HK1099395B - Improvement(s) related to particle monitors and method(s) therefor - Google Patents

Description

Particle monitor and method improvements

Technical Field

The present invention relates to the field of detection, analysis and/or determination of substances or particles suspended in a fluid.

In one particular form, the invention relates to a smoke detector for detecting unwanted pyrolysis or combustion in a substance. In another form, the invention relates to a smoke detector of the early detection type, which may be used for ventilation, air conditioning, or duct monitoring of a particular area. In yet another form, the invention relates to surveillance monitoring, such as building, fire, or security monitoring. In yet another form, the present invention relates to environmental monitoring, such as monitoring, detection and/or analysis of fluids, zones, areas and/or surrounding environments, including commercial and industrial environments.

It will be evident that the invention has a wide range of applications and thus the particular forms described above are given by way of example only, and the scope of the invention is not limited to these forms.

Background

The present inventors have determined a recognition that: the types of fumes produced in various pyrolysis and combustion environments are different. A fast burning flame tends to produce a large number of very small solid particles that can agglomerate into irregular shapes to form soot. Conversely, the early stages of pyrolysis tend to produce relatively small amounts of larger liquid particles (having high boiling points) that are typically present as suspended dust particles that can agglomerate to form larger, translucent spheres.

The inventors have also determined a recognition that: detection of relatively large particles slowly increasing in number throughout the duration is generally indicative of a pyrolytic or incomplete combustion (smoldering) condition, whereas detection of a rapid occurrence of a large number of small particles and no early pyrolytic or incomplete combustion is indicative of the use of a pilot fire involving the use of an oxidizer.

The inventors have also determined a recognition that: dust particles are produced by the abrasive or non-thermal decomposition of natural substances or organisms in the environment and are generally very large compared to smoke particles.

The inventors have also determined the following recognition:

conventional point-type smoke detectors were originally designed for ceiling mounting in protected areas. These detectors have a low sensitivity and are difficult to detect the presence of undesirable pyrolysis where large amounts of gas pass through the monitored area, thus impairing the ability of the detector to sense the presence of undesirable pyrolysis.

To overcome these drawbacks, highly sensitive suction smoke detectors have been developed and are often deployed on a duct for monitoring an area. These detectors provide detection that is hundreds of times more sensitive than conventional point-type detectors. These suction systems use negative pressure by means of an air pump and also a dust filter to reduce unwanted dust contamination which contaminates the detector or is indistinguishable from the detection of smoke causing a false alarm to be triggered.

The smoke detector preferably used in the suction system is a turbidimeter. This is a detector sensitive to particles of various sizes, such as various smoke particles generated in a fire, or in the early stages of overheating, pyrolysis, or incomplete combustion.

Prior art optical smoke (or airborne particle) detectors typically use a single light source to illuminate a detection zone that may contain such particles. The use of two light sources has been proposed for some detectors. The particles scatter a portion of this light to one or more receiver elements (or sensors). The signal output from the receiver element is used to trigger an alarm signal.

Other detectors utilize a laser beam, typically in the near infrared wavelength range, that provides a polarized monochromatic light source. However, these detectors are not considered true turbidimeters because they tend to be overly sensitive to a particular particle size (particle size) range at the expense of other size ranges.

A drawback of the above detector is that it is relatively insensitive to the very small particle characteristics of early pyrolysis and incipient fires, as well as certain fast-burning fires.

On the other hand, ionized smoke detectors utilize radioactive elements such as americium to ionize air within the detection chamber. These detectors are more sensitive to the very small particles produced by a flaming fire, but are less sensitive to the larger particles produced by pyrolysis or incomplete combustion. They have also been found to be more prone to replace the ventilation of ionized air in the detection chamber and thereby trigger false alarms. This creates a practical limit to the sensitivity with which it is useful.

Other smoke detectors use xenon lamps as the single light source. Xenon lamps produce a continuous spectrum of light similar to sunlight, including ultraviolet, visible, and infrared wavelength ranges. Particles of all sizes can be detected with this light source and the detector produces a signal proportional to the mass density of the smoke, which is characteristic of a true nephelometer. However, this detector cannot characterize the type of fire because it cannot distinguish (differentiate) a particular granularity. Furthermore, the xenon light (source) has a short lifetime of only about 4 years and its light intensity is known to vary, which affects sensitivity.

The present inventors have also realised that in order to provide a wide output range in terms of sensitivity, prior art detectors provide an analogue to digital converter (ADC) for applying smoke concentration data to a microprocessor. By careful design, substantially all of the ADC's capacity is used to represent the maximum smoke concentration, for example (typically) 20%/m. ADCs that operate at 8 bit resolution are efficient, while ADCs that are 10 bits or larger are expensive and require larger microprocessors. It has been found that a 10-bit ADC allows a 20%/m concentration to be divided into 1024 stages, each stage representing an increase (gain) of 20/1024 ═ 0.02%/m. So each stage is 0, 0.02, 0.04, 0.06, etc. without the possibility for finer increments. At low smoke concentrations, it is considered to be a very coarse resolution, making it difficult to set the alarm threshold finely. However, at high smoke concentrations, a resolution of 0.02%/m is not necessary, for example, even the ability to set the alarm threshold at 10.00%/m or 10.02%/m is of little benefit. The resolution of the prior art detectors is considered too coarse at low smoke concentrations and too fine at high smoke concentrations.

Any discussion of documents, devices, acts or knowledge in this specification is included to explain the context of the invention. It should not be taken as an admission that: any material forms part of the prior art base or the common general knowledge in the relevant art in australia or elsewhere on or before the priority date of the disclosure and claims herein.

It is an object of the present invention to provide a particle detection apparatus and method that can improve the detection, discrimination and/or analysis of particles, pyrolysis, incomplete combustion (smoldering), and/or fire events and dust, thereby providing a corresponding improvement in the detection of fluid-borne particles.

It is a further object of the present invention to provide a particle detection apparatus suitable for use in conjunction with a pipeline or as a stand-alone detector and/or monitor.

It is a further object of the present invention to mitigate at least one disadvantage associated with the prior art.

Disclosure of Invention

According to various aspects of the present invention, the monitoring, determining, detecting, and/or analyzing of particles, environments, fluids, fumes, zones, or areas may include the determination of the presence of particles and/or particle characteristics, depending on the requirements of a given particular application of the present invention.

In this regard, the present invention provides in one aspect a method and apparatus for determining the presence of particles of substantially a predetermined size or range of sizes in a fluid sample, the method comprising the steps of: illuminating the sample with light of a first wavelength to obtain a first response signal indicative of the first illumination, illuminating the sample with light of a second wavelength to obtain a second response signal indicative of the second illumination, and determining the presence of particles having the size or the size range by comparing the first and second signals.

Preferably, the illumination is horizontally and/or vertically polarized (polarized).

In another aspect of the invention there is provided a gain control apparatus adapted to provide gain control in a particle monitor, the apparatus comprising a first gain stage having a first amplifier, a second gain stage having a second amplifier, and voltage or current controlled feedback (means) from the output of the second gain stage to the input of the first gain stage such that the frequency response of the amplifier is unaffected by the feedback (signal).

In yet another aspect of the invention, there is provided a method of determining a service interval for a particle monitor, the method comprising the steps of: determining the presence of dust particles, providing detection of the presence of the particles, and providing a maintenance indication when the detection reaches a predetermined threshold.

In yet another aspect of the invention, a particle monitoring chamber is provided that includes a first lens operably associated with an illumination source, a second lens for focusing incident light onto a receiver element, and a main stop for primarily preventing light from being scattered (diverged) directly from the first lens to be incident on the second lens.

