CA 02275986 1999-06-23
WO 98/28858 PCT/AU97100873
-1 -
A TERMINATIION CIRCUIT
The present invention relates to a tE~rmination circuit for power distribution
lines, and a method for determining values of components of the circuit. The
present invention also relates to an isolation or conditioning circuit.
Power distribution systems provide an existing infrastructure which can
advantageously be used for the transmission of communication signals. A
difficulty
associated with establishing a communication system using a power distribution
system is providing an effective termination circuit for the lines of the
distribution
system. A further difficulty is also posed by a need to provide sufficient
isolation
between power signals traditionally distributed on the lines, and the
communications signals. Isolation may be required both at a customer's
premises,
and at connection points on lines of the system.
The configuration of multi-line power distribution systems varies
considerably depending on local factors, such as regulations, stage of
development
and physical characteristics. In Australia, power distribution lines are used
to
distribute to customers a high voltage and low frequency power signal, which
typically has a frequency of 50 hertz and a distribution voltage which may be
415
volt phase to phase (LV), 6.5, 11, 22 or 66 kilovolt phase to phase (HV) on
each of
the N lines of the system. The number of lines N usually ranges from 2 to 5
but may
be higher where both LV and HV systems are together in close proximity.
Typically
N=4 for a three phase 415 volt AC service, with one line being designated
neutral.
The wires of the lines may be open in that they are strung overhead with no
insulation and are separated by air. The wires may also be bundled, such as in
an
aerial bundle or for underground cables, where they are separated by
insulation
and covered.
When the lines reach an end point where they no longer need to be
continued, such as at the end of a street, they are not electrically
terminated. Other
CA 02275986 1999-06-23
WO 98128858 PCT/AU97I00873
-2-
open circuit points are also present where physical tension points are
included in
the lines or where power delivered from opposite directions meets an open
point.
Step down transformers, on the other hand, present an impedance discontinuity
which may be a short circuit for low frequency signals and a relatively high
impedance for high frequency signals. These open and short circuit
characteristics
inhibit efficient transmission of a low voltage radiofrequency (RF)
communication
signal. The problem could be addressed by including a termination circuit for
the
communication signal which impedance matches the lines at the open and short
circuit points so as to eliminate unwanted reflections of the communication
signal.
Yet it has proved particularly difficult to provide and correctly configure an
impedance matched termination circuit. The difficulties arise primarily
because any
RF pulse transmitted on the lines causes coupling between the lines, thereby
rendering it difficult to make effective impedance measurements to determine
component values for a termination circuit, particularly when a wide carrier
frequency band needs to be catered for.
The capacity of the distribution lines to transmit high frequency
communications signals is also inhibited by impedance discontinuities which
occur
at different points in the distribution system, and in particular occur at the
connection points made at most supporting power poles of an overhead system.
Communication services to customer premises would normally be delivered on
service wires to the premises and are typically connected to the rest of the
distribution system at the poles using junction boxes. Due to a shunting
affect
introduced by the impedance discontinuities at the junction boxes, a
considerable
amount of the communication signal power can be directed down the service
wires,
leaving tittle power for transmission further down the rest of the
distribution system
for other connection points. If not attended to, this can result in rapid
communication signal attenuation as it propagates down the fines of the system
to
other customers.
In accordance with the present invention there is provided a termination
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
circuit for N power distribution lines, having resistances r;~ connected
between
points P; on N-1 of said lines and between said points P; and a ground, which
is
connected to the remaining one of said lines.
The present invention also provides a method of determining values of the
components of the termination circuit, including:
determining matched termination values t;~ between said lines when at least
one of the lines is connected to a communication signal source and the
remaining
lines are connected to said ground;
setting said resistances r;~ to nominal values and measuring the resistance
between said points P; to obtain measured point impedances PI;~;
determining, on the basis of said t;~, final point impedances FPI;~; and
determining, on the basis of said t;~ and PI;~, sequence point impedances
SPI;~
which need to be set in sequence and measured to set the point impedances PI;~
in
the termination circuit to the final point impedances FPI;~.
