US3478269A

US3478269A - Directional antenna signal combining arrangement and phase shifters therefor

Info

Publication number: US3478269A
Application number: US408907A
Authority: US
Inventors: Robert B Enemark; Carson K H Tsao
Original assignee: Continental Electronics and Manufacturing Co
Current assignee: Continental Electronics and Manufacturing Co
Priority date: 1964-11-04
Filing date: 1964-11-04
Publication date: 1969-11-11
Anticipated expiration: 1986-11-11

Description

1969 R. B. EN-EMARK ETAL 3478'269 DIRECTIONAL ANTENNA SIGNAL COMBINING ARRANGEMENT AND PHASE SHIFTERS THEREFOR Filed Nov. 4. 1964 20! f g cf 5 i #9 N a, g :1 711T Z! 1-"- IV? 1 Q: 1 E

w mums/sic & Z

a a a 1 202 INVENTOR5 E1022 ,3. Zzzemrit 60171020 1f. E2300 9%, WW ,w W

United States Patent ABSTRACT OF THE DISCLOSURE Signals from directional antennas are combined by phase shifting one signal an angle corresponding to the spatial agle of the antennas, and summing it with the other signal. The phase shift network produces, through specific transfer functions, optimalized system performance. It is formed with a balanced symmetrical lattice network using only passive impedance elements, and terminated by a resistance, or with an unbalanced network employing amplifiers.

The field of the present invention is that of receiving systems for electromagnetic waves wherein several wave collecting means are combined for obtaining for final detection a signal of predetermined directional pattern.

It is well known to combine electromagnetic wave signals collected by several antennas of similar collecting pattern, such as loop antennas, for the purpose of obtaining predetermined patterns. If such arrangements depend on phase shifting circuitry, such circuitry has to fulfill requirements such as to precision and range dictated by the performance specifications for the system as a whole, and this presents problems that require particular consideration and for which solutions are not readily available.

Objects of the invention are to provide a radio receiving system having in effect an optimally omni-directional antenna pattern derived from several antennas having directional radiation patterns:

To provide such a system which includes directional antennas that are not detrimentally affected by operation within media, such as water, that attenuate the electric field component, and operates in such media as if it were fed by omni-directional, such as whip antennas that are with omni-directional response, such as the response of a whip antenna, but with responsiveness principally dependent upon the relatively unattenuated magnetic component of electromagnetic waves;

To provide such an omni-directional receiving system which includes antenna structures that are less susceptible to electrostatic interference, such as precipitation static, than omni-directional antenna structures;

To provide such a system which is especially suitable for use in conjunction with a pair of crossed loop antennas, which provides an optimally uniform omni-directional reception pattern, which is available for either the tuned or broad band modes of reception, which has a normally satisfactory operating range over carrier frequencies different by a ratio of about 2.5 while maintaining good circularity of the omni-directional receiving pattern to within 1 to 1.5 decibels, and which uses only linear circuit elements so that distortion of the receiving pattern does not develop from interfering signals outside the carrier frequency band of interest; and

To provide such a system which is comparatively simple, employs only easily obtainable standard components, is rugged, and simple to set up, calibrate and operate.

The substance of the invention can be briefly summarized as involving in a broad aspect the reception of an electromagnetic wave signal, coming from a given direction, by two directional antennas effectively displaced at an angle to each other, by shifting the phase of the two signals an angle corresponding as closely as possible to the angular displacement of the two antennas and by then summing the thus phase shifted signals in such a manner that the resulting output signal is virtually undistinguishable from a single signal of optimal omni-directional characteristics. In a presently most important practical aspect, the directional antennas are a pair of loop antennas crossed at right angles and the phase of one of the two antenna signals is accordingly shifted for combination with the other signal of unmodified phase.

In another important phase, the system according to the invention includes a phase shift network which is particularly suitable for purposes of the receiving system according to the invention as a whole. In its general aspect, this phase shift network is described by a physically realizable transfer function including terms which permit the derivation of practical network component ratings producing predeterminable and exactly prescribed tolerances for optimalized system performance.

In one specific practical aspect the phase shift network is realized as a balanced symmetrical lattice network using only passive impedance elements, with two equal series arms (each having a capacitance in shunt with a series combination of resistance and capacitance), with two equal shunt arms (each having an inductance in series with a parallel combination of resistance and inductance), and terminated by a resistance.

