U.S. patent number 3,618,133 [Application Number 04/826,149] was granted by the patent office on 1971-11-02 for symmetrical polyphase networks utilizing constant reactances.
This patent grant is currently assigned to International Standard Electric Corporaton. Invention is credited to Micahel John Gingell.
United States Patent |
3,618,133 |
Gingell |
November 2, 1971 |
SYMMETRICAL POLYPHASE NETWORKS UTILIZING CONSTANT REACTANCES
Abstract
Symmetrical polyphase networks are disclosed comprising N
single-phase networks each including a constant reactance, i.e, a
reactance whose value remains constant with changes in the
frequency of the signal applied thereto. Due to the use of this
constant reactance (controlled constant current source, such as
properly connected and controlled transistors, or impedance
transformers), the polyphase network responds differently to input
signals of negative and positive frequencies (a positive frequency
is a counterclockwise sequence of vectors representing polyphase
input signals and a negative frequency is a clockwise sequence of
vectors representing polyphase input signals). Also the polyphase
network has a different insertion loss characteristic depending on
the sequence of the polyphase input signals.
Inventors: |
Gingell; Micahel John
(Sawbridgeworth, EN) |
Assignee: |
International Standard Electric
Corporaton (New York, NY)
|
Family
ID: |
10255190 |
Appl.
No.: |
04/826,149 |
Filed: |
May 20, 1969 |
Foreign Application Priority Data
|
|
|
|
|
Jun 7, 1968 [GB] |
|
|
27,161/68 |
|
Current U.S.
Class: |
333/24R; 332/151;
333/215 |
Current CPC
Class: |
H03H
11/22 (20130101) |
Current International
Class: |
H03H
11/02 (20060101); H03H 11/22 (20060101); H03h
007/44 () |
Field of
Search: |
;333/1,80,8T,24,70,6
;307/295 ;328/155 ;324/107,108 ;336/5,10,12 ;330/107,109 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Lieberman; Eli
Assistant Examiner: Gensler; Paul L.
Claims
While I have described above the principles of my invention in
connection with specific apparatus, it is to be clearly understood
that this description is made only by way of example and not as a
limitation to the scope of my invention as set forth in the objects
thereof and in the accompanying claims.
1. A symmetrical polyphase network comprising:
N single-phase networks, a different one coupled to each phase of
an N-phase input signal, where N is an integer greater than
one;
each of said single-phase networks including
at least one constant reactance so that said polyphase network
responds differently to input signals of negative and positive
frequencies, the reactance value of said constant reactance being
different than zero.
2. A polyphase network according to claim 1, wherein
each of said constant reactances includes
an N-port symmetrical gyrator, each of said N ports being coupled
to a different one of said single-phase networks.
3. A polyphase network according to claim 1, wherein
each of said constant reactances includes
an N-phase l to j impedance transformer, where j is equal to
-1.
4. A polyphase network according to claim 3, wherein
said impedance transformer includes
N circuits each having
a transistor with its collector-emitter circuit connected in series
between the input and output terminals of its associated one of
said circuits, and its base coupled to the collector of said
transistor in another one of said circuits.
5. A polyphase network according to claim 4, wherein
N is equal to four; and
the base of said transistor in one of said circuits is directly
connected to the collector of said transistor in another of said
circuits receiving as an input signal a signal lagging in phase
with respect to the phase of the input signal coupled to said one
of said circuits.
6. A polyphase network according to claim 4, wherein
N is equal to two;
said transistor of one of said circuits includes
a first transistor; and
said transistor of the other of said circuits includes
a second transistor, the base of said second transistor being
directly connected to the collector of said first transistor; and
further including
a third transistor network coupling the base of said first
transistor to the collector of said second transistor.
7. A polyphase network according to claim 6, wherein
said third transistor network includes
a third transistor,
an electrical power supply,
a first resistor coupling the emitter of said third transistor to
one terminal of said power supply providing a given potential,
means directly connecting the collector of said third transistor to
the base of said first transistor,
a second resistor coupling the collector of said third transistor
to the other terminal of said power supply providing a potential
lower than said given potential, and
means directly connecting the base of said third transistor to the
collect or of said second transistor.
8. A polyphase network according to claim 4, wherein
N is equal to three; and further including
a fourth transistor network coupling the collector of said
transistor of each of said circuits to the base of said transistor
of an adjacent one of said circuits.
9. A polyphase network according to claim 8, wherein
said fourth transistor network coupled to two adjacent ones of said
circuits includes
an electrical power supply,
a fourth transistor having its base directly connected to the
collector of said transistor in one of said two adjacent ones of
said circuits and its collector directly connected to one terminal
of said power supply providing a given potential,
a first resistor coupled between the emitter of said fourth
transistor and the other terminal of said power supply providing a
potential greater than said given potential, and
a potentiometer coupled between the emitter of said fourth
transistor and the base of said transistor in the other of said two
adjacent ones of said circuits,
said potentiometer including
a second resistor having a first terminal connected to the emitter
of said fourth transistor and a second terminal, and
a third resistor having a first terminal directly connected to said
second terminal of said second resistor and a second terminal
connected to the emitter of a transistor included in said fourth
transistor network coupled to another two adjacent ones of said
circuits,
said second resistor having a value one-half of the value of said
third resistor,
the base of said transistor in said other of said two adjacent ones
of said circuits being directly connected to said second terminal
of said second resistor and said first terminal of said third
resistor.