In another aspect of the invention, a method and apparatus for determining the rate of fluid flow through a given area is provided, the method comprising the steps of: providing a first sensor at a location of lower flow velocity in the fluid flow path, providing a second sensor at a location of relatively higher flow velocity in the fluid flow path, the second sensor having temperature characteristics substantially similar to the first sensor, and determining the flow velocity based on detection of a cooling effect of the fluid flowing past the first and second sensors.

Further, according to another aspect of the present invention, there is provided a method and apparatus for installing a housing on a pipeline, the method comprising the steps of: providing at least one fitting associated with the housing, positioning the housing proximate to a mounting area of the pipe, trimming (profiling) the fitting to substantially conform to a contour of the pipe proximate to the mounting area, and attaching the housing using the fitting.

The invention also provides a monitor for monitoring the presence, concentration and characteristics of particles in a fluid medium.

The invention also provides a logarithmic signal as an output to trigger a threshold or alarm of the detector. Which refers to a signal whose amplitude can be compressed according to a logarithmic function or scale. The logarithmic signal may be indicative of various characteristics of the detected particles, such as presence, quantity, frequency, concentration, and/or duration.

In essence, in one aspect of the invention, different wavelengths, various wavelength ranges and/or polarizations are used to detect predetermined particles in a fluid.

In essence, in another aspect of the invention, the subtraction of the two signals or providing a ratio of the two signals makes the output indicative of particle and size detection easier to measure.

In essence, in another aspect of the invention, this output indicative of particle detection is amplified from both signals.

Other aspects and preferred aspects are disclosed in the specification and/or defined in the appended claims, which form a part of the description of the invention.

It has been found that the present invention can yield advantages such as reduced size, cost and energy consumption while achieving the highest industry standards for minimizing sensitivity, reliability, maintenance cycles and false alarms, and/or for monitoring the presence of smoke and/or dust particles in the environment so that a very high sensitivity to smoke can be provided without false alarms occurring due to dust.

Throughout this specification, reference is made to a plurality of different light sources having specific wavelengths. The light sources and wavelengths are mentioned only because they are now commercially available. It should be understood that the principles underlying the present invention have equal applicability to light sources of different wavelengths.

The monitor may comprise the mentioned detector or similar device.

Further scope of applicability of the present invention will become apparent from the detailed description given hereinafter. It should be understood, however, that the detailed description and the specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

Drawings

Further disclosure, objects, advantages and aspects of the present application will become better understood by those skilled in the relevant art by reference to the following description of preferred embodiments when read in conjunction with the accompanying drawings, which are given by way of illustration only, and not by way of limitation, in which:

FIG. 1 shows the results for a wavelength of 430nm for blue light and 660nm for red light for particles across the entire size range;

FIG. 2 shows the results for blue light 430nm wavelength and green light 530nm wavelength for particles across the entire size range;

FIG. 3 shows the results for a wavelength of 470nm for blue light and a wavelength of 940nm for infrared light for particles in the entire particle size range;

FIG. 4 shows the result of a relative subtraction of a red signal from a blue signal;

FIG. 5 shows the result of a relative subtraction of a green signal from a blue signal;

FIG. 6 shows the result of a relative subtraction of an infrared light signal from a blue light signal;

FIG. 7 illustrates the variation in particle size over time for various types of combustion agents;

FIG. 8 shows comparative responses of infrared and blue light channels to smoke from various stages of fire and/or fire growth;

FIG. 9 shows the relative ratio of channel B output and channel A output in response to airborne particles by a given combustion agent during tracking;

figure 10 shows a schematic block diagram of a smoke monitor according to an embodiment of the present invention;

FIG. 11 shows a circuit diagram of a gain controlled amplifier in accordance with one form of an embodiment of the present invention;

figures 12, 13 show the geometry of a preferred control chamber including an indicator light path;

FIG. 14 illustrates the use of a lenticular lens according to one aspect of the present invention;

FIG. 15 illustrates the relevant operation of an aspheric lens according to one aspect of the invention;

FIG. 16 illustrates the use of an aspherical mirror according to one aspect of the invention;

FIG. 17 illustrates the operation of a lenticular lens in accordance with an aspect of the present invention; and

figure 18 shows an example of mounting the detector unit to the pipe arrangement.

Detailed Description

In the described embodiment, at least two channels are mentioned, one being channel a, which uses wavelengths such as red or infrared wavelengths, and the other being channel B, which uses wavelengths such as blue wavelengths. Other channels may be used, such as channel C, which uses a wavelength such as the green wavelength. Other wavelengths may also be used in accordance with the present invention, as will become apparent from the following description. In general, it is preferred if the readings established by the longer wavelengths rival those established by the shorter wavelengths. More preferably, the longer wavelengths are subtracted from the shorter wavelengths. Ratios can also be used to compare wavelength readings.

Wavelength of light

In one aspect of the invention, the inventors have determined that the wavelength of light employed significantly affects the sensitivity of the device to particle size. Bohren CF and Huffman DR describe the Scattering (divergence) of Light from Particles in various size ranges in "Absorption and Scattering of Light by Small Particles", ISBN 0471-.

It has been determined that the michaelis equation (Mie equation) is suitable for examining particles in a size range consistent with conventional smoke and dust. Fast-burning fires tend to produce very large quantities of very small carbonaceous particles that can agglomerate into irregular shapes to form soot. Conversely, the early stages of pyrolysis tend to produce smaller quantities of larger liquid particles (having high boiling points), often present as aerosols, and may agglomerate to form larger translucent spheres or droplets (droplets). Dust particles are generally produced by mechanical abrasion and have an irregular shape that can be approximated by a larger sphere for modeling purposes. The source of smoke or dust is unlikely to be monodisperse (comprising a particle size) but more likely to be polydisperse, having a size range that can follow a gaussian distribution. The inventors have found that the typical standard deviation for the size distribution is around 1.8 to 2.

The airborne particle distribution in cities was also found to be bimodal, peaking at about 0.1 and 10 microns. Typically, smoke particles are in the range of 0.01 microns to 1 micron, while airborne dust particles are in the range of 1 micron to 100 microns. However, there is a partial overlap at the 1 micron boundary because the smallest dust is smaller than the largest smoke particle possible in nature.

The inventors have also determined that certain particle sizes are more readily discernable by light of a particular (different) wavelength. Assuming this, we use two wavelengths of incident light. The light may be in any range from blue to red (as well as infrared). One example is light in the range from 400nm (blue) to 1050nm (red). For example, 430nm (blue) and 660nm (red) may be used.

Fig. 1 shows the results for two wavelengths of incident light (430nm (blue) and 660nm (red)) each of which is unpolarized, vertically polarized or horizontally polarized and emits at the same angle relative to the optical axis, by applying Mie theory (Mie theory) to particle sizes having an overall average diameter in the range of 0.01 to 10 microns and using a standard deviation of 1.8.

In fig. 1, the resulting blue light system (B: unpolarized blue light, BV: vertically polarized blue light, BH: horizontally polarized blue light) is well suited for smoke and dust detection, while the resulting red light system (R, RV and RH) is also suitable for dust detection, but is weaker in detecting a wide range of smoke particles due to lack of response to small particles. All curves in fig. 1 cluster together above about 0.8 microns, while there is a significant difference between the curves for particle sizes less than 0.8 microns. An optimal separation of vertical blue light (BV) from horizontal red light (RH) is achieved. The curves cannot be effectively separated at larger diameters. The phase cancellation and intensification due to the interaction between a given wavelength and a given grain size causes periodicity (edge effects or resonances) in the curve.

If instead a combination of wavelengths 430nm (blue) and 530nm (green) is investigated, the results shown in FIG. 2 are obtained. Here, the graphs are more similar to each other, and it is difficult to separate the graphs above about 0.5 microns.

The wavelengths selected for exemplary illustration are limited to those of commercially available emitters. Based on the information obtained in FIG. 2(530nm), the results for orange light (620nm) are similar to FIG. 1(660 nm).

