How to Wire a Hardwired Zone With the Resistor at the Last Device

Why this matters

A hardwired zone that tests fine at the keypad can still be wired so it would miss a real intrusion. Put the end-of-line resistor at the panel instead of out at the last device, and a jumper or a staple across the pair anywhere in the walls looks exactly like a closed, normal zone. Splice a contact onto the wrong leg, and opening that door does nothing at all. Neither shows up on a quick door-open test at the keypad, because the door you test happens to be on the right leg. This card wires a series zone of normally closed contacts with the resistor where it belongs and proves it with a meter at the panel end, before the walls close and before the zone is landed. Why the resistor's position matters is owned elsewhere; in one clause, a resistor at the far end lets the panel tell a cut wire and a shorted pair from a normal loop. See related: End-of-Line Resistor Supervision: Open, Short and Normal.

The shape of the loop

Normally closed contacts go in series on one conductor. The other conductor runs through every device unbroken. At the last device, the resistor bridges the two conductors, closing the loop. In the normal state the panel sees the resistor. Open any contact and the loop goes open. Short the pair anywhere along the run and the panel sees near zero, which it reports as a short.

 PANEL             D1 door     D2 door     W1 window
                   contact     contact     contact
 Z3  --- red ----[ NC ]--red--[ NC ]--red--[ NC ]--+
                                                   |
                                          [ EOL resistor,
                                            inside W1 ]
                                                   |
 COM --- black --(splice)----(splice)--------------+
          40 ft         30 ft        25 ft

 Black passes through D1 and D2 spliced, no device on it.
 Cable: two-conductor 22 AWG, 95 ft panel to W1.

Read the drawing as one path: from Z3 out on red through D1, D2 and W1's contacts, through the resistor, and back on black to COM. Every contact is in that path once. The resistor is the last thing before the return.

Start state

You need: two-conductor alarm cable (22 AWG is common for contact loops; see related: Alarm Cable Types and Which Run Gets Which), the contacts, the resistor value and type the panel's documentation specifies, insulated crimp or gel-filled splices, a multimeter with a resistance range, and the panel's manual open to the zone wiring page.

Check before you begin: the resistor value. It is per the panel's documentation, never a number carried over from a different panel. Different makers specify different values, and some panels support single or double end-of-line on a programmable basis. A resistor of the wrong value reads as a fault or, on some panels, an alarm. The panel in this card's example specifies 2.2 kilohms.

The steps, in physical order

1. Run the cable as a daisy chain, device to device, not a star. Panel to the first device, first device to the second, and so on to the last. A star (a separate run from the panel to each device, joined in the can) also works electrically if wired in series at the panel, but it puts the "last device" back at the panel, which defeats the point. Hazard: drilling top plates and pulling through attics: check the wall cavity for wiring and plumbing before drilling, work from the joists in an attic, not the drywall between them, and limit attic time in heat. Where you drill through a top plate that serves as fireblocking, seal the hole with a fireblocking material as the residential code your jurisdiction adopted requires (IRC R302.11 in the model code).

2. At each intermediate device, break the red, splice the black through. Red from the previous run to one contact terminal, the other contact terminal to red on the next run. Both blacks spliced together and tucked in, not landed on the contact. Why: every contact has to interrupt the one path the current takes. Two mistakes defeat that and both happen in cramped door frames. Splice red straight through and land the contact on the two black ends, and the contact sits across a piece of wire that is already joined: opening it changes nothing. Land the contact across red and black, and it shorts the pair whenever the door is shut, which the panel reads as a short, not as normal. Keeping one rule, red is the switched leg on every device, makes the next tech's job readable too.

3. At the last device, land the resistor inside the device housing. Red to one contact terminal, the other contact terminal to one leg of the resistor, the other resistor leg to black. Inside the contact's housing or the junction behind it, not in a splice buried in the wall cavity, so the next tech can find it. Why: a resistor inside the last device means a short anywhere between the panel and that device bypasses the resistor and reads as near zero. A resistor at the panel means the same short reads as normal.