The present invention also provide; a circuit for use in delivering a
communication signal on a power distribution system which distributes power
signals, said communication signal having a high frequency relative to a
frequency
of said power signals, said circuit including a transformer which has windings
for
each phase of the distribution system, and has no net flux for the power
signals,
and has a net flux for the communication signal.
Preferred embodiments of the presE~nt invention are hereinafter described,
by way of example only, with reference to 'the accompanying drawings, wherein:
Figure 1 is a circuit diagram of a preferred embodiment of a termination
circuit;
Figure 2 is a flow diagram of an impedance determination program;
Figure 3 is a circuit diagram of bridged power distribution lines;
Figure 4 is a circuit diagram of a preferred embodiment of an isolation
circuit
connected in a customer's premises;
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-4-
Figure 5 is a circuit diagram of a preferred embodiment of a conditioning
circuit;
Figure 6 is a circuit diagram of a first equivalent circuit of the circuit of
Figure
5;
Figure 7 is a circuit diagram of a second equivalent circuit of the circuit of
Figure 5; and
Figure 8 is a block diagram of a junction box incorporating the conditioning
circuit.
A termination circuit 2 for N lines 4, 6, 8 and 10 of a power distribution
system includes three parts 12, 14 and 16, as shown in Figure 1.
The first part 12 is an isolation part which is used to isolate the high
voltage
level low frequency power signal on the lines 4 to 10 from the low voltage
level
radiofrequency (RF) signal handled by the second and third parts 14 and 16 of
the
termination circuit 2. The power signal typically has a frequency of 50 hertz
and
one of the distribution voltages, such as 415 volt phase to phase, 6.5, 11, 22
or 66
kilovolt phase to phase on each of the tines 4 to 10. The radiofrequency
signal is
typically less than 1 volt rms with a frequency in a range of 2 to 100
megahertz.
Accordingly. effective isolation can be achieved by placing isolation
capacitors 18
in the lines 4 to 10.
The second part 14 of the termination circuit 2 is a driving point network
which includes a drive transformer 22 having its secondary coil 24 connected
to
input/output points P,, PZ and P3 of N-1 of the lines 4, 6 and 8. Drive
resistances
26, 28 and 30, having values r,) rz and r3 are connected in parallel to the
secondary
coil 24 between the coil 24 and respective points P,, Pz and P3, as shown in
Figure
1. The primary coil 32 of the transformer 22 is connected to an RF
inputloutput
coaxial termination 20 for inputting and outputting the RF communication
signal,
which is either placed on the points P,, Pz and P3 or received from the
points. The
remaining line 10 is connected to RF ground 34, together with the opposite
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
terminals of the coils 32 and 24 of the transformer and the outer sheath of
the
coaxial termination 20. The line 10 which is connected to ground 34 would
normally
be the neutral line.
The third part of the circuit 2 is a termination network 16 which is able to
terminate the N-1 lines 4 to 8 so as to present a matched impedance to any RF
communication signal received on the line:> 4 to 8. Assuming the lines 4 to 10
exhibit low losses, to absorb all RF signals incident on the fines 4 to 8, the
termination network 16 comprises N(N-1 )/2 resistors connected to the lines 4
to 10
in all combinations to cater for coupling between the lines. This involves
connecting
resistors with appropriate resistance values between all possible pairs of the
drive
points P,, PZ and P3 and between each of the drive points P,, PZ and P3 and
the RF
ground 34. As shown in Figure 1 for N=4, r~esistances r;~ provided by
potentiometers
36, 38, 40, 42, 44 and 46 are connected bE~tween respective pairs of points
P,, PZ
and P3 and between respective ones of the drive points P,, Pz and P3 and the
RF
ground 34.
