GB1588518A - Microwave commutating hybrid networks - Google Patents

Description

(54) IMPROVEMENTS IN OR RELATING TO MICROWAVE COMMUTATING HYBRID NETWORKS (71) We, THE MARCONI COMPANY LIMITED, a British Company of Marconi House, New Street, Chelmsford, Essex CM1 1PL, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: This invention relates to a microwave commutating hybrid network for use at millimetric wavelengths.

A known commutating hybrid is constituted by two 3dB hybrid couplers each having a conjugate pair of arms interconnected by two transmission all-pass netrworks with specified characteristics, one of the networks being a reference network and the other network a delay line. The term "hybrid coupler" as used herein means an essentially lossless device having four terminals arranged in two pairs such that when a respective arm, to which each of the terminals is connected is completely matched no power flow takes place between the arms of a pair and such a pair of arms is herein termed a conjugate pair of arms. A further property of a hybrid coupler is that when power ebters one of the arms of either conjugate pair, it may be arranged divide equally, or with some other predetermined ratio, and flows out of the arms of the other conjugate pair; the relative phase angle of the two equal outputs being dependent upon the type of hybrid junction used. Such couplers are well known in the art, for example, the magic T coupler and the so-called 3dB Riblett coupler.

In one regime, the transmission characteristics of the microwave commutating hybrid network are such that a frequency multiplexed signal applied at one of the ports of one of the hybrid couplers is divided, in the frequency domain, to emerge as a series of comb-like channels alternating between the two output ports of the other hybrid coupler. Conversely, in another regime, because a microwave commutating hybrid network is a reciprocal device, similarly spaced channels applied to conjugate input ports of one of the couplers will emerge as a frequency multiplexed signal from a single port of the other coupler. Networks of this kind are described in the article "Review of Design and Performance of Microwave Multiplexers" by C.S. Barham, The Marconi Review, Vol. XXXV. No. 184 1972 pages 1 to 23 The channel spacing and hence channel width between the signals emerging from the output ports of the first mentioned regime is dictated by the relative phase delay introduced by the interconnecting all-pass networks and it is possible to produce a multi-channel separating/combining unit by cascading, in a binary fashion, commutating hybrid networks in which the differential delay introduced at each successive stage decreases/increases monotonically so that binary frequency division/combination takes place at each stage.

The passband response of each comb channel is a sin2x function and to produce a more acceptable "squared off" response the so-called "multipole multiplexer" has been devised as described in the Marconi Review volume XXXV No. 184 first quarter 1972 by C.S.

Barham. In such a multiplexer, conjugate pairs of arms of input and output hybrid couplers are interconnected by the conjugate arms of further hybrid couplers, the other conjugate pairs of arms of input and output hybrid couplers are interconnected by the conjugate arms of further hybrid couplers, the other conjugate pair of arms of the further couplers being terminated in lengths of waveguide of length L and 2L respectively. At the input port of the aforementioned waveguide lengths are positioned susceptances to provide the requisite degree of squaring off. Such a multipole multiplexer is dispersive, i.e. the change in phase with frequency is not constant, with the result that the centre frequency spacing between the channels is not linear. Thus. the difference in phase between the two waveguide lengths for increasing frequency is not nz radians, where n is an integer, and it is accordingly an object of this invention to produce a microwave commutating hybrid network in which the spacing between channels can be given desired predetermined values.

According to this invention a microwave commutating hybrid network includes two 4-terminal hybrid couplers each having a conjugate pair of arms interconnected by a conjugate pair of arms of at least one further 4-terminal hybrid coupler, each of the arms of the other conjugate pair of arms of each said further 4-terminal hybrid coupler being connected to a tapered length of waveguide, the taper being shaped in accordance with a desired predetermined channel spacing. Where the network is formed from rectangularly sectioned waveguide normally the broad wall dimension of said tapered lengths of waveguide is tapered. The tapered lengths of waveguide are such that in operation substantially arithmetically spaced channels are achieved at the network output.

Preferably, in such a case the tapered lengths of waveguide have a taper in accordance with an approximately cosine law.

