Multi-Function Load Controller for Carrier Load Control Subsystem

The invention disclosed herein relates to two-way communication and control systems. The following four commonly assigned United States patent applications were all filed in the United States Patent and Trademark Office on June 28, 1984 and relate to such communication and control systems: Ser. No. 625,747 filed by William R. Verbanets and entitled Multipurpose Digital IC for Communication and Control Network: Ser. No. 625,863 filed by William R. Verbanets and Theodore H. York and entitled Improved Digital IC-Microcomputer Interface; Ser. No. 625,862 filed by Leonard C. Vercellotti and William R. Verbanets and entitled Low Error Rate Digitial Demodulator; and Ser. No. 625,864 filed by Leonard C. Vercellotti, William R. Verbanets and Theodore H. York entitled Digital Message Format for Two-Way Communication and Control Network.

A. Field of the Invention

The present invention relates generally to two-way communication and control systems, and, more particularly, to communication and control systems in which a number of remotely positioned controllable devices such as circuit breakers, motor control relays, lighting systems, and the like may be controlled from a central or master controller over a common network line.

While the invention is of general application in such control systems, it is particularly suitable for use with and will be described in connection with a line carrier control subsystem intended for residential use in conjunction with a load management terminal associated with the watt hour meter of the residence and controlled by a public utility over its power distribution system. Specifically, the present invention is concerned with a low cost software based digital IC which is pin configurably to operate either as a central controller associated with the utility load management terminal, or as a slave unit associated with a particular circuit breaker controlling a predetermined residential load device, or as a customer service unit by means of which the residential utility customer can be informed of the existing utility rate schedule, i.e. the cost of AC power per kw hour, and can control certain slave units associated with different residential loads by selectively activating corresponding switches on the customer service unit.

B. Description of the Prior Art

Various communication and control systems have been heretofore proposed for controlling a group of remotely located devices from a central controller over a common network line, such as a power line. For example, in Feiker U.S. Pat. Nos. 4,173,754 and 4,185,272 an arrangement is disclosed wherein a redundant digital address and function code is generated at a control station which is transmitted over a power line to a group of receiver-decoders individually associated with different controlled apparatus. However, the central station requires a relatively complex and expensive CPU, such as an Intel 8080A, and one or more separate random access memories. Also, the receiver-decoder at each remote station is of the type shown in Eichelberger et al U.S. Pat. No. 4,091,361 wherein a large number of hardware components are employed to receive and decode a message and exercise a desired control function. With such an arrangement it is not possible to provide a common software based digital IC which can function as either a central controller or a remote receiver-decoder by simply selecting different pins on the digital IC to establish the desired function of the device. Furthermore, in the prior art arrangement of these patents noise immunity is said to be achieved by redundantly transmitting the digital address and function code and delaying performance of the desired control function until the correct address portion of the code has been received at least two times out of three. However, such an arrangement substantially reduces the throughput of the system because it requires transmitting each message a minimum of three times irrespective of the noise level on the common network line.

Eichelberger et al U.S. Pat. No. 4,213,182 and Miller U.S. Pat. No. 4,167,786 also disclose a communication and control system in which the central controller requires a relatively complex and expensive CPU and the load control processor at each remote station is hardware based. Also, the system disclosed in these patents requires the transmission of complementary bytes for both address and data to provide security against errors in transmission. In Miller, et al U.S. Pat. No. 4,367,414 a further type of communication and control system is disclosed in which different types of microcomputers are employed at the central station and at a large number of remote stations which are connected to the central stastion by means of a dedicated twisted pair line as the common network line. U.S. Pat. Nos. 4,213,182, 4,167,786 and 4,367,414 are not concerned with the transmission of modulated carrier signals over a power line or the reception of such signals in the noisy environment usually encountered on power lines.

There have also been developed various types of communication systems for electrical distribution networks wherein a power line carrier system is employed to provide communication between the central station and load management terminals which are utilized as end devices in the utility's distribution network communication system. Such communication systems and load management terminals are shown in Whyte U.S. Pat. No. 3,911,415, Whyte U.S. Pat. No. 3,942,168, York U.S. Pat. No. 4,427,968, Kennon et al. U.S. Pat. No. 4,402,059 and Kennon U.S. Pat. No. 4,429,366.

In the communication and control system of the present invention a low cost 4 bit microprocessor is used as the common building block of the system and is programmed to function either as a central controller or as a remote load control unit, these different functions being selectable by simply grounding certain pins of the software based digital IC device. This basic unit can also function as a customer service unit in load management control subsystems to inform the customer of existing utility rates and permit him to override an existing load management schedule if he so desires. When functioning as a central controller the software-based digital IC accepts a base band data message from an associated load management terminal which includes the relative address of the central controller and a function portion corresponding to a desired function to be executed at a specific remote station. The address portion of this LMT message is then changed to the absolute address of the remote station and a corresponding carrier modulated signal is generated which includes start bits of predetermined logic value, the new address, the function of the LMT message and a BCH error checking code which is calculated based on the preceding bits of the signal and is added at the end thereof to form the complete transmission signal.

In the central controller mode the digital IC also receives a modulated carrier reply message from the addressed remote station, calculates a BCH error code from the bits of the received message and compares the calculated code with the BCH error code at the end of the end of the reply message. If no error in transmission is indicated by such comparison, the absolute address of the addressed remote station is modified to the original relative address and a corresponding unmodulated base band data message is sent back to the load management terminal which includes the reply of the addressed slave. If no reply message is received from the addressed slave, the function code portion of the original LMT message is modified to indicate no reply from the slave and is retransmitted back to the LMT.

