Page IMSA Journal2Continued on page 43

The ABCs of Fire Alarm Systems - Section IBy Anthony J. Shalna 2009Principal IMSA Representative to the Automatic Fire Alarm Association

President: Southeastern Signalmen of MassachusettsApprovals Manager: Gamewell-FCI by Honeywell

A fire alarm system is used primarily to evacuate the prem-ises in the event of occurrence of a fire condition and then secondarily to report the fire to the proper authorities.

A fire alarm system differs somewhat from a security system. The security system only recognizes two states or conditions: normal or alarm, and cannot differentiate between a line break and the opening of an alarm switch. The fire alarm system recognizes four different states or conditions: normal, alarm, trouble and supervisory.

Simplistically speaking, a basic system consists of a fire alarm control panel (FACP) to which are connected initiating (input) devices, notification (output) appliances, a source of operating power, and a source of standby power in the event the operating power should fail.

The function of a fire alarm control panel is basically three-fold:

) Accept an alarm or supervisory input from an initiating device.

2) Provide an alarm output to the notification appliance(s).

) Monitor the integrity of the panel itself and also the wir-ing to the above devices.

MINIMUM BASIC SYSTEMFire alarm systems have changed dramatically over the past few years, primarily due to the ad-vent of the low priced micropro-cessor. Basically there are two different approaches used for the fire alarm control panel, conven-tional and addressable.

We are concerning ourselves in this installment with the con-ventional (hard-wired) system as opposed to an addressable system, which will be covered in a following installment.

The minimum basic components of a conventional system are:

) A locked fire alarm control panel listed for the purpose by a Nationally Recognized Testing Laboratory, (NRTL) as recognized by OSHA. The standard governing fire alarm control panels is ANSI/UL Standard 8, current-ly entering its ninth edition. OSHA currently recognizes Underwriters Laboratories, Factory Mutual Approvals and ETL-Semko as certified to test equipment per this standard.

2) A primary operating power sup-ply (20 VAC).

) A secondary or standby power supply. This is most often a rechargeable storage battery, although genera-tors are permitted subject to certain conditions.

) At least one initiating device circuit to which is wired at least one manual station, automatic heat or smoke detec-tors, waterflow switch activated by a sprinkler system, etc. These devices are located in one area, or zone, so an alarm condition in this zone can direct fire fighting personnel to the source of the alarm. Typically, a zone usually consists of a floor of a small building, or wing of a larger building, etc. with area limitations defined in the National Fire Alarm Code (NFPA 72).

) At least one (output) notification appliance circuit to which is wired at least one horn, bell, and strobe, if required.

The basic minimum system is shown in Figure .

The secondary power supply (usually a battery) automati-cally furnishes operating power to the system in the event of failure of the main 20 VAC supply or if the main supply voltage falls below 8% of normal (Brown-out condition). The battery must be of the rechargeable type, since dry cells are not permitted.

Figure 1

May/June 2009 Page

Continued from page 42

A gel- cell battery is a rechargeable battery.) The battery must operate the system for a specified period of time in a standby or quiescent condition, and have sufficient reserve at the end of the standby period to operate the panel in an alarm condition for a period of five () minutes. Batteries are required by the National Fire Alarm Code in all fire alarm systems not having multiple standby generators. The control panel must also be capable of recharging the battery within a specified period after a discharge.

BASIC SYSTEM OPERATIONInitiating devices employed in a su-pervised, conventional system usually have normally open, dry contacts which close on alarm. (Dry contacts are contacts that have no voltage ap-plied to them.) Exceptions to this are 2-wire smoke detectors, which receive their operating power from the (supervisory) current flowing through the circuit, and alter the characteristics of the circuit when they go into alarm. These will be covered in a separate article.

The act of operating a manual station or actuation of an automatic detector closes the contacts of the device and applies power to the alarm circuitry, causing the panel to go into alarm, light one or more red LEDs on the panel, and energize the notification appliance(s). This appears to resemble the classic operation of a doorbell system, but the BIG difference here is that the fire alarm control has the ability to monitor its own integrity, commonly referred to as supervision.

SUPERVISIONA supervised system (sometimes referred to as a closed circuit system) will create a trouble signal in the event of a break in the field wiring, disconnection or removal of an initiating device or notification appliance, failure of main operating power, discon-nection of the standby battery, or off-normal position of a panel switch. A trouble condition will light one or more yellow LEDs on the panel and cause an audible signal, (usually a piezoelectric device) to sound. The audible signal can be silenced by operating the Trouble Silence switch on the panel. Since the panel is locked, the trouble sounder can only be silenced by authorized personnel who have access to the key.

In a conventional system, supervision is made possible by use of an End of Line (EOL) device, usually a resistor, although other components may be used, depending on the designer.

TROUBLE CIRCUITRYMany years ago, manufacturers used relays to achieve supervision. Two relay coils, alarm and supervisory, were connected in series with the initiating device circuit. The supervisory relay was rated at a lower voltage and was continually energized by the reduced current flowing through the circuit via the EOL device. The alarm relay was rated at the operating voltage and would only energize when the current was increased by an initiating device that short-circuited the EOL device. If the circuit is opened by a break in the wiring, or unauthorized removal of a detector or station, or if the winding of either relay opened, the trouble relay contacts would fall out, applying voltage to the trouble LED and sounder. These

The ABCs of Fire Alarm Systems - Section I . . . relays were eventually replaced by solid state compo-nents, mostly microprocessors, that monitor the circuit supervisory current.

In Figure 2, we see an initiating circuit of a Fire Alarm Control Panel (FACP). Current flows out of the FACP and through the circuit, in and out of one contact of the initiating devices, through the end of line resistor (EOL), through the second contact of each initiating device and back to the FACP. (We will discuss the in and out wiring to the contacts later.)

Continued on page 44

Page IMSA JournalContinued on page 45

In the event of a break in the outside wiring, unauthorized removal of a detector or sta-tion, failure of the main power supply, or re-moval of the EOL device, the microprocessor will sense the change in circuit current and create a trouble condition, energizing the yellow System Trouble LED, a yellow Zone Trouble LED dedicated to that circuit, and an audible sounder inside the panel, signify-ing a trouble condition. The sounder may be silenced by operating a Trouble Silence switch, but the yellow LED(s) will remain lit. Once the trouble condition is rectified and the circuit in question is restored to normal, the audible signal will sound again, or ring back. The trouble silencing switch is then restored to the normal position, silencing the sounder and extinguishing the Trouble LED(s).

ALARM PROCESSINGIn the event of an alarm, the contacts in the initiating device close, shunting out the EOL, raising the circuit voltage to full operating voltage, and energizing the alarm circuitry. The alarm circuitry will then ap-ply operating power to the notification appliance circuit, sounding horns, flashing strobes, and performing other functions.

This type of circuit is referred to as a Class B, Style B circuit. The National Fire Code, NFPA Standard 72, makes references to both Classes and Styles for circuits. The Class A or B designa-tion has been traditionally used for discussion purposes, while the Style designations refer to a wider variety of cir-cuits having subtle differences which are beyond the scope of this article. (The 200 Edition of the National Fire Alarm Code will apparently do away with style classifications and adopt new definitions of Classes.)

CLASS B, STYLE B INITIATING CIRCUITThus, a Class B (or Style B) circuit is a two-wire circuit with external EOL. Any device electrically located beyond a break in the field wiring will be disabled. Any devices located electrically before the break will still be able to turn in an alarm. A Class B system is economical, since it only uses two wires, but has several drawbacks, such as surviv-ability, or inability to operate if a device beyond the break goes into alarm. Also, the EOL is often installed in the last initiating device, the location of which is often unknown if a less than competent installer doesnt document the location at the FACP. Therefore, we see the need to develop a better circuit. The Class A, Style D circuit is an answer.

CLASS A, STYLE D INITIATING CIRCUITA Class A (or Style D) initiating circuit uses four wires, and has the EOL device located on the FACP terminal board or at least, inside the cabinet. Figure shows a typical Class A circuit. The circuit operates in the same fashion as a Class B circuit, but the wiring returns to the FACP after the last initiating device. In the event of a wiring break, etc., the

Continued from page 43The ABCs of Fire Alarm Systems - Section I . . .

trouble circuitry operates and connects line A to line D, and line B to line C, thus effectively shunting out a single break anywhere in the circuit. Alarm operation otherwise is exactly the same as in the Class B circuit. A Class A circuit therefore, has the ability to turn in an alarm in spite of a single break in the circuit. This gives the circuit a greater degree of sur-vivability than the Class B circuit, eliminates the problem of lost EOL devices, and makes trouble shooting easier. The main disadvantage is that it requires twice the amount of wires and the codes require that the return pair be run in separate raceways or conduit from the outgoing pair to ensure survivability. One typical trouble situation is often caused by a different trade unknowingly cutting through a conduit. There are numerous horror stories regarding this.