The results for blue light (470nm) and infrared light (940nm) are shown in FIG. 3. In fig. 3, the wavelength separation is substantially one octave (octave). It can be seen that in the region below 1 micron (the standard boundary between smoke and dust) the curves are more clearly separated.

While there are some advantages to operating the monitor at even more widely separated wavelengths, currently available technology is a limiting factor. The receiver element for detecting scattered light is a PIN photodiode with an improved blue response. For practical purposes, the emitter wavelength is currently limited to this range due to the peak response at 850nm, which decreases by about 30% at 400nm and 1050 nm. Of course, if another receiver element is used, the wavelength of the light incident on the particles can be varied to obtain greater separation.

From the above conclusions, it can be seen that in one embodiment of the invention the wavelengths for the two emitters to illuminate the particles to be detected should preferably be in the range of 400nm to 500nm for blue/ultraviolet light and in the range of 650nm to 1050nm for red/infrared light.

In another aspect of the invention, it has been found that if the results of the received signals are compared to each other, for example by comparing the ratios or by subtracting one signal from the other (i.e. subtracting the other signal from one signal), a more reliable "trigger" or detection signal can be generated indicating the presence of particles of the particle size of interest in the application for which the monitor of the invention is suitable. Thus, for example, if the monitor of the present invention is set up as a "smoke" monitor, then smaller particles should be of greater interest than larger (dirt) particles. Thus, the inventors have recognized that for smoke monitors, for example, blue light has been found to respond to smaller and larger particle sizes, while infrared light responds only to larger particles. By acquiring a signal based on a "blue" response signal less than an "infrared" response signal, the monitor can be set to have a relatively high response rate for small particles and a low or zero response rate for larger particles.

For example, fig. 4 shows the result of subtracting red-horizontal (polarized) (RH), red-unpolarized (R), or red-vertical (polarized) (RV) data from blue (B) data. The monitor set up in these ways responds to particles smaller than 1 micron with higher sensitivity (with the best sensitivity from the B-RH combination). To avoid confusion, the results of BH and BV are not shown, but are identical.

For comparison with fig. 4, subtracting GH, G and GV from B yields the results of fig. 5. Although the edge effect is significant, relatively small particle sizes are still more readily discernable than larger (dirt-like) particle sizes.

Fig. 6 shows the result after subtracting IRH from B. Other results have been omitted for clarity. In addition, some published data on the average particle size obtained for fragrance, cotton wick, toast bread and portland cement (a dust substitute) are shown. It can be seen that the monitor used to implement this subtraction has a suitable sensitivity to conventional smoke types and is able to reject dust to a considerable extent (relatively).

In accordance with this subtractive aspect, a further aspect of the present invention has been developed in that a suitably configured gain amplifier can be utilized to provide an appropriate output signal for use by an alarm or other alert device or system. This aspect will be disclosed more fully below.

In addition to the two wavelengths disclosed above, if a third or other wavelength can be used, it is possible to identify not only small and large particles, but also other (intermediate) sized particles, depending on the wavelength used.

Dual channel design

According to one aspect of the invention, by using a dual channel design to provide another feature, namely by subtracting the a (reference) channel from the B (sample) channel (or vice versa) as described herein, we can achieve zero point balance. It has been found that the equilibrium does not change significantly if the background of the monitoring chamber changes over time. The present inventors have recognized that the background light level changes as the monitoring chamber ages or becomes contaminated over a long period of time (i.e., a significantly extended period of time through the use of dust filters). The benefit of channel subtraction is that since the response of the two channels (especially to dust accumulation) is substantially the same, their effects cancel out themselves, which minimizes any change in the output obtained by the summing circuit over time. It is worth noting that the signal obtained from dust is not dependent on whether it is airborne-or whether it can stay on a surface. The same is true for any material larger than dust-like condensation or even (wall) walls.

This zero-going offset due to contamination is considered a valuable feature by maintenance standards.

Signal level analysis

The invention is further disclosed with reference to the use of a smoke monitor. It is to be noted, however, that the invention is not limited to this application.

Conventional ceiling mounted "optical" smoke detectors are typically provided equivalent toA sensitivity of shading (obsuration) of about 10%/m (3%/ft) for generating an alarm. The criteria established for extremely high sensitivity smoke detection require at least two orders of magnitude higher sensitivity, corresponding to alarm set points with less than 0.1%/m shading over the entire range. Eccleston, King and Packham (Eccleston AJ, King NK and Packham DR, 1974: The Scattering Coefficient and Mass concentration of paint from one aspect Forest Forest fire, APCAjournal, v24 noll) have demonstrated that for eucalyptus Forest fire Smoke, this 0.1%/m level corresponds to a visible range of 4km and 0.24mg/m³The concentration of smoke of. Such high sensitivity enables detection of early pyrolysis and thus provides the earliest warning of a potential fire in a building with a low false alarm rate.

Most of today's ultra-high sensitivity smoke detectors utilize an optical (monitoring) cell with an infrared solid state laser diode. The long wavelength of infrared light is advantageous for detecting larger airborne particle characteristics of dust and aerosol particles from certain types of fires, but is less preferred for detecting very small particles contained in other fires. Conventional solid-state lasers that preferably operate at shorter visible wavelengths are expensive or do not operate reliably at elevated ambient temperatures (60 ℃). To overcome these difficulties, in a preferred embodiment of the invention applied to a smoke monitor, it was decided to use a Light Emitting Diode (LED) emitter operating at the blue end of the visible spectrum (470 nm).

As will be further explained below, the monitor arrangement incorporates such a blue light emitter arranged at an angle of 60 ° to the axis of the receiver element within the optical monitoring chamber. The monitor also includes a 940nm (infrared) reference emitter set at the same angle, but horizontally opposite the blue emitter. At an effective transmitter illumination cone angle of 10 deg., this configuration provides a relatively optimal setting that maximizes the sensitivity of the system while minimizing background light that may interfere with the receiver elements.

For a given Smoke density (0.1%/m), including, for example, Particles having a mass mean particle diameter of 0.3 μm (with a true geometric standard deviation of 1.8), Weinert (Weinertd, 2002: association of Light Scattering from Smoke Particles for an a project type Dual-mounted Smoke Detector, unpublished) has determined that in the monitor setting used, the signal intensity received by such Smoke by irradiation with an unpolarized blue Light source is of the order of about 4.5E-8 per unit of irradiation. Weinert data at 470nm and 940nm are plotted and shown in figure 3. Critically, this means that the "background" light intensity received by the element must be at least 8 orders of magnitude lower than the emitter beam intensity, due to unwanted residual reflections from the monitoring chamber walls, so as not to interfere with the desired light signal (scattered from smoke).

In one form, the blue emitter is specified to have a luminous intensity of 40 candelas (cd) at a drive current of 500 mA. By definition, a power level of 1cd is 1.464mW per steradian (sr), so the nominal power is 1.464 x 40 — 58.6 mW/sr. The 5 ° half angle is converted to 2 pi (1-cos (5)) -0.024 sr, so the output power is 58.6 x 0.024-1.4 mW. Incidentally, at this drive current, the emitter voltage drop is 4.0V, so if a 0.1% duty cycle is used, the input power to the emitter is 0.5 x 4.0 x 0.001 — 2.0mW, which is 1% less than its maximum power dissipation rating.

Thus, at a pulse emitter power output of 1.4mW, the scattered light signal directed to the element for the setup used is 1.4 x 4.5E-8 — 6.3E-5 μ W. This level of illumination is directed and focused to fall onto the receiver element, which is a PIN photodiode in the receiver module. The sensitivity of the element was specified to be 0.2A/W at 400nm and 0.31. mu.A/μ W at 470 nm. Thus, at a given 92% (uncoated) lens transmittance, the signal transduced by the illuminated element was 0.31 x 6.3E-5 x 0.92 ═ 1.8E-5 μ a.