In another specific practical aspect the phase shift network is realized as an unbalanced network having a first arm with a series resistor and a resistor and an inductor parallel thereto, feeding into a first amplifier which feeds into a second arm having two parallel branches one with a resistance and the other with an inductance, the branches each being earthed through a resistor; the potential difference developed across the resistances is amplified in a differential amplifier which has output reference to earth; the amplifiers have high input resistance, low output resistance, and low phase shift.

These and other objects and aspects of novelty of the invention will appear from the following description of its principle and mode of operation, and of several embodiments illustrating its novel characteristics.

The description refers to a drawing in which FIG. 1 is a block circuit diagram of the receiving system according to the invention;

FIG. 2 is a plot of the transfer function of the phase shift network incorporated in FIG. 1;

FIG. 3 is the block diagram of a practical phase shift network which is a component of FIG. 1;

FIGS. 4 and 5 are circuit diagrams of the arms of the network according to FIG. 3; and

FIG. 6 is the circuit diagram of another practical phase shift network suitable as a component of FIG. 1.

In FIG. 1, L1 and L2 are two loop antennas of conventional design which may be crossed at 90 or which may be separately mounted at a similar angle. These antennas are connected to amplifiers A1 and A2 which are conventional variable gain control amplifiers permitting the adjustment of the signals from L1 and L2 to equal amplitudes. The output of amplifier A1 is fed into a high quality phase shift network P which will be described more in detail hereinbelow. The output signal from the phase shift network P is fed to an amplifier A3 of conventional type, capable of adjustment to compensate for losses through the phase network and thus permitting derivation of an optimally omni-directional reception pattern. The phase shifted signal amplified at A3, and the signal from loop L2, amplified at A2 as mentioned above,

are combined in a summing amplifier of conventional design indicated at A4, which provides the composite omnidirectional output signal. Such an amplifier is for example described under adding Network on page 458 of Reference Data for Radio Engineers, International Telephone and Telegraph Coroporation, fourth edition, 1956. Only two inputs of this network, supplied from A2 and A3, respectively, will be used.

The composite signal A4 is fed to a cathode follower amplifier A5 of conventional design providing a low driving source impedance. Suitable amplifiers of this type are for example described on pages 448 to 450 of the above I'IT reference Data Handbook.

All amplifiers used in this circuitry are preferably vacuum tube devices which, it was found, provides for greater flexibility in circuit design and performance. It is however understood that transistorized circuitry of similar performance characteristics can be used if carefully designed for the purpose at hand.

It will be evident that, as indicated above, the quality of the receiving system as a whole depends on phase and amplitude linearity of the phase shift network P which should maintain a 90 phase shift to Within a few degrees, and a negligibly small gain change, between its input and output signals over a wide band such as 3 octaves centered about the chosen optimal carrier frequency. Referring to FIGS. 2 to 6, the practical construction of two precision phase shift networks P, fulfilling these conditions, will now be described. This description is based on the following theoretical considerations.

A constant amplitude transfer function can be written according to conventional circuit theory (compare for example E. A. Guillemin, Synthesis of Passive Networks, Wiley, 1957, page 194) for a terminated, symmetrical, constant resistance lattice network as follows:

12 a b) a b) For an all-pass transfer function with poles and zeros symmetrical about the j-axis, /Z /Z is a reactance function. With Z Z =l, S=jw, and Z =jX we then have wherein 5:2 tan X, is the transfer function phase with phase slope dfi/dw=(Sin [3)/w. Since X is a pure reactance it is a function of m with positive slope so that dfl/dw 0. Thus, in order to qualify as an all-pass network, the ideal phase shifter must have a phase characteristic satisfying dfl/dwzsin fi/w, and d 8/dw 0.

Assuming now, that, as in the present specific embodiment the desired phase shift is approximately 1r/ 2, then dfi/dw l/w and Ar3=ln (to /m For a frequency range as for example such that w =2w the numerical value of AB would thus be 0.7 radian. This is much greater than can be tolerated in a phasing network useful in the present system. However, if instead of an all-pass transfer function a network is accepted which merely approximates the ideal characteristic with a permissible amount of deviation over the entire frequency range, a practical solution can be found. For that purpose, conditions for a permissible amount of deviation must be established, for example as follows.

With v =v +v wherein v is the output voltage of loop L1 and v that of loop L2 through the phase shifter, it can be shown that wherein A is amplitude, 0 the phase shift, and for purposes of simplification,

The above formula for v indicates that, if a practical omni-directional pattern with an ellipticity of less than :1 db is admissible, the requirements of (AA/A) 0.l and 0.1 radian must be fulfilled.

