WO1997026635A1

WO1997026635A1 - A networked, distributed fire alarm system

Info

Publication number: WO1997026635A1
Application number: PCT/US1997/000662
Authority: WO
Inventors: Dale Fergus
Original assignee: Firecom, Inc.
Priority date: 1996-01-16
Filing date: 1997-01-16
Publication date: 1997-07-24
Also published as: AU1829797A

Abstract

A fire alarm system (200) includes a fire command station (202) that communicates with a plurality of intelligent communication cabinets (204) through a network riser (206). The intelligent communication cabinets (204) communicate with plural peripheral devices (208-210, 212, 214, 216). The fire command station (202) is connected to the network riser (206) through a network interface (220) and may communicate with the central station through a telephone link or other communication channel (218). Bidirectional communication between the fire command station and the intelligent communication cabinets (204) as well as independent operation of each intelligent cabinet are achieved. The system further provides a ground fault detection circuit (1328) and includes a line testing circuit that is responsive to a ground fault detection signal and configured to sequentially test lines for the ground fault.

Description

A NETWORKED, DISTRIBUTED FIRE ALARM SYSTEM Field of the Invention

The present invention is directed to a fire alarm system and, more particularly, to an improved fire alarm system having its program and database distributed throughout the system. Background of the Invention

Most cities in the United States and the world require buildings to have a fire alarm system. For example, in New York City all buildings are required to have a fire alarm system. New York City classifies buildings over 100 feet high as "class E" buildings, and has implemented rigorous fire alarm systems regulations for these buildings. Fig. 1 illustrates a typical high rise fire alarm system 100. A fire command station 102 resides in one part of the building, such as the basement or lobby. The fire command station ("FCS") typically includes a computer that controls all or most of the fire alarm system operation. The computer includes a program that controls the operation of the system and a database that contains building specific information such as the number of floors, location of fire alarm devices, etc. (collectively referred to as the program and database 103) . The FCS communicates with a number of nodes 104 installed at various locations throughout the building. The FCS 102 and nodes 104 are connected by a riser 106, which is a wire or set of wires which typically runs up the length of the building. The FCS may also communicate with a remote central monitoring station via a telephone line or other communication channel 118. The central monitoring station notifies the fire department if a fire is detected.

Each node 104 communicates with a number of peripheral devices . These peripheral devices are the sensing and activating portions of the alarm system 100. A typical node 104' is shown in Fig. 1. The node 104' is connected to several peripheral devices such as an alarm bell 108, a fire alarm pull station 109, a number of smoke detectors 110, an exhaust fan 112, a fire door release mechanism 114, and a loud speaker 116. It is the peripheral devices which detect an alarm situation, alert building occupants of an emergency, close fire doors, and activate exhaust fans. Th s, it is critical to the proper operation of a fire alarm system that the peripheral devices function at all times.

The operation of a typical high rise fire alarm system is as follows. The FCS 102 "polls" each node sequentially. "Polling" means obtaining status information from the node about the peripheral devices or other components connected to it. The FCS may, for example, poll the node 104 on the first floor, then the second floor, and so on until it polls the node at the highest floor. It then returns to the first floor. If a node indicates a problem -- an alarm is sensed, a peripheral device has been activated, etc. -- the FCS 102 then stops polling the nodes 104 and interrogates each peripheral device on that node to determine the problem. Once the problem is determined, the FCS acts accordingly. For example, if smoke has been detected, the problem is determined to be an alarm condition. If there is an alarm condition, the system may: (1) activate the bell 108 on the floor on which the smoke was detected and all floors above it; (2) activate the fire door releases to close the fire doors in the area near where the smoke was detected; (3) alert the building supervisor of the alarm condition; and (4) play a prerecorded announcement over the loud speakers on the floor where the smoke was detected.

The problem may not always be a smoke or fire condition. The node 104 may indicate that it is not receiving information from some of the smoke detectors connected to it, suggesting a malfunction or disconnected wire. This problem is determined to be a trouble condition. When the FCS 102 polls the node 104, it will stop polling the nodes 104 and interrogate the peripherals to determine the trouble. Once it determines the trouble, it may, for example, alert the building supervisor of the trouble. One problem associated with the prior art fire alarm systems is that this sequential polling method often results in slow alarm report times. For example, assume a fire alarm system in a forty floor building has a node on every floor. If the FCS 102 begins polling the nodes 104 at the 1st floor and continues upwards until it polls the node at the 40th floor (a node 104 may not be installed on every floor; in some instances one may be installed on every third floor and control peripherals on the floor above and below it as well as the ones on the floor on which it is installed) . Assume smoke is detected on the 40th floor at an instant after the FCS polled the node on that floor. The FCS will not poll that floor again until it polls the 39 floors below it. Valuable time -- seconds or minutes -- is lost before the FCS again polls the node, interrogates the peripherals, determines a smoke condition has been detected, and acts accordingly The more nodes m a building, the greater this delay is likely to be. This delay is aggravated if other nodes indicate trouble, which requires the FCS to interrogate the peripherals, determine the trouble, and act. This delay may result in serious property loss, injury, or death to the occupants.

Another problem associated with the prior art fire alarm systems is that if the FCS 102 is disconnected from some or all of the nodes 104, all or a portion of the fire alarm system 100 is defeated. For example if, in the building described above, the riser 106 connecting the FCS 102 with the nodes 104 was severed at the 10th floor, floors 11 through 40 would no longer be connected to the fire alarm system and could not be polled by the FCS 102. Thus, if an alarm was detected on the 30th floor, the FCS 102 cannot find out about it and cannot activate the peripheral devices, sound an alarm, or alert the central monitoring station.

One fire alarm system, a Firecom 8500 manufactured by the assignee herein, has partially addressed this latter problem. The 8500 operates in a "degraded mode", so that if the FCS 102 is disconnected from the nodes 104 above the 10th floor, each node operates independently of the fire alarm system 100. Using the example above, a smoke condition on the 30th floor will be detected by the 30th floor node 104 and the appropriate peripherals will be activated. The FCS 102 and the nodes 104 servicing the floors above and below this node are, however, unaware of the smoke condition.