CLASS B, STYLE Y NOTIFICATION APPLIANCE CIRCUITNotification appliances, such as bells, horns, strobes or combination horn/strobes are likewise installed in Class B or Class A configurations. However, the supervision of these appliances is made possible by use of blocking diodes wired internally in each device. As you probably already know, a diode only conducts DC voltage in one direction. If the polarity of the supply voltage is reversed, the diode will become an open circuit, and no conduction occurs. See Figure .

Every notification appliance is equipped with a blocking diode. Thus the appliance is supervised up to the point of connection to the circuit. The internal coils, etc. are NOT

May/June 2009 Page

1-888-GEN-TRANwww.gen-tran.com

Safe, convenient way to feed generator power to signals 20, 30 and 50 amp configurations available Optional pilot lights, switched neutral and service entrance Surface and flush mount units available

Generatortransfer switchesfor traffic signals,

now requiredby Code.

Protect importantintersections from

power outages.

A WBENC-Certified Womens Business Enterprise

supervised. Supervisory voltage is applied to the notification appliance circuit, which conducts the current to each appliance where it is blocked by the diode, travels through the EOL device, and back to each appliance and hence to the FACP. In the event of a break or removal of an appliance, the circuit will open and a trouble condition will occur, just as in an initiating circuit. During alarm, the polarity of the voltage is reversed by panel circuitry, and the blocking diodes conduct the current to the interior components of the appliance.

SUPERVISORY CIRCUITSupervisory circuits are basically used to detect off normal condi-tions in the sprinkler system, if one exists in the building. A super-visory circuit is basically an initiating type of circuit, Class A or B, to which are connected supervisory or tamper sprinkler system switches. These switches transfer if a gate valve has been operated to shut off a sprinkler system, if sprinkler pressure is dropping due to a leak, if the water level in a rooftop tank is too low, too high, or if a freeze up is imminent, etc., or in general, indicate a problem or tampering with the sprinkler system. In this case, the supervi-

Continued from page 44The ABCs of Fire Alarm Systems - Section I . . . sory switch process the signal in the same manner as an alarm, but the panel circuitry is programmed not to energize the notification appliances or transmit an alarm signal off premises. Instead, the circuit lights its associated LED and the trouble sounder may sound, as the sharing of the audible trouble signal by both trouble and supervisory circuits is permitted. Some-times a single notification appliance may be energized, such as a flashing yellow light, etc.

Thus, we see the operation of a simple, basic system. A larger system would have additional initiating cir-cuits, notification appliance circuits, circuits that make off premise notification (fire department) by various means, and control circuits to capture elevators, shut down air circulating equipment, and perform various required auxiliary functions.

Our next installment will cover the devices that place a conventional fire alarm system into alarm.

It seems like we just finished cleaning the desert sand off our clubs, and here we are already thinking about next years Outing. The Florida Section is already hard at work planning what we hope will be a great golf outing at one of the premier golf courses in Central Florida.

The National Course at Champions Gate was designed by Greg Norman and offers a great mix of American style bunkers to challenging tee shots for every skill level. As you drive towards a couple of the greens there's enough sand and water to make you think your over on Daytona Beach.

The Outing is going to be held on Saturday the 22nd at 8:00 AM. The price is $55.00 per person. This will include your round, cart and a catered lunch after in the Club House. Registrations can be downloaded from the IMSA web site or contact the Golf Committee for more information and pricing on club rentals.

We hope all of you can join us for one of the more unique events of the conference.

Sincerely, The Officers of the Florida Section

Golf Committee:John Lemonias 727-464-8887 JLEMONIA@CO.PINELLAS.FL.USJeff Dyar 727-464-8909 JDAYAR@CO,PINELLAS.FL.USTyson Evatz 727-464-8982 FLPRES@HOTMAIL.COM

2009 Golf Outing

August 228 a.m.

Page IMSA Journal36

The ABCs of Fire Alarm Systems - Section IIBy Anthony J. Shalna 2009Principal IMSA Representative to the Automatic Fire Alarm Association

President: Southeastern Signalmen of MassachusettsApprovals Manager: Gamewell-FCI by Honeywell

In our first installment, we discussed basic fire alarm con-trol panels that contain one or more initiating circuits and notification appliance circuits. We will go into greater detail about addressable panels in future installments, but now want to discuss some of the devices that place the initiating circuits in alarm.

INITIATING DEVICES

Initiating devices commonly used to activate the initiating circuit of a fire alarm control panel are: heat detectors, smoke detectors, water flow switches and manual (pull) stations.

In this installment, we will concern ourselves with heat detec-tors, which, like sprinkler heads are basically intended for property protection rather than for life safety.

Heat detectors fall into two basic styles of protection: Line, and Spot detection. Line detection protects areas over an elongated path. Spot detection protects an area resembling the area lit by a spotlight.

LINE DETECTION

Line heat detection is less common, but is invaluable for protecting certain hazards. One of the most common line detectors in use today consists of a twisted pair of wires insulated with a thermal coating that has a specific melting point. If excess heat is applied to the cable, the insulation melts, the wires short circuit together, and the control panel goes into alarm. The system is restored by cutting out the damaged section of cable and splicing in a new section. Figure 1 shows a typical line detection device.

Other types of line detection make use of eutectic salts or similar insulation that is non-conductive until it reaches a specified tem-perature and then conducts current from one conduc-tor to the other. U n l e s s m a j o r damage occurs, the insulat ion again becomes non-conductive when the temperature drops, thus making this type of detection essentially self-restoring. Some older systems use copper tubing installed throughout the area, filled with air or gas under pressure. Diaphragm arrangements then respond to increases in pressure caused by heat, and close contacts, creating an alarm.

Line detection is best suited to servicing conveyor belts, escalators, raceways, wire troughs, tunnels, grain elevators, silos, etc. Weatherproof versions of line detection cable are

also available. This line can be stapled under piers or wharfs, allowing out-door weatherproof protection where no other sensors would function properly.

SPOT DETECTION

Spot detectors cover a finite area that varies according to the rated sensitivity of the detector and the distance (height) of the detector from the floor.

The most commonly used types of spot heat detectors are: Fixed Temperature, Rate of Rise, and Rate Anticipation detectors.

Electronic (thermistor/microprocessor) detectors have been introduced fairly recently and may be used only with com-patible control panels, usually addressable panels. With the exception of the electronic versions, heat detectors are mechanical in nature, and contain contacts that close when the detector is in alarm, making them compatible with any conventional control panel.

FIXED TEMPERATURE DETECTORS

The fixed temperature detector goes into alarm ideally when the ambient temperature reaches a certain setpoint. The most commonly used fixed temperature detectors operate on two different principles: fusible alloy and bimetallic strip.

FUSIBLE ALLOY FIXED TEMPERATURE DETECTORS

The fusible alloy unit uses an alloy physically resembling sol-der, but with a much lower melting point. The most common temperature melting points are in the vicinity of 135o F and 190-200o F, depending on the manufacturer. The alloy holds a spring type mechanism in place. This mechanism holds a spring in an extended position keeping a set of contacts open. When the alloy reaches its melting point, the spring is released, allowing the contacts to close, placing the detector in alarm. The detector is usually non-resettable, and either the detector or fusible unit must be replaced after actuation. Figure 2 shows a popular fixed temperature detector with replaceable element.

Figure 1

Figure 2

Continued on page 38

BIMETALLIC STRIP FIXED TEMPERATURE DETECTORS

The bimetallic strip unit contains a strip of metal, plated on each side with a different metal, each of which has a different

Page IMSA Journal38

Continued from page 36The ABCs of Fire Alarm Systems - Section II . . .

Figure 3A

Figure 3B

Figure 4 shows a typical rate compensation heat detector

Continued on page 39

coefficient of expansion. This means that, when heat is ap-plied, the metals expand at different rates, causing the strip to bend, or warp. When the strip bends enough, it touches a contact, completing the circuit. Bimetallic strip detectors are mostly used in household applications, since their listed area of coverage is usually insufficient to meet Code require-ments for larger buildings. Advantages and disadvantages of the various types of heat detectors will be summarized at the end of the article.

