Smoke detector

Abstract

A smoke monitoring instrument having first and second chambers, each with a respective photodetector means, a light source common to both chambers and positioned to illuminate part of the interior of each chamber, means for conveying a first air sample from the space to be monitored into said part of the first chamber, means for conveying a second air sample from outside the space into said part of the second chamber, each photodetector means being arranged to receive light from the source which is scattered by its respective air sample but not to receive light directly from the source whereby the effects of suspended material in the first sample originating from outside the space to be monitored can be observed or removed from the output of the instrument.

Description

This invention relates to early warning fire detectors, and has for its object the provision of means for detecting the presence of smoke particles (or similar light scattering material -- for example, airborne dust -- hereafter included within the term "smoke") in gases such as air.

Existing early warning fire detection systems employ smoke sensing units which operate in a variety of ways, chiefly by ionization, light obscuration or light scattering. A common characteristic of these units is that they must be designed and operated in such a way that false alarms are minimized. They are intended for connection into automatic alarm systems, and false alarms can result in substantial inconvenience -- and occasionally even water damage to property which they are intended to protect. In practice, this requirement results in the setting of such units to be relatively insensitive, and substantial concentrations of smoke must be present to trigger them. For example, the specifications of one known device require it to trigger only at a light obscuration figure of 20 percent per meter, and to remain quiescent at an obscuration figure of 4 percent per meter. With the type of fire likely to develop in telephone exchanges, computer installations or the like, smoke concentrations of this order take a substantial time to accumulate, so that potentially valuable early warning time is presently being lost. However, merely increasing the sensitivity of the units is not a satisfactory solution to the problem of providing a very early fire warning because of the accompanying rising incidence of false alarms. Such false alarms may be triggered either by instability in the units, or by smoke not associated with a fire hazard -- for example, cigarette smoke within the area being monitored, or high outside air pollution levels. The invention has for its object the provision of means which can provide an earlier fire warning than is customary with the described known systems.

In one form, the invention provides a smoke monitoring instrument having first and second chambers, each with a respective photodetector means, a light source common to both chambers and positioned to illuminate part of the interior of each chamber means for conveying a first air sample from the space to be monitored into said part of the first chamber, means for conveying a second air sample from outside the space into said part of the second chamber, each photodetector means being arranged to receive light from the source which is scattered by its respective air sample but not to receive light directly from the source. The light source is preferably a flash tube operated periodically to illuminate the air samples with an intense flash of light, and the photodetector means are preferably photomultiplier tubes.

The source of light may be continuous rather than intermittent but the latter is preferred because less energy is required to power the lamp and less heat is developed in the lamp. In addition, advantage may be taken of the period when the source is inactive to monitor noise levels in the instrument so that more accurate output is possible with appropriate compensation.

It is particularly advantageous to use a single flash tube rather than a separate tube for each chamber because the light output from lamps is inclined to vary from one pulse of light to another and thus at one flashing instant the sources produce different amounts of light which would lead to errors. The use of a single flash tube eliminates this possible source of error. There still remains however, the possible error due to fluctuations in the output of the single source but this can be removed by taking an output from the difference of the outputs of the photodetector means. Further, it has been found that a flash tube of relatively low capacity and operated near its full capacity will produce output pulses which are more uniform in intensity than those obtained from a higher capacity flash tube operated at low capacity.

We have found that various "normal" causes of smoke in the air of a space to be monitored for fires exhibit characteristics over a period of time different from those associated with real fires, and accordingly, in a particularly advantageous embodiment of the invention, the outputs from the detectors are recorded directly, or subtracted one from the other and the difference signal recorded. The record should be such that short-interval peaks of smoke, typical of cigarette smoke or the like can be distinguished from the steady build-up of smoke which accompanies the outbreak of a serious fire. With such a record, it is possible to set the sensitivity of the instrument to detect smoke levels well below those typical of conventional level-triggered devices. An alert may be set to trigger at light obscurations as low as 0.02 percent per meter. A supervisor may then investigate the incident by inspecting the record -- for instance, in the form of traces on a chart from a pen recorder -- and take appropriate action. Alternatively, automatic circuitry may be set into operation to interpret the record -- suitably in the form of stored charges on capacitors or the like.