In one form, the receiver module includes a three-stage AC-coupled pulse preamplifier including a current-to-voltage converter followed by two voltage amplifiers. The converter is an operational amplifier with a PIN photodiode differentially connected between inverting and non-inverting inputs, ignoring series resistance. The feedback resistor may be 3.9M (shunted with 3.9 pF), so at mid-band frequencies, the output from this stage will be 3.9E6 x 1E-6 — 3.9V/μ a for a1 μ a input signal. In response to the specified element illumination, the output became 3.9 × 1.8E-5 ═ 7.0E-5V or 70 μ V.

In one form, the next two stages are operational amplifiers each with a mid-band gain of 10, so the receiver module output at a given illumination should be 7.0 mV. The standard full scale output level for signal processing may be 3V, so the main amplifier voltage gain would be 3/7.0E-3-429. With two similar stages, such an amplifier would require a gain of 21 per stage. In practice it has been found that a gain of 17 per stage is sufficient to produce a nominal 0.1%/m sensitivity required to meet the full range.

Clearly, the sensitivity of all smoke detectors depends on the particle size, and a meaningful standard would require specifying this size (or range of sizes). However, a well established international standard for performance is the VESDA MK3 monitor recently manufactured by Vision Systems Australia using a xenon light source. In fact, this light source is comparable to a blue emitter because of the spectral characteristics of a xenon lamp, combining the spectral response of a PIN photodiode with light from small suspended particles or molecules (which is advantageous, for example, as 1/λ @)⁴Short wavelength of xenon-based monitor), the characteristic wavelength that determines the reference for a xenon-based monitor is 470nm, as is the case with a blue emitter. For this reason, reliable gases such as nitrogen and FM200 can be used continuously as standards (which is not possible with infrared laser based detectors).

As described above, the monitor employs two transmitters operating at different wavelengths. Referring to fig. 3, for larger particles (> 1 μ), the design goal is to produce the same signal level at the element from the infrared signal, as is the case for the blue signal. At an infrared light wavelength of 940nm, the receiver element has a sensitivity of 0.55 μ A/μ W (comparable to 0.31 μ A/μ W at 470 nm). Since the lens transmittance remains 92% at 940nm, and since all correlation equations are linear and the geometry is relatively uniform, the infrared emitter output power can be reduced by a factor of 0.31/0.55 to 0.56. Since the infrared light emitter has a power level of 343mW/sr (comparable to 58.6mW/sr of the blue light emitter) at a current of 500mA, the required driving current of the infrared light emitter is 500 x 0.56 x 58.6/343 ═ 48 mA. If a polarizing filter is used, the drive current needs to be increased in order to overcome the losses in such a filter.

At the required transmitter drive setting, as seen from the receiver elements, the small background signal caused by the accumulated reflections from the monitored chamber walls should be at approximately the same (very low) level for any one transmitter. This requires monitoring the reflection (or absorption) of the chamber walls largely independently of the difference in wavelength used. Therefore, the differential voltage between the two channel outputs should be close to zero (or can be adjusted to be so) when there is no smoke in the monitoring chamber.

By introducing smoke into the monitoring chamber, the voltage on each channel should be increased, but the differential voltage between the channels may often be non-zero. This differential voltage provides an indication of the characteristics of the airborne particles. Fig. 6 shows the sensitivity obtained when subtracting the infrared light channel from the blue light channel. This result can be used to highlight the presence of particles having a mass average particle size of less than 1 μ. Lines are included in figure 6 consistent with published data on mass average particle size of particles produced from some prior art materials (portland cement "dust", toast bread loaf, cotton wick and fragrance). The differential voltage should be zero or slightly negative in the first example (large particles), but significantly positive in the other three examples (small particles). This demonstrates the possibility of distinguishing dust while maintaining good smoke detection.

The particle size in the aerosol-suspended particles can vary substantially depending on the combustion agent used, the temperature and time period (cycle), and the air flow conditions that determine the oxygen supply, cooling and aerosol dilution. In FIG. 7, data from clean, Weinert and Mulholland (clean TG, Weinert, DW and Mulholland GW, 2001: Moment Method of inhibiting particle sizes Measures of Test brushes, NIST) were averaged to produce a plot of suspended particle Size generated from four combustion agents, ready-to-eat oil (glass on stove), toasted bread (oven), polyurethane foam (incomplete or smoldering), and beech wood blocks (stove). It can be seen that in each case the average particle is initially small, increasing in size with complete consumption of the combustion agent and then falling. To summarize, it can be said that the detection of small particles is important for the earliest possible alarm of an early fire. Other data indicate that the suspended particle mass concentration reaches a maximum in the second half of each cycle plotted and decreases at the end.

Fig. 8 provides a broader comparison of the relative responses of the two channels, expected to be arranged in published order of particle size for a large number of substances. Here, the response of portland cement (dust substitute) has been normalized by reducing the infrared light emitter signal by a factor of 0.64. The data for Douglas fir and hard polyurethanes (Bank ston et al; Bank ston CP, Zinn BT, Brown RF and Powell EA, 1981: accessories of the mechanics of the stamp Generation by Burning Materials, Combustion and Flame no 41 pp273-292) show the progress of three different phases of the radiant heat release rate, which should produce corresponding differential voltage signals.

Roughly and for the reasons stated earlier, figure 8 can be taken as a comparison of the expected performance between standard xenon-based and current laser-based (infrared) detectors.

Furthermore, with respect to the dual channel monitor, fig. 8 demonstrates that the sensitivity to early fire events including pyrolysis and incomplete combustion is improved (by a factor of 4 or 5) while the likelihood of false alarms from dust is greatly reduced compared to these infrared detectors. This means that no dust filter is required. In contrast, dust filtration is desirable to minimize contamination and thereby maximize the maintenance cycle and overall operating life of the monitor. Given that a good filter of dust can also capture smoke, dust discrimination capability can be exploited to avoid unwanted false alarms caused by small amounts of dust that inevitably pass through the actual filter.

Moreover, because channel a responds primarily to dirt, the output from channel a over time (measured in months or years) can be integrated (accumulated) to record the actual exposure of the monitoring chamber and filter element to dirt when distinguished from smoke, thereby enabling the maintenance interval to be determined and forewarned based on the (often unpredictable) surrounding environment. For example, a filter service interval may be determined based on an accumulated or calculated number of detected dust readings. Once the count reaches or exceeds a predetermined threshold, the service indicator lights up or otherwise communicates. Preferably, the maintenance indicator circuit should integrate the actual dust level and its duration.

Logarithmic output

As mentioned above, in order to provide a wide output range of sensitivity, prior art detectors provide an analog to digital converter (ADC) for providing smoke concentration data to a microprocessor. By careful design, the maximum smoke concentration is expressed using substantially all of the ADC's capacity, for example (typically) 20%/m. ADCs that operate at 8-bit resolution are useful, while ADCs of 10 bits or more are expensive and require larger microprocessors. It has been found that a 10 bit ADC allows to divide this 20%/m concentration into 2¹⁰1024 stages, each representing 20/1024 in increments of 0.02%/m. So each stage is 0, 0.02, 0.04, 0.06, etc. without the possibility for finer increments. At low smoke concentrations, it is considered to be a very coarse resolution, making it difficult to set the alarm threshold finely. However, at high smoke concentrations, a resolution of 0.02%/m is not necessary — for example, even the ability to set the alarm threshold at 10.00%/m or 10.02%/m is of little benefit. The resolution of the prior art detectors is considered too coarse at low smoke concentrations and too coarse at high smoke concentrationsAnd too fine.