RATE OF RISE HEAT DETECTORS

Another type of widely used heat detector is the Rate of Rise detector. This detector contains a chamber with a calibrated vent hole and diaphragm at the top. An actuator strip is located above the diaphragm, just below a set of contacts. When the air outside the detector rises in temperature, the air inside the chamber likewise gets warmer, and as we all know, it expands. If the air expands gradually, it escapes through the calibrated vent. If the air heats rapidly and expands too fast to be vented, pressure is exerted on the diaphragm, causing it to bulge, pushing the contacts closed and placing the detector in alarm. The rate of temperature rise required to place a detector in alarm is 15o F in one minute, or equivalent, such as five de-grees in 20 seconds. Therefore, this detector does not depend upon high temperatures to go into alarm, but senses a rapid rise in temperature. The rate of rise detector is self-restoring, since the diaphragm returns to normal as the ambient air cools. The ROR detector often has a fixed temperature feature as a back-up in the event high temperatures are reached, while the temperature rises too slowly to activate the rate of rise feature. This detector is referred to as a combination Fixed Temperature and Rate of Rise detector. Figures 3A and 3B show typical rate-of-rise/fixed temperature detectors. Figure 3A shows a high profile detector while Figure 3B shows a low profile version developed for use in finished interiors. Both operate identically.

RATE ANTICIPATION DETECTORS

The fixed temperature detector depends upon heat absorp-tion to activate it, and in some instances, a rapidly increasing temperature could conceivably reach a hundred or more degrees higher than the setpoint of the detector before the

fusible alloy could absorb enough heat to melt it. This is referred to as thermal lag. The rate anticipation detector was designed to eliminate ther-mal lag. The rate anticipation detector is cylindrical (cigar shaped), sealed, and contains a pair of bowed struts each containing a contact. In normal operation, the struts are bowed away from each other, separating the contacts. The cylindrical case is made from a special alloy with a coefficient of expansion that allows it to expand rapidly. When the am-bient temperature rises, the detector case expands, (actually stretching or elongating) until the internal struts are likewise stretched, causing their contacts to close, placing the detector in alarm. This detector has very little thermal lag, and will go into alarm as soon as the ambient temperature reaches the setpoint of the detector regardless of the rate of rise. Since this detector is sealed, it is a simple matter to weatherproof or make it explosion proof. See Figure 4.

ELECTRONIC HEAT DETECTORS

The electronic detector depends upon thermistors or similar components that change value when exposed to heat. This detector requires operating voltage in order to measure the change in value, and must also be listed by a Nationally Recognized Testing Laboratory (NRTL) as being compatible with the particular control panel initiating circuit, since it does not contain dry contacts, but alters the characteristics of the circuit, placing it into alarm. This will be discussed in greater detail in a future installment.

APPLICATIONS

The rate of rise detector (ROR) responds to rapid increases in temperature. Therefore, these detectors can be used in any normal ambient and are rated for greater spacing than the fixed temperature detectors. They are especially suited for cooler ambients where they can detect a developing fire long before a fixed temperature unit can actuate. These de-tectors commonly incorporate a fixed temperature element for reliability and are referred to as combination FT/ROR detectors.

July/August 2009 Page 39

Largest Selection Highest Quality Best Customer Service Special Orders

405-340-3434 pelcoinc.com

Continued from page 38

The ABCs of Fire Alarm Systems - Section II . . .

The fixed temperature detector (FT) alarms when the ambient temperature reaches a certain setpoint, commonly 135 or 200o F, and ignores any fluctuations or sudden increases in temperature. Therefore, the fixed temperature detector is best suited for applications where rapid fluctuations in temperature can be encountered, such as attics, boiler rooms, kitchens, bathrooms and loading platforms with forced hot air heaters. The 200o F version should al-ways be employed in attics, boiler rooms, garages or kitchens. The most common fixed temperature detectors are usually destroyed upon activation and must be replaced after the alarm.

The rate anticipation detector outperforms the fixed temperature detector since it will alarm faster than the FT detector in the event of a rapid temperature increase up to the setpoint. It is costlier than the FT de-tector, but is also self-restoring, so doesnt require replacement after actuation. This is a great advantage in installations where staging or scaffolding would be required to replace the detector.

ROR and FT detectors are available in weatherproof and explosion proof ver-sions, but require fairly costly housings for these versions. The rate anticipation detector, being sealed, lends itself to weatherproof or explosion proof applica-tions at a moderate cost.

The rated spacings of these detectors vary. The ROR detector mounted on low ceilings can be spaced up to 50 feet on centers, depending on its listing, while the fusible element FT detector is typically rated at only 15 feet on centers. The bimetallic strip is rated for even less. Electronic/thermistor detectors may be rated for greater spacings. For spacing information, refer to NFPA Standard 72, National Fire Alarm Code. This publica-tion contains all types of information regarding installation, layout and spacing of detectors on all types of ceilings and elevations. Most state codes are now based on this standard.

Now that we have a basic understand-ing of the devices that place a fire alarm control panel into alarm, our next install-ment will concern itself with addressable (microprocessor based) fire alarm control panels.

Page IMSA Journal28

Addressable Fire Alarm SystemsUntil now we have been concerned with conventional fire alarm systems. These systems require specialized wiring practices using in and out connections to devices in order to maintain supervision of the devices. T tapping is an-other practice also not permitted in conventional systems. These will be covered in detail in a later installment.

Addressable systems are a different matter. No specialized methods of wiring are required. These systems are controlled by microprocessors and their operation relies on communi-cation between the microprocessor in the central processing (control) unit, and connect-ed devices which have their own microprocessors. Ad-dressable systems feature an LCD or similar visual display that may give loca-tion and specific informa-tion about the connected devices, depending on how it is programmed. Some systems show detailed in-formation via an alpha-nu-meric display while other smaller, economy type sys-tems might display only a code number assigned to the device. These systems have been nicknamed smart or intelligent systems.

Addressable systems are obviously much more complex than conventional systems, but they have an infinitely greater flexibility. Oddly enough, even though they may be exceed-ingly complex, describing them is a relatively simple matter. The initiating devices and notification appliances are either connected to or incorporate a transponder that has a specific address assigned to it. These transponders are connected to a circuit of the central processing unit which interrogates each transponder in sequence. When interrogated, a device may respond that it is normal, or if an initiating device, that it is in an alarm condition. If a device is inoperative, disconnected, damaged, etc., it will not respond. The central processor then creates a trouble condition and a device missing message will be displayed. Present standards require a maximum of five () seconds for a CPU to report an alarm from a device. The speed of interrogation has increased dramatically in recent years with the development of better and faster microprocessors. Now there exists a signaling protocol for certain models of smoke sensors where the CPU interrogates a large cluster of sensors instead of individually. This cluster will instantly report an alarm condition when interrogated and then indi-cate the address of the specific device in alarm. This saves a significant amount of time that would be otherwise required to interrogate each device individually.

The ABCs of Fire Alarm Systems Part IIIBy Anthony J. Shalna 2009Principal IMSA Representative to the Automatic Fire Alarm Association

President: Southeastern Signalmen of MassachusettsApprovals Manager: Gamewell-FCI by Honeywell

The addressable system operates over a circuit known as a Signaling Line Circuit (SLC). Depending on the manufacturer and design, the circuit can be wired in various methods in regard to fail-safe operation. These methods are presently referred to as Styles, but I understand that the 200 Edition of NFPA 72, National Fire Alarm Code, will revert to an older Class designation. At any rate, I recom-mend consulting the charts of the current edition of NFPA 72 for further information.

The SLC does not resemble conventional initiating circuits since it is a data gathering circuit, while conventional devices, having normally open alarm contacts, place a short circuit across their initiating device circuit. A short on an SLC will cause a trouble condition instead of alarm.

Connected to this SLC are addressable smoke or heat sensors, monitor modules, output modules and/or data gathering panels. Dif-ferent manufacturers all

have their own various designations for these modules, but generically, they are usually referred to as transponders and in common conversation as modules.

The sensors and modules all may be intermingled on the SLC regardless of their (input or control) function. (See illustration) As stated above, they have their own unique address that is assigned when the system is programmed. This address is assigned to the transponder at installation via a DIP or rotary switch. Again, the variety of modules, sensors and data gathering panels is limited only by the manufacturers imagination and technical expertise. Addressable smoke or heat detectors are usually referred to as sensors to distinguish them from conventional detectors. These sensors may either contain an integral address switch, or in older or economical systems, they may simply be conventional detectors installed in an addressable base. Conventional detectors installed in an ad-dressable base would constitute an Addressable System. Smoke sensors containing microprocessors now report their status to the CPU. While a conventional detector has only two conditions: Alarm or Normal, the newest sensors can notify the CPU that they smell smoke, or are approaching an alarm condition, or signal that they are dirty and require cleaning, etc. Their sensitivity can be varied by the control panel and new features are being announced continually.