In the remote load control mode, the digital IC receives a modulated carrier transmission signal over the power lines and calculates a BCH error code from the bits of the received message and compares the calculated code with the BCH error code at the end of the received message. If no error in transmission is indicated by such comparison, the absolute address of the remote load control device is compared with the address portion of the received message. If they are equal the function portion of the received message is decoded and a control function, such as the closing of a circuit breaker, is performed in accordance therewith. If the decoded message includes a request for a reply, a modulated carrier signal is developed by the digital IC which includes start bits of predetermined logic value, the absolute address of the remote device, the status of the controlled load, and a BCH error checking code which is calculated based on the preceding bits of the signal and is added at the end thereof before it is supplied to the common power line.

When the digital IC is employed in a carrier load control subsystem and is operated in the customer service mode, the function code includes instruction bits for controlling a plurality of indicating devices which are selectively energized to inform the customer of the then existing peak or shoulder condition of the utility load curve, this information being supplied from the utility to the residential load management terminal. In the customer service mode, the reply message which is generated includes the status of a plurality of customer controlled load switches on the customer service unit which load switches correspond to different residential loads. The reply message also includes the status of a customer controllable override switch by means of which the customer can request that the loads he selects by the customer controllable load switches be shed or restored in contravention to the then existing utility load schedule.

In accordance with a further aspect of the communication and control system of the present invention, the signals transmitted in both directions over the power line comprise an on-off keyed carrier signal in which the presence of carrier corresponds to a logic "0" and the absence of carrier a logic "1". With such an arrangement, noise on the power line is always falsely detected as a carrier tone, i.e. transforms a data "1" to a data "0". To provide improved noise immunity, each data bit includes a first carrier-on portion of predetermined duration and a second data portion in which carrier is present or absent depending upon the logic value of the data bit. In detecting each data bit the incoming carrier is sampled a number of times beginning in the middle of the carrier-on portion of the first data bit to the middle of the carrier-on portion of the next bit. If any of the samples are logic "1", i.e. absence of carrier, the data bit is treated as a "1". Only if all of the samples are "0", i.e. presence of carrier, is the data bit treated as a "0".

The invention, both as to its organization and method of operation, together with further objects and advantages thereof, will best be understood by reference to the following specification taken in connection with the accompanying drawings in which:

FIG. 1 is an overall block diagram of the described carrier load control subsystem;

FIG. 2 is a block diagram of the multipurpose load controller employed in the system of FIG. 1;

FIG. 3 is a diagram of the message bit format employed in the system of FIG. 1 to transmit messages from a load management terminal to a central load controller.

FIG. 4 is a diagram of the message bit format employed to transmit carrier modulated messages to and from the central load controller to a remote unit;

FIG. 5 is a diagram of the message bit format employed to transmit a message from the central load controller to the load management terminal;

FIGS. 6 to 9, inclusive, when arranged in the manner shown in FIG. 6A, comprise a detailed schematic diagram of the multipurpose load controller of FIG. 2;

FIG. 10 is a circuit diagram of a portion of the switch matrix employed in the controller of FIG. 2 and shown without back-biased diodes to illustrate how incorrect signal paths may be generated;

FIG. 11 is a circuit diagram similar to FIG. 10 but shown with the diodes included;

FIG. 12 is a timing diagram of a portion of the message format of FIG. 4;

FIG. 13 is a logic flow diagram showing the power up sequence employed in the controller of FIG. 2;

FIG. 14 is a logic flow diagram showing the operation of the controller of FIG. 2 in an addressed slave mode;

FIGS. 15 and 16 are flow diagrams of subroutines employed in FIG. 14; and

FIGS. 17a and 17b are a logic flow diagram showing the operation of the controller of FIG. 2 in a central load controller (CLC) or master mode.

Referring now to the drawings, and more particularly to FIG. 1 thereof, the present invention is therein illustrated in conjunction with a utility distribution network load management system. More particularly, the secondary 20 of the distribution transformer 22 is connected to a number of residences, or factories, one of which is shown in FIG. 1 as having the watt-hour meter 24. A load management terminal LMT includes suitable logic circuits indicated generally at 26, and is coupled to the distribution secondary 20 through a coupler 28 and a transmit/receive modem 30. The load management terminal LMT, which is preferably microprocessor based, is arranged to receive messages from and transmit messages to the central utility over the power distribution system. These messages may include commands to shed or restore load at any one of a number of remote stations associated with different loads in the home or factory, one of which is indicated at 40 in FIG. 1 and includes the circuit breaker 42 for controlling one of these loads from the distribution secondary. More particularly, the load management terminal 26 is arranged to develop command messages having the format shown in FIG. 1A which are supplied over a bi-directional serial data line 44 to a central load controller indicated generally at 46, under the control of the clock line 48 from the LMT 26. The central load controller (CLC) 46 includes a controller 50 which is preferably a software based digital IC which is pin configurable to operate in any one of several different modes, as described in detail hereinafter, the controller 50 being coupled to the distribution secondary through the coupler 52 and the transmit-receive modem 54. At the remote station 40 a secondary carrier controlled remote module (SCCR) is employed, this module including the controller 50a which is similar to the controller 50 but is pin configured to operate as a remote load control slave, the module 58 also including the coupler 52 and modem 54, as in the central load controller 46 module.