However, according to this aspect of the invention, the above-mentioned drawbacks of the prior art are overcome by providing an output range of logarithmic or decile numbers. According to the invention, it has been found that the resolution is suitable for a given smoke concentration, i.e. fine at low smoke concentrations and coarse at high smoke concentrations. As illustrated, with the present invention, by using a logarithmic output range, the alarm threshold can be set at 0.010 or 0.011%/m at low smoke concentrations, and at 10%/m or 11%/m at high smoke concentrations with the same ease.

In other words, the use of logarithmic output provides advantageous sensitivity resolution for an overall wide range of smoke concentration and/or threshold settings, since it is recognized that smoke is a very variable substance, with little benefit to the accuracy with which its density (concentration) is measured to greater than 2 significant figures.

Results of the Smoke test

A series of traces were made using a smoke monitoring device constructed in accordance with the present invention and constructed and assembled in accordance with the above disclosure of signal level analysis. The monitor was mounted on a 200mm diameter ventilation duct and a probe was inserted into the duct to collect an air sample passing through the duct. The inlet fan maintains a relatively continuous flow within the duct and ensures thorough mixing of airborne particles with the incoming fresh air. The outlet of the duct is discharged through a flue. An electric furnace operating at about 350 ℃ was placed at the fan and duct inlet so that a small sample of the combustion agent could be placed on the electric furnace.

This device can undergo considerable dilution because the fumes are entrained and mixed with the main fresh air flow which is continuously drawn into the duct from the laboratory. This situation is used to simulate a real protected environment in which a high level of dilution is expected early in the growth of an early fire. Several different samples of the combustion agent were heated separately on an electric furnace to produce aerosol suspended particles. In addition, some samples of dust released at the fan and duct inlets were also evaluated not with an electric furnace but by stirring.

The outputs of the two monitor channels a and B are measured to provide a voltage excursion beyond a quiescent state (clean air) after airborne particles are introduced into the monitor.

It was observed that various types of combustion agents produce aerosol-suspended particles at different rates and concentrations. As the various combustion agents are heated and consumed, it is expected that the suspended particle size will vary over time, and thus the associated outputs from channels a and B should vary accordingly. Fig. 9 shows channel B output in response to several particle sources (after correction for the detected stable transient) expressed as a ratio of channel a output. These data are presented in ratios to illustrate the different airborne particle densities involved (assuming we are now concerned with particle size). The length and position of each horizontal bar represents the ratio of the ranges that occur during each tracking trial. In many cases, the ratio increases quickly to a maximum value and then decreases slowly. In some cases, the ratio increases again at a lower value after one cycle. Some of these observed models (signals) are clearly bimodal.

Figure 9 also shows the relative sensitivity of the monitor to these combustion agent sources and dust sources (in apparent order of average particle size). Accordingly, the nylon tube initially produced the smallest particles (peak ratio 5.3). After half the test, the ratio slowly dropped and the combustion agent actually melted on the furnace and produced suspended particles over a relatively long period of time. Styrofoam has similar results. The burning agent further down in the graph tends to char and produce solid carbonaceous residue.

The heating wire test consisted of a 2m length of PVC insulated wire heated by high current delivered by a 2V AC "range" transformer to simulate an overheated cable leading to early pyrolysis.

The solder resin results from the melting of the short length of rosin core solder, the position of which in the table indicates the production of rather large particles (high melting droplets).

The steam anomaly is due to the output reading from the boiling water source being of a very small order of magnitude and not producing an alarm condition, but the ratio includes the granularity set at the lower end of the graph. In contrast, in the case of various dust sources (including talc), all other sources produce large output readings and only the channel output ratio is small.

There is clearly a large difference between smoke aerosol particles and dust based on particle size, so it is possible to use this embodiment to distinguish between a source of smoke required for an alarm and a source of dust which is not required in generating the alarm.

Subtracting channel a from channel B (e.g., blue light) will result in a large reduction in readings, and thus may avoid undesirable alarms caused by these sources.

Moreover, these results are believed to be consistent with published data representing multiple combustion agents, with the first particles released by pyrolysis being quite small. Thus, the type of monitor used here may provide the earliest warning of pyrolysis.

Description of the Circuit

Figure 10 schematically illustrates in block diagram form one form of the invention for detecting smoke. The circuit drives a pair of light emitters 1 and 2, each having a different wavelength (color) and/or polarization characteristic. Each emitter is driven independently to provide light pulses of short duration (e.g. 0.4ms), optionally at intervals of (say) 150ms and 350 ms. This enables the air quality to be updated twice per second, a high sample update rate commensurate with low power consumption.

Part of the light scattered (diverged) from airborne particles passing through the monitoring chamber 3 is received by a photocell (not shown) within the receiver module 4. The signal is amplified in the receiver module 4 and passed to a main amplifier 5 having a gain controller 6. The amplified signal then passes through a discriminator (discriminator) comprising a pair of synchronous detectors 7, 8 and a pair of buffered sample and hold circuits 9, 10 which separates the signals originating from the two respective transmitters into two channels, channel a being denoted by numeral 9 and channel B being denoted by numeral 10. The dual channels provide information about the type of particles in the air. Channel a is particularly responsive to dust particles, while channel B is primarily sensitive to smoke and somewhat sensitive to dust. This is because dust and smoke particles each cover a wide range of sizes, which may overlap to some extent. In the subsequent circuit, therefore, the dirt reading for channel a is subtracted from the smoke reading for channel B by means of the adder 11 to provide a signal indicative of substantially only smoke density (concentration).

The smoke density signal is applied to a threshold sensitive line 12 which operates a series of three lights and relays 13 in response to the level of fire hazard detected. These lamps and relays are, for example, represented as: a1 (warning, or level 1), a2 (active, or level 2), and A3 (fire, or level 3). Typically these three alarm levels represent smoke densities corresponding to light shades of approximately 0.03, 0.06 and 0.10%/m, although the monitor may be adjusted to other settings, it will be appreciated that the signals and settings may be set as appropriate for the particular application of the invention.

Furthermore, the direct output 14 from channel a is used to indicate when the dirt level is higher independent of the smoke concentration level. This may also facilitate testing, commissioning and demonstration. The output also indicates when the monitor is in the process of recognizing dust.

An additional lamp and relay 13 may be provided as a "fail-safe" circuit applied to the adder 11 to provide a fault alarm in the event that the monitor is not functioning correctly with sufficient sensitivity. The analog output from the summer 11 may also be provided for remote processing of fault and alarm notifications. Alternatively, the analog output may be provided by each of the slave channels a and B to allow remote signal analysis and processing of fault and alarm notifications.

A clock generator 15 may provide appropriate timing signals when required, and a power supply section 16 may distribute power to the various parts of the circuit at appropriate voltages.

It is necessary that the output signal from the discriminator channel is not saturated when very high smoke or dust concentrations are encountered. This saturation can lose information about the relative signal levels produced by the two transmitters, thereby overwhelming the discrimination function. First, the amplifier is provided with a large "headroom" to enable full-scale operation when, for example, the signal level is half saturated. Second, an automatic gain controller is provided. The DC output voltage from the discriminator channel is fed back to the gain control means to ensure that the saturation concentration is not reached.

Gain controller

Referring to fig. 11, the mid-band gain of the operational amplifier is determined by the ratio of the feedback resistance to the input resistance. For IC3a in FIG. 11, the voltage gain is R4/R3, while for IC3b, the voltage gain is R6/R5. High frequency breakpoints were determined by C4. R4 and C6. R6, while low frequency breakpoints were determined by C1 (R1/R2), C3. R3 and C5. R5. The amplifier is DC coupled and DC bias is set by R1 and R2.