Systems employing these sensors are known as Analog Continued on page 30

Page IMSA Journal0

The ABCs of Fire Alarm Systems Part III . . . Continued from page 28Addressable since their condition is dis-played in an analog fashion rather than a digital normal or alarm condition.

Many people erroneously use the term Analog Addressable to describe all devices connected to an SLC, but technically this term does not apply to devices such as manual pull stations or electronic addressable heat detec-tors that have been employed in these systems until now. Even though heat sensors may be addressable and signal an alarm via the SLC, they can only report the two states, alarm or normal. Now appearing on the horizon are heat sensors that can indicate their chang-ing condition in the same manner as a smoke sensor, such as reporting an increase in ambient heat, so nothing is etched in concrete regarding the ability and features of these sensors.

Monitor modules have their own as-signed addresses and commonly fea-ture a conventional initiating circuit complete with an end of line resistor. Conventional dry contact initiating devices are connected to this circuit. These may be electromechanical heat detectors, manual pull stations, water-flow switches, etc. When any of these devices go into alarm, the monitor mod-ule provides a collective address for the devices on its circuit. Newer modules are now available featuring multiple initiating circuits and provide different addresses for each of these circuits.

Some modules are small enough to fit inside a device housing or backbox. These modules are usually intended for connection to a single initiating device. An example of this would be a module installed inside a pull station, providing an address for the station. Some manufacturers preassemble the monitor modules into pull stations, etc. while others prefer to sell the module only, and allow the installer to connect the modules in the field.

Other monitor modules have an initiat-ing 2-wire circuit that may be extended considerably, and can accommodate a specified number of conventional 2-wire compatible conventional smoke detectors in addition to a number of dry contact initiating devices. This module provides one address for each circuit, and is useful in large areas, such as auditoriums, atriums or gymnasi-

ums where only a collective address is needed to guide fire-fighting forces to the source of the fire.

Data gathering panels have been used since the earliest days of addressable systems. These panels resemble a small conventional fire alarm control panel and contain a number of conventional initiating circuits, battery standby, etc. These panels in effect provide a subsystem with its own address, but I wouldnt be surprised if some of the latest ones provide multiple addresses, one for each initiating circuit.

Output or control modules may also be installed anywhere on the SLC. These modules likewise have their own unique address and may be pro-grammed to perform a function in the event of an alarm or trouble on any individual or combination of monitor modules or sensors. These control modules may contain dry contacts that transfer on command, but again, new configurations are arriving on the mar-ket on a daily basis. Some will accept an audio input and provide supervised loudspeaker appliance circuits, while others will supervise and operate notifi-cation appliance circuits. Still others can be used to energize releasing solenoids or provide smoke damper operation in smoke control systems.

System programming also varies with the manufacturer. In the earlier days some small system manufacturers used burn-in chips, or PROMS for the programming. Larger systems are programmed via software. In these sys-tems, the programming is often done in the office on a computer and then down-loaded to the control panel in the field via lap-top computer. Some software programs are user friendly while others are not, and usually require specialized training. The easiest ones to use contain drop-down menus and the programmer need only to literally fill in the blanks in order to complete the program.

When the system is programmed, a description of the device can also be included (in systems with alphanu-meric display) such as ionization sensor, main lobby, right wing. Thus, when the device goes into alarm, this message will display on the readout, giving firefighters explicit information as to the location of the fire.

OPERATION SUMMARYThe microprocessor in the main control interrogates or polls the transponders in sequence, with only nanoseconds being required for each device or cluster of sensors. When a transponder is polled, it responds that it is either in a normal state or in alarm. In Analog Address-able Systems, a transponder installed in a smoke sensor will also indicate the condition of the sensor, such as dirty, approaching alarm condition, etc. If the transponder is disconnected for any reason (malfunction or break in the wir-ing) the panel will show a system trouble and show a device missing message (or code) on the display, which will give a specific location of the device. Other fault messages can be for an SLC break, circuit shorted, etc. Some systems may indicate a circuit break by displaying a list of missing addresses located beyond the break or short.

One disadvantage of the addressable system is that polling or interrogation consumes a fraction of a second per device polled. If a substantial number of devices are installed in an individual signaling line circuit, a considerable amount of time could elapse until an alarm is processed. Present control panel standards require a maximum of seconds for a panel to process an alarm after it is received from an initiat-ing device. The development of sensors that will operate in a cluster mode has made this an easier matter. However, it is common to use a larger number of SLCs with fewer devices installed on them rather than have one huge SLC with numerous devices.

The addressable system is also pro-grammed to make specific responses depending on the device activated. Thus the system can process an alarm from a sensor with complete alarm re-sponse including notification of the fire department and operation of elevator capture devices, operation of smoke dampers, etc. It may also process a su-pervisory signal from a tamper switch with sounding of dedicated appliances only, or indicate that a control module has been activated and has performed its required function.

Now that we have described the basic operation of both conventional and ad-dressable fire alarm controls, our next installment will concern itself with smoke detection.

THE ABC'S OF FIRE ALARM SYSTEMS - IV (2009)

Anthony J. Shalna

SMOKE DETECTION

Unlike heat detectors or sprinkler heads that rely on detection of excessive heat for their operation, the basic purpose of smoke detectors is to save lives. The secondary purpose is to save property, especially in conditions where there is considerable smoke build-up and the smoke could do as much or more damage than fire before sufficient heat is generated to actuate sprinkler heads or heat detectors.

With the advent of microprocessor-based smoke detectors, the science of smoke detection has expanded with no end in sight. Information published today will definitely be surpassed in the near future, limited only by the ingenuity of the manufacturers and cutting edge microprocessor technology.

RUDIMENTARY SMOKE DETECTION

The smoke with which we originally were concerned for many years had as its signature, both the visible and invisible particles emanating from combustion. The particles given off nearest to the flame are hot and invisible. As the distance from the flame increases, the particles cool and combine, or agglomerate, so they become visible to the naked eye. Thus they possess less "thermal lift" than the hotter particles. Thermal lift causes smoke to rise. Additional signatures that were beyond the state of the art until recently, were thermal output, carbon monoxide components, infrared radiation and carbon dioxide components.

CONVENTIONAL SMOKE DETECTION TECHNIQUES

Conventional smoke detection falls into two basic techniques at present, ionization and photoelectric. Conventional smoke detectors at present make use of these techniques. However, the expanding use of microprocessors is rapidly changing this picture and Im sure it will change drastically in the future. Conventional smoke detectors latch-in at alarm and require resetting by interrupting the power source momentarily.

IONIZATION PRINCIPLE

As stated above, one of the basic signatures of smoke is the generation of particles of combustion. The conventional ionization detector is concerned with these. This detector contains a radioactive source, usually a minute amount of Americium 241, which bombards the air between two plates, or electrodes in a detection chamber, ionizing the air in the chamber, making it conductive. Since this ionized air is conductive, if a voltage is applied to the plates, a current will flow between them. When particles of combustion enter the chamber, they adhere to the ionized air molecules, neutralizing them so they are no longer conductive, and, simplistically speaking, interrupt or decrease the current flow in the sensing chamber. Detector circuitry senses the decrease in current and puts the detector into an alarm condition. Note the use of fail-safe technology, which is the keystone of fire alarm systems. Ionization detectors respond best to the invisible particles of combustion. In simplistic terms, ionization detectors smell smoke, rather than see it.

PHOTOELECTRIC PRINCIPLE

The conventional photoelectric detector uses an entirely different detection approach. This detector contains a smoke chamber that houses a light source, usually an infrared LED, which emits a light beam. The chamber also contains a photocell that is placed at an angle to the light source so it cannot normally see the light.

If you go outside at night and shine a flashlight into the air, the flashlight beam will be invisible unless there is smoke, fog, mist or dust in the air. If any of these factors are present, the light beam becomes visible to the eye because these conditions scatter the light so it can be seen. This is also the case with the photoelectric detector. Introduction of visible smoke (usually 3%/ft. obscuration of gray smoke) in the chamber scatters enough light so the photocell can see it and register an alarm. Most photoelectric detectors use verification techniques of some sort. The most common technique is to flash the LED at a fixed rate. If the photocell sees light, the flash rate usually increases. If a predetermined number of accelerated flashes still reveal scattered light, the detector goes into alarm. Some detectors may reset themselves between verification flashes while

others may not. The verification (flash) rate varies widely between manufacturers and models.

Recent advances in microprocessors have resulted in one manufacturer developing a detector that is normally set to very low sensitivity to minimize unwanted alarms. Upon detection of traces of other signatures, such as carbon monoxide, etc. it automatically increases its sensitivity to a high level placing it in a high state of readiness in the event of alarm conditions.

In simplistic terms, photoelectric detectors see smoke rather than smell it.

