Lossless arbitrary output dual mode network

Abstract

A lossless and matched dual mode network (10) in which the maximum voltage amplitudes (a, b, and c, respectively) appearing at three output ports (11, 12, 13), are preselected and are arbitrary subject only to the constraint that the sum of the squares of any two elements of the set (a, b, c) must be equal to or greater than the square of the third element of this set. The set of complex voltages (A, B, and C, respectively) appearing at the three output ports (11, 12, 13) when an input signal is applied to one of the input ports (1 or 2) is conjugate with the set of output voltages (AA, BB, and CC, respectively) appearing at the three output ports (11, 12, 13) when an input signal is applied to the other input port, which is isolated from the initially selected input port (1 or 2). The network (10), which may be used as a feed network in an antenna (25) system, e.g., as an even/odd mode network, comprises three 90.degree. couplers (31, 32, 33 ) and three phase shifters (41, 42, 43). The couplers (31, 32, 33) have preselected characterizing angles (T1, T2, and T3, respectively), which are specified herein. Similarly, the three phase shifters (41, 42, 43) impart preselected phase shifts (P1, P2, and P3, respectively), which are similarly specified herein. In a first embodiment, the first input port (1) is coupled to a first input of the third coupler (33); in a second embodiment, the first input port (1) is coupled to a second input of the third coupler (33).

Description

DESCRIPTION

1. Technical Field

This invention pertains to the field of distributing electromagnetic energy, typically at microwave frequencies, by a "dual mode network", i.e., a network in which the maximum amplitudes of the voltages appearing at each of several output ports are the same regardless of which of two input ports is excited.

2. Background Art

U.S. Pat. Nos. 2,614,170; 3,843,941; and 4,223,283 disclose dual mode networks in which the maximum voltage amplitudes at the output ports must be equal to each other. In the present invention, on the other hand, the output voltage maximum amplitudes are arbitrary, subject to a sole constraint that results from the desirable isolation of the input ports (1 and 2) and lossless nature of the network (10).

U.S. Pat. No. 3,219,949 is a dual mode network in which the power distribution at the three output ports is fixed in the ratios 1:2:1.

U.S. Pat. No. 3,742,392 is not a dual mode network; and it is a 1:2 network (one input port and two output ports), not a 2:3 network as described herein.

U.S. Pat. No. 3,988,705 is a 1:4 network in which the voltages at the four output ports are always equal.

U.S. Pat. No. 4,231,040 discloses a dual mode network in which the output voltage distribution can be unequal. However, the network depicted in this reference is inherently lossy because some of the input power is forced to flow through resistors if an unequal output voltage distribution is to be accomplished. The present invention, on the other hand, is lossless in network design; the only possible losses are nominal losses in the components constituting the network (10).

DISCLOSURE OF INVENTION

The invention is a dual mode network (10) having two isolated input ports (1, 2) and three output ports (11, 12 and 13). As used herein, "dual mode" means that the distribution of maximum amplitudes (a, b, and c, respectively) of voltages appearing at the three output ports (11, 12, and 13) remains unchanged whether an input signal is applied at the first input port (1) or the second input port (2).

a, b, and c are preselected based upon the user's needs, and are arbitrary subject only to the constraint that the sum of the squares of any two members of the set consisting of a, b, and c must be equal to or greater than the square of the third element of this set.

The network (10) is theoretically lossless. By this is meant that none of the power applied at the input ports (1 and 2) is forced to flow through resistive elements as an incident to accomplishing the goal of arbitrary voltage distribution at the output ports (11, 12, 13). The only possible source of loss occurs in the components that comprise the network (10). These components, which can be made with insubstantial loss, are three 90.degree. couplers (31, 32, 33), three phase shifters (41, 42, 43), and transmission media (e.g., waveguide, coaxial cable, microstrip, or suspended substrate) interconnecting these six components and the ports (1, 2, 11, 12, 13). A resistor (28) is used to terminate one of the couplers (33), but no power flows therethrough.

Given the preselected values of output voltage maximum amplitudes (a, b, c), this specification gives values of the requisite characterizing angles (T1, T2, and T3, respectively) of the couplers (31, 32, 33), and the amount of phase shift (P1, P2, and P3, respectively) that must be imparted by the phase shifters (41, 42, 43).