The load management terminal 26 is also arranged to send messages to and receive messages from the CLC 46 which are intended for a customer service unit (CSU) 60. The CSU 60 includes an SCCR module 58 which is indentical to the module 58 used in the remote station 40. However, the controller 50a of the SCCR unit at the CSU station is employed to control a pair of LED's indicated as "P" and "S" which are employed to indicate periods during which a peak load condition "P" or a shoulder condition "S" of the utility load curve is in force. The customer service unit also includes an override switch the status of which is supplied to the controller 50a to indicate the desire of the customer to override an existing load condition in selected ones of the utility controlled loads of that customer.

In FIG. 2 there is shown in block diagram form the basic module which, in accordance with the present invention, can be used either as the central load controller 46, the remote station SCCR 58 or the customer service unit SCCR 58 of FIG. 1. More particularly, this basic module 70 comprises a microprocessor 72 which is arranged to be controlled by the mode switch 74 so that the module 70 can function either as a central load controller (CLC) or as a slave breaker control unit (SCCR) at a remote station, or as a customer service unit. The microprocessor 72 is arranged to receive base band data messages from and transmit base band data messages to the LMT 26 over the conductor 44 under the control of the clock line 48 from the LMT 26. The microprocessor 72 also developes a modulated carrier signal which is supplied through the transmit-receive modem 54 and the AC line coupler 52 to the distribution secondary line 20. A 4.times.4 switch matrix 76 is provided in the module 70, 12 nodes of the matrix 76 being employed as a field configurable offset address when the module 70 is operated as a central load controller. The other four nodes of the matrix 76 are employed as customer operated load switches for four loads which may be selectively controlled by the customer when the module 70 is used as a customer service unit (CSU). When so used the microprocessor 72 also supplies signals to "S" and "P" LEDs 78 to inform the customer of the existing utility load curve and rates existing at that time. A function (F1) switch 80 is also connected to provide an input to the microprocessor 72, the switch 80 functioning as an override switch when the module 70 is employed as a customer service unit.

When the module 70 is employed in a secondary carrier controlled remote (SCCR) 40, the microprocessor 72 provides a control signal to a breaker control unit 82. Also, when the module 70 is used as an SCCR unit, the 12 nodes of the switch matrix 76 are employed as the absolute address of this remote station. In the SCCR mode of the module 70 the other four nodes of the switch matrix 76 are not utilized. A power supply 84, which is energized from the AC line 20, is employed to supply power to the elements of the module 70, as will be described in more detail hereinafter.

Considering now the general functioning of the communication and control system described thus far, the format of the messages transmitted between the LMT 26 and the master or central load controller CLC 46, and the modulated carrier messages transmitted from the master module 46 over the power line 20 to the SCCR units which are operated either to control a remotely located breaker or to serve as a customer service unit (SCU), the LMT 26 initiates communications over the carrier subsystem in accordance with instructions received from the utility over the primary carrier control system and the CLC 46 idles in a receive mode. In the idle state the clock line 48 is asserted high by the LMT 26 and a valid data bit is present on the serial data line 44 when the LMT 26 causes the clock line to make a high-to-low transition. Additional data bits follow on subsequent clock high to low transitions until the message from the LMT 26 to the CLC 46 is completed. The period of the clock line 46 is from four milliseconds minimum to 1.0 seconds maximum for each state (either high or low). In the event that prior to the completion of a message in progress, the CLC 46 finds the clock period to exceed the 1.0 second maximum, the CLC 46 immediately reverts back to the idle state waiting for a new command word from the LMT 26. After the LMT 26 sends a command word, it delays the clocking in of the reply word from the CLC 46 for a minimum of 3.0 seconds and a maximum of 15.0 seconds. For delays greater than this maximum time, the CLC 46 immediately reverts back to the idle state waiting for a new command word.

The format of the LMT to CLC message is shown in FIG. 3 and consists of 22 bits, the first two bits being start bits having logic values "0" and "1" respectively. The start bits are followed by a 12 bit SCCR relative address and an 8 bit function code. The 12 bit SCCR relative address comprises the actual address of the slave, or CSU for which the message is intended minus an offset address which is field programmable by the switches in the switch matrix 76 so as to provide a hardware settable offset (modulo 4096) of the targeted SCCR.

In the 8 bit function code of FIG. 3 the numbers 0-3 and 8-255 are unused although they may be interpreted as a "send status" command, if desired. The 8 bit function code has different meanings depending upon whether the command message from the LMT 26 is intended for an SCCR operating as a breaker control, or an SCCR operating as a customer service unit (CSU). Considering first the active numbers of the 8 bit function code, when intended for a breaker unit function code number 4 commands the addressed slave to close the associated circuit breaker and send the status thereof back to the central load controller 46. Function code number 5 commands the SCCR to open the circuit breaker, disable a 12 minute timer and second status. Function code number 6 commands the addressws slave to open the associated circuit breaker, enable a 12 minute timer and send status. Function code number 7 requires no operation by the addressed slave but merely commands it to send the existing status of the circuit breaker.

Considering the 8 bit function code when sent to a customer service unit (CSU), function code number 4 commands the CSU to turn both the "S" and "P" LEDs 78 off and send the status of four customer load switches associated with loads A, B, C and D and the state of the customer override switch 80 (FIG. 2). These four load switches are the first four consecutive bit addresses identified in the 16 bit switch matrix 76. Function code number 5 commands the CSU to turn the "P" LED off, turn the "S" LED on and send the status of the customer load switches and the customer override switch. Function code number 6 commands the CSU to turn the "P" LED on, turn the "S" LED off and send the status of the customer load switches and customer override switch. Function number 7 commands no operation by the CSU but merely requests that it send status back to the LMT 26.