DETECTION STYLES As in the case of heat detection, smoke detection styles in common use today are "Spot" detection and "Line" detection. With "Spot" detection, a detector is placed so that it protects an area or "spot", while the "line" detector protects an elongated path. Other methods of smoke detection, such as air sampling and video smoke detection are in use. However since the above-mentioned techniques comprise the vast majority of smoke detector installations today, we will presently concern ourselves only with these.

SPOT DETECTION Spot detectors typically cover an area up to 900 square feet, (30 foot centers) based on smooth ceilings 15 feet high, and with minimal air movement. This spacing is based on the smoke detector manufacturers recommendation. Extensive engineering studies are presently being conducted to determine the validity of this spacing and results of the studies may alter these spacings.

LINE DETECTION Line detection smoke detectors at present consist of projected beam smoke detectors. These detectors rely on a slightly different approach from spot detectors inasmuch as the beam detector depends on a gradual obscuration of the light beam with a sudden obscuration causing a trouble condition. These detectors will be covered in greater detail in a future article.

TEST REQUIREMENTS All photoelectric and ionization detectors used with fire alarm control panels meet requirements of ANSI/UL Standard 268.

Smoke detectors intended for residential use are tested to a different standard, ANSI/UL217, and differ in operation slightly. While system detectors usually latch or lock in when alarmed, requiring reset at the control panel, residential detectors usually self-restore. These detectors usually contain an audible alarm signal, and so are better described as smoke alarms.

ADVANTAGES AND DISADVANTAGES Ionization detectors, because they detect small or invisible particles of combustion that by nature, are hot, respond more readily to fast, flaming fires or fires in their incipiency. One of the weaknesses of this type of detector is that over-heated metal cookware, exhaust from internal combustion engines (gasoline or diesel powered), aerosols from sprays, even strong fumes, such as from ammonia, all produce particles similar to products of combustion, causing unnecessary alarms. There are legendary tales 25 or 30 years ago, of alarms caused when ionization detectors were originally installed in racing stables. When a horse urinated, it caused the ionization detectors to go into alarm! These weaknesses caused no end of serious problems also around the same time when ionization detectors were widely used in household applications. Cooking odors caused numerous unwanted alarms causing occupants to disconnect these detectors or remove the batteries and then forgetting to place them back in service. Inevitably, in many instances, when actual fires later broke out fatalities resulted due to detectors being disabled. Different manufacturers have been able to overcome some of these difficulties to various extents but the problem remains, overall, a generic one. Additional problems also encountered in past years which have been overcome to an extent, resulted from high air velocity and also from burning polyvinyl chloride (PVC) which gives off large amounts of dense, cold smoke but contains few invisible (hot) particles.

Photoelectric detectors, because of their ability to detect scattered light, react to visible (gray) smoke, and are quite immune to problems caused by invisible particles resident in their environment. However, they are also affected by anything that scatters light in the same manner as visible smoke, such as steam, fog or dust. In addition, miniscule insects in some areas have had a tendency to set up housekeeping in the smoke chamber despite use of insect screens. These insects may grow rapidly, often in a period of hours. Their physical presence, cobwebs, etc. scatter light and can be a source of unwanted alarms. Photoelectric detectors also have difficulties with black smoke, since black smoke obviously doesn't scatter light.

CONFIGURATIONS Conventional smoke detectors are also sub-divided into two configurations, 2-wire and 4-wire. Both configurations have their advantages and disadvantages. The 4-wire detector requires an extra pair of conductors which furnish operating power to the detector. These are in addition to the initiating circuit wiring. The detector may operate on either the ionization or photoelectric principle, and contain a relay that energizes upon alarm. The dry contacts of this relay connect to the system initiating circuit in the same manner as a pull station. Since the power to the detector is unsupervised, an end-of-line (EOL) relay is connected to the last detector on the voltage supply circuit. (This may not necessarily coincide with the initiating circuit wiring, so care must be exercised here.) If the voltage supply wires are tapped off another supply circuit, an end-of-line relay must be connected to the end of each branch. This EOL relay contains contacts which are normally closed while the relay is energized and open when the relay de-energizes (due to power failure). The contacts of this relay are connected in series with the end of line device on the initiating circuit so a power failure, breaking of the power circuit, or removal of a smoke detector will cause an initiating circuit trouble signal. The 2-wire detector also connects to the system initiating circuit but uses the initiating circuit supervisory current as its source of operating power. In alarm, the detector draws enough current to cause the circuit to go into alarm. It cant place a short across the circuit since this would leave no current to light the alarm LED in the detector and absence of current would cause the detector to reset itself, then again energize into alarm, and reset itself, causing it to act like a vibrator or buzzer. Because the detector draws current from the initiating circuit, usually in the range of microamperes, it also stands to reason that there is a limitation as to the quantity of detectors that can be supported by a circuit. If the detector impedance is too low, the total impedance of a large number of detectors could draw enough current to put the circuit in alarm. Another possibility is that the impedance could also be low enough that a break in the circuit could go undetected, as the impedance would fool the circuit into thinking that an EOL resistor was still connected to it. The situation becomes more complex when one realizes that the amount of current the detector draws in alarm may not be enough to place the initiating circuit in into alarm, since both detector and initiating circuit possess their own peculiar response curves. This presents us with the problem of compatibility.

COMPATIBILITY Since 4-wire smoke detectors require a separate source of current, until lately any NRTL Listed four-wire smoke detectors may be used with any NRTL Listed control panel with similar operating voltage if the failure of the operating power is supervised. Lately however, the NRTL Standards are in the process of being changed since a situation was uncovered not long ago, where a control panel by one manufacturer, in worst case conditions, didnt provide sufficient voltage to power a 4-wire smoke detector. Care should be taken, also, that the particular model of detector does not require current filtering or regulation which the panel may not be able to provide.

2-wire detectors, because of considerations listed above present an entirely different problem. NRTL requirements state basically that two-wire detectors cannot be connected to control panels unless they are Listed as compatible with that model of panel. Compatibility testing is performed or confirmed by the NRTL, and compatibility identifiers must be displayed both on the detector and initiating circuit module in the panel. The suitability of two-wire smoke detectors for use with any control panel is determined very simply. The manufacturer lists all compatible detectors either in the Installation/Operating Manual or in an official Compatibility Document, with the admonition to use only the detectors listed in the document. If the detector is not listed in these publications, it is not compatible.

Another means of establishing compatibility is by the back door compatibility method. A smoke detector manufacturer purchases a number of different control panels, tests his detectors with these panels and has the test results confirmed by an NRTL. The smoke detector manufacturer then will also publish a compatibility document. The writer feels that the safest course to take in this instance is to use only detectors specified by the control panel manufacturer. In the event of a major disaster, the question of liability will eventually be established by the courts. The panel manufacturer will disclaim all responsibility if the detectors installed are not those specified in his manual. This will lead to any other parties having to bear a much greater burden of proof.

2-WIRE ELECTRONIC HEAT DETECTORS Electromechanical heat detectors have been previously covered in this series, but we avoided discussing thermistor and

microprocessor type heat detectors until now. Electronic heat detectors operate on the ability of an electronic component, the thermistor, often in conjunction with a microprocessor chip, to change resistance when exposed to heat. The resistance change is measured in a bridge type circuit in the detector, which places the detector into alarm when a specific temperature is reached. This detector operates in the same manner in the control panel initiating circuit as the 2-wire smoke detector. Therefore, these units must also meet compatibility requirements.

ADVANTAGES AND DISADVANTAGES

Four-wire detectors have the disadvantage of requiring an extra pair of voltage supply wires and installation of an end-of-line relay at the end of each voltage branch. The detectors and the EOL relays also consume current, and this has to be taken into consideration when calculating the size of the standby batteries. However, compatibility with the control panel is not presently a problem, although this may change as noted above. Since dry contacts are employed, there is virtually no limitation to the number of detectors that can be connected to an initiating circuit. In addition, the detector alarm relay can contain auxiliary contacts which will reliably perform any intended auxiliary function, regardless of how many detectors on the same circuit are in alarm condition. The four-wire detector is preferred for use in duct detectors since these detectors are often used for a number of smoke control functions in addition to just turning in a single alarm. Two-wire detectors draw miniscule amounts of current and dont require end-of-line relays, so they have minimal effect on the size of standby batteries. Installation costs are lower since only two wires are required, and they lend themselves to retrofit (replacing heat detectors) using existing wiring. However, NRTL compatibility requirements come into play now, and because of the loading on the initiating circuit in an alarm condition, their ability to perform auxiliary functions is severely limited. Also, because of this loading, there is a definite limitation regarding the quantities of detectors which can be installed on a single initiating circuit, and in many cases, the number is reduced further when Class A, Style D circuits are involved. Some models contain auxiliary relay contacts or have them in their bases, but these contacts may not transfer if other devices on the initiating circuit are in an alarm condition, especially dry contact devices. Therefore, two-wire detectors are recommended mainly for putting an initiating circuit into an alarm condition, and not to perform auxiliary functions.