Another drawback of current fire alarm systems is that if a grounded wire exists in the installed system, the only way to determine which wire or wires are grounded is to physically disconnect a set of wires and reconnect one wire pair at a time and test it to determine if it is grounded. This is a labor intensive task which may takes hours or days to complete, especially in buildings having many nodes.

Therefore it is an object of the present invention to provide a fire alarm system that has a rapid response time, even in buildings that have a large number of nodes.

It is a further object of the present invention to provide a fire alarm system that allows nodes to communicate with one another even in the event that the nodes are disconnected from the FCS. It is yet a further object of the present invention to provide a fire alarm system that can detect the location of short-circuited wires without having to physically disconnect and test each wire pair. Suimnarv of the Invention These and other objects are achieved by the present invention. The present invention provides a fire alarm system having its operating program and database spread throughout the system, rather than concentrated in a single location. Distributing the program and database allows each node to operate independently of the FCS without relying solely on the fire command station to control the system's operation. This allows nodes connected to each other to communicate, even if these nodes are disconnected from the FCS and other nodes.

Because the program and database are distributed, the peripheral devices need not be polled by the FCS. In a preferred embodiment of the present invention, the peripheral devices are polled by system nodes called Intelligent Transmission Cabinets or ITCs. If an ITC detects that one or more of its peripheral devices senses an alarm or trouble condition, the ITC transmits a message over the network "announcing" the problem. Each node (such as the FCS and other ITCs) connected to that ITC receives the message, reads it, and acts accordingly. Thus, the system does not need to wait until a node is polled before an alarm or trouble condition may be acted upon. The FCS may receive the message and display an alert to a building supervisor to the alarm condition or trouble. Preferably, the FCS may poll an ITC when the FCS has not received a message from that ITC within a predetermined time period. This polling is to determine that the ITC is functioning properly.

A preferred embodiment of the present invention comprises a FCS connected via one or more network risers to a number of ITCs. Each ITC communicates with a number of peripheral devices. The ITCs receive information from the peripheral devices, package the information into network messages, and transmit the packets on the network. The FCS includes a control panel module that is the master polling board and system operator's primary interface with the system. As the master polling board, it verifies that the ITCs are operating and communicating throughout the system. As the system operator's interface with the system, it alerts the operator to alarm conditions, malfunctions, or other problems on the system.

ITCs are the data gathering points for the peripheral units installed throughout the building. The ITC is the interface between the peripheral units and the network riser. The ITC includes a communications card, a monitor card, and a number of option cards. The communications card communicates with each card on in the ITC and transmits and receives messages on the network riser. Some of the monitor card functions are to monitor telephone communication, power supplies, and control the release of fire doors. The option cards are selected to communicate with the peripheral devices. Depending on the peripheral device and method of installation, different option cards may be selected.

In a preferred embodiment of the present invention, the communications card, monitor card, and option cards are configured to detect and locate ground faults. A ground fault detection circuit in the monitor card may notice a voltage change in the building ground, indicating a ground fault in the system. The communications card instructs the option cards to sequentially remove, test, and replace wire pairs until the grounded wire or wires are detected. In another preferred embodiment, the system may have two network risers.

In the event that one or more ITCs cannot communicate over a first network riser (i.e., the wire is severed or damaged) , these ITCs may switch to the second network. This may be done by providing the communications cards with timers monitoring when messages are received from the FCS. If a message is not received within a predetermined time period, the ITC may switch over to the second network. A router, preferably located in the FCS, connects the first and second networks together so that messages transmitted on the second network riser are received by nodes connected to the first network riser.

Brief Description of the Drawings

These and other features of the present invention will become apparent from the following detailed description, in conjunction with the following drawings, wherein:

Fig. 1 is a block diagram of a typical high rise fire alarm system;

Fig. 2 is a block diagram of a preferred embodiment of the fire alarm system according to the present invention; Fig. 3 is block diagram of a network data packet according to a preferred embodiment of the present invention;

Fig. 4 illustrates a preferred fire command station according to the present invention;

Fig. 5 is a block diagram of a preferred fire command station module according to the present invention; Figs. 6A and 6B are a flowchart of a preferred method of the fire command station control panel processing data received from the network;

Fig. 7 is a flowchart of a preferred method of the fire command station processing data received from a manual input according to the present invention;

Fig. 8 is a block diagram of a preferred network A/B router according to the present invention;

Fig. 9 iε a block diagram of a preferred ITC according to the present invention;

Fig. 10 is a block diagram of a preferred ITC communications card according to the present invention;

Fig. 11 is a flowchart of a preferred peripheral polling method performed by the communications card according to the present invention;

Fig. 12 is a flowchart of a preferred network message handling protocol performed by the communications card according to the present invention;

Fig. 13 is a block diagram of a preferred monitor card according to the present invention;

Fig. 14A illustrates a standard D-style wiring; Fig. 14B is a block diagram of one type of option card according to the present invention; and

Fig. 14C illustrates the trouble mode switching performed by the option card of Fig. 14B.

Detailed Description of a Preferred Embodiment I. Overview of the Present Invention A. The Parts of the System !• Introduction Fig. 2 illustrates a preferred embodiment of a fire alarm system 200 according to the present invention. A fire command station 202 resides in one part of the building, such as the basement or lobby. The FCS 202 communicates with a number of intelligent communications cabinets (ITC) 204 installed at various locations throughout the building. The FCS 202 and ITCs 204 are connected by at least one network riser 206, which is a wire pair or set of wire pairs, preferably a dedicated twisted pair, which run the length of the building. The FCS 202 is connected to the network riser 206 via a network interface 220. The FCS 202 may communicate with a central monitoring station via a telephone line or other communication channel 218.