A particularly advantageous feature of the instrument according to this aspect of the invention resides in its ability to allow for the effects of outside air pollution. Rising air pollution by smoke, which may occur during the morning in industrialized localities, or be caused by bush fires, can bring the scattering coefficient of outside air into the range where the instrument will signal an alert. "Outside air," for the purposes of this specification, refers to the air used to ventilate the space being monitored, and will normally be the open air surrounding the building. While smoke is building up in the outside air, it will -- in the absence of a fire inside -- have a larger light scattering coefficient than the inside air. When the outside smoke level drops again, however, smoke will have accumulated inside, and will register as having a scattering coefficient relative to the now cleaner outside air sufficiently high to trigger an alert. But reference to the record, which will show a considerable period during which the outside air had a higher scattering coefficient than the inside air, will explain the alert; and unless extra smoke is being generated inside the space to be monitored, pushing the difference signal above the level appropriate for the effects of external air pollution, no action need be taken on the alert.

The instrument is particularly effective if it is coupled into a forced air circulation or conditioning system. In this case, the first air sample is drawn from the exhaust air duct of the system, and the second air sample directly from outside, or from the inlet air duct of the system. The instrument may be made sufficiently sensitive that, despite the diluting effect of large volumes of air passing into the exhaust duct, small quantities of smoke (concentrated at their source, but invisible to the naked eye at the entry to the exhaust duct) may be reliably detected.

The invention also provides a method of detecting the generation of smoke associated with a fire hazard in a space to be monitored, comprising the steps of drawing a first sample of air from the space to be monitored, drawing a second sample of air from outside said space, exposing both samples to a source of light and generating two signals, each related to the amount of light scattered by a respective one of said samples, and storing a record of said signals over a period of time.

Preferred features of the method according to the invention will become apparent from the following description of particular embodiments of the invention, in which:

FIG. 1 is a diagrammatic representation of one form of the instrument according to the invention;

FIG. 2 shows a cross-section of the instrument at 2--2;

FIG. 3 is a diagrammatic representation of another form of the instrument according to the invention;

FIG. 4 is an airflow diagram showing how the instrument is connected to the space to be monitored;

FIG. 5 is a block diagram of a circuit for operating the instrument;

FIG. 6 is a more detailed diagram of part of the detecting circuit shown in FIG. 5;

FIG. 7 is a graphical representation showing various features of the output from the instrument.

The instrument shown in FIG. 1 comprises two light-tight chambers, 1 and 1', the internal surfaces of which are coated with optical black paint or the like to minimize reflections. Inside each chamber are located a series of knife edge light baffles, 2-9 and 2'-9', which restrict the field of view of photomultipliers 10, 10', to volumes 1, 1'. The light baffles are also coated with optical black paint to limit reflections, and have central apertures graded upwardly from baffles 3-8 (3'-8'). Baffles 2, 2' have apertures the same size as baffles 3, 3', and baffles 9, 9' have smaller apertures than baffles 8, 8', to form respective light-traps at the end of each chamber.

FIG. 2 shows baffles 6 and 6' in a cross-section of the instrument taken at 2--2 in FIG. 1. Central apertures 11, 11' define the fields of view of the photodetectors 10, 10', and allow air to pass from inlets 12, 12' to outlets 13, 13'. Further apertures 14, 14' in light baffles 6 and 6' are provided for extra air circulation to allow a more even distribution of air within the chambers. Similar air circulation apertures are provided in light baffles 5, 5', 7 and 7'. The light baffles may be replaced wholly or in part by one or more tubes positioned to define similar fields of view for the light detector. Xenon flash tube 15 is provided in a common wall 16 between the chambers in such a way that flashes of light, transmitted through light diffusing windows 17, 17', for example made of opal glass, can be scattered by smoke in the air samples in both chambers into respective photodetectors 10, 10'. Casing 18, also in common wall 16, houses a capacitor bank which in operation discharges through lamp 15.

The instrument shown in FIG. 3 is essentially the same in operation as that described previously but its physical arrangement and construction include a number of features to improve performance and reduce manufacturing and maintenance costs. Parts in this form which are common to that described previously have the same reference numerals.