Upon receipt of a command from the LMT 26, CLC 46 adds its 12 bit address offset to the 12 bit SCCR relative address of FIG. 3 to provide the 12 bit SCCR absolute address. The CLC then develops a modulated carrier message of the format shown in FIG. 4. Refering to this figure, the first start bit of "0" consists of a carrier on interval followed by a "1" start bit in which the carrier is turned off. The carrier is then modulated in accordance with the 12 bit SCCR absolute address and the 8 bit function code shown in FIG. 4. As these bits are being transmitted, the CLC 46 also computes a 6 bit BCH error code for the message which is being transmitted from the CLC to the designated SCCR over the power line, this 6 bit BCH error code being added at the end of the modulated carrier signal as shown in FIG. 4. Preferably, this BCH error code is calculated from first bit to last using the root polynomial X.sup.6 +X.sup.5 +1.

The modulated carrier message shown in FIG. 4 is impressed upon the power line 20 through the coupler 52 of the central load controller 46 and is received by all of the SCCR units 58 connected to the line 20. Each SCCR unit compares its absolute address with the 12 bit SCCR absolute address of the received message. If these addresses compare equally the addressed slave SCCR unit then calculates its own BCH error code based on the bits of the received message preceding the final 6 bit BCH error code thereof and compares it with the last 6 bits of the received message. If the calculated BCH error code is equal to the received message BCH bits the SCCR then executes the commanded function. Thus, if the addressed SCCR is a breaker unit and the function code is number 4 the SCCR closes the breaker contacts 42 (FIG. 1). If the function code is number 5 the SCCR unit opens the breaker and disables a 12 minute timer. If the function code is number 6 the breaker contacts 42 are opened and a 12 minute timer is enabled at the end of which the breaker contacts are closed.

If the addressed SCCR is a customer service unit and the function code is number 4 both the "S" and "P" LEDs 78 are turned off. If the function code is number 5 the "P" LED is turned off and the "S" LED is turned on and vice versa if the function code is number 6.

After the commanded function has been executed the addressed SCCR assembles a reply message which consists of the 0, 1 start bits of FIG. 4, the 12 bit SCCR absolute address and an 8 bit status reply in place of the 8 bit function code which was sent from the CLC 46 to the addressed SCCR. The addressed SCCR also computes a 6 bit BCH error code which is added to the end of the reply message, this reply message being modulated on a carrier and transmitted over the power line 20 back to the CLC 46.

The CLC 46 receives the status reply message of FIG. 4 and computes its own 6 bit BCH error code which is compared with the last 6 bits of the status reply message. If the two BCH error codes compare equally, the CLC assembles a message to be transmitted to the LMT 26, in the format shown in FIG. 5. More particularly, referring to FIG. 5 the 12 bit SCCR absolute address of FIG. 4 is modified to subtract the address offset assigned to the CLC module 46 and the 8 bit SCCR status reply is added thereto and sent to the LMT as a base band message. The 8 bits of the status reply are identified as bits F.sub.7 -F.sub.0, the F.sub.0 bit being the last bit sent in the SCCR-LMT reply message. If, during the initial transmission of a modulated carrier signal from the CLC 46 to the addressed SCCR, no reply message was received from the SCCR, the CLC 46 sets the F.sub.7 bit to 1 and maintains the bits F.sub.6 -F.sub.0 with the same logic values as the corresponding function bits in the command message shown in FIG. 3.

If, on the other hand, the addressed SCCR receives the message and assembles a reply message the status reply F.sub.7 will be zero. Assuming that the addressed SCCR is associated with a customer service unit (CSU) the F.sub.6 status reply bit is unused. The status reply bits F.sub.5 -F.sub.2 are set in accordance with the states of the CSU customer load switches A, B, C, and D, it being recalled that these switches comprise four switches of the matrix 76. The F.sub.1 status reply bit represents the state of the customer override switch 80 (FIG. 2) and the F.sub.0 status bit is unused by the CSU unit. Assuming that the addressed slave is associated with a circuit breaker, the status bits F.sub.6 -F.sub.1 are unused. The F.sub.0 status reply bit is a 0 if the circuit breaker is closed and a "1" if the circuit breaker is open.

Considering now the hardware employed in the communication and control system of the present invention, and referring to FIGS. 6-9, inclusive, the controller 50 (FIG. 1) comprises a microprocessor 90 which is preferably a commercial type COPS 402 microprocessor manufactured by National Semiconductor Inc., or equivalent. All logic functions of the multipurpose module of FIGS. 6-9, are carried out by the microprocessor 90. In the illustrated embodiment the microprocessor 90 is customized through a use of the memory chip 108 and an address latch 106. With the specific ROM code set forth below in Appendix A, the microprocessor 90 is transformed into a customized multipurpose unit which can perform all of the necessary functions to provide both master and slave units in the communication and control system of the present invention.

The microprocessor 90 is controlled by a 3.58 MHz crystal 92 for accurate repeatable time interval measurements. Resistors 94 and 96 and capacitor 98 bias the crystal for proper oscillator operation. The initialization circuit of the microprocessor 90 comprises the diode 100, the capacitor 102 and a 4.7K ohm pullup resistor 104. With this external circuit arrangement the initialization input signal is delayed by the RC time constant so that the power supply 84 (FIG. 2) has time to stabilize before the microprocessor 90 begins operation.