4. Before you land anything at the panel, read the loop at the panel end with a meter. The zone wires are free in the can, the walls still open where you fished, the device covers still off. Readings are listed under the worked example. Measure resistance only on the lifted pair: a zone landed on a powered panel carries the panel's loop voltage, and a resistance reading on a powered circuit is wrong and can damage the meter.

5. Open each device in turn and watch the meter. Open D1: open circuit. Close it: back to the resistor. Then D2, then W1. Why: this is the test that catches a device spliced past instead of through. Opening a door at the keypad only proves the doors you happened to try.

6. Check each conductor to earth ground. Red to the panel's ground terminal or a known ground: open circuit. Black to ground: open circuit. Why: a staple through the jacket into a metal stud, or a pinched cable at a metal box, grounds a conductor. Many panels flag a ground fault; some just misbehave intermittently in wet weather.

7. Only then close up: patch, insulate, snap covers, and land the pair on Z3 and COM. Program the zone type and name, and confirm the keypad shows the zone ready.

8. Put the account on test and walk the zone with the central station. See related: How to Put an Alarm Account on Test and Take It Off.

Worked example: two doors and a window on Z3

The zone runs from the panel to the front door (D1), the side door (D2), and a dining room window (W1), 40 ft, 30 ft and 25 ft of 22 AWG, 95 ft of cable in all. The panel's documentation specifies a 2.2 kilohm end-of-line resistor, and the resistor fitted is a 5 percent tolerance part.

What the meter should read, worked out before reading it:

  • Cable resistance: 95 ft of cable is 190 ft of conductor in the loop (out on red, back on black). 22 AWG copper runs about 16.1 ohms per 1,000 ft, so 190 / 1,000 x 16.1 = 3.1 ohms.
  • Resistor band: 2,200 ohms plus or minus 5 percent, 2,090 to 2,310 ohms.
  • Expected normal reading: the resistor's actual value plus 3.1 ohms. The wire adds about one seventh of a percent, less than the resistor's own tolerance, so the reading is effectively the resistor.

What the meter read:

Test Reading Meaning
All closed 2.18 kilohms Inside 2,090 to 2,310, normal
D1 opened OL D1 is in the loop
D2 opened OL D2 is in the loop
W1 opened OL W1 is in the loop
Red to ground OL No ground fault
Black to ground OL No ground fault

The finish test is the first row: the closed loop reads within the resistor's tolerance band of the value the panel's documentation specifies. Every other row proves a device or the insulation, not the resistor.

The two readings that would have meant a wiring fault:

  • About 3 ohms with everything closed. The resistor is not in the field loop. Either it was left at the panel, or it was landed across the pair somewhere ahead of a device so it is not the last thing on the loop. Find it before closing up.
  • 2.18 kilohms with D2 opened. D2 is not in the loop: red was spliced straight through and the contact landed on joined wires. A door that opens without faulting the zone is a door the system does not protect.

What changes the wiring

Normally open devices on the same zone. Some devices close on alarm (certain glass-breaks, panic buttons, older heat detectors on burglary zones). Those go in parallel across the pair, not in series, and mixing them on one single end-of-line zone needs care. See related: Normally Closed vs Normally Open Devices on One Zone.

Tamper supervision on the zone. Panels that support double end-of-line add a second resistor across the alarm contact, so the panel reads a tamper, an alarm and a normal loop as three different values. The resistor placement changes with it. See related: Double End-of-Line and Tamper on One Zone.

A takeover with resistors at the panel. Common on older installs. Moving them to the field means finding the last device, which may be hidden in a finished wall. Where that is not practical, record that the zone is unsupervised against a short, tell the customer what it means, and note it on the account.

Long or mixed runs. On very long zone runs or thinner conductors, wire resistance stops being negligible against the panel's loop window. Recompute with the actual length and gauge and check the panel's maximum loop resistance.

References

  • The panel manufacturer's installation manual: end-of-line resistor value and type, zone wiring and maximum loop resistance
  • International Residential Code R302.11, fireblocking, as adopted by your jurisdiction
  • See related: End-of-Line Resistor Supervision: Open, Short and Normal
  • See related: Alarm Cable Types and Which Run Gets Which
  • See related: Normally Closed vs Normally Open Devices on One Zone