In order to accommodate the above conditions leading to a phase shift network that is practical for present purposes of an omni-directional receiving system, the restriction of using an all-pass transfer function is abandoned; instead a more general algorithm is now adapoted which is indicated by Guillemin,'supra, at page 316 et seq., at follows.

With P(7\) a rational polynominal function of x, and with Q( another rational polynominal function of A, it is possible to Write Z =P()\)/Q( Therefore, the phase function must be specified in such a way that its tangent is a rational function of polynominals, and then P( \)Q()t)=M+N and the zeros of M+N can be assigned to P()\) and Q( to form the transfer function.

As a new concept in this algorithm, it can now be specifield that tan B=a(w/w )/[(w/w 1]. The constant a must be sufiiciently large so that the phase deviates from rr/Z by no more than the permissible amount for the operating frequency range. For example, if we require that and 0.45w/w g25, then the constant a must be greater than 21.

Quite generally, the transfer function can be written as where G is the gain of the phase shift network, for example the ratio v /v of the output and input voltages, or i where G'=[G|e with and where and 4: are the input and output phase angles, respectively. In this transfer function there is further X (S/S where S =-w '=(211'f) with f the frequency of the input sinusoidal wave form, in cycles per second. Further, S is a constant which depends on the chosen carrier frequency f at which optimalized performance is desired; in general, f =S 211-. If fd= /21r c.p.s., then S =1, further \=S, and still further a =a =a =a =L This optimalizing procedure corresponds to conventional so-called frequency scaling techniques, compare for example E. A. Guillemin, Introductory Circuit Theory, Wiley, 1963, pages 309 ff.

With the above-mentioned newly introduced constant a chosen arbitrarily as a=25, we have, with the above substitutions, N=25S and M=S +1 which latter it will be noted appears in the general transfer function G. There is further The assignment of zeros to P and Q is guided by the restriction that the degree of P must not exceed that of Q by unity. With P(S)=1 and Q( (s+0.04 (SH-24.96) then Z =1/[(S+0.004) (SA-24.96)].

It can be shown that |1z is far from unity for 0.4 w 2j. Therefore, some compensation for amplitude gain as well as for phase shift is required.

For purposes of amplitude compensation, we can assign a pair of complex poles or zeros, or a pair of real poles or zeros symmetrical about the j-axis. Introduction of these poles and zeros will not modify the phase shifting operation in any way.

Further for purposes of amplitude compensation, to obtain a nearly constant response in the frequency range of interest, there can be put in the numerator of the transfer function the factor (S 1); a numerator was introduced into the transfer function G above for this very purpose. The transfer function thus becomes This function is plotted in FIG. 2 where performance has been optimalized for carrier frequencies near /21r c.p.s.

For purposes of a practical embodiment, the last mentioned transfer function can be realized in a physical, balanced, symmetrical lattice network, either directly from this formula or from the plot FIG. 2. Such a network, PI, is shown in FIGS. 3 to 5. The component impedances are Z =(1-Z )/(l+Z and Z Z =L Using the above formula for Z the lattice arm values are obtained as FIG. 3 shows the symmetrical lattice arms Z,, and Z of the network PI and FIGS. 4 and 5 indicate the ratings of the individual elements of the two blocks Z and 2 As a second practical embodiment of the phase shifter P, an unbalanced network PII (FIG. 6) will now be described.

An unbalanced phase shifting network is sometimes desirable, but the circuitry according to FIGS. 3 to 5 cannot be rendered unbalanced without nearly ideal transformers. If available transformers are not sufficient for that purpose, if active circuits are permissible, and if constant input impedance is not perticuiarly important, a network such as PII according to FIG. 6 is quite satisfactory.

1116 ratings of the passive elements are written into MG. 6. The active elements consist of conventional amplifiers A6 and A7. The amplifier A6 can for example be a grounded cathode-grounded plate pair as described in the above referred to ITT Reference Data on page 446. The amplifier A7 can for example be a dilferential amplifier as described on page 447 of the same Reference Data book. These amplifiers are required to have effectively infinite input and effectively zero output impedances. In the example according to FIG. 6, the input impedance is Z =24.96 -|-S ohms, and the gain with S=jw as referred to above. In this context it will be remembered that for arithmetical simplicity, the phase shift network has been optimalized for carrier frequencies near /211- c.p.s.

The above described broad band omni-directional radio receiving system has in actual performance been optimalized for an operating range from 14 to 35 kc. producing circularity to within one to one and one-half decibels, and beyond that range still useful operation to about 65 kc. without serious pattern distortion. The above described phase shift network which is of primary importance in systems of this type provides phase and amplitude linearity of the 90 phase shift within a few degrees over a 3 octave band if contained in the above described system with directional loop antennas crossed at 90.

