The first addressable systems appeared among fire alarms about 30 years ago.
Soon, addressable security systems appeared.
They are gradually gaining popularity.
Now there is a widespread opinion that addressable systems are very, very good, but expensive.
I will try to dispel both of these stereotypes.
First, let's clarify what we mean by addressable systems.
These are systems in which the terminal peripheral devices — detectors, alarms, fire extinguishing modules, etc. are connected uniformly to one, usually two-wire line, receive power from this line and exchange information along the same line with a certain central device, and the device individually distinguishes each terminal device by the address assigned to this device.
For fire alarms, it is customary to subdivide into addressable and addressable-analog, but this division made sense in the distant past.
Nowadays, all addressable devices are to some extent addressable-analog, because they are capable of transmitting extended information to the central device, for example, about the degree of smoke or temperature at the measurement point.
When these terms were invented, the sensors were primitive and could hardly transmit their address.
Now, all such detectors contain a microprocessor and can transmit anything.
On the other hand, when the term «addressable-analog» arose, it was implied that all sensors transmit the value of the measured analog quantity to the central device, and the device itself, having compared all the information, makes a decision.
In fact, no reasonable algorithms for comparing information from different sensors have been developed, and therefore the decision is made (no matter whether by the detector itself or by the control and monitoring device) for each measurement individually.
Therefore, all existing systems do not fully correspond to the original definition of addressable-analog.
By the way, you may have noticed that I quite freely mix the concepts of «sensor» and «alarm».
Of course, according to the existing domestic regulatory literature, they should all be called alarms. But all this literature comes from the last century and is simply not true.
Addressable analog detectors, by definition (given in the same NPB and GOSTs), are not detectors, because they do not transmit a “fire notification,” but merely measure and transmit the value of the controlled parameter.
The decision that a fire is taking place is made (in theory) by the control and monitoring device.
In addition, the word “detector,” in my opinion, is typical Soviet-military bureaucratic jargon.
There is no similar word in any other language, more general terms «sensor», signaling device, detector» are used, but no one has thought of inventing a special word specifically for fire alarm components.
I have been working in this industry for a long time, I am used to using the word «detector», but from time to time it still makes me cringe and some synonym pops up on its own.
But let's get back to addressable alarm systems. The most common are fire addressable systems, although, in fact, addressability is much more important for security systems.
Fire alarms, with rare exceptions, are uniform, always «on guard», and it doesn't really matter which detector gave the alarm or even in which specific room.
The result is always the same — the entire building must be evacuated, and it doesn't really matter which room to run and unroll your sleeves on the floor where the fire is, it will be clear there on the spot.
In a security alarm system, one room may contain perimeter detectors (glass break sensors and reed switches on vents), which are constantly on guard, as well as infrared motion detectors, which are armed only when no one should be in the building.
Of course, both must be distinguished (they cannot be hung on one non-addressable loop).
Neighboring rooms are often armed separately, as the staff leaves, at different times. This means that even identical detectors in neighboring rooms cannot be connected to one non-addressable loop.
Before moving on to criticism, let us briefly (a lot has been written about this) list the advantages of addressable systems.
Addressable systems allow you to get much more information.
In addition to the individual identification of the triggered detector, addressable devices (at least modern addressable-analog ones) provide additional information about the cause of the alarm and extended self-diagnostics in standby mode.
All this allows for a much more accurate determination of the cause of the alarm and selection of appropriate response measures, as well as the ability to eliminate faults in advance during operation before they lead to system failure or, even worse, to false alarms.
Addressable systems allow for simpler cabling: there is no need to run a separate cable to each separately identified device.
Addressable systems allow you to easily reconfigure the system or update it by adding new devices to the same line.
Is it all true? Yes, it's all true. But is it as good as it seems?
Addressable systems allow you to get much more information.
Moreover, they force you to get much more information.
Where previously the control panel had only one red “fire” light, now a text display is inevitably added, showing at least the address of the fire alarm.