The effectiveness of the termination network 16 can be shown by
considering an arbitrary voltage travelling vvave on the power distribution
system,
together with its corresponding current travelling wave, linked by the
inductance
per unit length and capacitance per unit length matrices of the distribution
lines 4 to
10. To provide a matched termination, the i:ermination network 16 needs to
maintain and appear to continue the relationship between the voltage and the
current travelling waves when the waves arrive at the network 16. The waves
can
be shown to be in phase so the termination network 16 has an admittance matrix
which represents a network of positive resistances interconnecting every line,
which is the form of the termination network 16 described above. Correct
determination and setting of the resistance values r;~ is described
hereinafter.
A unique set of resistance values r;~ (ij=0 to N-1 } need to be established
for
each particular configuration of N fines 4 to 10 being terminated. Existing
CA 02275986 1999-06-23
WO 98128858 PCT/AU97/00873
-6-
configurations of power distribution lines 4 to 10 vary not only in the number
N of
lines but also the size and spacing of the conductors for the lines 4 to 10.
The lines
4 to 10 may also be, as described above, closely bundled and twisted metal
cables
which are covered with insulating material.
Initially a set of matched termination values t;~ is determined by conducting
a
series of N(N-2)I2 experiments on the N lines 4 to 10, which may be an actual
section of the lines to be used or a simulation which constitutes a scale
model. If a
scale model is used, the dimensional proportions of a cross-section of the
fines 4 to
10 of the distribution system needs to be preserved. Each experiment involves
the
determination of a matched termination value t;~ for a particular bridge
configuration
on each end of the line 4 to 10. In this context "bridging" means connecting
an RF
short circuit between certain combinations of lines, at both source and load
ends.
Each bridging combination provides an RF "ground" line or set of lines and an
RF
"active" fine or set of lines. Each bridging combination is also independent
of the
others and identical at the source and load ends in a particular experiment.
For
these experiments a suitable pulse generator is connected at the source end,
via
.an impedance transformer if impedance mismatches justify, to the RF ground
and
RF active lines or sets of lines. The source end is monitored with an
oscilloscope.
A single adjustable termination resistor is connected at the load end between
the
RF ground and RF active lines or sets of lines. Each experiment then consists
of
setting the variable termination resistor whilst monitoring the reflected
signal at the
source end, such that no reflected signal is seen coming back from the load
end.
The value of this resistance is measured and comprises the matched termination
value t;~ for that experiment. The procedure is repeated for the other
experiments.
An example of the bridging conditions and symbols representing the resulting
matched terminations is shown in Table 1.
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
_7_
Table 1: Example of Bridge Conditions <RF active><RF ground>
[Matched Termination Symbol]
N 2 3 4 5
No. of Expts.1 3 6 10
Expt. No.
1A <1 ><p> <1 ><0._2> <1 ><0_2_3> <1 ><0_2_3_4>
[t1a] [t1a] [t, a] [tta]
1 B * <2><p..1 <2><0_1 _3> <2><0_1 _3_4>
> [t1 b] [t1 b]
[t1 b]
1 C * * <3><0-1-2> <3><0-1-2-4>
[t,~] [t,~]
1D * * * <4><0-1-2-3>
[t, d]
etc.
2A * <1 _2><0> <2_3><0_1 <2-3_4><0_1
[t2a] > >
[t2a] [t2a]
2B * * <1-3><0-2> <1-3-4><0-2>
[t2b] [t2b]
2C * * <1-2><0-3> <1-2-4><0-3>
[tz~] [tz~]
2D * * * <1-2-3><0-4>
[t2d]
etc.
3A * * * <1-2><0-3-4>
[t3a]
3B * * * <3-4><0-1-2>
[tab]
etc.
*Experiment not applicable for the value of N
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97100873
_g_
For example for N=4, referring to Tabie 1, in the first three experiments 1 A,
1 B and 1 C one of the lines 4, 6 and 8 including the corresponding drive
point P,, P2
and P3 is respectively connected to the RF active, whilst the remaining lines)
including the neutral line 10 are connected to RF ground 34. In each
experiment
the variable resistance is connected between RF ground and the fine which is
selected to be the RF active. In the three remaining experiments, 2A, 2B and
2C
the neutral line 10 and one of the other lines 4, 6 or 8 are connected to RF
ground
with the remaining two lines being used as RF active. The remaining two lines
are
connected to one other and connected to RF active, and again the variable
resistor
is connected between the RF active lines and the RF ground lines. A set of
matched termination values obtained for a set of open wire powerlines for N=4
is
shown in Table 2.