BRIEF DESCRIPTION OF THE DRAWINGS

These and other more detailed and specific objects and features of the present invention are more fully disclosed in the following specification, reference being had to the accompanying drawings, in which:

FIG. 1 is a sketch of the dual mode network 10 of the present invention used as a feed network in association with an antenna 25;

FIG. 2 is a schematic of a first embodiment of the present invention;

FIG. 3 illustrates specific values of complex voltages occurring at certain points within the FIG. 2 embodiment when an input signal is applied at input port 2;

FIG. 4 illustrates specific values of complex voltages occurring at certain points within the FIG. 2 embodiment when the FIG. 3 input signal is applied at input port 1 rather than input port 2; and

FIG. 5 is a schematic of a second embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

FIG. 1 illustrates a typical use of dual mode network 10 of the present invention: as a feed network for a communications antenna system. The output ports 11, 12, 13 of network 10 are coupled to feed elements 21, 22, 23, respectively, comprising feed array 20. Array 20 is disposed towards antenna 25, which may be a paraboloidal reflector. In the case where this antenna system is used as part of a communications satellite, it is common for dual mode network 10 to be an even/odd mode network. By this is meant that a bandwidth of frequencies to be radiated by antenna 25 is divided up into a group of typically equally-wide frequency suballocations, which may be numbered consecutively 1, 2, 3 . . . n. The odd-numbered suballocations, e.g., 1, 3, 5, etc., are combined and become a first input, which is fed to one of the input ports. The even-numbered frequency suballocations are combined and fed to the other input port. When the input ports 1, 2 are isolated from each other, adjacent frequency suballocations are thus also isolated from each other. Therefore, this technique compensates for less than ideal isolation between adjacent frequency suballocations, such as may be caused by less than ideal filtering.

It is often desired for the output voltage maximum amplitudes a, b, c to be preselectable and arbitrary. In the case of the antenna 25 application described above, this permits arbitrary illumination of antenna 25, and thus flexible control of the radiation pattern emanating therefrom.

The present invention accomplishes this arbitrary preselection of a, b, and c, subject only to the constraint that the sum of the squares of any two of a, b, and c must be equal to or greater than the square of the third of a, b, and c. A second way of phrasing this same constraint is that a solution must exist to the design of network 10, given the preselected values of a, b, and c. A third way of phrasing this same constraint is as follows: Let V1 be a vector in three-dimensional space whose three co-ordinates are the complex (i.e., amplitude and phase) voltages appearing at output ports 11, 12, and 13, respectively, when an input signal is applied at input port 1. Let V2 be the three-dimensional vector whose co-ordinates are the complex voltages appearing at output ports 11, 12, and 13, respectively, when an input signal is applied at input port 2. Then V1 and V2 must be orthogonal, i.e., their dot product must be zero. However it is phrased, this constraint follows from the fact that input ports 1 and 2 are isolated, and network 10 is theoretically lossless.

As used herein, "theoretically lossless" means that there are no losses attributable to the design of network 10 itself, because no power is forced to flow through resistive components. Another way of saying this is that network 10 is substantially lossless. The only possible losses are I.sup.2 R losses in the components 31, 32, 33, 41, 42, 43, and transmission media interconnecting these components and the ports 1, 2, 11, 12, 13. These components can be chosen to exhibit insignificant loss.

Network 10 is also matched, i.e., there are no standing waves, no reflected power, and no impedance mismatches attributable to the design of network 10.

When an input signal is switched from input port 1 to input port 2, or vice versa, the second set of complex voltages AA, BB, and CC, respectively, appearing at output terminals 11, 12, 13 is conjugate with the initial set of complex voltages A, B, C, appearing thereon. By "conjugate" is meant that a, b, and c remain the same, while the phase differences between the voltages at any two adjacent output ports 11, 12, 13 change sign. ("Adjacent" means one of the pairs of output ports 11,12; 12,13; or 13,11).

Conjugateness is illustrated on FIG. 2 as follows: Assume that an input signal is applied to one of the input ports (1 or 2). The resulting voltages appearing at output ports 11, 12, and 13 are A=a.angle.q, B=b.angle.q+k, and C=c.angle.q+k+p, respectively. (In this standard notation for representing complex voltages, a.angle.q means that the maximum amplitude of the sinusoidal voltage A is a volts, and its phase angle is q.degree..) Then, when the input signal is switched to the other input port, the output voltages appearing at the output ports 11, 12, and 13 are AA=a.angle.r, BB=b.angle.r-k, and CC=c.angle.r-k-p, respectively. q and r can be any values, since the antenna radiation pattern is not affected by q and r.

The fact that the sets of output voltages are conjugate and not equally phased is of little detriment in an antenna 25 system application, because amplitudes have first order effects on radiation patterns, whereas phase differences have merely second order effects.

It is possible to apply input signals at each of ports 1 and 2 simultaneously, in which case network 10 keeps the inputs isolated from each other, and the output voltages are composites equivalent to input signals being separately applied to input ports 1 and 2.

