GB2149223A - Phase switches - Google Patents

Abstract

An input and an output are connected by selectable transmission lines all of equal pathlength (i.e. as measured in wavelengths) and each comprising an E-plane rotating element all rotators being of equal pathlength but differing in rotation capabilities. As shown, an input and an output are connected by a pair of finlines either of which may be selected by means of the switches 23. The lines are of equal length and contain rotators 14B, 14T of equal length. Rotator 14T is constructed to provide a crossover of the line elements and achieves a rotation of 180 DEG . Rotator 14B is of symmetrical construction and rotates the incident wave first through + 90 DEG then through - 90 DEG . It is apparent that the lengths of the rotators, expressed in terms of wavelengths, are equal. The phase-shifts of the live paths differ by 180 DEG . <IMAGE>

Description

SPECIFICATION Phase switches This invention concerns phase switches, and relates in particular to wideband phase switches useful in microwave circuitry.

Microwave transmission lines-that is to say, electrically conductive lines carrying alternating current electromagnetic energy with a wavelength in the region of 0.3m and smaller (and thus a frequency in the neighbourhood of 1 GigaHertz and greater3 are now commonplace in all sorts of communications systems.These lines may take the physical form of a conventional waveguide (a tubular-hol- low elongate---conductor of an appropriate cross-section), a stripline (currently in the form known as microstrip, having a conductive strip parallel to and spaced from a conductive ground plane), a slotline (a slot in a conductive sheet), or one of the newer finlines (a conventional waveguide within which is mounted a length of dielectric material supporting spaced conductive fins extending along its length and inwardly of opposed sides of the guide).

In the use of such microwave transmission lines it may be necessary to insert a switch to change the phase of the transmitted wave energy. One example when such a device is required is in communications systems using Bi-polar or Quadrature Phase Shift Keying modulation (in which the energy wave phase is shifted by + 90 and/or + 180 to denote information carried by the wave); another is in systems employing phased array antennas. Microwave phase switches as presently realised are one of two types; the reflection type (often based upon a circulator or a 3 dB coupier) and the transmission type (often using the switched line principle).Unfortunately, though many designs of these types are available, they all suffer to a greater or lesser extent from relatively narrow operational bandwidths and degradation in amplitude and phase balance. The present invention suggests a new type of phase switch, not suffering to any significant extent from these drawbacks, in which a matched pair of lines (one or other of which may be selected for input/output) contains two matched devices for rotating the E-plane one in each line; the device in one line preferably rotates the Eplane to the full extent required to achieve the desired phase shift, while that in the other line preferably rotates it half way and then rotates it back again (to achieve zero rotation but with all the other effects matched).

In one aspect, therefore, the invention provides a microwave phase switching device comprising: an input line, an output line, and selectively connectable therebetween a plurality of equal pathlength intermediate lines, wherein the intermediate lines include, one in each, a like plurality of equal pathlength E-planerotating elements the rotating capabilities of any pair of which differ by an amount equivalent to a desired phase change.

Before discussing in detail the phase switch of the invention, it may be useful to comment upon the operation of such a switch-and, for convenience, one containing two only intermediate lines as follows: The device comprises a common input line feeding two intermediate lines both of which feed in turn a common output line, and an input signal may at will be routed to one or other intemediate line and thence to the output line. The intemediate lines are of equal path length so that their treatment of an input signal should (apart from the induced E-plane rotation) be equal regardless of the signal's frequency.Within (and foming a part of) each intemediate line is an E-plane rotator; these two rotators are of equal path-length, so that their treatment of an input signal should (again, apart from the E-plane rotation) be equal regardless of the signal's frequency.

Each E-plane rotator rotates by a predetermined amount (which may include zero) the signal's E-plane, and the two rotations are so predetemined that the difference between the two is equivalent to the required phase change. The consequence of the two equal pathlength E-plane rotators one each in the two equal pathlength intermediate lines is that regardless of the signal's frequency the phase of the signal when transmitted via one intermediate line differs from that when transmitted via the other by an amount equivalent only to the difference in induced Eplane rotation.In theory, then, the operation of the phase switch device is essentially independent of signal frequency-though, as pointed out hereinafter, this independence will only be achieved in practice if each rotator is of a suitable length, and if the selection of either intermediate line is itself effected in a frequency-independent manner.