Table 2: Example of Matched Termination values for N=4 obtained from
experiments defined in Table 1
Matched Termination Value (ohms)
(tea) 363
(t~ b~ 333
(t~~~ 370
(tZa] 281
(t2b) 242
(t2~~ 224
Once the matched termination values t;~ have been determined, they can be
used in a procedure; which can be executed by a computer program 50 as shown
in Figure 2, to determine the final point impedances FP1;~ which need to be
seen
between the points P;. and also between the points P; and the RF ground 34, to
render the termination network 16 effective. The matched termination values
t;~ are
inputted at step 52 of the program 50 and at step 54 are transformed to
termination
network admittance elements g;~ for the network 16 using a linear transform
LT1. An
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
_g_
admittance matrix [YTN) is then obtained using a second linear transform LT2,
in
step 55, from the matched termination values t;~. Next, at step 56, the
admittance
matrix [YTN) is used to obtain the final point impedances FPI;~ using a third
transform
LT3. The final point impedances FPI;~ are the ultimate desired measured
resistance
values across the points P; (i=0 to N-1 ) for the termination network 16.
Given a set
of termination values t;~ expressed as admittances with the end conditions set
in
turn as described in Table 1, the values of the final point impedances FPI;~
can be
determined. For example, for N=4 in experiment 1 A the second and third lines
6
and 8 bridge to the RF ground 34 with the neutral line 10 at the source end,
and
the remaining first line 4 is used as the RF active to receive the RF test
signal. At
the opposite load end, the test signal will arrive in the same form that it
left source
end, and therefore bridging the load end will not affect the reflected wave.
Hence t,a
can be expressed in terms of the resistances r;~, or the corresponding
admittances
g;~, which are not short circuited by bridging for this experiment, as shown
in Figure
3. Hence it follows that t,a = 1/ro, + IIr,Z + 1!r,3 = go, + gt2 + gt3 . Other
expressions
follow similarly. Solving the g;~ in terms of tf~~e t;~ defines the transform
LT1. For
example for N=4, the linear transform LT1 is defined by the equations
g01= (t2b ttb _
~~2 + t2c t1c)
_
got- (t2c ttc _
i/z + t2a t1a)
_
g03= (t2a tta _
~~z + t2b t1 b)
_
g12= (t1a t2c)
~/2 + t1b
_
g23= (ttb _ t2a)
~~z + t1c
g31- (ttc t2b)
~~z + t1a
_
which are the values for g;~ used in step 66, as described hereinafter. The
transform
LT2 used in step 55 can be obtained by standard circuit analysis techniques.
The
linear transform LT2 for N=4 is defined by equations relating the terminating
network admittance matrix elements YTNij to the t;~ as follows
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-10-
YTN11- t1a
YTN22t1 b
YTN33t1 c
YTN12YTN21 1/ ~t2c t1 a
t1 b~
YTN23YTN32 ~~2 ~t2a t1
b t1 c~
YTN31YTN13 ~~2 ~t2b t1c
tla~
These nine numbers form the admittance matrix [YTN] relating the three
voltages at
the points P;, i=1 to 3 relative to RF ground Po 34 to the currents into these
points.
Inverting gives the impedance matrix elements ZTNij, which for the values of
Table 2
is as follows:
440.3 155.5 169.7
(ZTN] - [YTN) 1 = 155.5 443.0 214.1
169.7 214.1 496.7
The final point impedances FPI;~ are related to the ZTN;;, as determined in
step 56, are as follows
FPIo1= ZTN11
FPIo2= ZTN22
FPlo3= ZTN33
FPI12= ZTN11 + ZTN22
2ZTN12
FP123= ZTN22 + ZTN33
- 2ZTN23
FP131= ZTN33 + ZTN11
2ZTN31
which give the final values in Table 3 described below. The last two sets of
relationships connecting the YTN;~ with the FP I;~ define the transform LT3.