In a preferred embodiment one arm of said conjugate pair of arms of both the first mentioned couplers is connected to one arm of a conjugate pair of arms of a further 4-terminal hybrid coupler so that said couplers form a multipole arrangement, and in such an embodiment the tapered lengths of waveguide connected to the other conjugate pair of arms of one of the two said further 4-terminal hybrid couplers are twice the length of the tapered lengths of waveguide connected to the other conjugate pair of arms of one of the two said further 4-terminal hybrid couplers are twice the length of the tapered lengths of waveguide connected to the other of said further 4-terminal hybrid couplers.

Advantageously, susceptance loading means are positioned at the inputs of tapered waveguides which are connected to the input ports of said one of the further 4-terminal hybrid couplers so as to provide the network with a squared-off response.

The invention will now be described, by way of example, with reference to the drawing accompanying the Provisional specification which shows in schematic form a commutating hybrid network which provides substantially arithmetically spaced channels.

The commutating hybrid network shown in the drawing is formed from rectangularly sectioned waveguide and has a 3dB 4-terminal hybrid coupler 1 having two pairs of conjugate ports 2, 3 respectively, one port 2 being connected to receive a broadband frequency input and the other port 2 being connected to a load and the ports 3 being connected to another 3dB 4-terminal hybrid couplr 4 having conjugate pairs of ports 5, 6 respectively. One of each of the ports 3 and 5 is connected to a conjugate pair of ports 7, 8 of further 3dB 4-terminal hybrid couplers 9, 10 respectively. The other conjugate pair of ports 11 of the coupler 10 are each connected to a tapered length of waveguide 12 having a length L for cut-off of a predetermined frequency and the other conjugate pair of ports 13 of the coupler 9 are each connected to tapered lengths of waveguide 14 of length 2L for cut-off of said predetermined frequency. All the tapered lengths of waveguide have the broad wall dimension tapering in accordance with an approximately cosine law.

The network described thus far, by suitable dimensioning of the tapered length of waveguide 12 and 14, is such as to provide arithmetically spaced channels of a band of frequencies applied to one of the ports 2 at the ports 6 over a 25% bandwidth. Such a network is, thus, substantially non-dispersive with the result that the phase difference between the channels at each of the ports 6 is substantially nx radians. The passband characteristic produced is of the form sin2x and to provide a squared-off response susceptances 15 are introduced at the input ports of the longer waveguide lengths 14. Where the susceptances are inductive, shims are introduced which extend between the outer extermities of the broad wall dimension, but where the susceptances are capacitive then the shims extend between the outer extremities of the narrow walls of the waveguide lengths 14 input ports.

For some applications, it may be desirable to provide a network response in which output channels are not arithmetically spaced apart but which are, instead, spaced apart in a different predetermined manner. Given the required network response in terms of pass bands (i.e. channels) interleaved with guard bands, the characteristic phase function between the unloaded short waveguides 12 and the loaded long waveguides 14 must be arranged to provide approximately a 1800 phase shift for frequencies in the guard bands whilst maintaining a constant phase in the pass band. This may be achieved approximately by calculating a phase function for the taper of the loaded waveguide which, when loaded by the appropriate discontinuity, ensures resonances at the centre of each of the guard bands. The phase function for the taper of the unloaded waveguide is calculated such that there is an integral number of 1800 in the centre of the first pass band and 1800 between each of the centres of the pass bands. Additionally, it is necessary that in the pass band the derivative of the phase matches as near as possible the derivative of the phase of the loaded waveguide. This procedure does not give a unique phase function for the waveguide tapers and it is desirable to choose phase responses which yield the smallest second derivative of phase whilst still giving the required networks response. The profile k(a) of the waveguide taper in the waveguide transverse dimension 'a' is given by the equation:

Where p = vm/2f and tits the velocity of the internal medium of the waveguide and m is the mode number for the TEmo mode, and l(p) is the required group delay response for the taper.

The terms t(p) for the loaded waveguide and l(p) for the unloaded waveguide can be found from the phase functions as indicated previously, thereby enabling the taper profiles to be determined for arbitrarily spaced pass bands.