VERIFICATION CIRCUITRY (CONTROL PANEL) Verification circuitry in a fire alarm initiating circuit introduces a delay period into the smoke detector operation that could diminish unwanted alarms caused by transient conditions. ANSI/UL Standard 864, which covers Fire Alarm Control Panels, contains the regulations governing use of verification circuits. Per this standard, initiating circuit verification can only be used with smoke detectors that have a verification (flash rate) time of less than 10 seconds. Verification can be used with

either 2-wire or 4-wire smoke detectors, although the use of verification with 4-wire detectors is unknown, since verification cannot be used with open contact devices. Many control panels with verification circuitry can differentiate between two-wire smoke detectors and open contact devices. With open contact devices, the verification sequence is usually automatically by-passed. Per ANSI/UL 864, a verification circuit has a sequence that operates like this: When one detector goes into alarm, the cycle begins with a 20 second pre-alarm window. The panel shows a trouble condition during this period, with the trouble light usually flashing. This window is followed by a 4 second automatic reset time (the trouble light usually stays lit constantly at this time), followed by a 100 second alarm verification window (flashing trouble light). The alarm verification window holds the circuit in the pre-alarm state, waiting for the first alarm to repeat itself. If a subsequent alarm comes in during this period from the pre-alarmed detector (or any other detector in the circuit), a system alarm will occur. If an alarm is not received within the 100 second period, the panel returns to normal condition. Its my opinion that the expanding application of microprocessors will eventually increase detector reliability to the extent that the verification requirement will no longer be necessary.

DUCT DETECTORS Duct detectors are used primarily to prevent the spread of smoke in a ventilation system by shutting down the HVAC system in case of fire. They should not be used as a substitute for area detectors because smoke may not enter the ducts if the HVAC system is shut down. Duct detection techniques usually involve either ionization or photoelectric spot detectors mounted in a housing fastened to a wall of the duct. Perforated or slotted tubes extend from the duct housing through the wall of the duct and extend across the duct. These tubes, commonly referred to as "sampling tubes", allow air flow from the duct to be channeled into the detector housing and detector. Duct detectors are used in systems with airflow greater than 2,000 cubic feet per minute capacity and are usually rated for air velocities of roughly 500 FPM to 3,500 FPM, although recent advances in technology may alter this. Spot detectors also have a velocity rating that can range up to 3,500 FPM depending on the model, etc. In the event of physical difficulties (duct too small) in mounting a duct detector, or if the air flow in a duct is too low for sampling tubes, a spot detector may be mounted inside the duct as long as it has a listed air velocity rating compatible with the air velocity in the duct, has a remote alarm indicator, and it can be readily tested in place or remotely. Care should be taken in determining the air velocity rating of a detector, as different manufacturers use different terminology. For example, a detector rated at 300 FPM might be capable of turning in an alarm at that velocity, while another manufacturer might rate a detector at 2,000 FPM meaning that the detector will be stable (will not false alarm) at that velocity. The detector might not be able to turn in an alarm at that velocity, however as the smoke may not linger long enough to create an alarm. A spot detector mounted in a duct is also subjected to contamination and therefore the maintenance intervals should be reduced.

APPLICATIONS The comments given in the advantages and disadvantages paragraph above give guidance in regard to applications. As in the case of suppression systems, the detectors should be chosen in regard to the fire load, or what type of fire would be expected to occur in the area being protected. If a smoldering fire could be reasonably expected to occur, such as in the case of dormitory or hotel occupancies, a photoelectric detector might be the best choice. If a fast flaming (or arson) fire could be expected, an ionization detector would provide the quickest response. In the case of computer rooms, a mix of both types is often used.

LIMITATIONS "Spot" type smoke detectors are limited by Standard to indoor operation between 32oF and 120oF, and maximum relative humidity of 90%. Areas of installation should be relatively clean. Detectors in an ambient outside these parameters may or may not function properly as high ambient temperatures may cause detectors to "go to sleep" or conversely, cause the unit to go into alarm. Any environments outside the prescribed parameters are referred to as "hostile" environments.

TESTING According to Standard, smoke detectors should be tested in accordance with manufacturers' instructions. Some of these manufacturers are very insistent that testing be performed according to their instructions, and in many cases, prohibit use of other methods or application of foreign substances, going so far as to invalidate warranties and disclaim all responsibility and liability if instructions are not followed explicitly.

ADDRESSABLE AND ANALOG ADDRESSABLE SMOKE DETECTION. Addressable or analog addressable Smoke detectors are commonly referred to as smoke sensors since they are microprocessor based and used only with intelligent or analog addressable fire alarm controls. Recent advances in detection technology make this a major topic in itself that will be reviewed in a future article.

January/February 2010 Page 7

The ABCs of Fire Alarm Systems - Part V Putting It All Together

By Anthony J. Shalna 2009 Principal IMSA Representative to the Automatic Fire Alarm AssociationPresident: Southeastern Signalmen of Massachusetts

Approvals Manager: Gamewell-FCI by Honeywell

In previous articles we discussed the fire alarm control panel, devices that place it into alarm and the most com-mon devices that are turned on by the panel. The main feature that distin-guishes a fire alarm control panel from burglar alarm or switching panels is that the fire alarm panel is supervised, which means that it has the ability to monitor its own integrity.

Unlike the burglar alarm panel, which has only two conditions, normal and alarm, the fire alarm control has a number of conditions or states. These are: normal, alarm, trouble, and (fairly recently) supervisory. The normal and alarm states are obvious.

The supervisory state monitors sprin-kler devices so the panel can indicate that a waterflow device, such as gate valve, is in an off-normal condition. It is desirable to know that someone turned off the water supply to a sprinkler sys-tem (or forgot to turn it back on after service), but there is no need to create an alarm condition. Other supervisory devices can monitor water tanks for freezing, low or high water levels, etc. The supervisory condition results in a signal that differs from both alarm and trouble conditions, although the supervisory condition may share the trouble sounder.

The last state or condition is the trouble condition. This condition is character-ized by a yellow light on the panel accompanied by the sounding of an au-dible device, such as a buzzer or piezo-electric sounder, which may be silenced or acknowledged temporarily. Upon correction of the trouble condition, the sounder will re-energize, indicating that the panel is back to normal. Re-turning the silencing switch to normal or pressing the acknowledgment button will silence the sounder and return the panel to a quiescent condition. This is known as ring back, a phrase that was common in the past, but not used very frequently nowadays.

Trouble signals are caused by numer-ous things. Some of these are a break

in the field wiring, AC power failure, battery disconnection or failure, ground faults, open fuses, removal of plug-in detectors, disarrangement of panel switches, etc.

We have seen how supervised circuits of conventional fire alarm panels oper-ate with the aid of end of line devices that maintain a current flow through the supervised circuit. Addressable systems operate in a different man-ner and will be the subject of a future article. The current flow through the supervised field circuit must be maintained through the field wiring. This is why conventional fire alarm systems must be wired in a prescribed manner.

Terminations to a detector or appliance must be made by cutting the field wires at their respective terminals. In other words, one wire must bring current into a terminal of the detec-tor or appliance, and a second wire must exit from the same terminal and connect to the next device. See Figure 1. If a field wire is not cut, but looped around a screw, there will be no interruption of the su-pervisory current should the head of the screw shear off. The device will be disconnected from the circuit but there will be no trouble indication. The panel will never know that the device is disconnected. If the field wire is cut and both ends connected to two separate terminals, the shearing of a screw head or loss of a wire crimp lug will cause the ends to separate and a trouble condition will immedi-ately result.

One termination method is to have two screw terminals on the same

metal terminal plate with the plate providing continuity between the screws. See the illustration at the bot-tom of Figure 1. Thus the in wire will connect to one screw and the out wire to the second. Pressure plate terminals are also widely used. Two wire ends are stripped and inserted under a pressure plate. A screw holds the plate down and maintains continuity between the wires. See Figure 2 for a description of the proper and improper terminations with this type of terminal.

Another method of termination is with four () pigtail connections, two for the in wiring and two for the out wiring. See additional illustrations in Figure 2 for both improper and proper methods of connecting the pigtails. An X shows the unsupervised wire which, if cut, will remain undetected.