Each ITC 204 communicates with a number of peripheral devices. A typical ITC 204' is shown connected to an alarm bell 208, a fire alarm pull station 209, a number of smoke detectors 210, an exhaust fan 212, a fire door release mechanism 214, and a loud speaker 216. Each ITC 204 also has a network interface 222 which connects it to the network riser 206.

Preferably, the fire alarm system 200 includes a second network, which comprises a second network riser 206' and a router 224. This second network is configured to provide communications with the FCS 202 and some or all of the ITCs 204 in the event that the first network riser 206 is severed or broken. This is similar to a known "style-7" wiring configuration in which a second riser 106 may be connected to a prior art fire alarm system 100. 2. The Polling

Method

Unlike prior known fire alarm systems, the FCS 202 of the present invention does not contain all or most of the fire alarm's program and database. The present invention distributes the program and the database of the system 200 between the FCS 202 and each of the ITCs 204. According to the present invention, each ITC 204 polls the peripheral devices connected to it. If trouble exists or an alarm condition is sensed in a peripheral connected to an ITC 204, the ITC 204 transmits a message on the network. The message is read by all of the ITCs 204 and the FCS 202. Each ITC and the FCS react accordingly. The FCS only polls an ITC if it has not received a communication from it within a predetermined time. For example, assume a fire alarm system 200 according to the present invention is installed in a forty floor building having an ITC 204 installed on every floor. Assume smoke is detected on the 30th floor. The ITC 204 transmits a message along the network which is received by each ITC and the FCS. In response to this message, the ITCs on floors 31 - 40 may activate an alarm bell 208 and play a prerecorded announcement over the loud speakers 216. The ITC on the 30th floor may also activate the bell 208, the loudspeaker 216, activate the fire door release mechanisms 214 in the area near the detected smoke, and turn on a fan 212 to evacuate the smoke from the floor. The FCS may also receive the message and may alert the building supervisor to the smoke condition. Note that there is no delay in activating the peripherals or informing other floors and the building supervisor as in the prior art polling method.

If an ITC 204 indicates that it is not receiving information from some of the smoke detectors connected to it, suggesting a malfunction or disconnected wire, it transmits a message over the network. The other ITCs 204 may ignore the message because it does not affect them, and the FCS 202 may alert the building supervisor to the problem. Assume that a fire alarm system 200 according to the present invention has only one network riser 206, and the riser 206 connecting the FCS 202 with the ITCs 204 is severed at the 10th floor. Floors 11 through 40 would no longer be connected to the FCS 202. The ITCs between floors 11 and 40 will continue to communicate with each other, but not with the FCS 202 and the ITCs on floors 1 through 10. The ITCs on floors 1 through 10 still communicate with the FCS 202 and each other because they are still connected. Thus, if smoke is detected on the 11th floor, the ITCs on floors 11 through 40 will receive a transmission over the network and respond accordingly. At the same time, the FCS 202 determines that it has not received a transmission from the ITCs on floors 11 through 40 within a predetermined time and will attempt to poll them to determine their status. When the FCS cannot poll the ITCs on those floors, it may alert the building supervisor of the trouble. 3. The Network Protocol

An important feature of the invention is that each component is part of a communications network. The present invention may be used with any suitable network protocol, such as asynchronous transfer mode (ATM) , Ethernet, or the like.

In a currently preferred embodiment, a network according to the present invention uses an Echelon proprietary network. The network interfaces 220, 222 may include an Echelon TP/XF- 78 Twisted Pair Control Module (described in Echelon's Lonworks Products 1994 catalog, pages 43-46) and available from Echelon Corp., 4015 Miranda Avenue, Palo Alto, Ca.

94304) . The Echelon module uses a Motorola Neuron 3150 Chip

(described on page 5 of Motorola's Neuron Chips brochure) , available from Motorola, Schaumberg, Illinois. In a preferred embodiment of the present invention, network nodes (every network component having a network interface is considered to be a network node) share information with each other using different types of messages. Which of the message types is sent depends on the network components communicating and the type of information being conveyed. Certain network components read only certain types of messages and ignore the others. This allows different types of messages to be sent over the network, but the components only read messages having information relevant to them.

These messages carry information unique to the inventive fire alarm system. In the preferred embodiment of the present invention, this information is "packaged" for transport using the Echelon protocol. Fig. 3 illustrates a typical message. Fig. 3 illustrates a network data packet 300. A network data packet 300 may include a command byte 304; a length byte 306; a state byte 308; and a number of data node 310, 310' ; data slot 312, 312'; and data address 314, 314' bytes. The command byte 304 contains information about the type of information is contained in the packet. For example, the command byte 304 may indicate if the packet is reporting a state change, is issuing a command to test a device, or is some other type of communication. The length byte 306 identifies how many total bytes are in the packet 300. The state byte 308 contains information about the state of a device. Typically, the state byte 308 will indicate one of alarm, alarm reset (returning to a normal state after an alarm) , test, or test reset (returning to a normal state after a test) . The data node bytes 310, 310' identify a particular node in the system 200 to which the packet 300 pertains. The data slot byte 312, 312' identifies a particular set of peripheral devices within the identified node. The data address bytes 314, 314' identify the address of particular peripheral devices within the identified set. If an ITC is reporting an alarm or trouble condition, a packet 300 will contain only one node/slot/address. If the packet is from the control panel, it may contain more than one node/slot/address in a packet (i.e., a command to test a number of peripherals) .

II. The Fire Command Station

Fig. 4 illustrates a fire command station 202 according to a preferred embodiment of the present invention. The FCS 202 comprises several components, including a control panel module 402, a NY-100 module 404, a fire sign module 406, a forty device display/control unit 408, a strip printer module 410, a thirty-two device display/control unit 412, and a multi-audio control (MAC) module 414. The FCS may also have a serial interface board for communicating with serial data devices such as printers, CRTs, and modems.