The gain control IC4 typically includes an LDR (light dependent resistor) and an LED (light emitting diode) tightly coupled in a light-tight enclosure. The LDR provides an adjustable resistance whose value is determined by the current delivered through the LED, which is externally controlled by R7. The resistance of LDR is virtually infinite when no current passes through R7, and decreases to a range of 10k Ω to 100k Ω when the current is 10mA to 20 mA. Typically such LDRs will be connected across R4 or R6. This has the advantage in operation that the high frequency break point (C4 · R4 or C6 · R6) is increased, thereby enhancing (upsetting) the desired frequency response and phase characteristics of the amplifier. In addition, such devices have been found to produce gain control of incomplete dynamic range.

Since the two stage circuit is non-inverting for the amplified signal, it is possible to connect the LDR from the output of the second stage (IC3b) to the input of the first stage (IC3 a). This greatly increases the available effective dynamic range. Furthermore, when IC4 is in effect, neither the breakpoints C4 · R4 nor C6 · R6 are affected.

The current drive R7 derives from the sample and hold voltage signals (high and low) for channel a and channel B, through zener diodes (zener diodes) D5 and D6 to ensure that the gain control action does not come into effect until the signal level is significant.

Importantly, the characteristics of the LDR, LED and zener diode combination are neither abrupt nor linear. It is non-linear and has the effect of providing a logarithmic gain function. Abrupt changes in gain can cause instability or irregular behavior because high signal levels can cause an abrupt decrease in gain, which can cause an abrupt decrease in output, which in turn can reduce the drive to IC4, causing the gain to increase again. Also, this may cause the alarm output relay to vibrate. The non-linear design allows for a small increase in output when the input reaches a high level and provides wide dynamic range control.

The standard full-scale sensitivity of the monitor corresponds to a shading rate of 0.1%/m, corresponding to the highest alarm threshold ("fire"), below which an intermediate alarm threshold is available. By using this logarithmic characteristic, it is possible to vary the set alarm output threshold so that higher level alarms can be in the non-linear range. In this way, sufficient resolution can be provided to provide a first level of warning ("warning") at very low smoke densities (e.g., 0.01%/m), while the highest level of warning can be up to 1%/m, 10%/m, or even higher.

Monitoring room optics

Fig. 12 shows a ray diagram of an emitter operating at different wavelengths and/or polarizations. For clarity, the sample rays are shown according to their position at the center 1201, left or right end 1202 of the beam. Alternatively, the beams are actually operated for a short pulse duration. It can be seen that the beam is formed by lensed emitters 1203, 1204 and is confined by diaphragms 1205, 1206 so as to pass through a center, monitoring area or zone 1207 of the monitoring chamber. If smoke or dust is passing through this area 1207, these particles scatter a small portion of the beam energy in multiple directions. This portion of the energy is dispersed in the direction of the primary receiving diaphragm 1208 and is thus scattered to the lens 1209, which focuses the energy onto the photocell in the receiver module 1210. It is worth noting that intermediate diaphragms are avoided in this path, since stray light reflected by the monitoring chamber members and thus originating from inappropriate directions may be reflected by these intermediate diaphragms to enter the lens.

The direct light beams 1201, 1202 then enter an absorption channel 1211, where multiple reflections from highly absorbing walls 1212 consume the light energy. The channel is designed to direct multiple reflections towards the far end of channel 1213 so that multiple reflections occur before any residual light occurs. This absorption, in combination with the geometry of the main diaphragm for the monitoring chamber and the beam diaphragm, avoids interference of the residual starting beam with light scattered from smoke or dust particles.

Ray 1214 represents the area that is made sensitive to the photo cell by the receive lens and the primary stop. It can be seen that the sensitive volume is concentrated within the monitoring region 1207, but the photocell 1210 maintains sensitivity along the optical axis beyond this region. The sensitivity of this expansion is limited by the absorption zone 1215 at the distal end of the monitoring chamber. The design is intended to ensure that negligible light energy from the emitters 1203, 1204 can fall on this absorption region, which tends to interfere with the light scattered by the particles. This unwanted (undesired) light originates mainly from reflections of the emitter stops 1205, 1206. The combination of shielding (shielding) the absorbing area and reflecting stray light out of the area minimizes this disturbing light. In addition, the walls of the absorption region are preferably colored black to absorb incident light.

Fig. 13 shows typical unwanted radiation produced by reflection from the emitter stops 1205, 1206 that is blocked from reaching the central absorption region 1215. The figure also includes unwanted radiation 1216 passing through the primary stop 1217 and being absorbed in the receiving channel 1218. Additionally, as shown, unwanted rays 1219 reflected from the primary stop 1217 are focused off the central axis of the photocell within the receiver module 1210 and are avoided by the photocell within the receiver module 1210 (shown as 1401 in fig. 14).

A combination of all these methods is used to avoid interference with the light scattered from the airborne particles. The scattered light density is typically 1 hundred million times lower than the emitter light, thus recognizing the difficulty of this task.

Referring again to fig. 12, the brightness within the central cone of light 1202 from the emitter is considered the first level brightness within the monitoring room. The bright light is directed to an absorption channel 1211 along which it is efficiently absorbed after multiple reflections. Outside of this central cone angle is a second level of brightness 1220 caused by the emitter optics and the reflection of the emitter stop. Thus, it is believed that the entire emitter diaphragm area must be illuminated in multiple directions. Accordingly, the emitter diaphragm must be obscured from view of the receiver or lens diaphragm, which can be achieved by the positioning of the main diaphragm 1217. To achieve this shadowing, the geometry of the monitoring chamber is set by a straight line 1221 (shown in dashed lines in fig. 13) from the outermost ends of the emitter diaphragms 1205, 1206, to the innermost end of the primary diaphragm 1217, to the outermost end of the lens diaphragm 1222. Given that the purpose of embodiments of the present invention is to produce a monitor with the smallest available size and the highest possible sensitivity, the geometry is considered to be a defined geometry.

Being outside of the central emission cone 1202, the primary diaphragm 1217 is exposed to light rays of the second level brightness 1220 from the emitter diaphragms 1203, 1204. Thus, the primary diaphragm 1217 reflects light of the third level of brightness 1219 in multiple directions. It is noted that in the present discussion, the "order of magnitude of brightness" does not necessarily mean ten times. Assuming that a black surface can absorb 99% of the incident light, reflecting only 1%, and this 1% is further reduced by scattering due to non-specular reflection, the reduction in brightness may be on the order of 1000 times or more. Thus, the third level of brightness is not an accurate measurement, but provides only a relative representation. A small portion of this tertiary brightness light 1219 will be reflected towards lens stop 1208 and lens 1209. As shown in FIG. 14, lens 1209 will focus this unwanted light 1219 off the central axis of receiver element 1210 and is blocked by receiver stop 1401. The use of a lenticular lens, a longer focal length, and a wider primary diaphragm enables unwanted rays (off-center) reflected from primary stop 1217 to fall onto the sides of receiver element 1210 and be attenuated by receiver stop 1401.

It is expected that a relatively precise control of the focusing of the lens is necessary in order to control the separation of unwanted (harmful) light from wanted light. An aspherical lens 1501 having a shorter focal length is proposed (as shown in fig. 15). Such a lens provides precise control of focus over the entire surface of the receiver element, avoids spherical aberration and forms a picture-quality image. Fig. 15 illustrates the operation of such a lens 1501 in focusing scattered light received from particles detected in the monitored region 1207 (fig. 12). FIG. 15 also shows the position of lens 1501 relative to primary stop 1217, and element 1210. However, fig. 16 shows that a part of harmful light reflected from the main stop with such an aspherical lens falls on the element. This may interfere with the desired signal.