10. A polyphase network according to claim 3, wherein
N is equal to two; and
said impedance transformer includes
two circuits each having
a transistor with its emitter-base circuit connected in series
between the input and output terminals of its associated one of
said circuits;
the base of said transistor in one of said circuits being directly
connected to the collector of said transistor in the other of said
circuits; and
a third transistor network coupling the base of said transistor in
said other of said circuits to the collector of said transistor in
said one of said circuits.
11. A polyphase network according to claim 10, wherein
said third transistor network includes
a third transistor,
an electrical power supply,
a first resistor coupling the emitter of said third transistor to
one terminal of said power supply providing a given potential,
means directly connecting the base of said third transistor to the
collector of said transistor of said one of said circuits,
a second resistor coupling the base of said third transistor to the
other terminal of said power supply providing a potential higher
than said given potential, and
means directly connecting the collector of said third transistor to
the base of said transistor in said other of said circuits.
12. A polyphase according to claim 1, wherein
each of said constant reactances includes
an N phase 1 to -j impedance transformer, where j is equal to
-1.
13. A polyphase network according to claim 12, wherein
N is equal to four; and
said impedance transformer includes
four circuits each having
a transistor with its collector-emitter circuit coupled in series
between the input and output terminals of its associated one of
said circuits and its base connected to the collector of said
transistor in an adjacent one of said circuits receiving as an
input signal a signal leading in phase with respect to the phase of
the input signal coupled to said associated one of said
circuits.
14. A polyphase network according to claim 12, wherein
N is equal to three; and
said impedance transformer includes
three circuits each having
a transistor with its collector-emitter circuit coupled in series
between the input and output terminals of its associated one of
said circuits, and a fourth transistor network coupling the base of
said transistor of said associated one of said circuits to the
collector of said transistor of said associated one of said
circuits.
15. A polyphase network according to claim 14, wherein
said fourth transistor network includes
an electrical power supply,
a fourth transistor,
means directly connecting the base of said fourth transistor to the
collector of said transistor of said associated one of said
circuits,
means directly connected to one terminal of said power supply
providing a given potential,
a first resistor coupling the emitter of said fourth transistor to
the other terminal of said power supply providing a potential
higher than said given potential,
a second resistor coupling the emitter of said fourth transistor to
the base of said transistor of said associated one of said
circuits, and
a third resistor coupling the junction of said second resistor and
the base of said transistor of said associated one of said circuits
to the emitter of a transistor included in said fourth transistor
network of an adjacent one of said circuits,
the value of said third resistor being one-half the value of said
second resistor.
16. A polyphase network according to claim 1, wherein
N is equal to three; and
said constant reactance includes
a three-phase impedance transformer of one of a three-phase l to h
impedance transformer and a three-phase l to 1/h impedance
transformer, where h is equal to e .sup.j2 3.
17. A polyphase network according to claim 16, wherein
both of said l to h and l to 1/h impedance transformers include
three circuits each having
a transistor with its collector-emitter circuit coupled in series
between the input and output terminals of its associated one of
said circuits and its base directly connected to the collector of
said transistor of an adjacent one of said circuits;
a source of voltage V coupled to the first of said circuits;
a source of voltage Vh coupled to the second of said circuits;
and
a source of voltage Vh.sup.2 coupled to the third of said
circuits.
18. A polyphase network according to claim 17, wherein
said 1 to h impedance transformer includes
in each of said circuits the base of said transistor in said
associated one of said circuits being directly connected to the
collector of said transistor in an adjacent one of said circuits
receiving as an input signal a signal leading in phase with respect
to the phase of the input signal coupled to said associated one of
said circuits.
19. A polyphase network according to claim 17, wherein
said l to 1/h impedance transformer includes
in each of said circuits the base of said transistor in said
associated one of said circuits being directly connected to the
collector of said transistor in an adjacent one of said circuits
receiving as an input signal a signal lagging in phase with respect
to the phase of the input signal coupled to said associated one of
said circuits.
Description
BACKGROUND OF THE INVENTION
The invention relates to polyphase networks and more particularly
to symmetrical polyphase networks.
SUMMARY OF THE INVENTION
The term "constant reactance" as employed herein is defined as a
reactance whose value remains constant with changes in the
frequency of the signal applied thereto.
The term "positive frequency" as employed herein is defined as a
counterclockwise sequence of vectors representing polyphase input
signals.
The term "negative frequency" as employed herein is defined as a
clockwise sequence of vectors representing polyphase input
signals.
An object of the present invention is to provide a symmetrical
polyphase network including constant reactances.
Another object of the present invention is to provide a symmetrical
polyphase network including constant reactances such that the
symmetrical polyphase network responds differently to input signals
of negative and positive frequencies.
A feature of the present invention is the provision of a
symmetrical polyphase network comprising N single-phase networks,
one for each phase of an N-phase input signal, where N is an
integer greater than one; each of said single-phase networks
including at least one constant reactance so that the symmetrical
polyphase network responds differently to input signals of negative
and positive frequencies.