The driving point network 14 may not be included if all that is required is to
terminate the lines 4 to 10. However if the driving point network 14 is
present it wilt
affect the impedances seen from the N-1 lines 4 to 8. For the N=4
configuration as
shown in Figure 1, the driving point network 14 appears as an equivalent
termination network in parallel with the actual termination network 16 as seen
from
the points P,, PZ and P3. The driving point network 14 can be represented by
an
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-11-
admittance matrix [YppN] and therefore the effective termination network has
an
admittance matrix given by [YTNeff] equal to [YTN] + [YppN]. With or without
the driving
point network 14 the final point impedance~s FPI;~ will remain the same,
although the
settings of the potentiometers 36 to 46 will be different. Under DC
measurement
conditions, a resistor 49 is placed in parallel to the transformer 22 with a
resistance
value ro to represent the input impedance presented by the secondary coil 24
of the
transformer 22 in use, because in DC measurement conditions the coil 24
represents a short circuit. The transformer 22 is wound accordingly as an
impedance transformer with an RF source impedance, which is typically 50 ohms,
being on the primary coil 32. The impedance transformation is only an
approximation so the actual value of ro is chosen to be the resistance
measured
looking into the secondary coil 24 at the R1~ frequency of interest, i.e. the
frequency
of the carrier of the communication signal. For realisability, i.e. to produce
positive
resistance values, elements yTNeff ij and yppr, i; of the last two matrices
must therefore
satisfy yopN ij ~ yTNeff.j for every I~ =1 ... 3 for N=4 because yppN ij + yTN
ij - yTNeff ij~ TI11S
is used as an aid in determining the values for the drive resistances 26, 28
and 30.
The values of the drive resistances 26 to 30 are chosen to maximise power
delivery
to the tines 4 to 8 for a given RF input. The' drive resistances 26 to 30 can
be given
the same resistance value for similar signal levels on all lines 4 to 8, and
the sum
of the drive resistance values r, + rz + r3 is chosen to be greater than the
maximum
resistance r;j for realisability because a star-to-delta transformation of the
drive
resistances 26 to 30 puts r, + r2 + r3 in parallel with each of the interline
resistances
r;j. The r;j are set, as described above, to give a matching termination
network 16
and if the drive resistances are too high insertion loss will be excessive so
the
selection procedure should be repeated with smaller drive resistance values
until
the r;j are just realisable according to the program 50.
The final part of the program 50 involves determining a sequence of point
impedances SPI;j which can be measured and set in sequence to finally arrive
at
the desired final point impedances FPI;~. The sequence is important because
adjusting any of the resistance values r;j wlill affect the point impedances
Pl;j. At step
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-12-
58, the resistances r;~ are initially set by placing the potentiometers 36 to
46 in a
centre position and initial point impedances PI;~ are measured. As discussed
previously, during the measurement conditions the transformer 22 is out of
circuit,
but is represented by the resistor 49 having a value ro, which allows DC
measurements to be taken simulating actual impedances prevailing at operating
radiofrequencies. At step 60, the measured point impedances P I;~ are
transformed
to actual or measured element admittances g;~. A sequence determination loop
62 is
then entered at step 64 for k iterations. The number of iterations k is the
number of
impedances between pairs of the points P; and the points P; and the RF ground
34.