Returning again to fig. 12, a thicker lenticular lens (having two convex surfaces) is used and is shown in more detail in fig. 14 and 17. As shown in fig. 14, spherical aberration of this type of lens 1402 helps to improve the separation of the two sets of rays, since the unwanted light 1219 arrives from a direction that is off the central axis. This separation is further facilitated by the use of longer focal lengths (and has been found to be proportional to the focal length). In fig. 17, it can be seen that it is possible to use a lenticular lens 1402, since the focus point is less important than the included light path, since there is no need to form an accurate image, such as a photograph, at the receiver element 1210, but only to collect the light. Thus, the geometry of receiver element 1210 and lens 1402 are preferably arranged such that the maximum amount of scattered light from the detected particles can fall on the receiver element (as shown, where the light illuminates substantially the entire surface of element 1210) while unwanted light is either blocked by the above-described receiver stop 1401 or allowed to pass through the sides of the element.

Fluid dynamics

The design of the monitoring chamber is important from a hydrodynamic point of view. One embodiment of the present invention includes a micro duct detector for collecting successive small but representative samples of air flowing through a vent duct, such as the detector disclosed by the present inventor in co-pending U.S. patent application 2003/0011770.

Referring to fig. 13, a fluid sample, such as air, collected from the environment is drawn into the monitoring chamber of the present invention through inlet 1301, flows through the detection chamber and monitoring region 1207 (fig. 12), and exits through outlet 1302. This makes it possible to use a larger filter 1303 which can effectively remove dust over a long period of operation without causing significant head loss (pressure drop). The preferred filter type in use is a large pore, open cell foam filter with a large depth. The smallest dust particles that the filter is designed to remove are typically at least 10 times smaller than the average pore size of the filter. The removal of dust is achieved as a result of brownian motion (rapid thermal vibration) by which the dust particles appear to be many times larger than their physical size. As the fluid flows through the deep filter, the dirt is statistically removed such that substantially all dirt believed to be harmful is removed before the fluid exits the filter outlet 1314. This has been found to minimize the build up (contamination) of dirt within the monitoring chamber, thereby greatly extending the maintenance cycle. However, the open structure of the filter avoids the serious problem that occurs in prior art inhalation smoke detectors, namely that the sensitivity of smoke particle removal decreases with time. Moreover, such filters are of the following type: the head loss in the filter does not increase significantly as the filter is loaded with dust.

Typically, smoke particles are present in the range of 0.01 microns to 1 micron, while airborne dust particles are present in the range of 1 micron to 100 microns. However, there is a partial overlap at the 1 micron boundary because the smallest dust particles in nature are smaller than the largest possible smoke particles. Therefore, it is not appropriate to consider that the filter should be a good cleaner. To avoid a reduced sensitivity to smoke, a small fraction of the dust particles must therefore pass through the filter, which needs to be adjusted in another way (as disclosed later).

Each side of the filter 1303 has a mirror image diffuser 1312, 1313. The outlet face of the filter leads to a diffuser 1313, which effectively recombines the fluid, makes a 90 ° turn, and directs the fluid to the channel 1304. In the preferred embodiment of the invention, this channel narrows to a cross-sectional area still 5 times larger than the inlet tube, thereby maintaining very low losses, but the local air velocity is about 8 times faster than the velocity at the outlet face 1314 of the filter.

In a preferred embodiment, two sensing devices 1305, 1306 may be installed, one sensing device 1306 at the filter outlet and one sensing device 1305 in the throat 1304. In this arrangement, the sensor 1306 is subject to a relatively low velocity air flow out of the filter, so that the sensor is less subject to cooling. The sensor 1306 may be further prevented from cooling by means of a tube cap 1307. Conversely, sensor 1305 is more fully exposed to a significantly higher rate of airflow and is therefore more easily cooled than sensor 1306. Both sensors 1305, 1306 are preferably exposed to the same ambient air temperature. It may be preferred to use matching devices with a known temperature dependence whereby the different cooling rates caused by the different air flow rates to which they are exposed may be used to generate different voltages across each sensor, thereby providing a measure of air velocity in a manner largely independent of ambient air temperature.

The sensors may be of the type disclosed in US4,781,065, but the positioning of the sensors in the device of the invention is uniquely different.

Also, in the present device, the sensor is exposed to the airflow after the airflow passes through the dust filter 1303, thus minimizing contamination. Contamination may affect the cooling characteristics of the sensors 1305, 1306, thereby reducing the accuracy of the airflow detection circuit.

The fluid then enters another diffuser 1308, which is also a light absorbing channel 1308 for the emitter 1203 (FIG. 12). When the air stream reaches the inlet of the absorbing channel 1308, it is redirected to slow it to about 25 times less than the velocity at the inlet tube. Thus, very little loss is caused in the flow of air through the channel 1308, through the monitoring area 1207 (FIG. 12), and into the secondary channel 1309. Because of the slow speed here, any remaining dust particles that may be present in the airflow that are small in number and size (because of the filter 1303) have a very low impulse and are therefore not thrown out of suspension in the fluid by centrifugal force, thereby minimizing potential contamination in the vicinity of the monitored area 1207. In the case of a tendency to centrifugally separate dust particles, the momentum is in a direction such that these particles are deflected harmlessly away from the primary diaphragm 1217.

The air flow is sucked into the second suction passage 1309 and is gradually and efficiently accelerated by the diffusion action so as to become matched with the exhaust gas outlet 1302. As described in the above-referenced US4,781,065, the exhaust gas, e.g. dust, is then effectively returned to the environment of the sample.

It has been explained how the gas flow passes through a series of stages in a manner that minimizes losses and promotes laminar flow. Thus, the monitoring chamber is very effectively and quickly purged with a fresh air sample, while retaining only a minimal amount of smoke. Although the large cross-sectional area gives rise to low local velocities, the response of the monitoring chamber components to changes in smoke concentration has proven to be very rapid and suitable for smoke monitoring alarm purposes.

Since there is only a very small pressure drop in the monitor of the present invention, the absolute pressure at any location within the monitor is similar to that inside the pipe. Since there can be a large pressure differential between the interior of the pipeline and the ambient environment in which the monitor is located, the monitor must maintain a good pressure seal to avoid leaks anywhere. The possibility of leakage is minimized by the design of the monitoring chamber, which includes two similar halves of a planar connection, the docking flange 1310. Thus, only one flat gasket is required for sealing the monitoring chamber. In one embodiment, a thick sealing foam gasket is preferred because it readily accommodates changes in the flange plane within the monitoring chamber, thereby overcoming the small amount of bending and warping that may occur during injection molding. By extending the small edge 1311 that overlaps the central joint of the two monitoring chamber halves, the area of the monitoring chamber, especially the area close to the monitoring area 1207 that is sensitive to the light absorbing capacity of the monitoring chamber walls, is hidden from the gasket. Preferably the actual contact between the two halves of the monitoring chamber is only at these edges, which greatly simplifies the need to manufacture the interface plane.

The foregoing description has discussed in principle the use of a duct probe, however, in other embodiments of the invention the probe may be replaced by other means for acquiring a fluid sample to be monitored, such as air. Such other devices (disclosed in US4,781,065) may be venturi devices in small bore pipes of e.g. 20mm diameter. Such a conduit may be connected to a suction pump or fan (aspirator) placed either upstream or downstream of the venturi. If placed downstream, multiple monitors may be connected to a single aspirator. Upstream of each monitor, the small-bore pipe may extend throughout the fire zone (fire zone). The sampling pipe may be provided as a reticulated pipe or a branched pipe extending into the fluid region or area to be monitored or detected. Each of the conduits may comprise a branch conduit. Each of the duct and branch duct may include a plurality of small holes to draw air in the duct adjacent each hole. The components of the air sample from all of these orifices are then drawn into the venturi intermittently or relatively continuously. The venturi is arranged so that a portion of the air within the duct is drawn past the monitor so that the presence of smoke or dust can be sensed before the monitor airflow returns to the duct. All of the air is then drawn into the aspirator and expelled.