It should be understood that the present disclosure is for the purpose of illustration only and that this invention includes all modifications and equivalents which fall within the scope of the appended claims.

6 We claim: -1. An essentially omni-directional radio receiving system comprising:

tWo loop antennas placed at right angles;

means for shifting the phase of the signal from one of said antennas at right angles; and

means for summing the signal from the other antenna and the phase shifted signal to furnish a single signal which essentially conforms to a single unidirectional system,

said phase shifting means comprising a balanced symmetrical lattice network having:

two equal impedance series arms each having a capacitor in shunt with a resistor and a capacitor in series;

two equal impedance shunt arms each having an inductor in series with a resistor and an inductor in parallel; and

a resistor terminating the network.

2. An essentially omni-directional radio receiving system comprising:

two loop antennas placed at right angles;

said phase shifting means comprising an unbalanced network having:

a first arm having a series resistive branch, and a shunt branch having a resistor and an inductor in series;

a first amplifier having its input connected to said first arm at the junction of the branches of said first arm and having an input resistance appreciably higher than the output resistance, and low phase shift;

a second arm in series with said first amplifier and having a resistance branch and an inductance branch, the output of the first amplifier being connected to the junction of said branches in the second arm;

earthed resistance means across said branches in said second arm; and

across said resistance means a differential amplifier having output reference to ground, an input resistance appreciably higher than the output resistance, and low phase shift.

53. A phase shifting network in lattice formation comprrsmg:

' two equal impedance series arms each having a capacitor in shunt with a resistor and a capacitor in series; two equal impedance shunt arms each having an inductor in series with a resistor and an inductor in parallel; and a resistor terminating the network. 4. A phase shifting network comprising: a first arm having a series resistive branch, and a shunt branch having a resistor and an inductor in series; a first amplifier having its input connected to said first farm at the junction of the branches of said first arm and having an input resistance appreciably higher than the output resistance, and low phase shift; a. second arm in series with said first amplifier and having a resistance branch and an inductance branch, the output of the first amplifier being connected to the junction of said branches in the second arm; earthed resistance means across said branches in said second arm; and across said resistance means a differential amplifier havlng output reference to ground, and input resistance appreciably higher than the output resistance, and low phase shift. 5. A system for deriving an essentially omni-directional receiving pattern from two directional antennas placed substantially at a right angle in space, comprising:

balanced symmetrical lattice network means for relatively shifting the phase of the signals collected by said antennas an angle 3 approximating said angle in space and related to the radian frequency w of said signals by the expression tan 6:

and two shunt arms of impedance where Z is given by the expression and where 7. A system according to claim wherein the constant a is selected to be greater than about 20.

8. A system according to claim 7 wherein a is substantially equal to 25.

9. A system according to claim 5 wherein said phase shifting means is further characterized by a gain G given by the expression (a3)\) +a4 \+a5 where x =(w/w the constants a1, a2, a3, a4, and a5 are selected so that a4=a, and a1, a2, a3, and a5 are selected to furnish an acceptable amplitude deviation at a selected operating carrier frequency.

10. A phase shifting device in the form of a balanced lattice network, comprising means for shifting phase an angle 5 given by the expression tan 5:

where w is the radian frequency of the input signal applied to said network, ca is the radian frequency of the signal carrier, and a is a constant selected sufficiently large to produce a correspondingly small deviation in 5 over the desired operating frequency range.

11. A system according to claim 10 wherein said balanced symmetrical lattice network has two series arms of impedance.

and two shunt arms of impedance 12 4 1-21.

where Z is given by the expression and where 2 2 EL (a) 12. A system according to claim 10 wherein the constant a is selected to be greater than about 20.

13. A system according to claim 10 wherein a is substantially equal to 25.

14. A system according to claim 10 wherein said phase shifting means is further characterized by a gain G given by the expression the constants a1, a2, a3, a4, and a5 are selected so that (14:0, and a1, a2, a3, and a5 are selected to furnish an acceptable amplitude deviation at a selected operating carrier frequency.

References Cited UNITED STATES PATENTS 2,529,117 11/1950 Tompkins 33329 2,606,966 8/1952 Pawley 333-29 2,623,945 12/1952 Wigan 33329 2,951,152 8/1960 Sichak et a1. 325-369 XR 2,982,924 5/1961 Weged et al 33329 XR 3,242,430 3/1966 Chose 325369 XR ROBERT L. GRIFFIN, Primary Examiner R. S. BELL, Assistant Examiner US. Cl. X.R.