FIG. 2 illustrates a first embodiment in which network 10 comprises a first coupler 31 having a first input coupled to input port 2, a first output coupled via phase shifter 41 to output port 11, and a second output coupled to a first input of coupler 32. Coupler 32 has a first output coupled to output port 12, and a second output coupled via phase shifter 42 to output port 13. Coupler 33 has a first input coupled to input port 1, a first output coupled to a second input of coupler 31, and a second output coupled via phase shifter 43 to a second input of coupler 32. Coupler 33 also has a second input which is terminated via load resistor 28 to ground. Resistor 28 has the characteristic impedance of network 10. If coupler 33 is functioning properly, no current flows through load resistor 28, and thus it does not cause any loss in the operation of network 10.

Couplers 31, 32, and 33 are each 90.degree. couplers, i.e., their output voltages are 90.degree. out of phase with respect to each other. The characterizing angles of couplers 31, 32, and 33 are T1, T2, and T3, respectively. A characterizing angle T of a 90.degree. coupler is that angle such that the following equations are satisfied:

out1=((sinT).angle.90)(in1)+(cosT)(in2) out2=(cosT)(in1)+((sinT).angle.90)(in2) where out1 is the voltage at the first output of the coupler (31, 32, or 33), out2 is the voltage at the second output of the coupler (31, 32, or 33), in1 is the voltage at the first input of the coupler (31, 32, or 33), and in2 is the voltage at the second input of the coupler (31, 32, or 33).

Couplers having prespecified arbitrary characterizing angles readily exist, in such forms as stripline couplers, waveguide couplers, etc.

P1, P2, and P3 are the angular phase shifts imparted by phase shifters 41, 42, and 43, respectively. Many phase shifters with preselected arbitrary phase shifts exist, such as Schiffman's phase shifter or a simple section of transmission line with the required phase shift.

In order to achieve the desired preselected a, b, and c, the values for the parameters of the six components of network 10 are as follows:

T1=sin.sup.-1 (a/(a.sup.2 +b.sup.2 +c.sup.2).sup.1/2) T2=sin.sup.-1 (b/(b.sup.2 +c.sup.2).sup.1/2) T3=sin.sup.-1 (a/(b.sup.2 +c.sup.2).sup.1/2) P1=k degrees P2=90-p degrees, and P3=2k+p degrees, where k=(1/2)cos.sup.-1 ((c.sup.4 -a.sup.4 -b.sup.4)/2a.sup.2 b.sup.2) and p=(1/2)cos.sup.-1 ((a.sup.4 -b.sup.4 -c.sup.4)/2b.sup.2 c.sup.2).

A numerical example will illustrate the design of network 10. Let a=1, b=0.9 and c=0.8. The sum of the squares of any two of a, b, and c is equal to or greater than the square of the third of a, b, c. Therefore, we know that a solution to the design of network 10 exists. Solving the above equations, we obtain:

k=70.15.degree. p=46.82.degree. T1=39.71.degree. T2=48.37.degree. T3=56.14.degree. P1=70.15.degree. P2=43.18.degree. P3=187.12.degree.

These values have been inserted in FIGS. 3 and 4. The phase angle of the input signal was arbitrarily assumed to be zero degrees regardless of which input port is excited. Working backwards from the output ports (11, 12, 13), intermediate values of voltages at the inputs and outputs of the couplers (31, 32, 33) were inserted in FIGS. 3 and 4 using the relationships given herein. Note that in FIG. 3 all the input signal appears at input port 2, and in FIG. 4, all the input signal appears at input port 1. Note further that a, b, and c remain the same; the two sets of complex voltages are conjugate; and no power flows through resistor 28. Finally, note that all of the input power (proportional to the voltage squared) appears at the output ports (11, 12, 13), i.e., no power is lost in network 10.

In an alternative embodiment (FIG. 5), input port 1 is coupled to the second input of coupler 33, rather than the first input thereof, and load resistor 28 is connected to the first input of coupler 33. In this case, the six parameters (T1, T2, T3, P1, P2, P3) of network 10 are the same as for the FIGS. 2-4 embodiment except for the characterizing angle of coupler 33 and the phase shift imparted by phase shifter 43. The new characterizing angle of coupler 33, T3A, is given by:

T3A=sin.sup.-1 ((b.sup.2 +c.sup.2 -a.sup.2).sup.1/2 /(b.sup.2 +c.sup.2).sup.1/2); and the new phase shift imparted by phase shifter 43, P3A, is given by: P3A=2k+p-180 degrees.

The above description is included to illustrate the operation of the preferred embodiments and is not meant to limit the scope of the invention. The scope of the invention is to be limited only by the following claims. From the above discussion, many variations will be apparent to one skilled in the art that would yet be encompassed by the spirit and scope of the invention.