BRIEF DESCRIPTION OF THE DRAWING
The above-mentioned and other features and objects of this
invention will become more apparent by reference to the following
description taken in conjunction with the accompanying drawings, in
which:
FIGS. 1(A) and 1(B), respectively, illustrate positive and negative
sequence four-phase vector diagrams;
FIG. 2 illustrates a positive sequence four-phase vector
diagram;
FIG. 3(A) illustrates the frequency response curve of a simple
third-order elliptic low-pass filter;
FIG. 3(B) illustrates the frequency response curve illustrated in
the drawing according to FIG. 3(A) transformed into an asymmetric
form;
FIG. 4 shows the gyrator realization of constant reactance;
FIG. 5 shows the controlled source representation of a gyrator;
FIG. 6 shows a method of realizing constant reactances in a
four-phase system;
FIG. 7 shows the circuit diagram of a single-phase
asymmetric-about-zero frequency filter;
FIG. 8 shows the circuit diagram of the polyphase realization of
the circuit diagram of FIG. 7 for a two-phase system with
quadrature inputs;
FIGS. 9(A) and 9(B), respectively, show circuit diagrams of the
theoretical and practical realizations of a two-phase 1 to j
impedance transformer of the voltage shift type;
FIGS. 10(A) and 10(B), respectively, show circuit diagrams of the
theoretical and practical realizations of another two-phase voltage
shift type 1 to j impedance transformer;
FIGS. 11(A) and 11(B), respectively, show circuit diagrams of the
theoretical and practical realizations of a two-phase 1 to j
impedance transformer of the current shift type;
FIGS. 12(A) and 12(B), respectively, show circuit diagrams of the
theoretical and practical realizations of another two-phase current
shift type 1 to j impedance transformer;
FIG. 13 shows the circuit diagram of a single-phase network which
utilizes a plurality of constant reactances;
FIG. 14 shows the circuit diagram of the polyphase realization of
the circuit diagram to FIG. 13 for a two-phase network with
quadrature inputs;
FIG. 15 shows part of the circuit diagram of FIG. 14 together with
negative impedance converters;
FIGS. 16, 17, 18 and 20, respectively, show the practical circuit
diagrams of different forms of three-phase impedance
transformers;
FIG. 19 shows the equivalent circuit diagram of one phase of the
impedance transformer shown in FIG. 18;
FIG. 21 shows the circuit diagram of another single-phase network
which utilizes a plurality of constant reactances;
FIG. 22 shows the circuit diagram of the polyphase realization of
the circuit diagram of FIG. 21 for a three-phase network;
FIGS. 23 and 24 show equivalent circuit diagrams of one phase of
the circuit diagram shown in FIG. 22;
FIGS. 25 and 26 show the circuit diagrams of different forms of
four-phase impedance transformers;
FIGS. 27(A) and 27(B) show frequency response curves;
FIGS. 28(A) and 28(B) show circuit diagrams of a single-phase
filter before and after transformation by image design
techniques;
FIGS. 29(A) and (B) show frequency response curves for an N-path
frequency translation system having low-pass filters connected in
each of the N-paths thereof;
FIGS. 30(A) and (B) show frequency response curves for an N-path
frequency translation system which utilizes the symmetrical
polyphase networks according to the invention;
FIGS. 31(A) to (C) show vector diagrams; and quadrature inputs.
FIG. 32 shows the circuit diagram of a two-phase network with
quadrature inputs.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
In order to understand the operation of the symmetrical polyphase
networks according to the present invention, the concept of
negative frequency must be introduced. If a four-phase system is
considered, which has as shown in FIG. 1(A) voltages of V, -jV, -V,
+jV applied to its four input terminals, then the input signal can
be called symmetrical, since all voltages are equal in magnitude
and spaced apart by steps of 90.degree. , and of say positive
sequence, since, conventionally, all vectors rotate anticlockwise
and the voltage on path 1 leads the voltage on path 2 by 90.degree.
, and similarly, the voltage on path 2 leads the voltage on path 3
by 90.degree. , etc. If now the vectors rotate the opposite way,
i.e., as shown in FIG. 1(B), the system is still symmetrical, but
is now of negative sequence (negative frequency), since the voltage
on path 1 lags the voltage on path 2 by 90.degree. instead of
leading as before.
Considering the voltage on path 1 it can be seen from FIG. 2 that
this voltage is V sin .omega.t, i.e., the projection of the vector
1 onto the imaginary axis when it is being rotated
counterclockwise. When the sequence of vectors is reversed, -V sin
.omega.t will be observed. Since -sin .omega.t= sin (-.omega.t ),
it can be said that, in a single-phase network, a positive sequence
represents a positive .omega. and a negative sequence represents a
negative .omega.. Thus, where positive and negative frequencies are
hereinafter referred to with reference to the characteristics of a
single-phase network, it means positive and negative sequence,
respectively, in a polyphase network containing N single-phase
networks.
In order to realize a polyphase network which has different
magnitude responses to positive and negative sequence (positive and
negative frequency) inputs, it is easiest to first consider a
single-phase network having different magnitude responses to
positive and negative frequencies. By its very nature such a
single-phase network will not be physically realizable, since it is
not possible to distinguish between positive and negative
frequencies in a single-phase network containing real components.
However, consideration of an asymmetric-about-zero frequency
single-phase network can be used as a stepping stone to the
synthesis of a polyphase network where positive and negative
frequencies have a real significance and it will be seen that
interconnections between the single-phase networks, so that each
network "knows" what sequence of input has been applied, is
necessary and sufficient for polyphase realization of the
previously unrealizable single-phase network.