The two chambers 11, 11' are located on opposite sides of a plate 50 which forms the upper part of a housing for the instrument. The upper part of the housing with the chambers attached can be readily removed to allow access to the flash tube 15 and photodetectors 10, 10' for replacement. The chambers 11 and 11' are formed by sheet metal structures about 55 cm in length, 5 cm across, and 8 cm in depth. Both the inside and outside of the structures are painted with optical black paint. Each chamber 11, 11' includes baffles 51, 51'; 52, 52'; 53, 53'; 54, 54'; 55, 55'; 56, 56'; 61, 61' extending transversely across the chamber 11, 11'. The baffles 51, 51' and 53, 53' are provided with central aperatures 57, 57' and 58, 58' which define the field of view of the photodetector 10, 10' as shown by lines 59, 59'. Light from the tube 15 enters the chambers 11, 11' through openings 60, 60' formed in the chambers, the openings 60, 60' being surrounded by short branch sections 61, 61' which project from the sheet metal structures defining the chambers 11, 11'. The branch sections are terminated by diffusing windows 19, 19' which each comprise a pair of spaced opal glass plates 62, 62' and 63, 63', normals to the planes of which make angles of about 40.degree. to the longitudinal axes of the chambers 11, 11'. The spaced plates are found to provide a better source of diffused light than a single plate or a single plate of equivalent thickness. The majority of light emanating from the plates 62, 62' is inclined relative to the longitudinal axis of the chambers 11, 11' and more scattered light for a given smoke concentration can be detected by the detector because more light is scattered in the direction of the incident light beam than in other directions.

The area illuminated by the lamp 15 through the plates 62, 62' and 63, 63' is determined by apertures 65, 65' and 66, 66' in baffles 56, 56' and 61, 61' respectively and extremities are shown by lines 64, 64', and the area between the intersection of lines 59, 59' and 64, 64' define the sample volumes 67, 67' from which the scattering is measured.

The position of the air inlets and outlets to the chambers is not critical but in this arrangement the inlets 12, 12' and outlets 13, 13' are at the top of the chambers 11, 11' and located either side of the sample volume 67, 67'.

The baffles 52, 52'; 54, 54'; 55, 55'; are provided to reduce internal reflections within the chambers 11, 11' without taking any part in the definition of the sample volumes 67, 67'.

Oblique baffles 68, 68' are provided and function as light traps at the opposite ends of the chamber 11, 11' to the photodetectors 10, 10'. The oblique baffles 68, 68' are disposed so as to be substantially perpendicular to the plates 62, 62' and 63, 63' to minimize scattering from the baffles 68, 68'. This disposition is chosen because scattering from a surface varies in accordance with the cosine of the angle of incidence to the surface.

Turning to FIG. 4, the instrument 19 is housed either inside or outside the space to be monitored, 20. Inlet 12 to chamber 1 is connected to the exhaust air duct 21 from space 20, and outlet 13 is exhausted (to atmosphere if convenient). A blower 22 may be connected to inlet 12 to force air from duct 2 into chamber 1. Inlet 12' to chamber 1 is connected to ventilation inlet duct 23, or directly to the outside atmosphere, again via a blower, 24, if desired. Outlet 13' is also exhausted to atmosphere if convenient. Blowers 22 and 24 may be combined into a single blower following outlets 13, 13' if negative pressure inside instrument 19 does not give rise to serious inaccuracies due to leakage of air into either chamber.

The block diagram of FIG. 5 shows how clock pulses from Clock Pulse Generator 25 govern the operation of Trigger and Gate Pulse Width circuit 26 which in turn causes the High Voltage Discharge Supply circuit, 27, to discharge through Xenon flash tube 28. The frequency of pulses (and thus flashes) is not critical; frequency of 1 pulse per second has been found to be satisfactory but frequencies of 10 per second may readily be used. The duration of the pulse from the circuit 26 is not critical but a pulse duration of 50 .mu.s has been found to be satisfactory. Light scattered into Photomultiplier 29 (corresponding to Photodetector 10 in FIG. 1) causes an output signal which is amplified in Amplifier 30 and synchronously detected and stored in Synchronous Integrator and Hold Circuit, 31. The latter circuit is governed by circuit 26 so that a charge proportional to the output from the photodetector during an interval corresponding to the flashing time of the Xenon lamp 28 is stored on a capacitor. This output will include a component of noise, and during the following period where the lamp is quiescent, noise from the photodetector is averaged over the whole period and stored in another capacitor in circuit 31. At the end of the period, the charges on both capacitors are combined in such a way that the average noise charge is subtracted from the charge representing the scattered light signal plus noise, leaving a closer approximation to the scattered light signal than is possible with an uncompensated detector. The resultant signal is fed into Buffer 32, and thence into a first channel of a Chart Recorder 33, or a Difference Amplifier 34 (or both). The other input to the difference amplifier, shown diagrammatically at 35, is from a second channel (not shown) derived from a photomultiplier corresponding to the photomultiplier 10' in FIG. 1 and brought through a similar circuit to that for photomultiplier 29. The output from the second channel may be recorded on a second channel of Chart Recorder 33, and the signal from the difference amplifier 34 may be recorded on a third channel of Chart Recorder 33.