Claims

1. A substantially lossless dual mode network having first and second input ports, and first, second, and third output ports, wherein the maximum amplitudes of the voltages appearing at the three output ports are the same regardless of which input port is excited, said maximum amplitudes being denoted a, b, and c, respectively;

wherein a, b, and c are preselected and are arbitrary subject only to the constraint that the sum of the squares of any two members of the set consisting of a, b, and c must be greater than or equal to the square of the third member of said set;

said network further comprising a first coupler having a first input coupled to the second input port, and a first output coupled through a first phase shifter to the first output port;

a second coupler having a first input coupled to a second output of the first coupler, a first output coupled to the second output port, and a second output coupled through a second phase shifter to the third output port; and

a third coupler having a first input coupled to the first input port, a first output coupled to a second input of the first coupler, and a second output coupled through a third phase shifter to a second input of the second coupler.

2. The network of claim 1 wherein the two input ports are isolated from each other.

3. The network of claim 1 wherein the characterizing angle of the first coupler is sin.sup.-1 (a/(a.sup.2 +b.sup.2 +c.sup.2).sup.1/2);

the characterizing angle of the second coupler is sin.sup.-1 (b/(b.sup.2 +c.sup.2).sup.1/2);

the characterizing angle of the third coupler is sin.sup.-1 (a/(b.sup.2 +c.sup.2).sup.1/2);

the first phase shifter imparts a phase shift of k degrees;

the second phase shifter imparts a phase shift of (90-p) degrees; and

the third phase shifter imparts a phase shift of (2k+p) degrees;

where k=(1/2)cos.sup.-1 ((c.sup.4 -a.sup.4 -b.sup.4)/2a.sup.2 b.sup.2)

and p=(1/2)cos.sup.-1 ((a.sup.4 -b.sup.4 -c.sup.4)/2b.sup.2 c.sup.2).

4. The apparatus of claim 1 further comprising a feed element coupled to each of the output ports, wherein the feed elements are directed at an antenna.

5. The network of claim 1 wherein a composite signal comprising alternating members of a set of frequency suballocations is fed to the first input port, and a composite signal comprising alternating but different members of said set of frequency suballocations is fed to the second input port, so that the network is an even/odd mode network.

6. The network of claim 1 wherein the set of voltages appearing at the three output ports in response to excitation of the first input port is conjugate with the set of voltages appearing at the three output ports in response to excitation of the second input port.

7. The network of claim 1 wherein V1 and V2 are orthogonal, where V1 is the three-dimensional vector having as co-ordinates the complex voltages appearing at the three output ports in response to excitation of the first input port, and V2 is the three-dimensional vector having as co-ordinates the complex voltages appearing at the three output ports in response to excitation of the second input port.

8. A substantially lossless dual mode network having first and second input ports, and first, second, and third output ports, wherein the maximum amplitudes of the voltages appearing at the three output ports are the same regardless of which input port is excited, said maximum amplitudes being denoted a, b, and c, respectively;

wherein a, b, and c are preselected and are arbitrary subject only to the constraint that the sum of the squares of any two members of the set consisting of a, b, and c must be greater than or equal to the square of the third member of said set;

said network further comprising a first coupler having a first input coupled to the second input port, and a first output coupled through a first phase shifter to the first output port;

a second coupler having a first input coupled to a second output of the first coupler, a first output coupled to the second output port, and a second output coupled through a second phase shifter to the third output port; and

a third coupler having a second input coupled to the first input port, a first output coupled to a second input of the first coupler, and a second output coupled through a third phase shifter to a second input of the second coupler;

wherein the characterizing angle of the first coupler is sin.sup.-1 (a/(a.sup.2 +b.sup.2 +c.sup.2).sup.1/2);

the characterizing angle of the second coupler is sin.sup.-1 (b/(b.sup.2 +c.sup.2).sup.1/2);

the characterizing angle of the third coupler is sin.sup.-1 ((b.sup.2 +c.sup.2 -a.sup.2).sup.1/2 /(b.sup.2 +c.sup.2).sup.1/2);

the first phase shifter imparts a phase shift of k degrees;

the second phase shifter imparts a phase shift of (90-p) degrees; and

the third phase shifter imparts a phase shift of (2k+p-180) degrees;

where k=(1/2)cos.sup.-1 ((c.sup.4 -a.sup.4 -b.sup.4)/2a.sup.2 b.sup.2)

and p=(1/2)cos.sup.-1 ((a.sup.4 -b.sup.4 -c.sup.4)/2b.sup.2 c.sup.2).