There are a variety of transforms that will change a
symmetrical-about-zero frequency network to one that is
asymmetric-about-zero.
By way of example, consider a simple third-order elliptic low-pass
filter, the response of which is illustrated in FIG. 3(A). By
making the transformation
where .omega. is the frequency scale of the original filter,
and
.OMEGA. is a new frequency scale,
the response is transformed into the asymmetric form illustrated in
FIG. 3(B). As shown in FIG. 3(B) .omega..sub.x =.omega. .sub.1 so
that the peak that was at -.omega. .sub.1 (see FIG. 3(A)) is
shifted to .OMEGA. = + .infin. (see FIG. 3(B)). It is now possible
to observe the effect of the above on realizability. The inverse
transform of equation (1) is given by
Since the original network would have been constructed of coils and
capacitors and resistively terminated, a coil which had an
admittance of 1/j.omega.L would, therefore, be transformed as
below:
The coil has, therefore, been transformed into a coil in parallel
with a constant reactance. Similarly, a capacitor C would be
transformed into a capacitor C in series with a constant reactance.
However, since constant reactances are physically unrealizable in
single-phase networks, the single-phase network is also
unrealizable.
In the polyphase network according to the present invention, which
contains N single-phase networks, it is possible to realize the
constant reactance elements by means of, for example, gyrators, or
sets of controlled sources, the latter being N-port gyrators.
Consider a two-phase system with quadrature inputs of V.sub.in and
jV.sub.in. At any point in one of the two paths of the system there
will be a voltage of V and a current of I. Since the two paths are
physically identical, there will be at the corresponding point in
the other of the two paths a voltage of jV and a current of jI.
Thus, if gyrator G1 is connected between the two phases, the two
single-phase networks, in a symmetrical way as shown in FIG. 4, for
example, it will have a voltage of V across port 5 and a voltage of
jV across port 6. The gyrator must be symmetrical.
A symmetrical gyrator has the chain matrix ##SPC1##
Therefore, it can be seen from equations (6) and (7) that looking
into either port of gyrator G1 a reactance of 1/jgm is observed.
Also, since the gyrator is a lossless element (theoretically), it
is possible to construct lossless networks with their attendant low
sensitivity (compared with active and passive R-C networks).
As shown in FIG. 5, gyrator G2 can be represented by two controlled
sources, i.e., the constant current sources CCS1 and CCS2.
By using this representation, it is possible to envisage ways of
extending the use of gyrators in systems having more than two
phases. FIG. 6 shows a way of realizing constant reactances in a
four-phase system with what is virtually a four-port gyrator.
For other than two or four phases, the arrangements are slightly
more involved due to the complex relationship between voltages and
currents of the single-phase circuits.
In practice the gyrator circuits are realized with transistors. As
an example of the application of the above principles, consider the
filter described hereinabove with reference to FIG. 3(B).
The single-phase version of this filter could take the form shown
in FIG. 7 wherein coil L1 which is connected between one input and
one output terminal is shunted by constant reactance X3 and
capacitor C2, constant reactance X2 and capacitor C3 connected in
series. The input terminals of the filter are shunted by capacitor
C1 connected in series with constant reactance X1 and the output
terminals are shunted by capacitor C4 connected in series with
constant reactance X4.
The polyphase realization for a two-phase system with quadrature
inputs is shown in FIG. 8 where it can be seen that four two-port
gyrators G3 to G6 are utilized for the constant reactances of each
one of the two phases.
One port of two-port gyrator G3 is used for one phase and is,
therefore, connected in series between capacitors C2.sub.1 and
C3.sub.1 across coil L1, while the other port of this gyrator is
used for the other phase and is, therefore, connected in series
between capacitors C2.sub.2 and C3.sub.2 across coil L1.sub.2.
Similarly, the two-port gyrators G4 to G6 each have each one of the
two ports thereof utilized for one of the two input phases, i.e.,
two ports of gyrator G4 are utilized with capacitors C1.sub.1 and
C1.sub.2 to shunt the inputs of the two phases, the two ports of
gyrator G5 are used to shunt coils L1.sub.1 and L1.sub.2, and the
two ports of gyrator G6 is utilized with capacitors C4.sub.1 and
C4.sub.2 to shunt the outputs of the two phases.
The gyrators and N-port gyrators used to realize constant
reactances may be called polyphase-constrained networks, since the
N terminals thereof are forced to carry voltages and currents which
are always in N phases.
A further class of polyphase-constrained networks are those which
include 1 to j impedance transformers to provide the constant
reactances, where j is the conventionally employed mathematical
symbol equal to -1. (See page 479, "The International Dictionary of
Physics and Electronics," 1956).
By way of example, consider the circuit diagram shown in FIG. 9(A)
which shows a two-phase (quadrature) l to j impedance transformer
of the voltage shift type having a constant current source (CCS3 or
CCS4) in each phase thereof and the voltages and currents
associated with each phase are as indicated. The phrase "l to j
impedance transformer" means if impedances Z are placed on the
output of each phase, in other words, looking into the inputs of
the impedance transformer, impedances of jZ will be seen looking
into either input from the output of the impedance transformer.
These networks which may be utilized as circuit elements to
transform resistors into constant reactances are in practice
realized with transistors as shown in FIG. 9(B). Each one of the
two phases contains a transistor, i.e., transistors VT1 and VT2,
having their collector-emitter circuits connected in series between
the input and output terminals of the network.
