The Alarm Service Call SOP

Purpose

Most alarm service calls are decided before the tech reaches the driveway, because the station has already recorded what the panel did: which zone, what event, at what times, and whether it restored. A tech who rolls without reading that history arrives to a quiet panel, wiggles a few contacts and leaves, and the trouble comes back on the next cold night. A tech who works without the account on test sends the station an alarm the operator has to act on. This procedure puts the history first, the account on test before any device is touched, readings on the ticket before and after, and a station-confirmed walk test of everything the repair could have disturbed.

Scope

Covers any scheduled service visit to a monitored or unmonitored intrusion system: intake, history review, the visit, diagnosis, repair, testing and the ticket.

Does not cover night calls handled by phone, which belong to 'The After-Hours Alarm Trouble Call SOP'; the annual inspection, which is 'The Annual Security System Inspection SOP'; or the morning-after review of a dispatched alarm, which is 'The False Alarm Follow-Up SOP' and usually books the visit this procedure then runs. Any fire device on a combination panel is touched only under 'The Fire-Side Work and License Routing SOP'.

Roles and responsibilities

Procedure

1. Take the complaint in the customer's words. The office records what the keypad says, what the customer heard, when it started, whether it happens at a time of day or in weather, and whether anything changed at the house recently (painting, new windows, a new router, a carrier letter). Acceptance: the keypad text copied exactly, a start date, and the answer to "what changed". Wrong looks like "alarm acting up"; the stop rule is the office calls back for the keypad text before scheduling. Hazard: none, a phone call. If the complaint involves a smoke, heat or CO device, the ticket is routed under the fire-side SOP before it is scheduled.

2. Read the signal history before rolling. The tech reads at least 30 days of history (the shop default; 90 days for a problem the customer says is seasonal) per Reading Event Codes on a Signal History, and writes a working hypothesis on the ticket: which zone or component, what pattern, what part might be needed. Acceptance: a written hypothesis and the parts it implies on the truck. Wrong looks like a history showing troubles on 3 zones at once, which points at a shared cable, the panel or the power supply rather than 3 devices; the stop rule is the hypothesis is rewritten to the shared cause before parts are pulled. Hazard: none, desk work.

3. Put the account on test on arrival. Call the station, request a test window (the shop default is 2 hours for a service call), and record the confirmation number and operator ID per How to Put an Alarm Account on Test and Take It Off. Acceptance: confirmation number on the ticket before any device, panel door or keypad is touched. Wrong looks like an alarm memory already on the keypad from the customer trying things; the stop rule is the tech reads and records the memory, then clears it with the customer's code, before anything else. Hazard: none, a phone call.

4. Confirm the complaint on site. Read the keypad's trouble and alarm memory and the panel's event log if it keeps one, and compare them with the station history. Acceptance: the complaint either seen on site or matched in the panel's own log, written on the ticket. Wrong looks like nothing in either log; the stop rule is the tech does not replace parts on a symptom that cannot be found, records the readings as a baseline, and sets a follow-up for the next event. Hazard: none, keypad work.

5. Diagnose from the history and from readings. Take the readings the hypothesis calls for: the loop's resistance at the panel against the EOL value in the panel's documentation, with the loop's devices closed and then opened (see related: End-of-Line Resistor Supervision: Open, Short and Normal); AC voltage at the panel's AC terminals; battery voltage; wireless signal and battery status on the panel's diagnostics. Acceptance: one reading that explains the history, written on the ticket with its expected value beside it. Wrong looks like a reading that wanders as a wire is moved; that is the fault, and the tech traces it by section before cutting anything. Hazard: the meter goes on the low-voltage side only; the transformer is unplugged from its receptacle, never opened, and any 120 V fault at the receptacle goes to an electrician where local rules require it.

6. Repair the cause. Replace, realign, re-terminate or rewire what the reading found. Acceptance: the reading from step 5 now matches its expected value, written beside the before value. Wrong looks like a part on the counter and no change in the reading; the stop rule is the old part goes back in and the diagnosis is reopened rather than a second part being tried. Hazard: drilling a wall to re-run cable is done after scanning the stud bay for wiring and pipe; ladders are set on level floor; the battery is disconnected with the transformer unplugged when the board is worked on, reconnected red then black, and the transformer plugged back in until the keypad shows no AC or battery trouble.