For N=4, k=6. In the first step 66 of the loop 62 one of the actual
admittances g;~ is
replaced by the desired admittance g;~ determined in step 54. All of the
admittances
are then transformed to point impedances PI;~ at step 68 using the
relationship
between g;~ and t;~ described previously and the transforms LT2 and LT3. At
step 70,
the kth point impedance in the sequence to be set SPI;~ is taken to have the
value
PI;~ corresponding to the impedance obtained in the matrix derived by step 68
and
corresponding to the element g;~ chosen in step 66. For example if g,o had
been set
to its desired value in step 66, then SPI,o is the first impedance in the
sequence
and is taken to have the value PI,o in step 70 which is derived in step 68. At
the
decision step 72 a determination is made as to whether all iterations of the
loop 62
have been completed. Once all iterations of the loop 62 have been completed
the
sequence values SPI;~ are recorded in order together with the final FPI;~
values at
step 74 and the program 50 completed. Table 3 below sets out the results
produced by the program 50 for N=4, where the first column specifies the point
impedances between pairs of lines, with Neutral corresponding to the RF ground
line 34, Red corresponding to the first line 4, White corresponding to the
second
line 6, and Blue corresponding to the third fine 8. The initial measured point
impedances PI;~ inputted in step 58 are set out in the first column, the
sequence
point impedances SP I;~ determined by the loop 62 are set out in the second
column
and the final point impedances FPI;~ are set out in the third column.
Therefore to
achieve the final point impedances FPI;~, firstly the impedance between the
red line
4 at point P, and ground needs to be set and measured at 381 ohms. Next the
CA 02275986 1999-06-23
WO 98/28858 PCT1AU97/00873
-13-
impedance between the white line 6 point PZ and ground needs to be set and
measured at 411 ohms, etc. until the impedance between the blue line 8 point
P3
and the red line 4 point P, is finally measuired and set at 598 ohms.
Table 3 Termination Network Impedances
Points ij Initial PI;~ Sequence SP(;~Final FPI;~
Red-Neutral 340 381 440.3
White-Neutral 352 411 443.0
Blue-Neutral 359 488 496.7
Red-White 445 563 572.3
White-Blue 460 496 511.5
Blue-Red 454 598 597.5
The termination circuit 2 can be used in a customer's premises, as shown in
Figure 4, to receive signals inputted on the distribution lines 4 to 10 from a
source
103. The distribution lines 4 to 10 also provide, according to their normal
function,
power to the customer's premises which constitutes a power load 100. In order
to
isolate the termination circuit 2 from the load 100, an isolation circuit 102
is used
which includes a toroidal core 104 placed in series with the lines 4 to 10
connected
to the load 100, and capacitors C,, CZ and C3 connected between the neutral
line
10 and the red, white and blue lines 4, 6 and 8, respectively. An impedance Zo
is
also connected across the coil of the toroidal core 104 for the neutral line
10, 160.
The isolation circuit 102 provides isolation for the termination circuit 2
from
spurious noise and impedance effects of the load 100 at RF frequencies whilst
allowing maximum demand current to pass to the load 100 at the mains frequency
of the power distribution system. The toroidal core 104 has different
characteristics
for the mains frequency and the RF frequE~ncies. At the mains frequency the
coils
for each phase are wound on the ring 104 such that the magnetic fluxes add. A
coil
for the neutral line 10 is also wound but in such a way that it cancels the
flux
produced by the phases of the other lines 4 to 8 so that net flux at mains
frequency
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-14-
in the toroidal core 104 is 0, guaranteeing that the ring wilt not saturate
due to the
mains current. This ensures the isolation circuit 102 presents a very low
impedance
to the source 104 at mains frequency. At RF frequencies, the neutral line 10
is
substantially bypassed by the impedance Zo to produce a net RF flux in the
ring
and hence introduce an inductance which is used as part of an RF filter of the
circuit 102. The impedance Zo includes a capacitor Co and resistor Ro in
series. Ro
is included to prevent a magnetic short circuit for the active phases. The
capacitors
C,, CZ and C3 form the remainder of the RF filter of the circuit 102. This
ensures the
circuit 102 presents a high impedance for the termination circuit 2 at RF
frequencies. The capacitors C,, Cz and C3 shunt any RF signals, such as RF
noise.
output by the load 100.