The jV.sub.1 voltage input to one of the two phases is also applied
to the base of transistor VT1 in the other of the two phases by
means of transistor VT3 having its base connected to the jV.sub.1
voltage source, its collector connected to the base of transistor
VT1 and to a negative electrical power supply via resistor R1 and
its emitter connected to ground potential via resistor R2. The
input voltage V.sub.1 is applied directly to the base of transistor
VT2.
Alternatively, the two-phase voltage shift type 1 to j impedance
transformers may take the form shown in FIGS. 10(A) and 10(B). The
equivalent circuit diagram of FIG. 10(A) shows the voltages and
currents associated with each phase and the practical circuit
diagram of FIG. 10(B) comprises a transistor in each one of the two
phases, i.e., transistors VT4 and VT5 having their
emitter-collector circuits connected in series between the input
and output terminals of the network.
The jV.sub.2 voltage output of one of the two phases is applied
directly to the base of transistor VT4 and the output voltage
V.sub.2 of the other of the two phases is applied to the base of
transistor VT5 in said one of the two phases by means of transistor
VT6 having its base connected to the collector of transistor VT4,
its collector connected to a negative electrical power supply via
resistor R11 and to the base of transistor VT5 and its emitter
connected to ground potential via resistor R10.
Examples of two-phase (quadrature) current shift types of 1 to j
impedance transformers are shown in FIGS. 11(A) and 11(B) and FIGS.
12(A) and 12(B).
FIGS. 11(A) and 12(A) show the theoretical circuit diagrams and
indicate the voltages and currents associated with each phase and
do not require any further explanation.
FIG. 11(B) shows the practical circuit diagram for the theoretical
circuit of FIG. 11(A) and comprises a transistor in each one of the
two phases, i.e., transistors VT7 and VT8 having their emitter-base
circuits connected in series between the input and output terminals
of the network.
The base of the transistor VT7 is directly connected to the
collector of transistor VT8 and the base of transistor VT8 is
connected to the collector of transistor VT7 via transistor VT9
having its emitter connected to earth potential via resistor R12,
its collector connected to the base of transistor VT8 and its base
connected to the collector of transistor VT7 and to a positive
electrical power supply via resistor R13.
FIG. 12(B) shows the practical circuit diagram for the theoretical
circuit of FIG. 12(A) and comprises a transistor in each one of the
two phases, i.e., transistors VT10 and VT11 having their
base-emitter circuits connected in series between the input and
output terminals of the network.
The base of transistor VT10 is connected to the collector of
transistor VT11 via transistor VT12 having its emitter connected to
ground potential via resistor R15, its base connected to a positive
electrical power supply via resistor R14 and to the collector of
transistor VT11, and its collector directly connected to the base
of transistor VT10. The base of the transistor VT11 is directly
connected to the collector of transistor VT10.
FIG. 13 shows the circuit diagram of a single-phase network which
utilizes a large number of constant reactances X5, X6, X7, X8,
X9...XM. The input to the network is shunted by capacitor C5, the
output of the network is shunted by capacitor CN and constant
reactances X6, X8...X(M-1) are respectively connected in series
with capacitors C6, C7...C(N-1) between ground potential and the
junction of the constant reactances X5 and X7, X7 and X9...X(M-2)
and XM.
By utilizing any one of the l to j impedance transformers outlined
in the preceding paragraphs, the polyphase version of the circuit
diagram of FIG. 13 is shown in FIG. 14. This is a two-phase
(quadrature) network having input phases .phi..sub.1 and
.phi..sub.2 and input voltages V.sub.in and jV.sub.in.
The input to the single-phase circuits for .phi..sub.1 and
.phi..sub.2 are, respectively, shunted by capacitors C5.sub.1 and
C5.sub.2 while their outputs are respectively shunted by capacitors
CN.sub.1 and CN.sub.2.
Since the 1 to j impedance transformers can be utilized as circuit
elements to transform resistances into constant reactances, the
T-networks of resistors R5.sub.1, R7.sub.1, and R6.sub.1, and
resistors R5.sub.2, R7.sub.2 and R6.sub.2 are utilized in
conjunction with the j to l impedance transformer 2 which is
interposed between resistors R6.sub.1, R6.sub.2 and capacitors
C6.sub.1, C6.sub.2 to provide the constant reactances X5, X6 and X7
of FIG. 13 in each phase of the two-phase network of FIG. 14.
Similarly, the j to l impedance transformer 3 is utilized in
conjunction with the T-networks of resistors R(M- 2).sub.1,
RM.sub.1 and R(M- 1).sub.1 and resistors R(M- 2).sub.2, RM.sub.2
and R(M- 1).sub.2 to provide the constant reactances X(M- 2), XM
and X(M- 1) of FIG. 13 in each phase of the two-phase network of
FIG. 14.
The l to j impedance transformer 1 interposed between the inputs to
the network and resistors R5.sub.1 and R5.sub.2 is utilized to
phase shift the inputs by j before they pass through the network
and the l to j impedance transformer 4 interposed between resistors
RM.sub.1 and RM.sub.2 and the outputs from the network corrects
this phase shift before the signals are passed to the outputs of
the network.