In an alternative embodiment (not shown) the coupler 10 is omitted and the lowr ports 3, 5 (as orientated in the drawing) are interconnected by a suitable delay line.

WHAT WE CLAIM IS: 1. A microwave commutating hybrid network including two 4-terminal hybrid couplers each having a conjugate pair of arms interconnected by a conjugate pair of arms of at least one further 4-terminal hybrid coupler, each of the arms of the other conjugate pair of arms of each said further 4-terminal hybrid coupler being connected to tapered length of waveguide, the taper being shaped in accordance with a desired predetermined channel spacing.

2. A microwave commutating hybrid network as claimed in claim 1 and in which the tapered lengths of waveguide are such that in operation substantially arithmetically spaced channels are achieved at the network output.

3. A network as claimed in claim 2 and wherein the tapered lengths of waveguide have a taper in accordance with an approximately cosine law.

4. A network as claimed in claim 1 2 or 3 and wherein one arm of said conjugate pair of arms of both the first mentioned couplers is connected to one arm of a conjugate pair of arms of a further 4-terminal hybrid coupler so that said couplers form a multipole arrangement.

5. A network as claimed in claim 4 and wherein the tapered lengths of waveguide connected to the other conjugate pair of arms of one of the two said further 4-terminal hybrid couplers are twice the length of the tapered lengths of waveguide connected to the other of said further 4-terminal hybrid couplers.

6. A network as claimed in any of the preceding claims and wherein susceptance loading means are positioned at the inputs of tapered waveguides which are connected to the input ports of said one of the further 4-terminal hybrid couplers so as to provide the network with a square-off response.

7. A microwave commutating hybrid network substantially as illustrated in and described with reference to the drawing accompanying the Provisional Specification.

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (7)

**WARNING** start of CLMS field may overlap end of DESC **. and it is desirable to choose phase responses which yield the smallest second derivative of phase whilst still giving the required networks response. The profile k(a) of the waveguide taper in the waveguide transverse dimension 'a' is given by the equation: Where p = vm/2f and tits the velocity of the internal medium of the waveguide and m is the mode number for the TEmo mode, and l(p) is the required group delay response for the taper. The terms t(p) for the loaded waveguide and l(p) for the unloaded waveguide can be found from the phase functions as indicated previously, thereby enabling the taper profiles to be determined for arbitrarily spaced pass bands. In an alternative embodiment (not shown) the coupler 10 is omitted and the lowr ports 3, 5 (as orientated in the drawing) are interconnected by a suitable delay line. WHAT WE CLAIM IS:

1. A microwave commutating hybrid network including two 4-terminal hybrid couplers each having a conjugate pair of arms interconnected by a conjugate pair of arms of at least one further 4-terminal hybrid coupler, each of the arms of the other conjugate pair of arms of each said further 4-terminal hybrid coupler being connected to tapered length of waveguide, the taper being shaped in accordance with a desired predetermined channel spacing.

2. A microwave commutating hybrid network as claimed in claim 1 and in which the tapered lengths of waveguide are such that in operation substantially arithmetically spaced channels are achieved at the network output.

3. A network as claimed in claim 2 and wherein the tapered lengths of waveguide have a taper in accordance with an approximately cosine law.

4. A network as claimed in claim 1 2 or 3 and wherein one arm of said conjugate pair of arms of both the first mentioned couplers is connected to one arm of a conjugate pair of arms of a further 4-terminal hybrid coupler so that said couplers form a multipole arrangement.

5. A network as claimed in claim 4 and wherein the tapered lengths of waveguide connected to the other conjugate pair of arms of one of the two said further 4-terminal hybrid couplers are twice the length of the tapered lengths of waveguide connected to the other of said further 4-terminal hybrid couplers.

6. A network as claimed in any of the preceding claims and wherein susceptance loading means are positioned at the inputs of tapered waveguides which are connected to the input ports of said one of the further 4-terminal hybrid couplers so as to provide the network with a square-off response.

7. A microwave commutating hybrid network substantially as illustrated in and described with reference to the drawing accompanying the Provisional Specification.