US2814782A - Waveguide switch - Google Patents

Description

Nov. 26, 1957 J, F. ZALESKI 2 WAVEGUIDE SWITCH Filed Aug. 6, 1954 2 Sheets-Sheet 1 INVENTOR. JOHN E ZALESK/ ATTORNEY 2 Sheets-Sheet 2 Filed Aug. 6, 1954 DEL-4% 55 l e3 m INVENTOR JOHN E ZALESK/ ATTORNEY United States Patent WAVEGUIDE SWITCH John F. Zaleski, Valhalla, N. Y., assignor to General Precision Laboratory Incorporated, a corporation of New York Application August 6, 1954, Serial No. 448,250

2 Claims. (Cl; 333-7) This invention relates to a waveguide switch for directing microwave energy from an input waveguide into either one of two output waveguides.

Waveguide switches are, in general, analogous to conductive switches such as double-pole double-throw switches, which are familiar in circuits carrying low frequency energy, in that both types of switch serve to direct electrical energy from a source or input channel into a selected one of two or more output channels. However, the structure of a waveguide switch is radically different from that of a conductive switch since the transmission path, instead of comprising two discrete conductors, comprises a dielectric medium bounded by a conductive material.

A waveguide switch should have a number of desirable features, such as low voltage standing wave ratio over a wide band both while at rest and during switching, low inertia of the moving elements, and good isolation between output channels. A number of types of switches have been proposed in the past, many of them employing a resonant element arranged to be excited by the input energy and movable so as to radiate the energy into one or another output channel. Such switches have not been entirely satisfactory because the resonant element must have narrow band characteristics in order to be efi'icient and because a considerable portion of the input energy leaks into the nominally closed output channel.

It-is an object of this invention to construct a-waveguide switch which provides good'isolation between output channels and which has broad band characteristics.

It is another object of this invention to construct a waveguide switch which exhibits a low voltage standing wave ratio (VSWR) in the input channel both when passing energy to one or the other output channel and during switching intervals.

Another object is to construct a waveguide switchin which the movable elements have a low moment of inertia.

In accordance with the invention, the switch is constructed around a four armed ninety degree E-plane'junc tion. The corners within the junction are cut away to form concave cylindrical surfaces. The switching. element is a conductive plate, the central plane ofwhich extends diagonally across the interior of the junction so as to form a ninety degree double mitered: E-plane bend between the input arm and one output arm. The radius of curvature of the cylindricalsurfaces and the thickness of the conductive plate are selected so that the mean microwave path length between discontinuities is aquarter of a guide wavelength. Switching is: accomplished by rotating the plate through ninety degrees so as to; form an E-plane bend between the input arm and the other output arm. Isolation between ouput arms is attained by providing the plate with chokes cut into'opposite edges. The arm opposite the input arm may be closed by a shorting plate the position of which is adjusted for minimum VSWR during switching.

For a clearer understanding of the inventiomreference may be made to the following,detaileddescriptionandthe accompanying drawing, in which:

ice

Figures 1 and 2 are diagrams useful in explaining the invention;

Figure 3' is an isometric view of a switch according to the invention;

Figure 4 is a sectional view taken along the plane 4-4 of Fig. 3;

Figure 5 is a sectional view taken along the plane 55 of Fig. 3;

Figure 6 is an elevational view of the switching element.

Figure 7 is a plan view of the switching element; and

Figure 8 is a schematic diagram indicating one use of the switch.

Referring now to the drawing, Fig. 1 shows how a switch in accordance with the invention may be designed. The lines 11 and 12 are first drawn to represent the inside surfaces of a waveguide, spaced apart a distance equal to that t f the narrow dimension. The line 13 is next drawn perpendicular to lines 11 and 12, intersecting them at points 14 and 15, and represents the plane of a discontinuity encountered by microwave energy flowing toward the switch. At the intersection 16 of the line 13 with the axis of the waveguide, a line 17 is drawn making an angle of 45 with the line 13. A distance equal to a quarter guide wavelength is laid off on line 17 which determines the point 18, through whichpoint a line 19 is drawn perpendicular to line 13. The lines 21 and 22 are next drawn perpendicular to line 19 at points 23 and 24 equidistant from point 18 to represent the inside surfaces of a second waveguide. A line 25 joins points 14 and 23 while a line 26 joins points 15 and 24.