If, as is common, the reactances X6, X8, etc. are of opposite sign
to the reactances X5, X7, X9, etc., it would be necessary to add
negative impedance converters to the circuit diagram of FIG. 14 as
shown in the circuit diagram of FIG. 15. Referring to FIG. 15, part
of the circuit diagram of FIG. 14 is shown therein and includes
negative impedance converters 5 and 6 which are, respectively,
interposed between resistors R6.sub.1 and the junction of resistors
R5.sub.1 and R7.sub.1 and resistor R6.sub.2 and the junction of
resistors R5.sub.2 and R7.sub.2. Due to the use of the negative
impedance converters which would be included in each of the
resistance T-networks, it is necessary to change the j to 1
impedance transformers to 1 to j impedance transformers, e.g.,
transformer 7 (FIG. 15) for transformer 2 (FIG. 14).
The circuit diagram shown in FIG. 14 without the capacitors can be
considered as an N-port gyrator which is lossless and passive
although it may have to contain active devices to enable it to be
realized.
FIGS. 16, 17, 18 and 20 show examples of practical circuit
arrangements for three-phase constrained networks.
The three-phase constrained network according to FIG. 16 comprises
a transistor in each of the three phases, each of the single-phase
networks, i.e., transistors VT13 to VT15 having their
collector-emitter circuits connected in series between the input
and output terminals of the network. The base of transistor VT13 is
connected to the collector of transistor VT15, the base of
transistor VT14 is connected to the collector of transistor VT13
and the base of transistor VT15 is connected to the collector of
transistor VT14.
This network which can be considered as a l to h impedance
transformer has the chain matrix ##SPC2##
where Z.sub.1 = input impedance
Z.sub.2 = output impedance ##SPC3##
and 1+ h+ h.sup.2 = 0. (12)
The three-phase constrained network of FIG. 17 is basically the
same as the network of FIG. 16 except the bases of each of the
transistors VT13 to VT15 are connected to the collector of the
transistor in the subsequent instead of the preceding adjacent
phase to provide a l to 1/h impedance transformer i.e., the base of
transistor VT14 is connected to the collector of transistor VT15,
the base of transistor VT15 is connected to the collector of
transistor VT13 and the base of transistor VT13 is connected to the
collector of transistor VT14.
This network has the chain matrix ##SPC4##
FIG. 18 shows the practical circuit diagram of the three-phase
constrained network which includes a transistor in each of the
three phases, i.e., transistors VT16 to VT18 having their
collector-emitter circuits connected in series between the input
and output terminals of the network.
The interconnections between the three phases in order to provide a
l to -j 3 impedance transformer is effected by three networks which
each include a transistor, i.e., the transistors VT19 to VT21
having their collectors connected to a negative electrical supply
and their emitters connected to ground potential, respectively, via
bias resistors R17, R18 and R19.
The base of transistor VT19 is connected to the input voltage
V.sub.1 of one of the phases, i.e., to the collector of transistor
VT16, the base of transistor VT20 is connected to the input voltage
hV.sub.1 of another one of the phases, i.e., to the collector of
transistor VT17 and the base of transistor or VT21 is connected to
the input voltage h.sup.2 V.sub.1 of the other of the phases, i.e.,
to the collector of transistor VT18.
The emitter of transistor VT19 is also connected to the emitter of
transistor VT20 via resistor R16 connected in series with resistor
R16/2 and to the emitter of transistor VT21 via resistor R21/2
connected in series with a resistor R21. The junction of resistors
R21 and R21/2 is connected to the base of transistor VT18 and the
junction of resistors R16/2 and R16 is connected to the base of
transistor VT16.
The emitter of transistor VT20 is also connected to the emitter of
transistor VT21 via resistor R20 connected in series with resistor
R20/2 and the junction of resistors R20 and R20/2 is connected to
the base of transistor VT17.
The value of the resistances R16, R20 and R21 are arranged in
conjunction with the transistors VT19 to VT21 such that a voltage
of jV.sub.1 / 3 is applied to the base of transistor VT16, a
voltage of hjV.sub.1 is applied to the base of transistor VT17 and
a voltage h.sup.2 jV.sub.1 is applied to the base of transistor
VT18.
This network has the chain matrix ##SPC5##
The equivalent circuit diagram of one phase of the circuit of FIG.
18 is shown in FIG. 19 from which, by way of example, the voltage
applied to the base of the transistor which forms part of the first
of the three phase will be established.
Consider transistors VT19 and VT20.
From FIG. 19 it can be seen that ##SPC6##
The three-phase constrained network of FIG. 20 is basically the
same as the network of FIG. 18 except the three networks which each
include one of the transistors VT19 to VT21 are arranged such that
a l to + j 3 impedance transformer is provided, i.e., resistors
R16/2, R20/2 and R21/2 are, respectively, replaced by resistors
2R16, 2R20 and 2R21, the junction of resistors R16 and 2R16 is
connected to the base of transistor VT17, the junction of resistors
R20 and 2R20 is connected to the base of transistor VT18 and the
junction of resistors R21 and 2R21 is connected to the base of
transistor VT16.
This network has the chain matrix ##SPC7##
and Z.sub.1 =V.sub.1 /I.sub.1 = + j 3Z.sub.2 (23)
The electrical supply arrangements for each of the phases
(single-phase networks) of the circuit diagrams of FIGS. 16, 17, 18
and 20 are not shown, but in practice would be arranged such that a
predetermined potential difference exists between the collector and
emitter, and the emitter and base of each of the transistors which
form part of each of the phases. An example of how this may be
achieved in practice will be outlined in a subsequent
paragraph.