Continued on page 48

Page IMSA Journal8

The ABCs of Fire Alarm Systems Part V . . . Continued from page 47Wiring to initiating devices in a conventional system or notification appliances in addressable or conventional sys-tems must be made in an in and out fashion. Branching or T-tapping is not permitted in these circuits. Again, the reason for this is that the current must flow in and out of each device and finally through the end of line device so supervi-sion is maintained. Therefore, removal of a device from the circuit or a break in the wiring will interrupt the supervisory current and create a trouble condition. If a T-tap is used, a break in the T-tap branch will go undetected, since the supervisory current will not be interrupted. Figure 2 also shows both proper and improper methods of connection. An X indicates a break that will go undetected.

Note that T-tapping is permitted in an addressable system, since the microprocessor polls all addressable devices and will promptly detect a missing device.

A third method of wiring is commonly used with plug-in style two-wire conventional smoke detectors. One wire goes in and out of one base terminal, usually a pressure plate type of terminal. The other wire connects to a second terminal in the base and exits from a third. The detector has a built-in jumper that maintains contact between the second and third terminals, so if the detector is unplugged, it will cause a break in the wiring resulting in a trouble signal.

The type of wire used for field wiring should conform to the codes in effect in the area. These codes are almost always based on Article 70 of the National Electrical Code, NFPA 70, but some states make additional requirements. One state not only specifies the acceptable types of wire, but also speci-fies the insulation color, with DC power, initiating circuit wiring, initiating circuit return (Class A) wiring, notification appliance circuits, etc. all having different insulation color requirements! The prudent thing to do is to consult with the Authority Having Jurisdiction, such as wiring inspector, fire marshal, etc.

Until recently, solid or bunch-tinned stranded wire, 18 gauge minimum, UL Listed for fire alarm use, were the only types of wire acceptable for fire alarm. The reasoning behind this requirement was that solid or bunch-tinned stranded wire would be most likely to break cleanly, giving an instant trouble indication. If a stranded cable were to be damaged, leaving only one strand intact, the one strand would conduct supervisory current and maintain normal operation. During alarm, a notification appliance circuit could draw enough current to burn out the single strand, with a resultant failure at the most critical time. Now the NEC contains exceptions allowing stranded wire under certain circumstances. In addition, communication cable is also allowed in certain instances.

Power limited and non-power limited wiring also comes into account. There is no hard and fast simple rule about what types of circuits are power limited. The only way an installer can make this determination is that power limited and non-power limited designations are printed by the manufacturer on the control panel door label or on the ter-minals themselves. A typical label might state: All circuits are power limited with the exception of the AC input, battery

and city box connection.

Again, consult the local Authority Having Jurisdiction about the wire hierarchy chart in the NEC, if your State Code is based on the NEC.

Improper Termination Proper Termination

Figure 1

Figure 2

Fire Alarm Control Panel

Fire Alarm Control Panel

End of Line Device

End of Line Device

Smoke DetectorSmoke Detector

Smoke DetectorSmoke Detector

Pigtail ConnectionsCorrect Wiring Method Pigtail ConnectionsIncorrect Wiring Method

Correct Wiring Method

Incorrect Wiring Method

Page IMSA Journal2

The ABCs of Fire Alarm Systems - Part VIBy Anthony J. Shalna 2009 Principal IMSA Representative to the Automatic Fire Alarm Association

President: Southeastern Signalmen of MassachusettsApprovals Manager: Gamewell-FCI by Honeywell

NOTIFICATION APPLIANCESSo far, we have covered the basic operation of control panels and the initiating devices that place the pan-els in an alarm condition. We are now going to review the notification appli-ances that are energized or turned on by the control panel.

The term notification appliance cir-cuits has undergone somewhat of an evolution in the past few years. These were designated for many years as signal circuits, alarm circuits, and also indicating circuits. Revi-sions to NFPA Standards a few years ago adopted the term notification appliances and notification appliance circuits (NACs) to avoid confusion and misunderstanding and also to differentiate them from the commu-nication circuits used by addressable fire alarm systems.

Notification appliances are divided into two very basic categories, au-dible and visual. Obviously, the audible appliances make loud noises and the visual appliances flash bright lights. Recently, tactile devices have been developed for the hearing im-paired, but are used at the present only in very specialized applications and only as a secondary or ancillary unsupervised device. One example of this would be a bed shaker which is used to awaken the hearing impaired.

AUDIBLE APPLIANCESInstalled audible notification ap-pliances should differ from other audible devices in use in the area. For example, if a school uses bells to signal the start of classes, bells should not be used as notification appliances in that building.

The most widely used audible appli-ance is the horn. Horns presently on the market employ both electro-mechanical and electronic designs.

Electromechanical horns have been used for many years and are con-structed using a set of breaker points and diaphragm. These horns are capable of providing a surprisingly loud sound output, but have the dis-advantage of the mechanical breaker points oxidizing, corroding, or going out of adjustment. In addition, the loud output requires a substantial amount of current to do the job. Other disadvantages are that only one tone is available, the breaker points create spikes which can be transmitted back to the control panel, and elec-tromagnetic interference (EMI) can be radiated from the area where the horn is sounding. Manufacturers have designed their horns to minimize these problems, but EMI in the past has been known to cause unwanted alarms from smoke detectors located immediately in front of the horn.

One of the newest developments in this area is the electron- mechanical horn which uses the old familiar dia-phragm, but has an electronic circuit in place of the breaker points, much in the same manner as electronic ignition replaced the old automobile breaker points. This type of horn has the advantage of generating substan-tial sound output without the inher-ent disadvantages of breaker points.

Electromechanical disadvantages led to the development of solid state horns, which mostly operate on a piezo-electric principle. These horns offer good sound output combined and have a much lower current draw. They also offer a variety of tones, with siren, steady or warbling tones being commonly available, often by merely selecting the proper jumper position inside the enclosure. The disadvantages of electronic horns presently on the market are their loudness and the fact that many electronic horns are not suitable for weatherproof use. The loudness of these horns is measured at a higher frequency than the electro-mechanical horns. While the decibel measurements are comparable, in actual applications the electronic horns sometime are not as audible to the average human ear, or especially to people with deteriorating hearing. Recent research indicates that a lower frequency tone, in the vicinity of 20 Hz is heard much more readily by people with deteriorating hearing. Sound output, of course, is dependent upon building construction, layout, furnishings, etc.

The latest fire alarm requirements require a temporal pattern which requires the notification appliances to sound three rounds of three blows each, with specified intervals between blows and rounds of code. Temporal patterns are usually programmed at the factory by the panel manufacturer and the installer need not be not be concerned with the proper time du-ration of the blows and rounds. The most the installing technician need be concerned with is selecting the proper jumper or programming the panel to provide this sound pattern.

Bells have historically been associated with fire alarms and are still in use

Continued on page 53

Horn Strobe

March/April 2010 Page

www.solar-traffic-controls.com/imsInfo@solar-traffic-controls.com

Everything you need to know about solar-powered

flashers is at www.solar-traffic-controls.com

Wireless traffic systems solar-powered flashing beacons for school zones 24-hour applications pedestrian crosswalks road hazards stop signs radar speeddisplays wildlife crossings ITS sensors andcamera systems pre-emption repeater systems DC LEDs and specialty systems for DOTs;firefighters; EMS; police chiefs; industrial facilities

and public works departments view ourentire product line learn how to stretch yourbudget with affordable and effective systems

from Solar Traffic Controls

The ABCs of Fire Alarm Systems Part V . . . Continued from page 52today, although their use is rapidly decreasing due to the Temporal Pat-tern requirements. Bells presently in use usually range from to 10 inches in diameter.

Most bells encountered today are made in vibrating, motor-driven or single stroke variations with vibrating or motor driven bells in the majority. The single stroke bell is suited only for march time or coded applications (including tem-poral pattern) as it emits a louder, cleaner sound than a vibrating bell. The disadvantage is that it draws considerably more current than the vibrating or motor-driven bells and cannot operate on uninterrupted or non-coded current, as the bell would give one blow and then hang up. Vibrating bells draw moderate amounts of current, while motor bells usually draw less cur-rent. A disadvantage of the motor driven bell in the past was the ten-dency of the motor to stick or bind because it sat idle for long periods. To be sure, present manufacturers again have overcome many of the shortcomings, but these problems are inherent in the principle of op-eration. Vibrating bells do not lend themselves readily to Temporal Pattern signaling.

Chimes operate in the same man-ner as bells, and are usually con-structed similarly to bells, using the same mechanism, only with a chime kit mounted in place of the gong shell. Chimes do not have adequate sound output for use as general evacuation appliances, but are mostly used in hospitals, etc. as prealarm or presignal devices to alert available personnel that an emergency exists and a general evacuation may be imminent.