The control panel module 402 provides alerts to the building supervisor and polls ITCs from which the FCS control panel module has not received a network transmission from within a predetermined time period. The control panel has a display 415, such as a vacuum fluorescent display, for displaying the system status in text form. The NY-100 module 404 is required by New York City ordinance. It has a manual trip switch 417 to activate an alarm and a fan purge switch 416 and a fan shutdown switch 418. The fire sign 406 is also required by New York City ordinance. The fire sign 406 contains a light behind a translucent "FIRE" sign. The light is activated when the system has detected a fire. The forty device display/control device 408 contains tricolor LEDs and membrane switches. The LEDs indicate the status of a particular ITC or peripheral unit, and the membrane switches may turn peripheral devices on or off. The thirty-two device display/control module 412 contains 3 color LEDs and three position toggle switches. The 3-position switches allow devices to be turned on, off, or placed in the automatic mode. The membrane and toggle switches may have audio feedback to confirm that a switch has been pushed or thrown. The strip printer 410 is a printer which prints system status information. The multi-audio control (MAC) has a telephone 420 for two-way communication with various system stations and a microphone 422 for broadcasting announcements over system loudspeakers 216.

A. The Control Panel Module

The control panel module 402 is the master polling board and system operator's primary interface with the system 200. As the master polling board, it verifies that the ITCs are operating and communicating throughout the system. As the system operator's interface with the system, it alerts the operator to alarm conditions, malfunctions, or other trouble on the system 200. 1. The Control Panel Module Structure

Fig. 5 is a block diagram of a preferred embodiment of a control panel module 402 according to the present invention. This preferred embodiment of the FCS control panel module includes a processor 502; an address bus 504; a databus 506; a random access memory (RAM) 508; a first and a second read only memory (ROM) 510, 512; a real-time clock 514; a network interface 220; an RS-232 interface 518; several parallel input/output buffers 520, 522, 524; a data select register 526; several inputs and output buffers 528 - 540, a diagnostic indicator 542; and an LED latch 544. The processor 502, such as a Motorola model MC68306 microprocessor, controls the operation of the control panel module 402. The processor sends and receives addresses and data over databuseε 504, 506 respectively. The program and database controlling the processor 502 may be stored in the first ROM 510, preferably an electrically programmable ROM (EPROM) having at least 1 megabyte of memory. Building- specific data, such as tables containing node/slot/address information may be stored in the second ROM 512, preferably a flash EPROM having at least 1 megabyte of memory. Current status and variable data may be stored in a random access memory 508, preferably having at least 2 megabytes of memory.

The real-time clock 514 provides the time for the entire alarm system 200. The real-time clock is also a first- in/first-out device that keeps a running, time stamped record of a predetermined number of events stored in the RAM 508. When the number of events stored exceeds the capacity of the real-time clock 514, the first stored event is purged and the currently occurring event is stored.

The network interface 220 is a buffer between the network and the control panel 402. In a currently preferred embodiment, the network interface 220 is an Echelon TP/XF-78 Twisted Pair Control Module. The network interface 220 receives network packets (see Fig. 3) from the network. The packet is read to determine if the packet is carrying information in which the control panel is interested. The network interface 220 also receives information from the databuses 504, 506 and assembles the information into packets for transmission on the network.

The RS-232 interface 518 allows a laptop computer or other data device to connect to the control panel.

The parallel input/output buffers 520, 522, 524 are buffers between the control panel and related components, such as input keyboard decoders 528, 530, which decode information received from buttons pushed on the control panel; monitor inputs 532, 536 which accept commands from the control panel; and data buffers 534, 538 to the printer 410 and the display 415. The data select register 526 receives an address and data and selects whether the data is to be sent to the MAC latch 540, the control panel LED latch 544, or the diagnostic indicator 542. The MAC latch 540 receives data controlling the MAC module 414. The control panel LED latch 544 receives data controlling the LEDs on the control panel . The diagnostic indicator receives data that displays a code reflecting the operating status of the system. 2. The Control Panel Module Operation

Figs. 6A and 6B are a flowchart of an illustrative method 600 for the control panel 402 to process data received from the network. A network packet 300 is received at the control panel network interface 220 (step 602) . The network interface 220 translates the packets into a format which is understandable to the control panel and forwards the translated data to the appropriate location, such as processor 502 (step 604) .

The node, slot, and address to which the data is related are determined (step 606) . If the packet does not contain information in which the control panel 402 is interested (step 608) , this data is ignored, and the network interface 220 waits for the next packet to arrive. If the packet contains information in which the control panel 402 is interested (step 608) , the current status information stored in the RAM 508 is retrieved and compared with the new data (step 610) , and the node transmission timer is reset (step 618 discussed below) . If the status has not changed, no further action is required and the network interface 220 waits to receive the next packet (step 612) . If the status has changed (i.e., smoke or trouble is detected, trouble has been fixed, etc.) , the status in the RAM 508 is updated to reflect this new status; the information stored in the RAM is also stored in the real-time clock along with the time when it was stored (step 614) . Because the status has changed, the building supervisor must be alerted; thus the status change may be printed out on the strip printer 410, the vacuum fluorescent display 415, and/or the appropriate LED display 408, 412 (step 616) Note that if an alarm is detected, the control panel 402 does not activate the peripheral units -- this is done by the ITCs 204. The control panel's function is to alert the building supervisor of the alarm or trouble condition.

The control panel 402 is the master polling board. Thus, it maintains in the processor a timer for each node When a communication is received from a node (step 608) , the timer is reset for that node (step 618) . If no data transmission is received from a node within the predetermined time (such as 20 or 30 seconds) (step 620) , the control panel 402 polls the node (step 622) If the node does not respond to the poll (step 624) , trouble is detected and a status change is indicated (see steps 614, 616) If the node responds to the poll (step 624) , its status is determined If the node is not in trouble (step 626) , the timer is reset for that node (step 618) If the node is in trouble (step 626) , a status change is indicated (see steps 614, 616) .