It can be seen that the lines 25 and 26, together with waveguides 11, 12 and'21, 22 comprise a E-plane bend. Microwave energy propagated around the bend encounters a first discontinuity in the plane of the line 13 and a second in the plane of the line 19. The mean length of path between these discontinuities, thatis, the distance between points 16-and 18, is a quarter. of the guidewavelength, so that thereflections caused by thetwo discontinuities substantially cancel each other.

Fig. 2 shows how the basic structure of Fig. 1 can be extended to form a switch. The axes of the two guides 11, 12v and 21, 22 are extended to intersect at point 28 todeterminethe center of a circle'29 which passes through points 1 4, 15, 23, and 24. The are 31 of this circle between points 14 and 23forms one corner of the bend-instead of thechord 25.but the resulting change in mean path length' is insignificant. The switch element is defined bytheline 26, another line 32 parallel thereto equidistant. from the center 28, and. the two arcs 33 and34, so that its thickness is equal to the length of the chord of the arc 33. It is obvious from the drawing that when the switch element. is rotated about the'center Z-Sthmugh 90, asecond 90 E-plane bendwill be formed.

Referring now to Figs. 3,. 4 and, 5, there is shown a cubical block 41 of.conductive material such as brass. Two mutually perpendicular rectangular apertures are formed in the block each extending completely therethrough and, as best. shown in Fig. 4, defining an input Waveguide 42, to output waveguidesv 43 and 44 and fourth waveguide 45; A. circular aperture extends through the block perpendicular to both of the rectangular apertures.

As shown in. Fig. 4, the circular aperture cuts off. the corners of the four waveguides, forming concave cylindrical surfaces 46, 47,48 and49. As best shown in Fig. 5, the circular aperture is closed at the top and bottom of the block by means of annular members 51 and 52 at'the top and 53 and 54 at the-bottom which are arranged in the conventional manner to form choke joints. A generally rectangular conductive plate 55 is arranged within the space formed by the junctionof. all, of the aper- ,turesand is secured to a shaft 56 which lies along the axis of the. circular aperture. The shaft 56 extendsthrough 3. the choke joints and is provided with a knob 57 so that the plate 55 may be rotated from without the block 41.

Figs. 6 and 7 show the plate 55 in more detail. Its longer dimension, 1, is nearly equal to the longer dimension of the waveguide, it being, however, sufliciently shorter to provide mechanical clearance so as to prevent intermittent contact and consequent R. F. noise. As best shown in Fig. 5, there is left a space 58 at the top between the plate 55 and the annular members 51 and 52 and another space 59 at the bottom between the plate 55 and the annular members 53 and 54. For the same reason, the shorter dimension, 2, of plate 55 is sufficiently less than the diameter of the circular aperture to provide mechanical clearance. As best shown in Fig. 4, this results in small spaces 61 and 62 being left between the plate 55 and the block 41.

As best shown in Fig. 7, the plate 55 has two slots 63 and 64 formed in opposite edges. The width of these slots is not critical, but the depth should be such as to form a transmission path which is antiresonant at the frequencies normally propagated by the switch. This antiresonant depth may be obtained by filling the slots with a low-loss dielectric material, such as a tetrafluoroethylene resin, and making the depth equal to one quarter of the wavelength of the microwave energy within the slot.

When the switch is used simply to divert the input energy from arm 42 to one of the output arms 43 or 44, the fourth arm 45 may be closed by means of second conductive plate 66, the position of which may be adjusted by means of a rod 67.

With the position of the plate 55 as shown in Fig. 4, let it be assumed that microwave energy is flowing through the waveguide 42 toward the center of the structure. It is obvious that this energy will flow around the corner and enter waveguide 43. As explained in connection with Figs. 1 and 2, the mean path length, d, is made equal to one-fourth of a guide wavelength by properly selecting the radius, r, of the circular aperture. Therefore, reflections caused by the two discontinuities will substantially cancel each other. The spaces 58 and 59 (Fig. themselves constitute waveguides, but since the shorter dimension of these waveguides is perpendicular to the E vector of the incident microwave energy, the cut off wavelength of these waveguides is far below the band of wavelengths normally propagated through the switch. Therefore, substantially no energy leaks through these spaces into the waveguide 44. The spaces 61 and 62 (Fig. 4) also constitute waveguides, but the shorter dimension is parallel to the E-vector of the incident energy and the longer dimension is equal to the wide dimension used throughout the switch. Therefore, the spaces 61 and 62 constitute waveguides the dimensions and orientation of which render them capable of propagating the incident energy. Indeed, in the absence of slots 63 and 64 (Fig. 7), substantial amounts of energy would leak into the output guide 44. However, these slots constitute antiresonant waveguide transmission paths, since they are one-quarter of a wavelength long. Energy leaking into the spaces 61 and 62 is propagated into the slots 63 and 64, travels to the end of the slots, and is reflected back to the point of entrance 180 degrees out of phase with the incident leakage energy, thus effectively preventing energy from being propagated from the waveguide 42 into either of the waveguides 44 or 45.