The last audible appliance is the loudspeaker. Obviously, these are intended for voice evacuation systems, and must be listed by a Nationally Recognized Testing Lab-oratory (NRTL) for fire alarm use. Speakers are installed in supervised circuits and the great majority are used in high-rise buildings where

selective evacuation is necessary due to the difficulty of evacuating a building completely in an emergency. The results of the 9-11 disaster indicate that this prac-tice is far from being resolved and much effort is now being dedicated to mass notification measures. Emergency voice evacuation and mass notification systems will be discussed in a future article.

Continued on page 54Speaker Strobe

Page IMSA Journal

The ABCs of Fire Alarm Systems Part V . . . Continued from page 53VISUAL APPLIANCESVisual appliances employed in the past included flashing incandescent lights and various types of strobe lights. With the advent of the Ameri-cans With Disabilities Act, (ADA), only high intensity strobes are ca-pable of producing the light intensi-ties required. The ADA guidelines (ADAAG) have been updated in the recent past, are quite complex, and have been the subject of many lectures and much discussion. There was great confusion in the early days of the ADA, since the ADA is a federal law, enforceable basically in court, and local code enforcement officials had no authority to either approve or interpret the ADA. In fact, these requirements have been the subject of entire publications and are beyond the scope of this article. Many state codes have been harmonized in ac-cordance with ADAAG guidelines, so now meeting these codes also satisfies the ADA, but not all codes have been updated in this fashion. Anyone planning to become involved with the layout of fire alarm systems is well advised to consult one or more of these publications.

And now to return to strobe lights. These have a high intensity out-put, with some providing a light intensity as high as 110 candela. Light intensity levels are defined as CANDELA (effective candlepower) which is the measuring criteria per ANSI/UL Standards 1638 and 1971, and is defined as the average light output generated during one flash cycle. This should not be confused with peak candlepower which is im-mensely greater. Before strobes were standardized, some manufacturers used peak candlepower figures in their literature which led to some confusion.

Other strobes in common use are rated as 1/7 candela, as they emit a 15 candela flash when viewed from the side, but exhibit 7 candela when viewed directly.

Both the fire alarm designer and in-staller should be aware of one prob-lem with strobes. A strobe flash rate

greater than 1 Hertz per second could induce epileptic seizures in persons afflicted with this disorder. There-fore, the ADA mandates a flash rate between 1 and Hertz/second. If two or more strobes are observable from a single location, synchronization of the strobes is necessary to meet this requirement. Microprocessor based panels currently on the market are designed to meet this.

The other important thing to remem-ber when using strobes is that the high intensity light output is directly related to current consumption. In many cases, the quantity of strobes required for an installation may ex-ceed the capacity of a smaller control panel which might have a limited number of notification appliance circuits. Many of these circuits are commonly rated at 1.7 amp. maxi-mum per circuit with a total panel limitation being somewhat less, such as a maximum of amperes for both circuits combined. It doesnt take many strobes at .2 amperes each to exceed the capacity of such a circuit. In these instances, a larger panel or NAC extender panel will be required. These supply additional notification appliance circuits and correspond-ingly larger power supplies. Strobe lights depend on the charging and discharging of a capacitor to flash their xenon flash tube. If the current to the strobe is interrupted or pulsed, it could interfere with the proper operation of the strobe. Micropro-cessor-based controls and extender panels are designed with provision for synchronization of strobes and temporal patterns for the horns.

Strobes are most often mounted on horn and speaker covers so both the audible and strobe comprise a single unit. This makes for simpler wiring and combines both into a compact package. Strobes are also available in stand alone configurations and ceil-ing mount, although ceiling mount units do not at present completely meet ADA requirements, as the light intensity requirements are based on the strobes being observed directly, something difficult to do with an overhead device. Strobes are also

available on plates which accommo-date bells or chimes, so they can also be installed in one package.

Some jurisdictions require (errone-ously) that only the audible appli-ances can be silenced after an alarm with the strobes continuing to flash until the panel is reset. This practice has been criticized recently, as flash-ing strobes indicate an evacuation signal to the hearing impaired.

COMPATIBILITYUntil fairly recently, compatibility was not an issue as far as notifica-tion appliances were concerned. The devices merely had to have the same nominal operating voltage as the noti-fication appliance circuit. The advent of microprocessor panels has resulted in some control panels operating at somewhat higher voltage levels, resulting in instances, depending on the supply line voltage to the panel, where the operating voltage was outside the range of some notification appliance. This, in addition to flash synchronization requirements makes it necessary for strobes and horns to be tested for compatibility with indi-vidual control panels. If there is any question about device compatibility, the panel installation manual or man-ufacturers compatibility document will list the compatible devices.

The only exception to this compat-ibility requirement is for loudspeak-ers. There are no existing require-ments at present. The main thing to remember, however, in the case of strobe/speaker combinations, the strobe is still subject to compatibility requirements.

Page IMSA Journal2Continued on page 26

The ABCs of Fire Alarm Systems - Part VII Advanced Smoke Detection

By Anthony J. Shalna 2009 Principal IMSA Representative to the Automatic Fire Alarm AssociationPresident: Southeastern Signalmen of Massachusetts

Approvals Manager: Gamewell-FCI by Honeywell

Line detection techniques use a projected beam of light that covers an elon-gated path. Therefore, all line type smoke detectors are of the photoelectric projected beam design. The typical projected beam detector consists of a light source (transmitter), a light beam receiver and a re-ceiver control which may or may not be combined in the same housing with the receiver. The latest beam detectors now in-corporate both transmitter and receiver in the same housing, with a mirror or reflector placed at the far end of the path to be covered. The majority of units on the market are of the four-wire configura-tion. That is, the receiver control requires power from the control panel or power supply, usually contains both alarm and trouble dry contacts and is wired to an initiating circuit in the same manner as a four-wire spot detector except that an end-of-line relay may not be required, since the receiver/control contains its own trouble contacts.

The light source projects a beam across a protected area. The light beam is typically infrared, and in some cases is modulated, to eliminate the possibil-ity of extraneous infrared radiation interfering with the operation. If a smoke build-up causes a gradual obscuration of the beam over a period of several sec-onds, the receiver control causes the alarm contacts to transfer. An abrupt in-terruption of the beam will cause a trouble indication.

Thus the receiver can also perform the function of an end-of-line relay so a trans-mitter power failure will be immediately detected as an abrupt interruption.

At present, the projected beam detector protects ar-eas of up to 19,80 square feet (which is approximate-ly the coverage of 21 spot type detectors) as opposed to the 900 square foot area typically protected by spot detectors. Even though the width of the beam might seem relatively narrow es-pecially at extreme ranges, (typically up to 100 meters), the standard recognizes the tendency of smoke to billow or mushroom as it rises, hence the large area of protection. Thus the detec-tor is primarily concerned with visibility of the beam over a long path, rather than with the amount of smoke entering a spot de-tector smoke chamber.

Sensitivity of line detec-tors is expressed in terms of percentage obscuration. One such detector will go into alarm upon a 0 to 90% obscuration of the beam for a period of or more seconds. In comparison to spot detector sensitivities of 2-% per foot, this does not sound sensitive at all. However, remember this obscuration applies over an elongated path which could be over 00 feet long. Therefore when expressed in terms of obscuration per foot, the detector has a sen-sitivity which the standard lists as 0.2 to 2.% per foot depending on the length of the protected area. Because the projected beam detector works on an obscuration

basis and not on light scat-ter, the detector is color blind and responds read-ily to black smoke.

Since line detectors must project a light beam, it stands to reason that current consumption will be higher than that of a spot detector. Attempts to alleviate this situation have met with mixed success. One such method is to flash both the transmitter and receiver simultaneously and thus drastically reduce current consumption much in the same manner as spot detec-tors. Advanced technology using microprocessors have largely overcome this prob-lem, so both two-wire and intelligent addressable configurations are now becoming common.

EnvironmentSince projected beam detec-tors are not dependent on smoke chambers, most of the spot detector limita-tions do not apply. Depend-ing on the manufacturer, ambient temperatures can range from -22oF to 11oF, and some units are listed as waterproof, so the trans-mitters and receivers may be hosed down to remove contaminants. Most units are unaffected by presence of low levels of contami-nants in the surrounding atmosphere.

Virtually all projected beam detectors contain AGC (Au-tomatic Gain Control) cir-cuitry of one sort or another which makes compensat-ing sensitivity adjustments in the event of build-up of contaminants on the lenses. One such unit is capable

of making a number of periodic adjustments and then will cause an intermit-tent trouble signal after the final adjustment. Various manufacturers have shown different preferences in regard to the length of the adjustment period, which may range from minutes up to 8 hours.