Fig. 7 is a flowchart of an illustrative method 700 for the control panel 402 to respond to a manual input When, for example, a switch on the control panel is pushed, the instruction is received by a parallel mput/output buffer 520 and forwarded to the processor 502 (step 702) . The processor determines which device is to be activated, turned off, tested, or otherwise controlled according to this mput The current status of this device is retrieved from the RAM 508 and compared If there is a status change, RAM 508 is updated to reflect this new status; the information stored in the RAM is also stored in the real-time clock along with the time when it was stored (step 704) . The status change should be confirmed to the building supervisor; thus the status change may be printed out on the strip printer 410, displayed on the vacuum fluorescent display 415, and/or indicated on the appropriate LED display 408, 412 (step 706) . The processor 502 assembles the instruction to the peripheral (step 708) . The assembled instruction is forwarded to the network interface 220, which translates the instruction into a network programming message packet 320 (step 710) . The message programming packet is transmitted over the network to be received by the appropriate node (step 712) . B. The A/B Network Router Fig. 8 is a block diagram of a preferred embodiment of an A/B network router 224 according to the present invention. A first network riser 206 for network A and a second network riser 206' for a second network B are connected to an FCS/riser wiring terminal 802. Network A is connected to Net A portion 804 of router 224, which is connected to an FCS network 808. Network B is connected to Net B 806 portion of router 224, which is also connected to the FCS network 808. The FCS network 808 may be an interface with the network as it connects with the FCS devices such as the FCS control panel 402 and the display control modules 408, 412. The FCS network

808 is connected to the control panel network interface 220.

The purpose of the router 224 is to connect two network risers 206, 206' to a single network interface 220 and to protect the FCS from the risers if they are shorted. In a preferred embodiment, there is no switch in the router 224, but all network communications are transmitted over both network risers 206, 206' . The A/B routing is performed by the ITCs 204. This may be accomplished in a number of ways. Two are described below. One method may be for an ITC to be programmed to time how often a network packet 300 is received from the FCS 202. If such a packet is not received within a predetermined time, the ITC switches from the first network to the second network. A second method may be to program the ITC to time how often a special network verification packet is received from the FCS. If the special packet is not received, it does not issue any network communications until it switches to the second network.

Regardless of the method used, the communications transmitted on the second network (presumably from nodes located above a break in the first network riser 206) are received by the FCS through the router 224. Nodes below the break also receive the transmission because the router connects Net A and Net B together via the FCS network 808.

When the second network riser 206' is not being used, it may be occasionally checked to make sure the entire riser is viable by sending a network B checking packet which is to be received at the node furthest from the FCS. When the packet is received by the node, it may send a responsive node to the FCS verifying that the network riser 206' is intact. This node may need a second network interface in order to monitor both networks at the same time. C. Other FCS Components

Many of the FCS components are separate network nodes. That is, for example, the forty and thirty-two unit display control modules 408, 412 have their own network interfaces and respond to and generate their own network communications. If, for example, the forty unit display control module includes an LED representing the status of a particular smoke detector and that smoke detector issues an alarm, the module 408 will receive a network communication and may respond by changing the LED readout for that smoke detector.

III. The "Intelligent Transmission Cabinets" or ITCs

The intelligent transmission cabinets or ITCs 204 are the data gathering points for the peripheral units installed throughout the building. ITCs 204 are installed throughout the building. The ITC is the interface between the peripheral units and the network riser 206. The ITCs 204 receive information from the peripheral devices, package the information into network message packets 300, and transmit the packets on the network. The packets may be received by other ITCs and the FCS. The nodes may or may not act on the information, depending on the packet's content. A. The ITC Motherboard

Fig. 9 is a block diagram of a preferred embodiment of an ITC motherboard 204 according to the present invention. The motherboard 204 holds a number of printed circuit boards, or cards, including a communications card 902, a monitor card 904, and a number of "options" cards 906, which connect to the peripheral unit wiring terminals 908.

The ITC 204 also includes a network riser terminal input/output 910, an amplifier input/output terminal 912, a power terminal 914, a terminal to the amplifier input 916, a databus 918, a control bus 920, a display module 922, a door release terminal 924, and a tamper switch terminal 926.

The network riser 206 wires are connected to the ITC at the network riser terminal input 910, which may also include inputs for the telephone and microphone communications (see Fig. 4, elements 420, 422) . The phone and microphone inputs may also be sent to an amplifier terminal 916. Other ITC inputs/outputs include an amplifier input/output terminal 912 for communications with the loud speakers 216, and a power terminal 914, which receives a power input from, for example, a power supply. The power supply may receive the building's 120 volt AC power riser. In a preferred embodiment, the power supply converts the AC power into DC power and also charges a back-up battery which supplies power to the ITC in the event of a power outage or other power loss . The network inputs are received, translated, and acted upon by the communications card 902 (described in detail below) . The communications card outputs data and control signals on an ITC databus 918 and ITC control bus 920, respectively. The control bus may also send and receive serial data. The buses connect to the monitor card 904 and option cards 906 on the motherboard 204. The communications card also outputs status information to a display module 922 which, in a preferred embodiment, is a digital display such as an LED, LCD, or vacuum fluorescent display. The monitor card 904 sends and receives data and control information to and from the communications card 902 via the buses 918, 920. The monitor card controls the monitoring of the telephone communication, power supplies, and controls the release of fire doors through the door release terminals 924. The monitor card also receives information from a tamper switch terminal 926, which indicates if a panel door on the ITC is open. The switch may be a simple contact switch which indicates that the panel is open when contact is broken.

The option cards 906 are selected to communicate with the peripheral devices . Depending on the peripheral device and method of installation, different option cards 906 may be selected. For example, one type of option card may be selected for communicating with addressable devices and another option card may be selected for communication with conventional (non-addressable) devices. Each slot on the motherboard 204 in which an option card 906 is mounted has a unique slot ID 928, which identifies the slot for network communications. Each option card 906 also has a unique card ID. When the communications card polls a slot, it obtains the card ID for the option card in that slot. It compares the card ID with a card ID it has stored for that slot to make sure the card has not been inadvertently inserted in an incorrect slot . The ITC generates network message packets that include data node/slot/address bytes . Each ITC and the FCS has a unique node identifier; each option card is given a slot ID; and each peripheral device or group of peripheral devices is given an address.