Lines IP-0 through IP-7 of the microprocessor 90 are bidirectional address/instruction data lines. In the illustrated embodiment an address latch 106 is employed together with a programmable read only memory chip 108, the address latch 106 being preferably of the commercial type 74C73 and the memory chip 108 a commercial type 2716, manufactured by Intel Inc. For proper operation of the address latch 106, the lines IP-0 through IP-7 of the microprocessor must be pulled up by the pull up resistors RN1-1 to RN1-8. The output of the address latch 106, plus the lines IP-8 and IP-9 of the microprocessor 90 and the switch S3 form the address to the PROM 108. While the addressing range of the microprocessor 90 is only 1K bytes, and the PROM 108 has 2K bytes of memory, only one-half of this memory is used at a time, either the upper of lower 1K. The switch S3 is used to select which half of the PROM 108 the microprocessor 90 will address. With switch S3 closed, the lower half is selected which contains the operational program for providing both master and slave functions for the microprocessor 90, depending upon the position of the mode select switch 74, as discussed in detail heretofore. With the switch S3 open, the upper half of the memory 108 is selected which may contain a test program used to set up and check both the CLC module and the overall system operation. The A/D line of the microprocessor 90 is used to latch the address into the PROM 108 and enable the instruction to be output from the PROM 108 to the microprocessor 90. If desired, the microprocessor 90 may comprise a COPS 420 type microprocessor which has 1K of internal programmable read only memory, preferably mask programmable, in which case the address latch 106 and the PROM 108, together with the pull up resistors RN1-1 to RN1-8 are not required.

The 4.times.4 switch matrix 76 is comprised of a series of address switches SW4-1 to SW4-8 and SW5-1 to SW5-8 which are preferably single pole single throw switches in a dual in line package. Any given address bit is set to a "1" by closing the proper switch. The 12 address switches SW4-1 to SW4-8 and SW5-1 to SW5-4 are employed to establish the address of the controller, when operated in either the master or slave mode. In connection with the CLC module 46, these address switches establish the address offset which is added to the LMT message to provide the absolute address of the desired slave. in the case of the SCCR units, these address switches establish the absolute address of the remote station or customer service unit.

The four switches SW5-5 to SW5-8 are customer operable switches which correspond to four different loads of the residence or factory, identified as loads A, B, C and D which may be selectively controlled by the customer through the customer override switch 80. These four customer operable switches are combined withe the twelve address switches to form a 4.times.4 matrix which simplifies scanning of all of these switches by the 4 bit microprocessor 90. More particularly, the matrix 76 is scanned by a common row, column method in which the signal lines L0-L3 are row output lines from the matrix. These lines are pulled down to a low logic level by resistors 120, 122, 124 and 126. Lines D0-D3 are column input lines to the matrix. The matrix is scanned by setting 1 (and only 1) of the D lines high to a logic "1". This selects the column to which this D line is connected for reading. Any switch which is closed in that selected column will connect the D line to an L line bringing that line high. Thus, the L lines will be high wherever there is a switch closure. Since each switch has a unique row-address, all four switches in a particular column, which are closed, can be identified by successively reading the L lines.

Diodes D1-D16 are used to prevent "phantom" switch closures from appearing on an L line, when the switch for that selected row is open. Without the diodes D1-D16, sneak paths could appear since the switches are always open or closed, as opposed to a keyboard where the switches are normally open and are only momentarily closed. Also, without the diodes D1-D16 the D lines are shorted together. This is readily apparent by reference to FIG. 10 wherein a portion of the matrix is shown without the diodes D1-D16. Referring to this figure, when column 1 is selected, only S6 is closed and L0 is the only matrix output line that should be high. However, assuming that S1 and S3 are also closed in an assumed address, when S3 in column 0 is closed, matrix column input line D0 is connected (shorted) to line D1 through switches S3 and S6. Assuming that line D0 does not have sufficient drive to pull line L0 low, then line D0 will also be high. Since switch S1 is closed and D0 is now high, output line L2 is also high. Reading the L lines, it appears as if S4 and S6 are closed in column 1, while in actually only S6 is closed. The S4 closure is a "phantom closure". If the D0 line does have sufficient drive to pull line D1 low, then line L0 will be low because of the S3 closure between D0 and L0. In this case, no switch in column 1 will be read as closed, since D1 can never go high to drive any L line high.

FIG. 11 illustrates how the diodes D1-D16 prevent the above described sneak paths from occurring. When D1 drives L0 high through the closure of S6, D0 cannot be driven high, or D1 cannot be shorted low (depending on the drive capabilities of D0) because diode D3 being reverse biased acts as an open switch. Thus, no sneak paths and the related phantom switch closures can occur.

When the mode select switch 74 is closed the input line S1 to the microprocessor 90 is low, which informs the microprocessor 90 that the unit should operate as the master or central controller i.e. the CLC module 46 of FIG. 1. When the switch 74 is open, the line S1 is pulled high through the resistor 128 which causes the unit to operate either as an SCCR module 40 or a customer service unit 60 depending upon the external components connected thereto. More particularly, if the unit is to operate as a breaker control SCCR unit the G0 output line of the microprocessor 90 is supplied to the gate electrode of an external triac 130 through the resistor 132. The output of the triac 130 controls a relay coil 134 so as to cause the appropriate circuit breaker contacts to open or close. The G0 signal is also applied through the transistor 136 (FIG. 6) to an LED 138 so that this LED is energized during periods when the relay coil 134 is energized.