A more detailed diagram of one form of the synchronous Intergrate and Hold Circuit 31 is shown in FIG. 6. In this circuit a pair of field effect transistors F1 and F2 of opposite channel doping are used to control storage of charge on storage capacitors C1 and C2 respectively. The capacitors C1 and C2 are connected to the output of amplifier 30 through resistor R, the time constants RC.sub.1 and RC.sub.2 being much greater than the pulse duration to give an integrating effect. The capacitor C.sub.2 stores charge throughout the whole cycle period and thus its charge will be representative of the scattering and noise levels. The capacitor C.sub.1, on the other hand, does not store charge during the flashing period but stores at other times and hence its charge is representative of noise. During the non-sampling period, i.e. the greater part of the cycle, the capacitors C.sub.1 and C.sub.2 are effectively in series from the point of view of the output and hence their contributions to the output level will be subtracted, and thus the noise is effectively cancelled out.

FIG. 7 shows a series of traces depicting various conditions which may appear in the output from chart recorder 33, showing voltage on the vertical scale and time on the horizontal. The traces are diagrammatic only, and in practice the events shown will normally take place over a considerable time, with long periods of uniform signal levels being recorded between events. Three traces, corresponding to three pens in a multi-pen chart recorder, are shown: trace A corresponds to the inside air condition monitored in chamber 1; trace B corresponds to the outside air condition monitored in chamber 1'; and trace C corresponds to the difference Amplifier 34. Obviously, other parameters which may have a bearing on the interpretation of these traces may also be recorded, for example, mains voltage.

If desired, the traces may be recorded in situ, but in many circumstances it may be desirable to record them at a central monitoring station serving a number of installations. The signal levels may be transmitted along telephone lines, for example, as tones, the frequencies of which are respectively related to the amplitudes of the signals to be recorded.

In FIG. 7, traces A, B and C are all shown as beginning at their respective zero levels, indicating clean air inside and outside the space to be monitored. Events D and E, relatively sharp peaks in the measured scattering level of the inside air, are typical of the temporary rises causes by exhalations of smoke by a person smoking a cigarette, for example, particularly if he is near the ventilation exhaust duct of the air circulation system. Peak F indicates a calibrating step, which may be the firing of three fusees together -- each fusee generating about 20 mg of smoke -- in the centre of the space being monitored. In one working monitoring arrangement, it has been found desirable to actuate a "yellow" alert at this calibrating level. This alert is intended merely to draw a supervisor's attention to the instrument, as it does not necessarily indicate a fire. The triggering level may be set at a scattering coefficient [b (scat)] level of about 2.2 .times. 10.sup.-.sup.4 m.sup.-.sup.1, where b (scat) is the fraction of incident light scattered by the sample (integrated for all angles between incident and scattered light directions). This corresponds to a meteorological visual range of about 18 kilometres.

When a dark target is viewed against a light background, the contrast between target and background decreases with an increasing scattering coefficient between observor and target. When the amounts of light from the target and background respectively differ by less than 2 percent they can no longer be properly distinguished. This happens at the visual range, which for clear air is approximately 170 km (105 miles) and corresponds to a b (scat) of about 0.23 .times. 10.sup.-.sup.4 m.sup.-.sup.1.

More urgent alert signals may also be designed to operate at higher scattering levels than the "yellow." For example, an "orange" alert, initiating an investigation at the building being monitored, could be set at a b (scat) of 4.4 .times. 10.sup.-.sup.4 m.sup.-.sup.1 (visual range, about 9 km); and a "red" alert, calling the Fire Brigade and operating fire control systems, could be set at a b (scat) of about 7 .times. 10.sup.-.sup.4 m.sup.-.sup.1 (visual range, about 5.5 km). The "red" alert level, in particular, would have to be adjusted to suit the particular circumstances of the building being monitored.