It is worth noting that preferably in the case of a duct detector or venturi, only a portion of the available air passes through the monitor. This portion of air or air sample contains smoke and/or dust of the same density as the main fluid. However, by carefully minimizing the flow through the monitor, the rate of dust build-up in the filter can be minimized, thereby maximizing the service interval without affecting the sensitivity of the monitor.

In another alternative embodiment of the invention, instead of a venturi, the monitor may be connected directly to a small bore pipe, such as a 5mm bore pipe. This is suitable for running short distances such as a few meters. In this case, the entire air flow would pass through the monitor, but the flow rate would be low and therefore should not necessarily affect the service interval. To achieve a fast response time of a small-bore tube over a long distance, the pressure drop will be very high, necessitating the use of an aspirator having a high pressure and high energy consumption

Mounting of monitor

Referring to fig. 18, a monitor 1801 (e.g., a monitor according to the present invention) may be mounted on a surface having flat sides, rounded or other shapes, such as a pipe 1802 with a mounting fitting 1803. Monitor 1801 may be secured using, for example, screws, or other suitable means (not shown). During installation of the monitor, the tabs 1803 are simply bent until the tabs mate with the fixed monitor surface. For example, as shown in FIG. 18, when installed on a pipe, the fittings are bent until they grip or mate with the surface of the pipe. The tubing can be as small as 200mm (8 inches) in diameter. The fitting 1803 may be integrally formed with the housing of the monitor 1801, in which case a slot (not shown) formed in the housing may restrain the fitting and enable the fitting to be bent without deformation for gripping to a pipe surface or other mounting surface. While the invention has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications. This application is intended to cover any variations uses or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth.

While the invention has been described in conjunction with specific embodiments thereof, it will be understood that it is capable of further modifications. This application is intended to cover any variations uses or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth.

As the present invention may be embodied in several forms without departing from the spirit of essential characteristics thereof, it should also be understood that the above-described embodiments are not to limit the present invention unless otherwise specified, but rather should be construed broadly within its spirit and scope as defined in the appended claims. Various modifications and equivalent arrangements are intended to be included within the spirit and scope of the invention and appended claims. Therefore, the specific embodiments are to be understood as illustrative of the various ways in which the principles of the invention may be practiced. In the claims that follow, means-plus-function clauses are intended to cover the structures described as performing the defined function and not only structural equivalents, but also equivalent structures. For example, although a nail and a screw may not be structural equivalents in that a nail employs a cylindrical surface to secure wooden parts together, whereas a screw employs a helical surface to secure wooden parts together, in the environment of fastening wooden parts, a nail and a screw are equivalent structures.

When the specification uses the word "comprise/comprises", it is used to specify the presence of stated features, integers, steps or components but does not preclude the presence or addition of one or more other features, integers, steps, components or groups thereof.

Claims (28)

1. A method of determining the presence of particles having a predetermined size or range of sizes in a fluid sample, the method comprising the steps of:

illuminating the sample with light of a first wavelength,

obtaining a first response signal representative of the first illumination,

illuminating the sample with light of a second wavelength;

obtaining a second response signal indicative of said second illumination, an

Determining the presence of the particle having the size or range of sizes by comparing the first response signal and a second response signal, wherein the first response signal is subtracted from the second response signal.

2. The method of claim 1, wherein the second wavelength provides a response signal to both particles having the size or within the size range and particles not having the size or outside the size range, and the first wavelength provides a response signal to particles not having the size or outside the size range.

3. A method according to claim 1 or 2, wherein the first wavelength of light is of infrared wavelength and the second wavelength of light is of blue wavelength, the first response signal being subtracted from the second response signal to provide a "blue" response signal shorter than the "infrared" response signal.

4. A method according to claim 1 or 2, wherein the first response signal is subtracted from the second response signal so that smaller particles are more clearly distinguishable than larger dirt-like particles.

5. A method according to claim 1 or 2, wherein the first response signal is subtracted from the second response signal to balance the effect of contamination.

6. The method of claim 5, wherein the first response signal is predominantly responsive to dust.

7. A method according to claim 5, wherein the second response signal is the most responsive to smoke, but has some sensitivity to dust.

8. The method according to claim 1 or 2, further comprising the steps of:

triggering an alarm signal upon detection of particles having said predetermined size.

9. The method of claim 8, wherein the alarm signal is indicative of an alarm condition for a pyrolysis, incomplete combustion, and/or smoke event.

10. A method according to claim 1 or 2, wherein the first wavelength of light is in the range 650nm to 1050nm and the second wavelength of light is in the range 400nm to 500 nm.

11. The method according to claim 1 or 2, further comprising the steps of:

illuminating the sample with light of at least one other wavelength, wherein particles having at least one other size or range of sizes are relatively sensitive to the other wavelengths of light,

obtaining at least one further response signal representative of said further irradiation, an

Determining the presence of said particles having said other size or size range by comparing said first, second and/or other signals.

12. A method according to claim 1 or 2, wherein at least one of said illuminations is polarized.

13. A method according to claim 1 or 2, wherein at least one of said illuminations is horizontally and/or vertically polarized.

14. A method according to claim 1 or 2, wherein the first illumination is a horizontally polarized relatively long wavelength and the second illumination is a vertically polarized relatively short wavelength.

15. The method of claim 1 or 2, wherein the first illumination is horizontally polarized red or infrared light and the second illumination is vertically polarized blue wavelength light.

16. The method of claim 1 or 2, wherein the first illumination is horizontally polarized red or infrared light and the second illumination is unpolarized blue light.

17. A particle monitor adapted to determine the presence of particles having a predetermined range of sizes in a fluid sample, the monitor comprising:

first illumination means for illuminating the sample with light of a first wavelength, the first light having a wavelength to which particles having a first size are relatively sensitive,

first signal means for providing a first response signal indicative of said first illumination,

second illumination means for illuminating the sample with light of a second wavelength having a wavelength to which particles of a second size are relatively sensitive,

second signal means for providing a second response signal indicative of said second illumination,

logic means for comparing the first response signal and the second response signal to determine the presence of particles within the predetermined range, wherein the logic means is configured to subtract the first response signal from the second response signal.

18. The particle monitor of claim 17 wherein the second wavelength provides a response signal for particles having a predetermined size or within a predetermined size range and particles not having a predetermined size or outside a predetermined size range, and the first wavelength provides a response signal for particles not having a predetermined size or outside a predetermined size range.

19. A particle monitor as claimed in claim 17 or 18, wherein the first wavelength of light is of infrared wavelength and the second wavelength of light is of blue wavelength, the first response signal being subtracted from the second response signal to provide a "blue" response signal shorter than the "infrared" response signal.

20. A particle monitor as claimed in claim 17 or 18, wherein the first response signal is subtracted from the second response signal so that smaller particles are more clearly discernable than larger dust-like particles.

21. A particle monitor as claimed in claim 17 or 18, wherein the first response signal is subtracted from the second response signal to balance the effect of contamination.

22. The particle monitor of claim 21, wherein the first response signal is predominantly responsive to dust.

23. A particle monitor as claimed in claim 21, wherein the second response signal is most responsive to smoke but has some sensitivity to dust.

24. A particle monitor according to claim 17, wherein the monitor further comprises an output device adapted to provide a log-scaled signal as an indication of particles.

25. A particle monitor according to claim 24, wherein the scale signal is provided to an alarm.

26. A smoke detector including a particle monitor as claimed in claim 24.

27. An apparatus adapted to detect particles having a predetermined range of sizes in a fluid sample, the apparatus comprising:

processor means for operating in accordance with a predetermined instruction system,

the apparatus, in combination with the instruction system, for implementing the method of any one of claims 1 to 16.

28. An assembly comprising a lenticular lens and a particle monitor according to claim 17.