In the case of a customer service unit, the L5 and L6 lines of the microprocessor 90 are employed to energize the "S" LED 140 and the "P" LED 142 through the resistors, 144 and 146, respectively, the lines L5 and L6 being controlled in accordance with the function code of the message transmitted from the LMT 26, as described in detail heretofore. When the microprocessor 90 is functioning in the customer service unit mode, it also samples the four customer operable switches SW5-5 to SW5-8 to determine the state of these switches, this data being incorporated in the status portion of the reply message (FIG. 4) which is transmitted over the power line 20 back to the CLC module 46 and through this module to the LMT 26. The status of the customer override switch 80 is also read by the IN1 line of the microprocessor 90 and is incorporated in the F1 status reply bit (FIG. 5) of the reply from the customer service unit through the module 46 to the LMT 26.

Considering now in more detail the transmitter/receiver 54 and coupler 52 in the multipurpose module shown in FIGS. 6 to 9, which can function as either the module 46 or the module 58 in FIG. 1, the receiver portion of this unit comprises a preamplifier 150 and a phase lock loop receiver 152. Preferably the preamplifier is an operational amplifier of commercial type LM101 and the phase lock loop circuit is a commercial type NE567. The modulated carrier signal which is supplied to the AC power line 20 from the CLC module 46 and intended for one of the remote stations 40, is coupled through a nichrome link 154, a capacitor 156, a resistor 158 and a capacitor 160 to the inverting input of the operational amplifier 150. A resistor 162 is connected from the junction of the resistor 158 and the capacitor 160 to ground to provide a voltage divider, the resistors 158 and 162 preferably being 1% resistors and having the values of 750 ohms and 267 ohms, respectively.

The preamplifier 150 is operated as a pass band amplifier and is designed to have a low frequency cut off of 95 Khz, a high frequency cut off 125 Khz and a peak gain of 7. A feedback resistor 164 is connected from the output of the preamplifier 150 to its inverting input, the resistor 164 preferably having a value of 10.5 kilohms and a 1% tolerance. A feedback capacitor 166 is connected from the output of the preamp 150 to the junction of the resistors 158 and 162, the capacitors 160 and 166 each having a value of 0.001 microfarad. The resistors 158, 162 and 164 and the capacitors 160 and 166 form a feedback network around the preamplifier 150 so as to provide the desired frequency response and the gain of 7 in this preamplifier stage increases the overall input sensitivity of the module. Resistors 168 and 170 are employed to bias the DC output level of the preamplifier 150 to 2.5 volts which is half way between the supply voltage inputs to the amplifier. Thus, when a modulated carrier signal is applied from the power line 20, a full sinusoidal output appears at the output of the preamp 150. A capacitor 172 is used to stabilize the preamplifier 150 and preferably has a value of 120 picofarads. The filter capacitor 174 is employed to remove any undesired noise components in the five volt power supply. Since the preamplifier 150 drives the phase lock loop circuit 152 which also is energized by a +5 volt supply, the signal input to the phase lock loop circuit 152 is prevented from exceeding the supply voltage of the phase lock loop circuit 152 by running the preamplifier 150 from the same 5 volt power supply.

Considering now the phase lock loop circuit 152, the output of this integrated circuit becomes active low whenever there is an input signal of the proper frequency and at least a 100 millivolt level is supplied from the preamplifier 150. The phase lock loop circuit 152 has its own oscillator section which is used as an internal reference. This oscillator is also used as an input to the transmitter portion of the module, as will be described in more detail hereinafter. The output signal from the preamp 150 is coupled through the resistor 180 and the capacitor 182 to the IN line of the phase lock loop circuit 152. A transistor 184 is employed to switch the junction of the resistor 180 and the capacitor 182 to ground when the microprocessor 90 is in a transmitting mode so that there will not be any input signal from the power line 20 on which the phase lock loop circuit might try to lock during the transmission mode. The phase lock loop circuit oscillating frequency, or "center frequency" is determined by the RC network connected to pins 5 and 6 of the phase lock loop circuit 152. However, in accordance with a further aspect of the invention, the multipurpose module shown in FIGS. 6-9, is arranged to operate on either one of two carrier frequencies which may be selected by means of the switch 186. Accordingly, if the utility carrier control system is operating on one of these frequencies, the other frequency may be selected so as to provide selectivity against receiving the utility signal when it is being transmitted over the same power line. Also, in the event that both commercial control of loads in a residence or factory, is desired in addition to a utility load control system, the modules employed for the utility load control subsystem may employ one operating frequency and the commercial system can employ the other operating frequency. A control console coupled to the AC line will then be able to communicate with both types of components by switching between the two carrier frequencies, if desired.

More particularly, a potentiometer 188 is connected from one of the contacts of the single pole double through switch 186 to pin 5 of the phase lock loop circuit 152 and a potentiometer 190 is connected from the other contact of the switch 186 to this terminal. A capacitor 192 is connected from the common arm of the switch 186 to the input terminal 6 of the phase lock loop circuit 152. The two carrier frequencies may thus be selected by actuation of the switch 186 and the corresponding one of the potentiometers 188 or 190 adjusted to provide the exact desired center frequency. Preferably, the potentiometers have a resistance of 10,000 ohms each and the capacitor 192 has a value of 0.0047 microfarads. Preferably the potentiometer 188 is adjusted for a center frequency of 107 Khz and the potentiometer 190 is adjusted for a center frequency of 117 Khz. A capacitor 194 is connected to pin 2 of the phase lock loop circuit 152 to control the dynamic response characteristic of the phase lock loop. A capacitor 196, which is connected to pin 1 of the phase lock loop circuit 152 filters the output voltage of the phase lock loop portion which is supplied to its internal voltage comparator. The open collector output of the phase lock loop circuit 152 at pin 8 thereof is pulled up to five volts through the resistor 198 so as to generate the appropriate TTL input level of signal for the microprocessor 90. The incoming carrier signal on which the circuit 152 is locked, is transmitted to the receive data input line 27, (IN3) of the microprocessor 90 and the demodulation of the data bits of the received message is accomplished by the module software, as will be described in more detail hereinafter.