And if the system is addressable and analog, then the display also shows the value of the measured parameter, and, of course, the time the notification was issued.
Of course, you can also look at the device's self-diagnostics parameters, its serial number, the date of manufacture, the time since commissioning or the last scheduled maintenance, almost the last name of the stacker and the azimuth relative to the magnetic pole.
Now imagine a calmly dozing concierge (or a retired sergeant restlessly tossing and turning on stacked chairs) and what they will do with all this information.
Imagined it?
Realized that in addition to a highly informative display, it is absolutely necessary to install a simple panel with lights labeled «floor 1», «floor 2», etc.?
That is, in fact, for a real user the system should behave as a non-addressable one. All this additional information will be useful only for the operating engineer.
Accordingly, the system should be configured after installation so that the first light comes on upon a signal from any detector on the first floor, the second light comes on from detectors on the second floor, etc.
In a non-addressable system, to distribute detectors by zones, it is enough to draw in the project which detector to connect to which loop.
In an addressable system, you will have to add a section called «configuration» or «commissioning project» to the design documentation.
Specify which detector (physically, where it is located) should have which address, and then which address should be assigned to which zone (area, section, etc.). In particularly advanced systems, you also need to specify the operating parameters of each device.
What are you saying?
Have you ever done any design documentation other than a sketch on a piece of paper? Now you'll have to.
Or you'll have to send a highly qualified specialist to the site, and he'll figure out how to set it all up on the spot.
Obviously, such work is beyond the capabilities of a fly-by-night company with a temporary team of fitters made up of unskilled laborers who are ready to do anything.
Even the word «addressable» itself already contains additional work at the commissioning stage — assigning an address to each device.
In some systems this can be done almost automatically, in some it is necessary to set the address on each sensor with switches, in most systems you will have to tinker quite a bit.
The same in operation. It is not enough to replace a failed product with a similar new one.
You need to re-address it. Probably re-adjust the device's operating parameters. And as a result, discover that it is not fully compatible with the previously installed equipment.
These primitive discrete detectors with a dry contact at the output are all the same.
But addressable ones, even from the same manufacturer, can improve over time, and at the same time lose compatibility with previous versions. They have new functions, new parameters, and they are not always 100% compatible with the old ones.
The situation is even worse with the compatibility of devices from different manufacturers. Currently, all address system protocols are exclusively closed, and are the property of their manufacturers.
If several manufacturers produce similar devices using one protocol, then either one of them shamelessly uses the intellectual property of the other, or they are bound by complex obligations that limit competition.
The end user, if they want to upgrade or repair the system, will almost certainly have to go back to the same manufacturer they originally bought it from.
It's like video cameras.
They all used to have a BNC connector and would work with any video monitor (or TV).
Now they have Ethernet, each with its own encoding format, and a long list of secondary standards that the Ethernet switch must comply with in order for the whole system to work.
In general, if you want your equipment to be more informative, more modern, you will have to use more informed, more modern personnel both during installation and commissioning, and during operation and maintenance of this equipment. This cart can be repaired in any village. The carburetor can be cleaned in any garage.
Do you still want to drive a modern car with adaptive suspension and intelligent all-wheel drive?
It will have to be serviced not by Uncle Vasya's neighbor. The second main disadvantage of addressable systems is also the reverse side of their advantages.
Addressable systems, as a rule, use only one pair of wires for both power supply and data transmission. This is convenient for installation. This saves wires. But all devices are powered from one source in the central device.
Limited source. Even if you want to install an additional source, it is most likely impossible.
In addition, the circuitry that allows combining power and information inevitably imposes its own limitations. In most systems, the maximum power consumption of all devices on one loop should not exceed several watts.
Actually, this is not a small amount.
This is enough for a thousand fire alarms. But for some other devices, this is catastrophically little. A siren requires at least half a watt.
Each.
If there are many sirens in the system, they should not work simultaneously. Or they should have the ability to connect additional power. It is clear that the access control system locks in any of these systems cannot be powered directly from the addressable loop.