Event G denotes the onset of a real fire, and it can be seen that the "yellow" level (set, for example, at about 1 volt on the recorder) is passed shortly after the first traces of smoke are generated. A supervisor looking at the trace could immediately see its steadily rising form, and take prompt action.

Line H separates the traces for a later incident, shown for convenience on a much contracted time scale. At point I, outside smoke levels begin to rise, causing a negative swing in the difference signal. At a later point, J, the inside air begins to show significant signs of increasing scattering caused by smoky air from outside accumulating inside. The difference signal returns towards zero, but at point K, where the outside air has cleared, the difference signal goes positive until the smoky air inside is cleared. Of course, if a real fire breaks out inside during this interval of positive excursion, an additional signal, L, will appear and trigger the appropriate alerts.

It will be apparent that many modifications can be made to the embodiment of the invention described herein, and it is to be understood that the invention is not limited to the details of the construction illustrated, but includes all variations falling with its spirit and scope.

Claims

We claim:

1. A smoke monitoring instrument having first and second chambers, each with a respective photodetector means, a light source common to both chambers and positioned to simultaneously illuminate part of the interior of each chamber, means for conveying a first air sample from the space to be monitored into said part of the first chamber, means for conveying a second air sample from outside the space into said part of the second chamber, each photodetector means being arranged to simultaneously receive light from the source which is scattered by its respective air sample but not to receive light directly from the source.

2. An instrument as claimed in claim 1 wherein the light source comprises a flash tube and the instrument includes an energizing circuit to operate the flash tube periodically.

3. An instrument as claimed in claim 2 wherein each photodetector means comprises a light sensitive element and a detector circuit responsive thereto and operable to produce an output signal representative of the amount of light received by the light sensitive element.

4. An instrument as claimed in claim 3 including means to produce a differential output signal, being a signal derived from the difference of the output signals of the detector circuits.

5. An instrument as claimed in claim 3 wherein each detector circuit includes first and second storages, the first storage being operable to store a signal representative of the light received by its associated light sensitive element during the time the flash tube is operated and during the time the flash tube is not operated, the second storage being operable to store a signal representative of the light received by its associated light sensitive element during the time the flash tube is not operated, said output signal of the detector circuit being derived from the difference of the signals stored in the first and second storages, whereby noise signals are reduced.

6. An instrument as claimed in claim 1 wherein the means for conveying a first air sample from the space to be monitored is coupled to an exhaust duct of an airconditioning system for the space to be monitored.

7. An instrument as claimed in claim 6 wherein the means for conveying a second air sample from outside the space is coupled to an inlet duct of the airconditioning system.

8. An instrument as claimed in claim 3 including means to record the output signals of the detector circuits.

9. An instrument as claimed in claim 8 wherein the means to record is operable to record the difference between the output signals.

10. An instrument as claimed in claim 3, including alarm circuitry operable to produce alarm signals which are dependent upon the output signal levels of the detector circuit associated with the first chamber.

11. An instrument as claimed in claim 8 wherein the means to record is located remote from the detector circuits and is coupled thereto by a telecommunications link.

12. An instrument as claimed in claim 3 wherein the first and second chambers comprise a pair of spaced elongate chambers, the interiors of which are optically black, and said source of light is positioned between chambers, each chamber including a window adjacent to the source.

13. An instrument as claimed in claim 12 wherein the instrument includes opal glass diffusing surfaces interposed between the windows and the source.

14. An instrument as claimed in claim 13 wherein the opal glass diffusing surfaces are planar and their normals are inclined toward respective photodetector means.

15. An instrument as claimed in claim 12 wherein the chambers and the diffusing surfaces are mounted upon a removable housing for the instrument and the source and photodetectors are mounted on a base of the instrument.

16. A method of detecting the generation of smoke associated with a fire hazard in a space to be monitored, comprising the steps of drawing a first sample of air from the space to be monitored, drawing a second sample of air from outside said space, exposing both samples simultaneously to a source of light and generating two signals simultaneously, each related to the amount of light scattered by a respective one of said samples.

17. A method as claimed in claim 16 wherein the first sample is drawn from an exhaust duct of an airconditioning system for the space to be monitored.

18. A method as claimed in claim 17 wherein the second sample is drawn from an inlet duct of the airconditioning system.

19. A method as claimed in claim 16 including the step of storing a record of the signals over a period of time.

20. A method as claimed in claim 19 wherein the storing of a record is performed at a location remote from the space to be monitored.