FIG. 21 shows the circuit diagram of a single-phase network which
utilizes constant reactances X10 to X12. The input to the network
is shunted by capacitor C8, the output is shunted by capacitor C9
and the constant reactance X12 is connected in series with
capacitor C10 between ground potential and the junction of
reactances X10 and X11. It is assumed that the constant reactance
X12 is of opposite sign to the constant reactances X10 and X11.
By utilizing the l to h impedance transformers shown in FIG. 16,
the three-phase version of the circuit of FIG. 21 is shown in FIG.
22.
Referring to FIG. 22, constant reactance X10 is provided in each
phase by the l to h impedance transformer enclosed by the chain
dotted line 11A and resistors R10.sub.1/2 and R10.sub.1 for one of
the three phases, resistors R10.sub. 2/2 and R10.sub.2 for another
of the three phases and resistors R10.sub.3/2 and R10.sub.3 for the
other of the three phases. Constant reactance X11 is provided in
each phase by the 1 to h impedance transformer enclosed by the
chain dotted line 11B and resistors R11.sub.1/2 and R11.sub.1 for
one of the three phases, resistors R11.sub.2/2 and R11.sub.2 for
another of the three phases and resistors R11.sub.3/2 and R11.sub.3
for the other of the three phases.
Since constant reactance X12 is of opposite sign to constant
reactances X10 and X11, it is necessary, as will be outlined in a
subsequent paragraph with reference to FIGS. 23 and 24, to have two
1 to h impedance transformers, i.e., the transformers enclosed by
the chain dotted lines 11C and 11D, associated with the three
phases and resistors R12.sub.1/2 and R12.sub.1 for one of the three
phases, resistors R12.sub.2/2 and R12.sub.2 for another of the
three phases and resistors R12.sub.3/2 and R12.sub.3 for the other
of the three phases.
The input to each phase is shunted by a capacitor, i.e., capacitor
C8.sub.1, C8.sub.2 or C8.sub.3, the output of each phase is shunted
by a capacitor, i.e., capacitor C9.sub.1, C9.sub.2 or C9.sub.3 and
capacitor C10 of FIG. 21 is provided in each of the three phases by
capacitors C10.sub.1, C10.sub.2 and C10.sub.3.
The only difference between the 1 to h impedance transformers
enclosed by the chain dotted lines 11A to D and the transformer
shown in FIG. 16 is that resistor R22 is interposed between the
base of transistor VT15 and the collector of transistor VT14, and
constant current supply CCS5 is connected to the base of transistor
VT15. These additional components together with constant current
sources CCS6 which are connected between ground potential and the
junction of resistors R12.sub.1/2, R12.sub.2/2 and R12.sub.3/2 and
capacitors C10.sub.1 to .sub.3 provide the necessary electrical
supplies for the network.
Referring to FIG. 23, the equivalent circuit diagram of one phase
of the network according to FIG. 22 is shown wherein it can be seen
that resistors R10.sub.1 and R11.sub.1 are multiplied by h,
respectively, by transformers 11A and 11B and resistor R12.sub.1 is
multiplied by h.sup.2 by transformers 11C and 11D.
From equation (10), h= - 1/2 + j 3/2 ##SPC8##
The equivalent circuit diagram of FIG. 23 can, therefore, from
equations (24), (25) and (26) be reduced to the equivalent circuit
diagram shown in FIG. 24.
FIGS. 25 and 26 show examples of practical circuit arrangements for
four-phase constrained networks. The electrical supply arrangements
are not shown, but they may be provided in a similar manner to
three-phase networks.
The four-phase constrained network of FIG. 25 comprises a
transistor in each phase, i.e., transistors VT22 to VT25 having
their collector-emitter circuits connected in series between the
input and output terminals of the network. The bases of transistors
VT22 to VT25 are, respectively, connected to the collectors of
transistor VT23, VT24, VT25 and VT22.
The network without the dotted connections or the chain dotted
connections is of the voltage shift type having the chain matrix.
##SPC9##
When the dotted connections are added to the original network it
will be of the current shift type and the chain matrix becomes
##SPC10##
When the chain dotted connections are added to the original
network, i.e., crossing over the output leads of phases 1 and 3 and
phases 2 and 4, the matrices of equations (27) and (29) are
multiplied by -1.
The four-phase constrained network according to FIG. 26 is
basically the same as the network according to FIG. 25 except the
bases of transistors VT22 to VT25 are, respectively, connected to
the collectors of transistors VT25, VT22, VT23 and VT24 to provide
a l to -j impedance transformer, therefore, the network without the
dotted connections or the chain dotted connections is of the
voltage shift type having the chain matrix ##SPC11##
when the dotted connections are added to the original network
thereby providing a network of the current shift type.
As before, the chain dotted connections of FIG. 26 caused the
matrices of equations (31) and (32) to be multiplied by -1.
It should be noted that the design of the single-phase network may
be accomplished by the transformation of existing image network
sections to asymmetric polyphase sections which may be accomplished
by using known image design techniques, if the constant reactance
is included as an extra circuit element.