The inventive device includes an input line, an output line and a plurality of intermediate lines. Each line may be of any physical form appropriate, and indeed they may be of different forms (provided only that the path length of the intermediate lines is the same), though this would not normally be very convenient.

Typical types of line are conventional tubular waveguides, or striplines, slotlines, or finlines.

Examples employing the latter three are discussed hereinafter with reference to the accompanying Drawings.

Naturally, the input and output lines are suitably connectable to the remainder of the apparatus in which the phase switch is to be used. This can be done in any suitable way, and need not be discussed further herein.

The device of the invention includes a plurality of intermediate lines, and each line includes a rotator. "Plurality" is used, in its normal sense, to mean "two or more", and the most basic-and presently preferred form of the device has two only intermediate lines.

There may, however, be three or more, and a device with four, one being used as a "reference" line and the other three being paired up with it to give three possible different phase changes(+ 90", + 180 and + 270o, say), is particularly attractive. Nevertheless, for simplicity's sake this Specification discusses the invention mainly in terms of devices with two only intermediate lines.

The intermediate lines are selectively connectable to the input/output lines-in other words, any one may at will be connected between the input and output lines, so that a signal passing through the device may be routed via that one intermediate line. The selection is achieved by having one or more electrical switches associated with the junction of the input and intermediate lines (and the junction of the intermediate and output lines) such that the signal may be switched to (and from) the chosen intermediate line as required. The rate at which the phase switch is required in use to change a signal's phase can be quite high switching frequencies of up to 140 Mb/s are not uncommon--so that the switches employed need be fast acting.Semiconductor switches are the class preferred for this operation, typical types being field effect transistors (FETs) and positive-intrinsic-negative (PIN) or Schottky barrier devices. What the or each switch actually does, however, depends upon the nature of the lines. Using finlines, for example, the switches may be conveniently placed across the fins as though to "block" the intemediate lines; making all but one conductive (while the one is nonconductive) effectively joins the relevant fin sections and forces the signal to pass via the one. Where the lines are striplines however, the switches may be conveniently placed to "bridge" the gap to the intemediate lines; making one conductive (and each other nonconductive) joins the relevant stripline sections and forces the signal to pass that way rather than any other.These two examples are discussed further hereinafter with reference to the accompanying Drawings.

At this juncture it is necessary to point out that the available switches are not perfect. In their conducting state they still possess a finite residual resistance, while in their nonconducting state they exhibit a small but significant capacitance. Other parasitic elements may also be present. As a result, the switches themselves detract from the overall frequency independence of the device, and having regard to the exact nature of the components it may be necessary to allow for this in the basic design.

The intermediate lines forming part of the phase switch device are of equal (electrical) pathlength that is to say, so far as concerns an electrical signal passing therealong they are, or appear to be, of the same length.

Thus, apart from any effect caused by the Eplane rotators included within each line, the phase of a signal is not in any way changed dependent upon which line it is passed along.

In accordance with what might be said to be the inventive feature, the phase switch device includes, as part of each intemediate line, an E-plane-rotating element, these elements are of equal pathlength, and the Eplane-rotating capabilities of any two elements differ by an amount equivalent to a desired change (or switch) in phase of the signal passing through the device. It is convenient at this point to discuss the concept of the Eplane.

Electromagnetic wave energy travels through space (or any dielectric medium) as an electrical field, pointing transverse to the direction of travel, associated with a magnetic field, similarly pointing transverse to the direction of travel and at right angles to the electric field. Both fields vary sinusoidally in amplitude. When constrained to travel between two spaced conductive surfaces (as in a waveguide), the electric field points from one surface to the other exactly like the field between the plates of a capacitor, and the plane parallel to the electric field direction, extending in the direction of travel and positioned laterally in the strongest field (usually half way across) is the E-plane.If at any single instant a side view is taken of the energy progressing along between the two surfaces, and the electric field is represented vectorialy by an arrowheaded line lying across the direction of travel, pointing in the direction the field points and with a length indicating the field amplitude, there may then be visualised a sequence of arrowed lines (vectors) that sinusoidally increase in length to a maximum in one direction, decrease in length to zero, increase in length to a maximum in the opposite direction, decrease in length to zero, and repeat the cycle again and again over the length of the two surfaces.Each cycle represents the whole of one wave of the energy (the length of the cycle is the wavelength, the number of times it repeats in a second the frequency) and the "phase" of the energy at any point along the direction of travel is detemined by the wave's position in this cycle. The "zero amplitude and increasing" position is zero degrees phase, the "maximum positive amplitude position" is the 90' (or -4 radians) phase point, and the "zero amplitude and decreasing" position and the "maximum negative amplitude" position are respectively 180' ( 7r radians) and 270 ( 3'r/2 radians) phase points.

