, said second set being coupled to said first waveguide by being offset laterally from the center line thereof, a second waveguide provided with slot configuration and coupling identical to said first waveguide positioned collinearly with respect to said first waveguide, a cylindrical parabolic reflector positioned opposite said slots with its axis parallel to the longitudinal dimensions of said waveguide sections, and walls of absorbing material extending out from each end of said reflector past the remote ends of said first and second waveguides.