7. Walk test every zone the repair could have disturbed. That is the repaired zone plus any zone sharing its cable, its splice, its expander, or its wall cavity, each tripped with the operator reading back zone number, description and event, per How to Walk Test Every Zone with the Central Station. Acceptance: each received correctly and restored. Wrong looks like a neighboring zone that now reports a trouble; the stop rule is the tech fixes it before the account comes off test, because the repair caused it. Hazard: the same ladder and drilling rules as step 6; a siren that sounds during testing is silenced with the customer's code, and the household is warned first.

8. Take the account off test after a confirmed signal. Send one final signal, have the operator confirm it, take the account off test, record the operator ID. Acceptance: confirmed signal and live account. Wrong looks like an operator who does not see the signal; the stop rule is the communicator is checked before the tech leaves. Hazard: none, a phone call.

9. Close the ticket with readings and the cause. The ticket carries the hypothesis, the before and after readings, the cause in one sentence, the zones tested, and any quote for work not done. Acceptance: a second tech could read the ticket and repeat the test. Wrong looks like "fixed zone 7"; the stop rule is the office returns the ticket to the tech. Hazard: none, paperwork.

When the part is not on the truck. If the cause is found but cannot be repaired, the tech may bypass the faulted zone only with the customer's signed consent on the ticket, tells the customer which opening is unprotected, and books a return. A fire zone is never bypassed by the tech or by the customer.

The record this produces

The ticket: complaint in the customer's words, history window read, hypothesis, both confirmation numbers, before and after readings with expected values, cause, zones tested with the station's read-backs, and any bypass with the customer's consent. The office reads it to close and invoice; the next tech reads it before the next visit; it answers "what did you do" when the trouble returns.

A worked pass

Step 1. Keypad text "Zone 7 Trouble", basement window, started 3 weeks ago, mostly at night. The house had new gutters installed last month.

Step 2. Thirty days of history show zone 7 trouble and restore 9 times, all between 22:00 and 06:00, nothing on other zones. Hypothesis: an intermittent open on zone 7's cable, possibly disturbed by the gutter work where the cable rises outside the rim joist. No parts beyond a splice kit and contact. Why not the contact first: a contact that has drifted out of its gap fails the same way every time the window is opened, and the history shows no openings at all, only troubles with the house asleep and the outside air at its coldest, which is what a conductor nicked and pulled tight outdoors does as the cable contracts.

Steps 3 and 4. Account on test at 09:20, 2-hour window. The keypad's trouble memory shows zone 7 with a restore at 05:40, matching the last event in the station history, so the complaint is confirmed in the panel's own log even though the panel is quiet on arrival.

Step 5. At the panel, zone 7 reads the EOL value from the panel's documentation with the window closed, and goes open when the cable is flexed at the rim joist. Traced by section, the damage is a gutter strap screw driven through the cable.

Step 6. New splice made inside the basement with a gel-filled connector, cable re-routed clear of the strap. Zone 7 reads the EOL value, steady under flexing.

Step 7 fails its acceptance. Zone 7 is received and restored. Zone 8 (bedroom window) shares the same 4-conductor cable from the basement, and when tripped the operator reads back "zone 8 trouble" instead of alarm. The splice had landed one zone 8 conductor in the wrong connector. Stop rule taken: the account stays on test, the splice is redone, zone 8 reads its EOL value, and both zones re-test: zone 7 alarm and restore, zone 8 alarm and restore.

Steps 8 and 9. Final signal confirmed, account off test. Ticket cause: "Gutter strap screw through zone 7/8 cable at rim joist; spliced inside, re-routed."

References

• See related: Reading Event Codes on a Signal History
• See related: End-of-Line Resistor Supervision: Open, Short and Normal
• See related: How to Walk Test Every Zone with the Central Station
• See related: How to Put an Alarm Account on Test and Take It Off