The 360 450 , 540 and 630 (and so on) phase points are repetitions of the 0 , 90 , 180 and 270 points, and need no further consideration here.

It will be apparent that if electromagnetic wave energy is fed to a guide transmission line the relationship of the wave's phase at input to that at output at any one instant depends upon the length of the waves (the wavelength) and the length of the lint thus on how many waves "fit" into the path (the pathlength). Equally, if there are two such lines then any difference in phase between a wave passed along one line and the same wave passed along the other will be dependent both on the wavelength and on the difference in the two line's pathlengths and only if the two lines are of equal pathlength is there no such phase difference. However, if one of the two lines incorporates some means for rotating the E-plane, then there will be a phase difference despite the equal pathlengths.For example, if in one line the Eplane is rotated by 180 (sir/2) the relevant wave will arrive at the line end 180' advanced (or retarded; they are the same) in phase compared to an identical wave in the other line, and so there will be a phase difference of 1 80.

The phase switch device of the invention uses a plurality of intermediate lines of equal pathlength (so that, other things being equal, there will be no phase difference between wave energy passed along one and identical wave energy passed along any other), and in each line there is included an E-plane-rotating element. The resulting phase difference as between one line and each other is dependent upon the difference between the rotations effected by each rotator.

Just as the intermediate lines are of equal pathlength, in order that there should be no difference in phase, regardless of wavelength, except that caused by the E-plane rotators, so the rotators themselves must be of equal path length in order that the effect they have is only due to the E-plane rotation they cause and not in any way pathlength, and thus wavelength, dependent. If this constraint were to be ignored it might be possible to have a single rotator, in one intermediate line only (the other line not including any rotator).

However, the rotators at presently envisaged all have a finite pathlength that makes their "output" wavelength dependent; in order to allow for this there must be a rotator in each line, and they must be of equal pathlength.

The E-plane-rotating elements-the rotators-may be of any type and take any physical form, though conveniently they will be of the same general type. Perhaps the most simple type would be a length of tubular waveguide (of flattened rectangular cross-section, the E-plane extending between the two closest opposed guide walls for the dominant mode) twisted about its axis. A 180 twist would make what had been the guide "upper" surface into its "lower" surface, and any wave progressing through this twisted guide would have its E-plane rotated similarly. Naturally, other degrees of twist would rotate the E-plane by other (appropriate) amounts.

A different form of rotator involves the use of finline (and would normally be used with a finline guide system). A finline is a waveguide in the form of a rectangular cross-section tube but with a pair or pairs of conducting fins supported by a dielectric board situated edgeto-edge parallel to and in, or adjacent to, the E-plane and projecting towards each other from the two opposed guide walls. The electric field extends across the gap between the fins, and for reasons that need not be explained here the arrangement significantly relaxes the tolerances in the manufacture of the tubular guide.In the usual, unilateral finline the fins are coplanar with the E-plane; however, in the antipodal variation required to make a rotator for use in the invention the fins are parallel but spaced one either side of the E-plane (and commonly the two sets of fins are supported on either side of a single piece of dielectric board). Moreover, while in most usual finlines the fins extend, all along their length, from adjacent the walls of the tubular guide from which they project, in the variation required for the rotator the fins are so adjacent only at their ends, and between their ends they are spaced from the walls. Accordingly, in any area where, because of their projection into the tubular guide, the fins overlap so as to be face-to-face, they form a balanced line (the two opposed surfaces of a guide within a guide).Nevertheless, the electric field still extends across the gap between the two fins, and if the fins are so dimensioned as to "approach" each other as one moves along the finline segment, while becoming separate from their base portions, so the field and thus the E-plane itself---rotates in order constantly to point along the shortest line between the two.At some time the two fins are well spaced from their supporting walls and positioned side-by-side, the electrical field extending between them (and normal to each fin), and so the field-and the Eplane-has been twisted through 90'; eventually, if each fin approaches and becomes adjacent the opposite guide wall, the field-and thus the E-planc svill have been twisted through 1 80', pointing in exactly the opposite direction to that in which it pointed at the start of the finline segment.And so the segment-the rotator--has rotated the Eplane by 180 , and any wave energy travelling along any intemediate line with such a rotator within it will necessarily exit from that line switched in phase by 180 compared with the wave exiting from a line having the same overall pathlength but without such a rotator.

