Wireless Sensor Supervision, Signal and Battery Life

Why this matters

A wireless door contact that has stopped talking to the panel looks exactly like one that is fine: same white plastic, same magnet, no light. The only thing that tells you the difference is supervision, and the only place supervision shows up is the panel's trouble display and its diagnostics screen. Most repeat service calls on wireless systems are one of three things: a sensor that lost supervision, a sensor whose battery keeps dying early, or a sensor that was marginal on install day and failed the first cold night. All three are visible on the diagnostics screen before they become a trouble beep at 3 a.m., if somebody reads it.

What supervision actually is

A supervised wireless sensor does two kinds of transmitting. When it trips (a door opens, a motion sensor sees heat move), it sends an event. Between events it sends a supervisory check-in on a schedule set by the maker, a short "I am here, my battery is fine, my tamper is closed" packet. The panel keeps a clock for every enrolled sensor. If the clock runs past the supervision window without a check-in, the panel raises a loss of supervision trouble for that zone (some panels call it "check", "missing" or "supervisory").

Three things follow from how that works:

• Supervision catches a dead sensor late, by design. The check-in interval and the window are long compared with an event, because a sensor that transmitted every few seconds would flatten its battery. The interval and window are per the maker and the panel's listing. A sensor that died this morning may not show a trouble until the window expires, so "no trouble on the keypad" does not prove every sensor is alive right now.
• An event resets the clock but does not prove the path is good. A sensor at the edge of range can get one event through on a lucky retry and miss most check-ins. That is why the diagnostics screen's last check-in time matters more than "it beeped when I opened the door."
• A loss of supervision is a trouble, not an alarm, on most programming. Whether it reports to the central station, and whether it blocks arming, is programmed per account. If it blocks arming, the customer will call you. If it does not, the customer may live with a dead zone for months.

What the trouble does not tell you is why. The same trouble comes from a dead battery, a sensor knocked off the wall and thrown in a drawer, a new obstruction in the radio path, or a sensor that was deleted and re-enrolled in the wrong slot. The diagnostics screen narrows it.

Reading the diagnostics screen

Every panel family presents this differently, but nearly all show the same four facts per sensor: zone, signal level, last check-in and battery state. Here is one, taken on a January morning at a two-story house with an attached, unheated garage. Signal on this panel is shown as 0 to 5 bars; other panels show a number or a decibel figure, and the acceptable minimum is per the maker.

WIRELESS DIAGNOSTICS            01/14  07:42
ZN  NAME            SIG 0-5  LAST CHECK-IN  BATT
01  FRONT DOOR         5     01/14 07:31    OK
02  BACK DOOR          4     01/14 06:58    OK
03  SLIDER             4     01/14 07:05    OK
05  GARAGE OH DOOR     1     01/14 05:12    LOW
07  BSMT WINDOW W      -     01/12 22:40    --
09  KITCHEN MOTION     3     01/14 07:20    OK
RF JAM LOG: NONE        TROUBLE: ZN 07 SUPV

Read it top to bottom before touching anything. Zones 01, 02, 03 and 09 checked in within the last hour or so, with three or more bars and a good battery. Leave them alone.

Zone 05 is a sensor in trouble that has not failed yet. One bar, a low battery, and its last check-in was at 05:12, more than two hours older than the latest check-ins on the screen. On a steel overhead door in a garage that was well below freezing overnight, that combination is the common pattern: weak path plus cold battery.

Zone 07 has failed. No signal reading, no battery reading, and the last check-in was 22:40 two days earlier. That is the zone the panel's trouble is for. The dash in the signal column means the panel has nothing current to show, not that the signal is zero.

The jam log is empty, which matters: it rules out a transmitter in the neighborhood stepping on the panel's frequency as the cause of either problem.

What weakens the signal

A wireless sensor's transmitter is low power, and the path from sensor to panel receiver is rarely a straight line through air. The usual losses, roughly from worst to mildest in residential work:

• Metal in the path or on the mount. A steel entry door, a steel overhead garage door, metal studs, a refrigerator, a water heater, ductwork, an electrical panel. Mounting a sensor directly on steel is worse than having steel between the sensor and the receiver, because the metal right behind the antenna detunes it and reflects most of what it radiates.
• Foil-faced insulation and foil vapor barriers. Foil-backed rigid foam and radiant barrier in a finished basement or attic act like a metal sheet. This is the classic cause of a zone that worked for years and failed after a renovation.
• Mirrors and coated glass. Older mirrors are silvered on the back, and low-emissivity window coatings are a thin metal layer. A sensor on the far side of a mirrored closet door or a low-E window from the panel can lose a bar or more.
• Stucco and plaster on metal lath, masonry and concrete. Wire lath behaves like a screen. Poured concrete and block cost more than wood framing, especially below grade.
• Distance. Range is per the maker and is usually stated as open air. Indoors through several walls, real range is a fraction of it.
• Water and people. An aquarium, a water heater, a crowd. Minor for most homes but enough to drop a marginal sensor over the edge.