B. The Communications Card

1. The Communications Card Structure Fig. 10 is a block diagram of a preferred communications card 902 according to the present invention. The communications card includes the network interface 222 for the ITC and communicates with each card on the motherboard 204 via the ITC data and control buses 918, 920.

The communications card includes a processor 1002; two ROMS 1004, 1006; a RAM 1008; a network interface 222; a transceiver 1010; a CPU watchdog circuit and manual reset 1012, 1014; a clock crystal 1016; a memory decode logic PAL 1018; several buffers 1020, 1022, 1024, 1026, 1028; an inverter 1030; and a network interface manual service button 1032. The processor 1002 is preferably a microprocessor such as an Intel 80C32. The program and database controlling the processor 1002 may be stored in the first ROM 1004, preferably an EPROM having at least 32 kilobytes of memory Building- specific data, such as tables containing option card slot and peripheral device address information may be stored the second ROM 1006, preferably an erasable EPROM having at least 32 kilobytes of memory. Current status and variable data may be stored in a random access memory 1008, preferably having at least 32 kilobytes of memory.

The network interface 222 is a buffer between the network and the ITC 204. The network interface 222 receives information packets (see Fig. 3) from the network v a the transceiver 1010, which in a currently preferred embodiment is an Echelon TP/XF-78 Twisted Pair Control Module. The packet is translated and sent to the appropriate destination, such as the processor 1002. The processor 1002 may use this information to send control signals over the control bus 920. The network interface 222 also receives information from the ITC databus 918 via the transceiver 1010 and packages the information into packets for transmission on the network. The processor may send control data over the control bus 920 via a buffer 1024. The network interface may be manually reset (for example during servicing) by pressing the manual reset 1032.

A CPU watchdog circuit 1012 monitors the processor 1002 to ensure that it is operating. If it is not operating, the watchdog circuit may try to "wake" the processor 1002. If the processor cannot be "wakened", the watchdog circuit 1012 will output a CPU fail bit to a buffer 1028. The buffer sends the bit to the option cards 906 mounted on the motherboard 204.

If the ITC is being serviced, it may be desirable to reset the processor 1002. This may be done by depressing the manual processor reset 1014. If the processor is reset, a reset bit is sent to the buffer 1028 and sent to the option cards to reset them as well . The reset bit may also be inverted and applied to the network interface 222 and reset that as well . The clock crystal 1016 is applied to the processor 1002 to provide a timing source for it. The memory decode logic PAL 1018 decodes slot and address data (see Fig. 3) and outputs chip enable or card select bits on the control databus 920. A UART input/output in the processor 1002 sends and receives serial information to and from the monitor and option cards 906 via a buffer 1026.

The communications card can test the option cards for ground faults. The buffer 1020 receives the control signals for testing the cards. If a ground fault is detected

(explained below) , the communications card can initiate a test of the wires connected to an option card 906 that indicates a ground fault. When a ground fault is detected by the monitoring card 904, the communications card may initiate a ground fault test. This test physically removes or switches out wires from each field wiring terminal and then sequentially tests each wire pair for the ground fault. This allows a ground fault in the system quickly and accurately to be narrowed down to the wire pair (or pairs) that are grounded. This eliminates the hours and days of manual labor previously needed to detect and correct a ground fault in prior art fire alarm systems.

The display backlight buffer will turn on an LED backlight for the display module 922 when the backup power supply is activated. The display control discrete supplies control information for the display module, which is sent to a buffer 1022 before being sent to the display module 922.

The communications card checks each slot and verifies that the proper card is mounted in the slot. This prevents an option card from being inadvertently placed into an incorrect slot. The communications card will output an alert on the display module 922. The communications card 902 may also generate an FCS alert, which is sent to the network interface 222, which packages the message into a network data message packet 300 and transmits it over the network. The FCS control panel 402 receives the message, reads the command byte, recognizes that the information affects it, and displays an alert that an option card is inserted in an incorrect slot.

If the communications card CPU fails, the CPU fail bit indicates to the option cards that the ITC is not communicating with the network. If an alarm condition is detected when the CPU fail bit is active, the peripherals operate in a degraded mode and certain peripheral devices will activate (i.e., fire doors may be released and a fan may turn on if smoke is detected) .

2. The Communications Card Operation Fig. 11 is a flowchart of a preferred peripheral polling method 1100 performed by the communications card according to the present invention. The communications card 902 queries the first option card 906 (step 1101) . The communications card 902 obtains the card ID from the polled slot and compares it to the stored value to make sure the card is inserted in the proper slot (step 1102) .

If the card ID is incorrect, that slot is deleted from the list of slots to be queried (step 1104) . The processor 1002 on the communications card assembles a message for the display module 922 and the network interface 222 (step 1106) . The display module displays an alert that the card is in the incorrect slot, and the network interface 222 packages the message into a network data message 300 and transmits it over the network (step 1108) . If the card ID is correct, the status of the peripherals connected to that option card 906 is checked (step 1110) . If the status of the peripherals has changed (step 1112) , the communications card RAM 1008 is updated to reflect the new status (step 1114) . The status data for slot is assembled by the processor 1002 and sent to the network interface 222 (step 1116) . The network interface 222 packages the message into a network data packet 300 and transmits the packet on the network (step 1118) . If the status of the peripherals has not changed (step 1112) , nothing needs to be done for this card. If all the slots in the motherboard 204 have not been polled (step 1120) , the next slot is queried (step 1122) . If all of the slots have been polled (step 1120) , the communications card begins polling the first option card again.

An alternative to this method is to poll each option and check that the card IDs are correct. After each card ID is verified, then each card is polled for status. The communications card may also poll the monitor card 904. The monitor card is not optional and thus needs no slot or card IDs.