Of course, though the rotation of the E plane is not in any way dependent upon the wavelength of the transmitted wave energy (although, as discussed hereinafter, this type of rotator may require a certain minimum length that is wavelength dependent), the characteristics of the finline segment constituting the rotator may well cause other effects that are wavelength dependent. Accordingly, it is a requirement of this embodiment of the invention that each intermediate line contain a finline rotator, and that it is the relative rotation of any two that gives the required absolute rotation and thus phase switching-of the device.Obviously, the relative rotation could be achieved with a number of different pairs of absolute rotations; one pair might be + 90 and - 90" (giving a relative 180 ), or + 270 and + 90' (again a relative 180 ), but it is generally preferred at present so to engineer the rotators that all but one of them each provides the full rotation needed and the one provides overall a zero rotation that is, however it may treat the wave energy between its ends the net effect is no rotation at all.With a finline rotator this zero rotation is easily accomplished simply by arranging that each fin be adjacent the same guide wall at both ends thus, that instead of each fin extending, from the balanced line point, to the opposite wall it returns to the wall from which it started. Provided it has the same length, such a finline rotator segment should have exactly the same wavelength-dependent characteristics as the previously-described 180 rotator, so that when used therewith, in the other of the paired intermediate lines, it will ensure that the phase switching effect of the device is not depenent upon the wavelength of the transmitted wave energy.

Finline rotator devices are described in more detail hereinafter with reference to the accompanying Drawings.

A third form of rotator involves the combined use of striplines and slotlines, and isas might be appreciated-apprnpriate for employment in circuits which are themselves in stripline or slotline form. A stripline is a waveguide in which the guiding surface is a relatively narrow strip of conductive material parallel to and insulatively spaced between two relatively much wider conductive strips (so wide, in fact, that each is usually a sheet, and is referred to as a ground plane). In the version known as microstrip there is only one of these ground planes. Wave energy can travel along the strip, located between the strip and the (or each) ground plane. A slotline is a narrow but lengthy aperture, or slot, in a relatively much wider strip of conductive material (again, so wide that it is usually a sheet).Wave energy can travel along the slot, located between its edges. It is not uncommon to have both striplines and slotlines used together, and indeed to have the slotline formed as a slot in the ground plane of the stripline. In such a case, wave energy can be transferred from one line to the other by having the two line's ends crossing at right angles (with a degree of overlap appropriate to the wavelength), and whenever such a transfer occurs the electrical field direction (and the E-plane) is turned through the 904 between a line joining the sides of the slot and the line joining the strip to the ground plane.The actual field direction is that required to keep the field pointing at, or away from, the same conductive surface in the wave propagation direction, and this means that in the two strip portions at a junction either side of the slot the electric field (and the E-plane) is in opposite directions (that field in the non-propagation direction is then cancelled by the opposite phase field reflected off the strip end).From this it follows that if two strips are placed so as to overlap side-by-side, and if the two are then "joined" by a slot extending directly across between them, the electric field of the wave energy transferred from one strip to the other via the slot will undergo a 180 change in direction; 90' as it passes from one strip into the slot, and a further 90 , rotating in the same sense as the first, as it passes from the slot to the second strip. Clearly, such a strip/slot/strip arrangement is an E-plane rotator, rotating the Eplane through 180', and a transmission line containing such a rotator would cause wave energy transmitted there along to reach its output end 180 out of phase with the same energy transmitted along a line having the same overall pathlength but without the rotator.