Worse, there are many indirect consequences of the fact that addressable devices are initially designed for minimal power consumption.
For example, there are almost no addressable expanders that can connect non-addressable detectors powered by a loop.
Most of these expanders operate in pulse mode, only occasionally checking the loop status. Occasionally does not mean that you can manage to disrupt the loop while no one is looking.
No, occasionally means 20 times per second, but for several microseconds. This is unusual for traditional systems.
For example, this imposes certain restrictions on the capacity of the loop, which was never taken into account before.
All this is not bad in itself, but it brings us back to the same idea that addressable systems require more qualified designers, installers, and operation engineers. Gone are the days when you could not think, maybe it will work by itself.
The smarter the hardware, the smarter the people working with it should be.
Finally, let's remember that our regulatory documents supposedly allow installing fewer addressable fire alarms. Nothing of the sort.
Of course, you never need to install four, but to install fewer than three, the alarms must be, I don't remember literally, in general, «of increased reliability and improved quality.»
Does your alarm have a certificate that it is of increased reliability?
No?
You say there is no GOST standard by which quality can be certified as «increased»?
Well, install three detectors. Our GOSTs, codes of practice and NPB were originally designed for the undoubtedly disgusting quality of products, and to this day they do not provide for any possibility that detectors can be reliable.
In words, yes. But in fact, the standards are designed for the system to be built under duress, under duress, from the cheapest equipment that can be found.
People who voluntarily intend to install quality equipment are not taken seriously.
They probably have money to burn – so let them install expensive ones, but a lot of them.
Now let's dispel the second myth – about the high cost of addressable systems.
Of course, when the detectors consisted of simple microcircuits, in order for them to become addressable (not even addressable-analog), it was necessary to add a couple more microcircuits and the accompanying harness.
However, now almost all detectors (regular, non-addressable) have become microprocessor-based.
Instead of two comparators and several triggers, it has long been cheaper to install a single microprocessor. The price of any simple microcircuit (and microprocessors have long been classified as simple) is now determined not by the number of transistors in this microcircuit, but only by the number of gold-plated pins on its case.
Therefore, one microprocessor is cheaper than two or three old simple microcircuits.
And since the detector still has a microprocessor, the difference in circuitry between an addressable and non-addressable device for two-wire fire detectors is simply absent, and for security (and four-wire fire) detectors, the difference is not in favor of the old non-addressable ones.
Do you know what the most expensive thing in a typical detector is?
Do you think it's the microprocessor?
Or a super-sensitive infrared sensor?
Well, you guessed wrong.
The most expensive thing is the output relay, the contacts of which will be connected to the non-addressable loop.
The second most expensive thing is the terminal blocks for connecting the wires. Especially since non-addressable detectors have at least 4 terminal blocks, and another 2 terminal blocks for the tamper sensor.
As for control and monitoring devices, the difference is even more striking. Instead of a huge number of components for monitoring 20 or 24 non-addressable loops, they have only one or two loops, the circuitry of which is practically identical to non-addressable ones.
Why are addressable systems more expensive now? The reasons are purely marketing. Addressable ones are better (at least cooler), which means the price should be higher. The circulation of addressable systems is smaller, which means the price is higher.
Addressable devices are incompatible with similar devices from other manufacturers, which means that competition is weaker and the price is higher.
Addressable devices have been developed recently, a lot of intellectual effort has been invested in them and they are still paying off investments, which means that the price includes depreciation of the development cost. Let the experts in PBU and IFRS not throw slippers at me for this phrase.
Yes, formally, reimbursement of the development cost is not included in the cost price, but in fact it is included in the price — the end consumer always pays for everything.
In general, yes, in fact, addressable devices are now more expensive than similar non-addressable ones. But, firstly, this will quickly pass.
Especially when at least some kind of competitive environment is formed.
And secondly, informed people know that even now addressable devices have a significantly greater reserve for wholesale discounts than non-addressable ones, which have long been crushed by price competition.