Signal on install day is the best it will ever be. Furniture moves in, a new appliance lands in the path, foil goes up in a basement. A sensor enrolled at the maker's minimum has no margin for any of that. A sensible shop default is to want at least two bars better than the maker's minimum at every sensor on install day, or to fix the path then (move the panel or receiver, add the maker's repeater, relocate the sensor, or hardwire the opening) rather than ship a marginal zone.

Why batteries die early in cold places

Most supervised wireless sensors run on lithium cells of a type the maker specifies. Lithium chemistry holds charge well on a shelf, but two cold effects shorten its useful life in a sensor:

• Voltage sags under load when the cell is cold. A transmission draws a short pulse of current far larger than the sensor's idle draw. Cold raises the cell's internal resistance, so its voltage dips further during that pulse. The sensor measures battery voltage during or just after a transmission, so a cold cell with plenty of capacity left can report low.
• Retries multiply the load. On a two-way system, a sensor on a weak path does not get its packet acknowledged on the first try and repeats it. Every repeat is another pulse, so there a one-bar sensor transmits more than a five-bar sensor sending the same events. (A one-way sensor sends a fixed burst either way; on those, cold sag is the main effect.)

Put both together on a garage door and you get the pattern in zone 05: a battery replaced in the fall reads low by January, the customer replaces it again, and it reads low again in a few weeks. The battery is not the root cause. The path is.

Use the battery type the maker lists. A cell of the right size and the wrong chemistry can have a different voltage curve, and the sensor's low-battery threshold is set for the listed cell. A wrong cell can read low on day one or, worse, read fine and then drop out without warning.

RF jam detection

A jam is radio energy on the panel's receive frequency that lasts long enough to block sensor transmissions. It can be a malfunctioning device nearby, another system's transmitter stuck on, or deliberate interference. Many panels detect sustained in-band energy and log or report a jam event; whether a jam reports as a trouble or as an alarm, and whether it reports while disarmed, is programmed and per the maker.

Two practical points. First, an empty jam log is useful evidence, as at this house, because it eliminates a whole branch of causes for a supervision loss. Second, a jam that recurs at the same time of day usually points at a device on a schedule (a neighbor's equipment, a utility meter, a gate operator), and the fix is finding the device, not turning off jam reporting.

Worked example: the house with two bad zones

Before touching the system, the account goes on test with the central station, because opening doors and pulling covers on a monitored system sends signals. See related: How to Put an Alarm Account on Test and Take It Off.

Zone 07, the basement window. The customer mentions the basement was finished last summer. The install record shows the zone read three bars when it was enrolled. On site, the sensor is on the window and its battery tests fine on a meter, but tripping it produces no event at the keypad. The new wall between the panel closet and the window is foil-faced rigid foam behind drywall. The path that worked through bare framing now runs through a foil sheet.

The options, in order of what the tech checks first: whether the maker's wireless repeater can sit on the window side of the foil wall with a good path back to the panel; whether the panel's receiver can be relocated; whether the window can be hardwired without opening the new drywall. The tech places a repeater on the window side of the foil wall, re-tests the window, and reads four bars through the repeater. The zone is walk-tested to the central station, and the stale trouble clears.

Zone 05, the garage overhead door. The sensor is screwed directly to the steel door's top panel, and the garage reached single digits Fahrenheit overnight. One bar at the panel. The battery reads low at the sensor after a transmission and reads normal again at room temperature on the bench. That pattern says cold sag on a cell that is working too hard, not a failed cell.

The fix is the path, not the battery. The tech moves the sensor to a non-metallic spacer per the maker's mounting instructions and checks its path both direct and through the repeater, keeping the stronger. Signal comes up to three bars. A fresh cell of the listed type goes in anyway, since the old one has been cycled hard, and the date goes on the account record.

Restoring. Each repaired zone is tripped and confirmed received by the central station operator, the diagnostics screen shows a fresh check-in for both, and only then does the account come off test.

What would change the call

• A sensor in a heated space with a good signal that eats batteries. Cold and retries are ruled out, so look at the sensor itself: a stuck tamper or a contact that chatters as the door rattles produces a stream of transmissions. The event history will show far more activity on that zone than any other.
• A loss of supervision on several sensors at once. One failed path does not take out sensors in different rooms. Suspect the panel's receiver, its antenna, a jam, or a panel that was power-cycled and re-learned incorrectly before blaming every sensor.
• A sensor that checks in fine but misses events. Supervision only proves the check-in path. A motion sensor in a cold-start state or a contact with the magnet too far away (gap per the maker; see related: Door and Window Contact Types and Gap Ratings) can be supervised and still not trip.

How to verify you got it right

Leave the site with each repaired zone showing a check-in time from your visit, a signal level with margin over the maker's minimum, a battery state of OK, and an operator-confirmed signal at the central station. Note the signal reading on the account record; next year's inspection can compare against it and catch a slow decline before it becomes a trouble.

References

• Panel and sensor maker installation instructions: supervision interval, signal minimum, battery type, mounting
• See related: How to Enroll a Wireless Sensor and Confirm Its Signal
• See related: How to Put an Alarm Account on Test and Take It Off
• See related: Door and Window Contact Types and Gap Ratings
• See related: Security + Alarm Systems Reference