Equally, it follows that if two overlapping slotlines are joined by a crossing stripline then energy transferring from one slotline to the other in the stripline will undergo a 180 phase change. For simplicity the following discussion relates to two striplines joined by a single slotline.

The transfer of energy between strip and slot is not truely wavelength independent, and the design of each junction needs careful attention (as regards the amount of strip overlap and the shape and size of the slot end), due regard being paid to the wavelength or waveband of the energy to be transferred.

However, the transfer is even so somewhat indeterminate, and it is therefore necessary that in a phase switch device using such a strip/slot line rotator there should be a matching rotator (particularly matching as regards pathlength) in each of the intermediate lines, the E-plane rotational effect of one being different to that of each other while the wavelength dependent effects of one are the same as, and thus balanced by, those of each other.

The overlapping strip/slot/strip line junction just described causes a 180 change in the Eplane; a zero change can be accomplished by the comparatively simple expedient of curving one or other of the slot or a strip through 180 in the E-plane itself. Thus, in one case the slot is curved through 180 and then so positioned relative to the two strips that it extends not from the side of one to the facing side of the other but instead from the side of one to the same side of the other. This is perhaps best visualised by imaging the two strips are aligned (as though a single strip with a break), and the slot is U-shaped, crossing one strip in one direction and the other in the opposite direction.In the other case, one of the striplines is mostly out of the plane of the other, but is curved around in its E-plane (i.e., a plane nomal to the plane of the strip itself) into a hook shape the free end of which lies in the plane of, and alongside, the other strip; the slot then crosses directly between the ends of the two strips, as before (in the 180 rotation case discussed above), but because of the 180 rotation of the one strip the net result is a zero rotation.

It will be apparent, then, that a phase switch device of the invention having in one intermediate line a strip/slot/strip junction wherein each strip and slot itself has no overall rotation in the E-plane (or where the sum of all such rotations is an even multiple of 180 !) and having in another intermediate line a strip/slot/strip junction wherein one of the strips has a 180 rotation in the E-plane (or where the sum of all such rotations is an odd multiple of 180 !) can change the phase of the transmitted wave energy by 180 . This change should, over a limited wavelength range (about an octave), be effectively independent of the wavelength.A phase switch using two rotators each with two coplanar striplines, one rotator with a straight slot and the other with a 180 turned slot, is described in more detail hereinafter with reference to the accompanying Drawings.

The two presently-preferred major types of E-plane rotating element are the finline type, and the strip/ slotline junction type, in each case specifically that variety wherein there are two only intermediate lines and one rotator provides a 180 rotation while the other provides overall a zero rotation.

As has been implied hereinbefore, the actual length required for any rotator depends to a significant extent upon the physical nature of the rotator. One in the form of a conventional tubular waveguide, or of a finline, for example, needs a certain minimum pathlength in order to ensure that the rotation can be accomplished without mishap--specifi- cally, without the electric field experiencing a severe discontinuity as it progresses therealong. A typical minimum length is of the order of four wavelengths (at the lowest operating frequency).In a strip/slotline rotator, on the other hand, the actual transition from strip to slot and vice versa is predetermined but,nevertheless the rotator should have a length less than some maximum value (so as to reduce the effect of the interaction between the two transitions) but more than some minimum value (so as to prevent coupling between the rotator's input and output). It is not easy to recommend typical lengths for a strip/slot line rotator, and some fairly simple experimentation may be necessary, in any particular set of circumstances, to find a satisfactory length.

Circuits using a phase switching device of the invention may, of course, use one or more such device, in series or in parallel, to achieve the required degree and complexity of phase changing.

Various embodiments of the invention are now described, though only by way of illustration, with reference to the accompanying drawings in which Figure 1 is a schematic circuit representation of a 180 phase switch device according to the invention; Figure 2 is a diagrammatic representation of a finline realisation of the circuit of Figure 1; Figure 3 shows a series of cross-sectional views of the finline circuit realisation of Figure 2; Figure 4 is a diagrammatic view (partly in see-through) of the E-plane 180 rotator used in the finline embodiment of Figure 2; Figure 5 is a diagrammatic representation of a microstrip/slotline realisation of the circuit of Figure 1; and Figure 6 shows diagrammatically two physical arrangements of microstrip/slot line embodiment like the 180 rotator of Figure 5.