Fig. 12 is a flowchart of a preferred network message handling protocol 1200 performed by the communications card according to the present invention. A network packet 300 is received at the communications card network interface 222 (step 1202) . The network interface 222 translates the packet into a format that is understandable to the communications card and forwards the translated data to the appropriate location (step 1204) , such as the processor 1002. The processor determines node, slot, and address information of the packet (step 1206) . If the packet contains information in which the communications card 902 is not interested (step 1208) , this data is ignored, and the network interface 222 waits for the next packet to arrive.

If the packet contains information for this node (step 1208) , the processor 1002 determines if the message is a data or control message (step 1210) . If it is a data message, the processor determines if it is a message such as an alarm requiring an action such as activating or deactivating a peripheral (step 1212) . If it is not such a message, the data is saved in the RAM 1008 (step 1214) , and the communications card waits for the next network message. If the message was a data message (step 1210) and an alarm was sounded (step 1212) , the processor 1002 issues control messages to the appropriate peripheral units to activate or deactivate them (step 1216) .

If the received network message was a control message (step 1210) , the slot and address of the peripheral or other device to be controlled is determined (step 1218) . The control message is then sent to the peripheral and the control is performed (step 1220) .

Regardless of whether the message was an alarm or a programming message, any status change in the peripherals is recorded in the RAM 1008 (step 1222) . The status information is also assembled into a message and forwarded to the network interface (step 1224) . The network interface packages the message and transmits it onto the network (step 1226) and the network interface 222 waits for the next network message. C. The Monitor Card Fig. 13 is a block diagram of a preferred monitor card 904 according to the present invention. The monitor card includes a transceiver 1302; a monitor card databus 1304; a control signal buffer 1306; a decode logic circuit 1308; an audio supervision circuit 1310; a power monitor circuit 1312; a first sense discrete circuit 1314; a serial data buffer 1316; a control discretes latch circuit 1318; a degrade control and select circuit 1320; a power driver 1322; a door release relay 1324; a power monitor select 1326; a ground fault detection circuit 1328; a second sense discrete circuit 1330; and a power supply discrete circuit 1332.

When the communications card 902 polls the monitor card 904, the transceiver 1302 receives and sends data to the communications card 902. The data transferred among the various monitor card components are transmitted on the monitor card databus 1304.

The control signal buffer 1306 receives control signals, such as card select and write enable signals, from the communications card 902 via the ITC control bus 920. The buffer applies the control signals to the decode logic circuit 1308, which will decode the control signal and apply the signal to the appropriate location.

The audio supervision circuit 1310 receives audio input from the riser terminal input/output 910 on the ITC motherboard 204. Power inputs from the motherboard's power terminals 914 are received at the power monitor circuits 1312. The output of the audio supervisory circuit 1310 and the power monitor circuits 1312 are supplied to the first sense discrete circuit 1314. This circuit monitors the inputs for the presence of the signals. If an audio signal or power supply is not present, a trouble message will be issued on the data bus 1304 by the sense discrete circuit 1314 when the card is polled by the communications card 902 and the circuit is enabled by a chip enable control signal received from the control bus 920.

The serial data buffer 1316 sends and receives serial data between the communications card and the amplifier rack/cabinet.

The control discretes latch circuit 1318 enables a power driver 1322 when data is received on the data bus 1304 indicating that the display module 922 display light, service control mode, power fault, or LED indicators on the monitor card should be activated.

The degrade control and select circuit 1320 enables the power driver 1322 to activate the fire door release relay 1324 when the degrade control circuit 1320 receives a CPU fail bit from the communications card 902 and an alarm signal from an options card.

The power monitor select circuit 1326 indicates which power supplies the card should monitor. A ground fault detection circuit 1328 monitors the building ground and compares it with a reference ground voltage, preferably the system ground. If the building ground varies beyond the reference voltage, a ground fault is detected The output of the power select circuit 1326, the ground fault detection circuit 1328, the tamper switch 926, and a service switch are applied to the second sense discrete circuit 1330. If trouble is detected, the circuit 1330 will output a trouble signal on the databus 1304 when enabled. The communications card may also monitor the ground fault detection circuit 1328 when performing the ground fault test. As lines are sequentially tested, it will monitor the output of the ground fault detection circuit to determine if the tested line has a ground fault. A power supply discrete circuit 1332 monitors for low battery and AC failure of the option card's power supply and will output a trouble signal if the power supply is not detected. D. The Option Cards

Option cards 906 are connected to the field wiring terminals 908 for the peripherals located throughout the building. The option card described is for devices wired using style-D wiring. Fig. 14A illustrates style-D wiring. D-style wiring is well known in the art and is briefly described here to provide background for certain features of the present invention. A set of wires 1450 begin and terminate at a field wiring terminal 908. Several peripheral devices 1452 are placed across these wires. An end of line device 1454, such as a resistance or capacitance is located at the terminal end of the wires. A voltage dividing resistance 1456 and sensing circuit 1458 are located at the beginning negative terminal.

During normal operation, electrical current flows in the manner indicated by the solid arrow: the current flows from the beginning positive terminal to the ending positive terminal, through the end of line device 1454, through the ending negative terminal to the beginning negative terminal. The voltage divider divides the voltage and the sending circuit senses the voltage. If there is a change in the voltage, a trouble or alarm is sensed.

If a break 1460 occurs in the wires, the peripherals may still operate. Power is supplied to both positive terminals. The current flows as indicated by the dashed arrows . Current from the beginning positive terminal flows to the devices to the last device before the break. The current flows through this device back to the beginning negative terminal. Both negative terminals are connected to the sensing circuit 1458.

The peripheral devices on the other side of the break may have current flowing through them in a similar manner as indicated by the dashed arrows . Fig. 14B is a block diagram of one type of option card 906 according to the present invention. The option card illustrated in Fig. 14B has a transceiver 1402; a databus 1403; a control signal buffer 1404, a reset buffer 1406; a CPU fail/alarm buffer 1408; a degrade trouble control and select circuit 1410; a power control/trouble control circuit 1412, first and second OR gates 1414, 1415; a decode logic circuit 1416; a board type memory 1418; a card ID code 1420; an alarm sense/trouble sense circuit 1422; a power bus select and control circuit 1424; and a number of field wiring terminals 908.