The toroidal core 104 can also advantageously be used as part of a
conditioning circuit 150, as shown in Figure 5, for use in connecting service
cables
152 from a customer's premises to the overhead distribution lines 4 to 10. The
customer service lines 152 which run from the overhead lines 4 to 10 to the
customer's premises include red, white, blue and neutral lines 154, 156, 158
and
160 for a three phase service. The red, white, blue and neutral distribution
lines 4
to 10 are connected by respective series windings about the core 104 to the
red,
white, blue and neutral customer service lines 154 to 160, respectively. An
impedance Zo is again connected across the winding for the neutral lines 10
and
160, whereas respective conditioning impedances ZR, ZW and ZB are connected
across the windings for the remaining lines. The core 104 is again wound so
that
for the mains frequency, the magnetic fluxes add around the core for the red
and
white and blue phases, and the winding for the neutral line is such that it
cancels
the flux produced by the remaining phases, so that the net flux at the mains
frequency in the core 104 is zero. The value of the capacitance Co of the
impedance Zo is also selected such that at the RF frequencies, the neutral
line 10,
160, substantially bypasses the core 104, to thereby produce a net RF flux in
the
core 104. Ro is again used to prevent a magnetic short circuit for the active
phases.
The impedance value presented to the overhead lines 4 to 10 can therefore be
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-15-
varied between the mains frequency and the RF frequency, to provide impedance
conditions which are unchanged for power distribution on the service cables
152,
and which also prevent rapid signal attenuation of the communication signal at
the
RF frequency. The difference between the mains frequency and the RF frequency
used for the communication signal enables the conditioning circuit 150 to
present
an inductance value to the lines 4 to 10 wf rich has a very low reactance at
50 hertz
but a high reactance at RF. The circuit 15C) also does not present any
problems
with saturation due to the potentially large currents at the mains frequency,
as the
sum of the net flux in the core 104 will be zero.
With regard to the behaviour of the conditioning circuit 150 at the RF
frequencies, an equivalent circuit 170 is shown in Figure 6. At the RF
frequencies,
the inductances of the windings for each phase LR, LW and LB will all have
significant reactances. The leakage inductances L~R, LAW and LAB will also
have
significant reactances as the transformer 104 is not tightly coupled and only
a small
number of turns is used. There is also a falling off of magnetic permeability
at the
RF frequencies. It can be seen from the eq;uivaient circuit 170, that the
impedances
presented at each input for each active phase to the customer premises, i.e. R-
R~,
B-B~ and W-W~, wilf be high for the RF frequencies, thereby preventing
significant
loss or rapid attenuation of the communication signal for each set of service
cables
152 along the distribution system. It can be shown that the combined input
impedances for the active phases is the sum of the impedances of the overhead
distribution lines 4 to 10 and the service lines 152 all sharing a common
neutral line
10, 160, and accordingly will be high. The impedance seen at the input to red
phase of the core 104 due to the white and blue phases connected to the core
104,
will be of the same order as XR, being the reactance of LR.
It may however occur that the signal arriving on the service cables 152 for
each of the active phases 154 to 158 may be unfavourably out of balance. This
can
be addressed by including, within the circuit 150, the conditioning impedances
ZR,
ZW and ZB, which comprise resistors and capacitors connected in series across
the
CA 02275986 1999-06-23
WO 98/28858 PCT/AU97/00873
-16-
phase windings of the core 104, as shown in the RF equivalent circuit 172 of
Figure
7. The capacitors of ZR, ZW and ZB are chosen so that they present a short
circuit at
the RF frequencies, but provide a blocking impedance for power signals at the
mains frequency. This allows the resistors of ZR, ZW and ZB to be adjusted and
selected so as to balance the RF signals on the phases submitted to the
customer's premises. The circuit 150 also inherently acts as a signal
equalising
device via the coupled windings of the core 104. The circuit therefore
produces a
reactive isolation and conditioning effect which can be adjusted as desired
depending on the number of the turns of the windings, the size and the
material
used in the core 104, and the values chosen for the conditioning components
ZR,
ZW, ZB and Zo.
The circuit 150 can be incorporated into a junction box 180, as shown in
Figure 8, mounted on the supporting poles of an overhead distribution system
to
connect the overhead distribution lines 4 to 10 to the customer service lines
152 to
160.
Many modifications will be apparent to those skilled in the art without
departing from the scope of the present invention as hereinbefore described
with
reference to the accompanying drawings.