The circuit of Figure 1 is shown in conventional electrical terms; the conducting members are depicted as wire-like lines, though in practice for a microwave device these would be waveguides of some form (as in Figures 2 and 5).

As shown, the device has an input line (10), an output line (11) and two intermediate lines (1 2T, 1 2B) each of which is selected by ganged switches (1 3L, 1 3R). The two intermediate lines 1 2 are of the same electrical path length, and each contains a rotator an Eplane-rotating element ( 1 4T, 148). The two rotators are themselves of the same electrical pathlength (as each other), but one (1 4T) rotates the E-plane by 180 , while the other (14B) overall rotates the E-plane by 0'.

If there is imagined a particular instant in time at which wave energy is passing through the device along one or other intermediate path (as determined by the setting of the ganged switches), and if the electric field is represented by an arrow-headed line (as 1 5, and dotted in the 0" lines) but only at whole number multiples of a wavelength from where it is a positive maximum at the input end, then along the route including the intermediate line with the 0" rotator all the field arrows are upwardly pointing (as viewed) while along the other intermediate line with the 180 rotator the field arrows are downwardly pointing (as viewed) on the output side.A signal that would have had one particular phase at the output when sent via the 0" rotator would have had a 180 different phase at the output when sent via the 180 rotator-and because both the two intermediate lines and the two rotators are matched as regards path length this 180 phase difference is achieved quite independently of the wavelength/frequency of the transmitted wave energy.

A finline realisation of the Figure 1 circuit is depicted in Figure 2 (with arrows using the Figure 1 convention). The device has input and output lines (20, 21) and two equal pathlength intermediate lines (22T, B) connected thereto by ganged switches (23 TL, 23BL, 23TR, 23BR--the "ganging" is not shown). In each intermediate line is a finline rotator (14T,14B: the two have the same pathlength); the top (as viewed) one 14T is a 180 rotator, whilst the bottom (14B) one is, overall a 0" rotator.

As shown, the ganged switch pair 23BL, 23BR is conducting (with the other pair, 23TL, 23TR, non-conducting) so making the fins at the bottom (as viewed) of each junction continuous and forcing the wave energy to travel along the top (as viewed) intermediate line 22T and through the 180 rotator 14T, where its E-plane is rotated by 180 . The result is a 180 phase reversal of the wave energy (as depicted by the solid arrows).

The process of this rotation is depicted further in the sequence of cross-sectional views of Figures 3A to G (these are on the lines A to G in Figure 2, viewed in the direction of the arrows X), which are best considered together with the perspective view of Figure 4. Each view shows the inner wall of the tubular guide, the dielectric board on which the fin line fins are supported, and the fins themselves.

The fin line rotator is an antipodal finline (Figure 3B) which becomes a balanced line (Figure 3D) and then crosses over to end as an antipodal line (Figure 3F) but in the opposite sense. As the wave progresses through the rotator so the field direction - and thus the E-plane--rotates to keep the field pointing between the two nearest conductive fin pairs; because these cross over, the fin that was on top (as viewed) becoming the fin on the bottom (as viewed), so the E-plane is similarly changed over.

As just explained, the top (as viewed) intermediate line 1 4T rotates the E-plane by 180 ; the bottom (as viewed) line 14B, however, has an overall E-plane rotation of 0'. This is achieved in this particular embodiment-by first rotating the E-plane by 90 and then rotating it back again (as shown in Figures 3 C' and E' the structure of the rotator 14B, and its effect on the E-plane, starts off like that of the 180 rotator 1 4T, but then is different in that the fins return to their original guide supporting walls rather than crossing over to the opposite walls).

Figure 5 and 6 relate to microstrip/slotline realisations of the Figure 1 circuit. The view is from above, looking down upon the microstripline sections that are supported above a ground plane (not shown separately), and the electric field and E-plane direction is shown by the circled crosses and dots (representing respectively the tails and heads of arrows).