When the communications card 902 polls the options card 906, the transceiver 1402 receives and sends data to the communications card 902. The data transferred among the various option card components are transmitted on the option card databus 1403.

The control signal buffer 1404 receives control signals, such as card select and write enable signals, from the communications card 902 via the ITC control bus 920. The buffer applies the control signals to the decode logic circuit 1416, which decodes the control signal and applies the signal to the appropriate location.

Information from the peripherals is received from the field wiring terminals 908 and directed to the alarm sense/trouble sense circuit 1422. If an alarm is detected, when the communications card polls the option card, it will send a control signal over the control signal bus 920 to enable the read of the alarm sense/trouble sense circuit 1422. When enabled, a data signal is transmitted over the databus 1403 to the transceiver 1402 and to the communications card 902 via the ITC databus 918.

If the trouble sense circuit 1422 detects trouble in a peripheral, when the communications card polls the option card, a data signal is transmitted over the databus 1403 to the transceiver 1402 and to the communications card 902 via the ITC databus 918. The trouble data signal is also sent to the trouble control circuit 1412, which acts accordingly, as described below.

The outputs of the trouble control circuit and the degrade trouble control circuit are ORed together in OR circuit 1415, so that either condition will activate a trouble circuit 1480 A preferred embodiment of a trouble circuit 1480 is illustrated in Fig. 14C. The circuit is illustrated for the normal operating condition. The electrical current will begin at the beginning positive terminal, go across the end-of-lme device 1454 and return to the beginning negative terminal to the sense circuit If trouble is sensed, a set of single poll, double throw switches 1484 are activated This switch is controlled via relay coil 1486 which puts the return zone wires in parallel with the zone wires A current limiter 1488 is used to isolate the option card 906 from any short m the field wiring.

Returning to Fig. 14B, the power bus select and control circuit 1424 supplies power to the field wiring terminals 908 and the peripheral devices. If the communications card issues the CPU fail bit and an option card issues an alarm, the degrade power/trouble control and select circuit 1410 issues a signal which is ORed with the output of the power control circuit 1412. This activates certain peripheral devices during degrade mode operation. Thus, even if the communications card has failed, the peripheral devices on the option card may be activated Also, if a CPU fail bit and an alarm are both detected, the system will activate the trouble circuit 1480, so that it may be sensed from both sides if the circuit the event that a break exists.

If the communications card processor 1002 was reset, a reset bit is received from the communications card buffer 1028 at the option card reset buffer 1406 and applied to the power control/trouble control circuit 1412.

Each option card has a unique card ID, which may be stored in an ID code circuit 1420. The card ID is applied to a board type circuit 1418, which decodes the card ID into a description of the type of option card it is. This information is available when the communications card asks for it durmg the polling process, so that the communications circuit can determine if the card ID matches the value stored the communications card for that slot.

The option card shown in Fig. 14B controls peripherals wired a conventional "D-style". A person of ordinary skill in the art readily understands that, for example, a B-style field wiring uses half the number of wires as D-style and, thus, twice the number of wiring terminals (i.e., twice as many zones) may be connected to the options card 906. IV. Conclusion

A networked, distributed fire alarm system is described which has an improved polling method that greatly improves the response time for reporting an alarm or trouble condition, particularly in buildings having a large number of nodes. The fire alarm system also distributes the program and database throughout the network so that if the network riser is disconnected, the portions of the system above the disconnection may still communicate with one another. The portions below the disconnection may communicate with the FCS and the portion of the system below the disconnection. An optional feature of the invention provides a second network riser and routers to route network communications along the second riser if the first is disconnected. Another unique feature of the invention is its ability to automatically check for ground faults.

The present invention is not limited to the disclosed embodiment, but rather various modifications, substitutions, and structures may be used without departing from the scope of the present invention.

Claims

I claim :

1. A fire alarm system, comprising: a. a network riser; and b. a plurality of intelligent communications cabinets (ITC) , each ITC including:

(1) a monitoring device configured to monitor peripheral devices, and to output a signal indicating a status of the peripheral devices;

(2) an ITC network interface responsive to the monitoring device and configured to send messages on the network riser indicating the status of the peripheral devices, and to receive messages from the network riser controlling operation of the peripheral devices.

2. The fire alarm system of claim 1, wherein the fire alarm system further comprises a fire command station (FCS) connected to the network riser.

3. The fire alarm system of claim 2, wherein the FCS further includes : a. a FCS network interface configured to receive network messages indicating a status of peripheral devices; and b. a display device responsive to the FCS network interface configured to activate a display indicating the status of the peripheral devices.

4. The fire alarm system of claim 2, wherein the FCS further includes : a. remote peripheral device controls configured to generate signals for controlling the peripheral devices; and b. a FCS network interface responsive to the remote peripheral device control and configured to send network messages controlling the peripheral devices.

5. The fire alarm system of claim 4, wherein the signals generated by the remote peripheral device are configured to one of activate, deactivate, and test the peripheral devices.

6. The fire alarm system of claim 2, wherein the FCS further includes an ITC polling circuit configured to poll an ITC when the FCS has not received a network message from that ITC within a preselected time period.

7. An intelligent transmission cabinet (ITC) for a fire alarm system, comprising:

(1) a monitoring device configured to monitor peripheral devices connected to the ITC, and to output to the fire alarm system a signal indicating a status of the peripheral devices;

(2) an ITC network interface responsive to the monitoring device and configured to send messages to the fire alarm system indicating the status of the peripheral devices, and to receive messages from the fire alarm controlling operation of the peripheral devices .

8. A grounded wire locating circuit, comprising: a. a ground fault detection circuit configured to monitor a building ground and to output a ground fault detection signal if the building ground exceeds a reference voltage; and b. a line testing circuit responsive to the ground fault detection signal and configured to sequentially test lines for the ground fault.