Power’s out.  Laptop’s got a full charge, WiFi and cable modem are on UPS, and even my alarm system’s onboard NiMH battery has a 24-hour charge.  And I’m bored.

Really bored.

So …

How’s it goin?

Gonna talk about alarms.

Alarms are interesting little boxes of electrical wonder.  They’re not unlike octopi.  For more information than you’d ever want on an alarm system, check out homesecuritystore.com, browse their inventory, visit their forums, have fun.

The first alarm systems (mid 1800s) were basically circuits hooked up to a bell.  Door opens, bell sounds.  Door shuts, bell stops.  Very easy to bypass, very easy to kill.  Relays were added around the turn of the 20th century, so the alarm would keep ringing if the door was shut again.  Up until the 40s, alarms were “local only” — meaning the only way help would arrive was if a passerby heard your bell ringing.  The earliest monitored systems required a special connection from the phone company, and business owners would pay a ridiculous amount for the service.

You’ve probably seen old movies where Joe Criminal is breaking into a bank, and bypasses the alarm by cutting the wires.  There are two ways to wire an alarm zone — normally open (NO) or normally closed (NC).  NC means that when a door or window is opened, it breaks the circuit.  NO means the exact opposite - opening a door or window closes the circuit.  In either case, the control panel detects the change and sounds the alarm.  The problem with NO alarms is that cutting the wires to the sensor disables the alarm - it can’t possibly sound.  Most alarms today are NC, meaning that a wire cut will sound the alarm.

Since the early 90s, all alarms have multiple zones.  A “zone” is a way of identifying a sensor (eg, zone #1 = “Front Door”, zone #2 = “Living Room Motion”, Zone #3 = “Smoke Detector”), and in most modern systems, each sensor is in its own zone.  Each zone also has a “zone definition”, which determines what kind of sensor is hooked to the zone.  Zone #1 would be a “perimeter” zone, zone #2 is an “interior” zone, and zone #3 is a “fire” zone.  Zone definitions are then used to determine an appropriate response from the panel.

For example:  Modern alarms have several operating modes.  The most common are “disarmed”, “home”, and “away”.  When disarmed, the system will only respond to certain zones, such as a smoke detector or panic switch.  Everything else gets ignored.  “Away” mode is the exact opposite - every sensor will sound an alarm.  When armed “home”, only the perimeter zones sound the alarm (this allows you to move around in your home without setting off your motion detectors).

Most modern panels are intelligent enough to “adjust” your system based on what you’re doing.  For example, arming “away” will give you 30 seconds to leave the premises.  When you return, the alarm gives you an entry delay of 30 seconds to disarm before sounding.  When arming “home”, obviously you don’t need an entry or exit delay, so the system will immediately trip if a door is opened.

My system, a Visonic Powermax, can get pretty slick.  Say I give myself a 30-second entry delay.  I can configure the panel so that only the front door has the delay; the back door will sound the alarm instantly.  Additionally, I can configure the panel so that I can only move through the living room; going into another room will cancel the delay and sound the alarm.  I can even use a “vacation” mode, where if someone attempts to disarm the system during a specific timeframe (from minutes to months), the alarm will sound, even if they use a valid code.

Most door contacts today are of the magnetic-reed variety.  Basically, a magnet is mounted on the door (or inside the door, as is the case with Frost), and a sensor is mounted on (or inside) the door frame.  Putting the magnet near the sensor pulls a switch closed; moving the magnet away opens the switch and sounds the alarm.  This is the type of switch people talk about when they say, “OMG U CAN H4X0R T3H ALARM BY PUTTING A BUTTERNIFE IN T3H DOORFRAME N00B!!!!!!ZERGRUSH”.  Anyone who believes this is invited to try it at your local bank.

Motion sensors are another matter entirely.  You have four basic types of motion detectors:

PASSIVE INFRARED (aka “PIR”) - these are the most common and least expensive.  They register body heat and respond accordingly.  Analog PIRs are fairly prone to false alarms (leaving baloons up at night, etc).  Modern digital PIRs use ASIC signal processing to virtually eliminate falses.
MICROWAVE - these actively flood the protected area with radar beams, looking for any sign of movement.  Very common in commercial environments, and prone to false alarms.  Unlike PIRs, which register body heat, these trip for ANY motion at all.  Microwave detectors need to be “shaped” to each room, as microwaves can pass through walls.
ULTRASONIC - another kind of active detector, only this time using high-frequency sound waves instead of microwave.  These are rarely used outside of mission-critical environments where microwaves might cause problems, such as hospitals and data centers.
MICROWAVE IMAGING - these haven’t been used much since the 70s.  There are two components - a microwave emitter, and a microwave receiver.  Both are usually placed at a 45- or 90-degree angle from each other.  The emitter “pings” the room several times per second, and the receiver notes the pattern coming back.  Theoretically, since nothing should be moving when you’re away, any change in the pattern = an alarm.  Ridiculously prone to false alarms.

There are also combination detectors that use multiple technologies for maximum resistance to false alarms.  For example, Frost is using tri-tech sensors.  Each motion detector has two PIRs and one microwave built into the unit.  All three must concur before an alarm signal is sent.  As an added bonus, if one of the components is continually reporting a violation, the alarm may still sound if the other two concur that there could *possibly* be an alarm.  This prevents “creeping” to bypass the detector.

Most commercial systems are hardwired, meaning there’s twisted pair running to each sensor, each siren, each keypad, each holdup button, etc.  Most residental systems these days are wireless.  Though you might think a wireless system might be easily compromised, it just ain’t so.  There’s no lame-ass “rolling code” myth at work here.  Each wireless sensor has its own ESN, and there’s a disturbingly complex authentication protocol at work behind the scenes.  In addition, if the receiver detects tampering in the frequency (433mHz), such as interference or malformed packets, it will indicate a trouble status to the monitoring company.  The sensors themselves are supervised, so that conditions such as “low battery” or “tamper” are reported as well.  And if a sensor gets destroyed, the panel will realize that it failed to “check in” and send a trouble signal.

Most panels these days come with a cellular backup, so they can’t be disabled with a simple phone line cut.  Mine uses GSM, but most commercial panels use the control channel of the old analog cellular network.

There are several reporting formats in use today, some dating back to the early 60s.  4/2 Pulse is probably the oldest, and sends data at a painfully slow rate, generally taking several minutes to report a violation.  AdemcoSID and ContactID are the two most common formats, and can report the entire system status in under 20 seconds, including dial time.

Oh hey - power’s back on.

This officially marks the longest blog post I’ve ever done.  Neat.