Considering the symmetrical-about-zero frequency response shown in
FIG. 27(A) where it can be seen that
.omega. = .+-. .beta. (33)
or .omega. = .+-. 1/.beta. (34)
or .omega. = 0 (35)
By making the transformation .omega..sup.2 = .OMEGA., the frequency
response shown in FIG. 27(A) is transformed to the frequency
response shown in FIG. 27(B) where it can be seen that.
.OMEGA.=.beta. .sup.2 (36)
or .OMEGA.= 1/.beta. .sup.2 (37)
or .OMEGA.= 0 (38)
If all the admittances (Y) are multiplied by .omega. before
frequency transformation then
j.omega.C j.omega..sup.2 C j.OMEGA.C (39)
and 1/ j.omega.L 1/ jL Y constant (40)
By way of example, consider the single-phase filter shown in FIG.
28(A) wherein the image impedance at the input and output are,
respectively, represented by Y.sub.O (.omega.) and Y.sub.0. Then by
this transform technique the circuit diagram is transformed to the
circuit diagram shown in FIG. 28(B) where it can be seen that the
capacitances C11 to C13 remain the same but the inductance L10 is
transformed to a constant reactance represented by the block 10
i.e. Y=1/ j.omega. L10 1/j L10
jY=1/L 10. (41)
If, however, all the admittances (Y) are divided by .omega. before
frequency transformation then
j.omega.C jC Y constant (42)
and 1/j .omega.L 1/ j.omega..sup.2 L 1/j.OMEGA.L (43)
The insertion loss design of the single-phase network may also be
accomplished by using the Z-transform method providing it is
modified such that the transform takes account of both negative and
positive frequencies.
The symmetrical polyphase networks outlined in the preceding
paragraphs have a particular, but not necessarily an exclusive,
application in the N-path frequency translation system outlined in
British Pat. No. 1,098,250 and also in single sideband generation
in a manner similar, but superior to conventional quadrature
modulation.
The transfer function of the N-path frequency translation system is
defined by
V.sub.0 (p)= K.sup.. H(p-p.sub.1).sup.. V.sub.1 (p-p.sub.1
+p.sub.2)
where K is a constant.
H(p) is the transfer function of the network(s)
in the N paths
P.sub.1 = j2.pi.f.sub.1
P.sub.2 = j2.pi.f.sub.2
f.sub.1 is the input switching rate
f.sub.2 is the output switching rate
It can be seen that the transfer function H(p) is shifted along the
real frequency axis by an amount f.sub.1. Normally, in the N-path
filter system, where p.sub.1 =p.sub.2, this would result in a band
pass characteristic symmetrical about the frequency f.sub.1. If
low-pass filters are connected in the N paths, the resultant
characteristic will be that of a shifted low-pass filter (including
that at negative frequencies which is the mirror image of the
positive frequency response). This is shown in the drawing
according to FIGS. 29(A) and (B). Symmetrical characteristics are
often very wasteful when modulation processes are involved. In such
cases much more attenuation is needed on one side of the passband
than the other. By using the symmetrical polyphase networks
according to the invention, the characteristic can be made to fit
the requirement more efficiently. Also, it is no longer necessary
for the switching or carrier frequency to be at midband. FIGS.
30(A) and (B) illustrate this by way of example.
The symmetrical polyphase networks according to the invention may
also be used for splitting a single-phase signal into N phases.
According to the theory of symmetrical components any unbalanced
system of N vectors can be represented as the sum of N-symmetrical
vector systems. If, for example, a two-phase (quadrature) system is
considered with an input of V on one phase only, then this is
equivalent to applying two opposite sequence two-phase signals
simultaneously as shown in the drawings according to FIGS. 31(A) to
(C). If the transfer function of the system is H(p) to the vector
system of FIG. 31(B) then it will be H(-p) to the vector system of
FIG. 31(C). FIG. 32 shows a two-phase system with an input on one
phase only which includes a two-phase network 10 having the input
V.sub.1 for one phase thereof, i.e., phase 1 connected to voltage
source V and the input V.sub.2 for the other phase thereof, i.e.,
phase 2 connected to ground potential, i.e., V.sub.2 = O. The
voltage output V.sub.3 of phase 1 is connected via modulator 7 and
summing unit 9 to the output and the voltage output V.sub.4 of
phase 2 is connected via modulator 8 and summing unit 9 to the
output. A modulation signal sin (.omega..sub.C t) is applied to
modulator 7 and a modulation signal cos (.omega..sub.C t) is
applied to modulator 8.
At the output of phase 1, therefore,
V.sub.3 = V/ 2(H(p)+ H(-p))
and on phase 2
V.sub.4 = j V/2(H(p)-H(-p))
If quadrature modulation is then applied to V.sub.3 and V.sub.4 as
shown in FIG. 32 the resultant output is ##SPC12##
The effect is as if the modulation was done first, followed by a
normal type of filter with the response H(p+ p.sub.c). For this
purpose the characteristic of the polyphase network would be as
shown in FIG. 30. The lower sideband would then be suppressed while
the upper sideband V(p+ p.sub.c) would be passed. This basic method
can be used for any number of phases.
It should be noted that it is possible to use the network of FIG.
32 without modulators and, thus, simply as a circuit to provide a
two-phase output from a single-phase input. This is provided the
network offers sufficient attenuation to negative sequence inputs
and passes positive sequence inputs. FIG. 31 shows a suitable
characteristic. In a similar manner, it is possible to generate an
N-phase output from a single-phase input.
* * * * *