When it is desired to switch the energy into waveguide 44 instead of waveguide 43, the switch should be rotated counterclockwise, as viewed in Fig. 4, in order to avoid presenting a large reflective surface to the input guide 42, perpendicular to the direction of propagation. Such a surface would cause a very high voltage standing wave ratio in the input guide 42 during switching. Assuming the switch is rotated in the proper direction, the plate 66 may be adjusted to secure the lowest VSWR during the switching interval.

The plate 66 may be omitted, and, in certain usesof the switch, its omission is necessary. For example, as

shown in Fig. 8, the switch may be used to change the phase of the microwave energy in output guide 43. For such a use the plate 66 is omitted and the waveguides 44 and 45 are joined by a waveguide 72 of a selected length. When the switch is in the position shown, the energy in the output guide 43 will have a certain phase. With the switch in the opposite position, the energy in guide 43 will be retarded by an amount depending upon the length of the waveguide 72.

As a specific example, suppose it is required to design a switch for use at 9,000 me. The rectangular apertures may then be 0.4 x 0.9". The various dimensions shown in the drawing will then be as follows:

The depth, m, of the slots will depend upon the dielectric used. If a tetrafluoroethylene resin, such as that sold under the trademark Teflon is used, having a dielectric constant of approximately 2.10, then the depth, in, will be approximately .260.

It is apparent that a waveguide switch constructed in accordance with the invention has a number of desirable features. The switch is as broad band as a double mitered bend, since no resonant elements are required. Switching may be accomplished rapidly and with little applied power, because the only moving elements are the light weight element 55 and its supporting shaft. The quarter wave slots assure excellent isolation between output channels and low power loss. The position of the shorting plate 66 has no effect when the switch is at rest and therefore may be adjusted for the lowest VSWR during the switching interval.

Although a specific embodiment has been described, many modifications may be made within the scope of the invention. Construction of the switch from a solid block of conductive material is not essential. The chokes around the operating shaft may be modified. Many other modifications will occur to those skilled in the art.

What is claimed is:

1. A waveguide switch comprising, first and second rectangular waveguides joined to form a first double mitered ninety degree E-plane bend, the mean distance between miters being approximately equal to one quarter of a guide wavelength, a third rectangular waveguide collinear with said second waveguide and extending in the opposite direction from said bend, means for rotating the outer section of said bend about its midpoint whereby when rotated through ninety degrees a second bend comprising said first and third waveguides and said section is formed, and dielectric filled slots one quarter wavelength deep formed in opposite edges of said outer section.

2. A waveguide switch comprising four rectangular transmission paths bounded by conductive material the axes of which intersect in a common point to form a four arm E-plane junction, the interior corners of said junction being cut away along the surface of a cylinder the axis of which passes through said common point perpendicular to the axes of said transmission paths, a generally rectangular conductive plate within said junction the faces of which are parallel to said cylindrical axis and which is pivoted about said cylindrical axis, the thickness of said plate being selected so that opposite faces may lie in the planes joining the ends of opposite cylindrical surfaces, whereby when in this position said plate and the walls of two adjacent transmission paths define a double mitered ninety degree E-plane bend, the dimensions being selected so that the mean distance between miters is approximately one quarter of the wavelength within said transmission paths, and two dielectric filled slots one quarter wavelength deep formed in those opposite edges of said plate which are parallel to said cylindrical axis.

References Cited in the file of this patent UNITED STATES PATENTS 2,351,895 Allerding June 20, 1944 2,705,776 Starr et a1. Apr. 5, 1955 2,743,422 Muchmore Apr. 24, 1956 6 FOREIGN PATENTS Great Britain June 9, 1954 OTHER REFERENCES Microwave Transmission Circuits, Ragan, M. I. T. Radiation Laboratory Series, vol. 9, McGraw-Hill Book Co., copyright 1948, pages 203 and 204. (Copy in Div. 69.)

Images