The transmitter portion of the multipurpose module shown in FIGS. 6 to 9, inclusive, utilizes the center frequency oscillator signal developed in the phase lock loop circuit 152 as an unmodulated carrier signal which is then modulated in accordance with the data bits of the modulated carrier message format shown in FIG. 4 which are developed by the microprocessor 90. More particularly, when it is desired to transmit a modulated carrier message onto the power line 20, the microprocessor 90 develops an enabling signal on its G1 line (pin 29) which is supplied through the resistor 210 to the base of the transistor 184 so as to turn this transistor on and ground the junction of the resistor 180 and the capacitor 182 so that no received signal which is amplified by the preamplifier 150 is permitted to enter the phase lock loop circuit 152. As a result, the oscillator portion of this circuit operates at its center frequency and the 5 volt square wave output which is developed at pin 5 of the circuit 152 is supplied to an amplifier circuit indicated generally at 212 which includes the operational amplifier 214 which is configured as a pass band amplifier and is of the commercial type LM101. This amplifier also has low and high cut off frequencies of 95 Khz and 125 Khz, respectively, but has a gain of only 2. Resistors 216, 218, 220 and 222 and capacitors 224 and 226 form the feedback network for this response. The resistor 216 preferably has a value of 2.67 kilohms, the resistor 218 a value of 215 ohms, and the resistors 220 and 222 each have a value of 5.3 kilohms. The capacitors 224 and 226 each have a value of 0.001 microfarads.

The output of the operational amplifier 214 is connected to the base electrodes of a push pull pair of transistors 230 and 232, preferably commercial types 2N5192 and 2N5193, respectively, the emitters of which are connected through the capacitor 234 to the AC line 20. These emitters are also connected to the feedback portion of the amplifier 212. The push pull transistors 230, 232 are employed to increase the current capability of the operational amplifier 214 in driving the low impedance power line 20. The collector of a transistor 236, which is preferably of the commercial type 2N3904, is connected to the junction of the resistors 220 and 222 and its emitter is connected to ground. The transistor 236 is used to key the transmitter on and off for data transmission. With transistor 236 turned on the DC feedback path around the operational amplifier 214 is broken with the effect of turning this amplifier into a comparator. This causes the output of the operational amplifier 214 to saturate at its highest output level. When a data "0" is to be transmitted, the transistor 236 is turned off allowing the amplifier 214 to function normally as an amplifier and thus impressing the buffered phase lock loop center frequency onto the power line 20. The transistor 236 is controlled by the TX data signal developed at pin 30 of the microprocessor 90 which is supplied through the resistor 238 to the base of the transistor 236. The effect of the described transmitter circuit is to shape the five volt square wave from the phase lock loop circuit 152 into a 10 volt peak to peak sine wave which can be transmitted substantial distances over the AC power line 20.

As described generally heretofore, the communication and control system of the present invention employs an on-off keyed carrier type of system for transmitting messages between the CLC module 46 and the SCCR units 58, over the power line 20. The presence of carrier on the power line 20 is denoted as a logic "0" start bit, or data bit, whereas the absence of carrier denotes a logic value of "1". With such an arrangement the noise impulses on the power line 20 always evidence themselves as a falsely detected carrier tone in which a data "1" is transformed to a data "0". These noise pulses vary in width from approximately 10 microseconds to about 300 microseconds (1/3 of a bit time). In the microprocessor 90 the crystal frequency of 3.58 MHz is divided by 16 to provide instruction cycles which have a duration of 4.47 microseconds. Each bit time is made to have a duration of 400 instruction cycles or 1788 microseconds, which provides a baud rate of approximately 560 bits per second. Thus, as shown in FIG. 12, each of the modulated carrier messages transmitted between the modules 46 and 58 comprise a first start bit of 0, which has a duration of 400 instruction cycles during which carrier is impressed upon the power line 20. The second start bit has a logic value of "0" for 200 instruction cycles and then a logic value of "1" for 200 instruction cycles during which no carrier is present on the power line 20. In order to provide maximum noise immunity, the microprocessor 90 is programmed to detect the rising edge of the last half of the second start bit, indicated by the letter C in FIG. 12, which corresponds to the termination of carrier on the power line 20 at the end of the first half of the second start bit. Since there is no carrier signal continuously present on the power line, each date bit transmitted is also prefaced by a "0" initial portion which lasts for 200 instruction cycles after which the data bit is transmitted for the remaining 200 instruction cycles. Thus, referring to FIG. 12, the first bit time of 400 instruction cycles is divided up into a first initial portion of 200 instruction cycles during which carrier signal is impressed upon the power line (a logic "0" portion) followed by a second portion of 200 instruction cycles during which data bit 1 is transmitted. If this data bit is a "1" there will be no carrier transmitted during these 200 instruction cycles. On the other hand, if the first data bit is a "0" carrier will continue to be impressed upon the power line 20 for the full 400 instruction cycles of the first bit time.

