OPS: Impact of interference on security detectors.
False alarms are perhaps the most unpleasant phenomenon in any fire alarm system. Most users simply turn off the entire system when false alarms occur (after all, finding and turning off a single detector subject to interference is usually quite labor-intensive). The impact of interference on fire detectors has been discussed in articles quite often, in particular because the number of fire detectors in a system is usually large, and the importance of fire alarms is traditionally underestimated by users. In addition, fire detectors must work around the clock, including during working hours, when the amount of interference increases significantly. Security detectors are usually monitored only during non-working hours, which significantly reduces the level of interference.So, what are the sources of false alarms in security detectors? Let's consider the most common types of detectors today — passive infrared optical and sound glass break. The table shows possible sources of interference and their impact on the corresponding types of detectors.
Let's consider in more detail the features of the specified types of detectors, as well as possible sources of interference and their impact.
The IR passive detector responds to very low-frequency optical signals. The bandwidth of the measuring part of the detector is usually about 0.2-5 Hz, i.e. significantly lower than the frequency of the general industrial power supply network and incomparably lower than radio frequencies. However, if the interference is strongly modulated in the required range, it may well lead to false alarms. For example, some brands of detectors have been noted to react painfully to a call on a cell phone. You all know that often when establishing a connection with a cell phone, pulse interference is heard in the radio receiver. These are just a few pulses (more precisely, several data exchange packets), which occur with increased power until the cell phone and the cellular repeater establish a stable connection. Experiments show that some old (non-processor) infrared detectors could give a false signal when calling a phone located at a distance of up to 1 meter from the detector. In practice, this is an unlikely source of interference; to implement the described scenario, the phone must be forgotten in the protected room on the top shelf of the cabinet (after all, detectors of this type are installed high on the wall). At the time when 480 MHz phones were widespread, and especially phones of the Altai system, the problem was more pressing. Since then, both phones and detectors have changed, but such a source of interference is traditionally considered possible.
More likely sources of modulated radio interference are faulty electrical appliances. A starter lamp that periodically flashes on and off is a serious source of interference in the dangerous frequency range. In one known case, the source of interference was a slowly rotating ceiling fan, sparking with brushes in one position for each revolution. Again, it is comforting that during non-working hours, when the premises are under guard, both lamps and fans are usually turned off.
Optical interference, including infrared interference, can be a serious problem. Sources: heaters turning on and off, cold air flows from a window in winter or from an air conditioner in summer, light flows from lighting or from the headlights of a car passing outside the window.
Fortunately, heaters usually turn on and off with sufficiently long delays, so that modern analysis algorithms that require two changes (signal appearance and disappearance) in a short time exclude false reactions to such interference. Interference from fluctuations in convective air flows from heaters is small, and the need for protection from such interference is stipulated in the relevant GOST and is deliberately provided by all domestic manufacturers. I have also never heard of any real problems with interference from convective warm air flows when using equipment from foreign manufacturers.
Air flows from a window with a significant temperature difference and sudden changes in flows (a window flung open) are practically indistinguishable from the intrusion of a criminal — here technology is powerless. Just as a reed switch detector will honestly inform about an alarm from an unlocked flapping window, so an infrared detector is simply obliged to inform about the intrusion of air with a sharply different temperature. Just do not forget to close the windows.
Finally, powerful visible light sources. The aforementioned GOST R 50777-95 describes very strict requirements for protection against such interference. Inexpensive detectors from foreign manufacturers, as a rule, do not meet these requirements. In my opinion, the importance of protection against such interference in GOST is somewhat overestimated — flashes of a car headlight, aimed directly at the detector from a short distance, are possible only deliberately, to compromise the system by creating false alarms. Signals from direct sunlight falling on the detector are in any case much stronger and can lead to a false alarm, so no detector should be mounted in such a way that direct light from a window falls on it. However, I do not claim that this requirement is a disadvantage of GOST. An indirect consequence for detectors that comply with GOST is a general decrease in sensitivity, which also ensures reduced sensitivity to any other interference. Some domestic detectors have the ability to switch to increased sensitivity (however, it will not satisfy the GOST on the mentioned point — protection from sudden switching on of the headlight), but its resistance to other interference will remain acceptable (it is up to the user or the service organization to decide on the degree of acceptability, because there are many sources of interference, no GOST can provide for everything), and the sensitivity (range) will increase significantly (approximately twice).
Let's now move on to glass break detectors. The monitored physical parameter — the sound of breaking glass — is oscillations in the range of 100-10,000 Hz, with an envelope with approximately the same characteristic times of 0.2-2 seconds. Unlike an infrared detector, in this case the frequency range of sensitivity is significantly shifted upwards, sound detectors are more susceptible to interference from a 50 Hz network with all harmonics. However, in practice, electromagnetic interference is never complained about. There is a much more important source of complaints — acoustic interference, that is, sounds similar to breaking glass. The most dangerous thing is that glass break detectors are often assigned to the perimeter line and left on guard around the clock. Can you imagine such a detector in a canteen, where knives, forks sometimes fall, and even plates break?
The best detectors at the lowest sensitivity give false alarms about once a day. Of course, it is a mockery to install sound detectors in such a situation. But what about another situation — when glass doors slam, glass in old frames rattles from a passing tram, and in the neighboring room they are drilling holes in the walls with a hammer drill?The situation with the regulatory documents for acoustic glass break detectors differs from the GOST for infrared passive detectors. GOST 51186-98 describes sensitivity tests in great detail: it is required that the detector should generate an alarm signal under certain conditions. The conditions for testing interference (when the detector should not generate a signal) are, on the contrary, very soft and are easily met even by the simplest single-frequency devices without any analysis of the sound pulse shape. Imported glass break detectors, as a rule, at the upper sensitivity limit also approximately correspond to the GOST detection range of 5-7 meters. However, both imported and domestic ones necessarily have sensitivity adjustment, and experienced installers know that it is better to set this adjustment immediately, if not to the minimum, then in no case to the maximum. In good detectors, the adjustment changes the sensitivity by about 20 dB, which means a tenfold decrease in the detection distance. Considering that real glass is much larger than the minimum mandatory test glass «according to GOST» (30×30 cm), then even at minimum sensitivity the detector will give an alarm at a distance of at least 1 meter from the glass, but will not react to «heavy metal» in the neighbor's apartment.
Unfortunately, I am not aware of any objective studies of noise immunity of glass break detectors. I cannot call those conducted with my participation objective – the methodology has not been sufficiently developed. In addition, there can be a great many different types of noise, and different detectors using different detection frequencies can behave completely differently. The general statement that 3-frequency or multi-frequency detectors are more protected than 2-frequency ones is probably true. The technologies for analyzing the sound of falling fragments or other algorithms for analyzing the sequence of events, as far as I know, have also never been comparatively analyzed (at least, there are no open publications). My recommendation, alas, is banal: if you have problems with false alarms from a glass break detector, try another type, possibly more complex, and manually reduce its sensitivity to the minimum acceptable level.
A very typical and successful solution is to install glass break detectors behind curtains, near the glass. If there are heavy curtains, this is the only possible option — after all, when installed inside the room, no detector will hear through the thick brocade that the glass has broken. Of course, you have to install as many detectors as there are windows in the room, but at the same time the problem of noise immunity is solved — the sensitivity of the detectors can be set to a minimum, and the noise from inside the room behind the curtains will be much weaker.