The device depicted in Figure 5 has an input line (50), and output line (51) and two equal path length intermediate lines (52T, B) connected by ganged switches (53TL, TR, BL and BR-again, the "ganging" is not shown). In each intermediate line is a rotator (the microstrip/slotline combinations 54T, B); these two are of equal path length, and the top (as viewed) one 54T is a 1 80' rotator while the bottom (as viewed) one 54B is overall a 0 rotator.

Though it might not be immediately apparent, the path length of the upper intermediate line 52T is the same as that of the lower intermediate line 528, and the pathlengths of the two rotators 54T, B are also the same.

Moreover, in this particular case each "half" of each intermediate line is of the same pathlength as the relevant half of the other intermediate line--thus, from switch 53TL to rotator 54T is the same length as from switch 53BL to rotator 54B, and so on.

As shown, the ganged switched pair 53TL and 53TR is conducting (the pair 53BL and 53BR is non-conducting), so connecting the top (as viewed) microstrip 52T to the input/output lines 50, 51, and forcing wave energy to travel therealong between the device's input and output, and through the top (as viewed) rotator 54T, where the E-plane is rotated by 180 . The result is a 180 phase shift of the wave energy (as depicted by the circled crosses and dots).If, however, the switches were adjusted such that the ganged pair 53BL, 53BR was conducting (and 53TL, 53TR non-conducting), and the energy transmitted between input and output via the bottom (as viewed) intermediate line 52B, the rotator device 54B ensures first a 90' E-plane rotation in one direction and then a 90' rotation in the opposite direction and so a net rotation of 0'. Accordingly, a signal switched along one intermediate line arrives at the output end 180 out of phase with an identical signal switched along the other line.

Figures 6A and B show merely that the actual physical arrangement of the microstrip/slotline in a 1804 rotator need not be of the overlapping microstrip/slotline type shown in Figure 5, provided that the combination is manipulated to ensure that the energy travelling along the slot between the two strips does so in the correct sense. In each case the combination is in effect the same as that in Figure 5; energy transferred from microstrip 62TL to microstrip 62TR via slot 69T has its E-plane rotated by 180 .

Claims (10)

1. A microwave phase switching device comprising: an input line, an output line, and selectively connectable therebetween a plurality of equal pathlength intermediate lines, wherein the intermediate lines include, one in each, a like plurality of equal pathlength E-planerotating elements the rotating capabilities of any pair of which differ by an amount equivalent to a desired phase change.

2. A device as claimed in Claim 1, wherein all the lines are striplines, slotlines or finlines.

3. A device as claimed in either of the preceding claims, wherein there are two only intermediate lines.

4. A device as claimed in any of the preceding claims, wherein the selection of any particular intermediate line is achieved by having one or more electrical semiconductor switches associated with the junction of the input and intermediate lines, and the junction of the intermediate and output lines, such that the signal may be switched to (and from) the chosen intermediate line as required.

5. A device as claimed in Claim 4, wherein where fin lines are used the switches are placed across the fins as though to "block" the intermediate lines, while where striplines are used the switches are placed to "bridge" the gap to the intemediate lines.

6. A device as claimed in any of the preceding Claims, wherein the E-plane-rotating elements take the form of antipodal finlines, wherein in each the fins are adjacent the walls of the tubular guide from which they project only at their ends, and between their ends they are spaced from the walls.

7. A device as claimed in Claim 6, wherein there are two intermediate lines, in each rotator the fins are so dimensioned as to "approach" each other as one moves along the finline segment, while becoming separate from their base portion, and in one rotator each fin eventually approaches and becomes adjacent the opposite guide wall, while in the other rotator each fin returns to the wall from which it started.

8. A device as claimed in any of Claims 1 to 5, wherein the E-plane-rotating elements take the form of strip/slot/stripline junctions.

9. A device as claimed in Claim 8, wherein there are two intermediate lines, and in one rotator two strips are placed so as to be "joined" by a slot, and in one the slot extends directly across between them, while in the other the slot is curved through 180 and then so positioned relative to the two strips that it extends not from the side of one to the facing side of the other but instead from the side of one to the same side of the other.

10. A microwave phase switching device as claimed in any of the preceding Claims and substantially as described hereinbefore with reference to the accompanying drawings.