After the microprocessor 90 has synchronized to the incoming message on the rising edge of the second start bit "1", it then anticipates the location of the data bits and when a data bit is expected 22 samples are taken of the incoming carrier signal beginning with the middle of the "0" portion of the first bit and extending into the middle of the "0" portion of the second data bit. More particularly, as shown in FIG. 12, the sampling period starts at point A which is midway in the initial 200 instruction cycles during which carrier is impressed upon the power line 20, and ending at point B which is in the middle of the "0" carrier-on portion of the second data bit. The sampling period thus extends from a carrier on portion of the first bit through the data bit portion and into the carrier on portion of the second bit. If any of the samples taken during the sampling period are "1" i.e. absence of carrier, then the data bit is treated as a "1". Only if all of the samples are "0" i.e. carrier is present, is the data bit treated as a "0". The time from point A to point B in FIG. 12 is 393 instruction cycles. At time B+ 7 cycles the sampling period for the second data bit is started, and so forth. If the first data bit is a "1" then noise impulses must mask the entire data bit of 200 instruction cycles so that none of the samples are "1" in order for the data bit to be falsely interpreted. On the other hand, if the first data bit is a "0", in which carrier is present on the power line 20, noise impulses will simply be interpreted as "0" samples and the data bit will be treated as a "0". The software program in Appendix A is specifically designed to provide 22 samples during each total bit time as measured from the leading edge C of the positive portion of the second start bit, the number 22 being chosen because it was the maximum number of times the sampling routine could be repeated in the specified bit time. Obviously, the number of samples may vary with different bit times and the running time of a particular subroutine.

In FIG. 13 the logic flow of the power up sequence of the microprocessor 90 is shown. In this power up sequence the microprocessor 90 initializes the hardware, reads in its device address, and enters the specified operating mode i.e. either the CLC mode of the module 46 or the SCCR mode of the modules 40 and 60. The microprocessor 90 remains in the selected mode unless a full hardware restart is performed.

In FIG. 14 the logic flow of the microprocessor 90 when operating in the SCCR mode is shown. When a message is read from the carrier line the subroutine No. 4 (sub 4), shown in FIG. 15, is called. When a message is transmitted out on the carrier line the subroutine number 3 (sub 3), shown in FIG. 16, is called. From FIG. 4 it will be seen that messages transmitted in both directions over the power line 20 comprise 28 bits including the 2 start bits, a 12 bit SCCR absolute address, an 8 bit function or status group, and a 6 bit BCH error code. However, in order to simplify the coding of the subroutine No. 3 shown in FIG. 16, a total of 30 bits are transmitted and the last two bits are ignored by the receiving device. In a similar manner the subroutine No. 4 shown in FIG. 15, which reads the message from the carrier line in both the SCCR mode and the CLC mode, reads 30 bits but the last two bits are ignored.

With regard to the first line of the logic flow in the SCCR mode shown in FIG. 14, it is pointed out that a flag is kept which reflects the state of the breaker or "B" bit which is developed on the G0 output line (pin 28) of the microprocessor 90, for status purposes. When in the idle loops, i.e. when the microprocessor is waiting until the carrier data in is a zero (carrier on) or the microprocessor is waiting until the data is a "1", i.e. the rising edge of the second start bit "1", a twelve minute timer is kept as well as a "timer enable/disable" flag. If the timer is enabled, and it expires, the "B" bit is set to "0" and the status flag, which reflects the state of the "B" bit is changed. If the timer is enabled, offset values may be used to update it after a message is received, a function performed, or a false start bit is received. In the COPS 402 or 420 microprocessors, the internal timer may be used in many places, as shown in the listing in Appendix A.

In FIGS. 17a and 17b the logic flow of the microprocessor 90 when operating in the master or CLC mode is shown. With regard to the subroutine no. 1 (sub 1) which is referred to in line 6 of the flow diagram of FIG. 17a, this subroutine requires that the microprocessor wait for the LMT clock line 48 to go from low to high and then delays 2 milliseconds to be sure that it is really high. If the clockline has not gone high after 1 second, an error has occurred and the software waits for the clock line to go high and then restarts to step No. 1. With regard to subroutine No. 2 (sub 2), which is also referred to on line 6 of FIG. 17a, in this subroutine the microprocessor 90 waits for the LMT clock line 48 to go from high to low, and then delays 2 milliseconds to be sure that it is really low. If the clock line has not gone low after one second, an error has occurred and this subroutine restarts to step No. 1 of FIG. 17a.

The subroutines 1 and 2 are used whenever reading from or writing to the LMT 26 over the bidirectional serial data line 44. During all loops and delays the LMT clock is watched. If it goes low before the point labelled L6 in FIG. 17b, an error message is immediately sent to the LMT in which the most significant bit F7 of the status code shown in FIG. 5 is set to 1 and the software resets to the L1 step shown in FIG. 17a.

With regard to line 3 of the L5 portion of the logic flow shown in FIG. 17b, it is pointed out that subroutine no. 4 is entered at the step labelled "do 28 times" rather than at the beginning of this subroutine because the initial steps of subroutine No. 4 are shown in the main flow diagram at the top of FIG. 17b.

Obviously, many modifications and variations of the present invention are possible in light of the above teachings. Thus it is to be understood that, within the scope of the appended claims, the invention may be practiced otherwise than as specifically described hereinabove. ##SPC1##