6.2 Loop Installation, Lead-In, Splices, and Electrical Testing

6.2 Loop Installation, Lead-In, Splices, and Electrical Testing

A loop lives in a sawcut under traffic, in water, in freeze-thaw cycles, and under the plow. It is the least protected component in the entire signal system, and it fails more often than everything in the cabinet combined. The good news for a technician is that a loop's condition is entirely measurable — three instrument readings will tell you whether the fault is in the loop, the lead-in, the splice, or the electronics unit.

---

Installation Requirements That Determine Service Life

The Sawcut

The sawcut must be deep enough to keep the wire below the pavement surface and the traffic wear zone, with the corners either drilled or cut at an angle so the wire is not forced around a sharp 90-degree edge. A wire pulled tight around a square corner works against that corner under every axle until the insulation is cut through.

After the wire is placed, the slot is filled with a sealant appropriate to the pavement type and climate. The sealant is not cosmetic — it excludes water, which is the primary agent of loop failure.

Loop Wire

The loop and its lead-in are formed from a continuous, unspliced run of wire from one end of the lead-in, around the loop the required number of turns, and back. A splice inside the sawcut is a guaranteed future failure.

The Lead-In from Loop to Pull Box

• The lead-in sawcut should be at least 12 inches from adjacent loop edges so it does not couple into a neighboring loop.
• The two conductors must be twisted a minimum of 3 to 5 turns per foot. This is not a workmanship preference — twisting cancels the magnetic field of the two conductors, so the lead-in contributes no detection area of its own. An untwisted lead-in becomes a long, uncontrolled sensing zone that detects vehicles anywhere along its path.
• Loops should be labeled at the pull box with the loop number and the start and finish ends, so that series splicing preserves the correct polarity.

Splices in the Pull Box

The splice from loop lead-in wire to home-run cable is made in the pull box and must be soldered, insulated, and waterproofed. Wire nuts and crimp connectors are not acceptable: a mechanical connection in a pull box that fills with water corrodes into a high-resistance joint, which shows up as an intermittent detector that works in dry weather and fails after rain. Heat-shrink with adhesive lining, or an epoxy-filled splice kit, provides the waterproofing.

The Home-Run Cable

The cable from pull box to cabinet should be twisted, shielded, and waterproofed, with a polyethylene jacket. The shield is grounded at one end only — at the cabinet — and left insulated and floating at the pull box. Grounding a shield at both ends creates a ground loop that injects noise directly into the detection circuit. This one detail causes an enormous number of unexplained false calls.

---

The Three Field Measurements

Disconnect the loop pair from the detector card terminals before testing. Every measurement below assumes the loop is isolated from the electronics unit.

1. Series Resistance — Digital Multimeter, Ohms

Measure across the two loop conductors.

2. Insulation Resistance — Megohmmeter at 500 V DC

Measure from each loop conductor to a reliable earth ground (cabinet ground, a street-light pole, a hydrant).

This is the single most diagnostic loop test. A loop that measures fine on resistance but fails insulation resistance is a loop that will work today and fail with the next rain.

3. Inductance — Loop Tester or LCR Meter

Measure at the pull box to check the loop alone, then at the cabinet to include the home-run cable.

• Typical total system inductance falls in the 20 to 300 µH range depending on loop size, turns, and cable length.
• Compare the reading against the design value computed from L = K × feet of sawcut. A mismatch means the turns count is wrong.
• Measuring first at the pull box and then at the cabinet isolates the fault: if the loop reads correctly at the pull box but wrong at the cabinet, the problem is in the home-run cable or the splice, not the loop.

Many loop testers also report a quality factor (Q). A Q greater than 5 is acceptable with modern electronics units; a Q below 5 with a good insulation reading points to excessive series resistance.

---

Testing Before the Sealant Goes In

The order matters and it is examinable. Before filling the sawcut:

1. Test continuity with an ohmmeter to confirm an unbroken circuit.
2. Test insulation resistance with a 500 V DC megohmmeter at the pull box.
3. Test inductance with a direct-reading meter to confirm the number of turns actually installed.

Once sealant is placed, a defect can only be corrected by cutting the pavement again. A technician who signs off a loop installation without these three readings has accepted a repair he cannot see.

---

Replacing a Failed Loop

When a loop is beyond repair, the standard practice is not to reuse the sawcut:

• Install the new loop offset about 6 inches from the old loop. Cutting a new slot in the exact path of the old one runs into the same failed pavement that killed the first loop.
• Cut each abandoned loop in at least two places. An intact abandoned loop remains a closed conductive turn sitting inside the new loop's field, where it acts as a shorted transformer secondary and degrades the new loop's sensitivity. Two cuts guarantee it cannot form a closed circuit.
• Replace the lead-in with the loop if the fault was in the lead-in or if its insulation reading was marginal. Reusing a suspect lead-in on a new loop reproduces the original failure.

Interim Measures

While a loop is out of service, the phase must still be served. Maximum recall on the affected phase is the correct temporary action — it guarantees service at the cost of running that phase every cycle. Placing the affected phase on recall must be entered in the maintenance log with the reason and the date, and it must be removed when the loop is restored. An intersection quietly left on recall for a year is a permanent capacity loss that no one is tracking.