A New Signal Goes to Flash After the First Rain
A new span-wire intersection was energized on Monday. By Thursday's rainstorm the signal had dropped to flashing red on the arterial and flashing yellow on the side street. The MMU log shows a conflict on the phase serving the arterial through movement, time-stamped during the storm. The cabinet is a NEMA 332 with a pad-mounted ground rod; a fall-of-potential test reads 40 Ω. The inspector also measured one overhead signal head at 14 ft above the crown, and found a nick in the insulation of a signal conductor inside the span-wire messenger where it passes a thimble. The contractor wants to clear the MMU and return the intersection to stop-and-go.
1. What is the likely cause of the intermittent flash, and what is your diagnostic sequence?
Read the worked answer
The pattern - a conflict that appears only during rain - points to a moisture-driven fault path, and the nicked conductor at the thimble is the prime suspect. Wet insulation at a sharp metal edge can leak current between conductors or to the messenger, producing a false indication that the MMU reads as a conflict.
Sequence:
- Read and record the MMU log - channels, event type, and time stamp - before clearing anything.
- Inspect for water: cabinet interior, door gasket, conduit entries, head lenses and housings, and splices in the span.
- Insulation-resistance test the suspect conductors (megohmmeter) with the circuit isolated. A wet chafed conductor will read far below a healthy value.
- Check load switches and the conflict-monitor card for moisture or corrosion.
- Repair the fault - replace the damaged conductor section, add a proper thimble/bushing, re-seal entries.
- Field-verify every indication through the full phase sequence from the ground, then clear the MMU and monitor through at least one full cycle.
Do not clear the MMU first. Clearing returns conflicting greens to service with the fault still present.
2. Is a 40 Ω ground reading acceptable? What do you require before acceptance?
Read the worked answer
No, 40 Ω is high. The NEC baseline is 25 Ω or less, or a supplemental electrode; many agencies specify 10 Ω or tighter for signal cabinets. A 40 Ω reading indicates either a poor electrode, a bad connection, or dry/high-resistivity soil.
Before acceptance:
- Inspect the connection at the rod and the cabinet ground bus; a loose clamp can account for much of the reading.
- Add electrodes (additional rods spaced at least a rod-length apart, or a chemically enhanced/ground-plate system) until the reading is within spec.
- Bond the messenger and poles to the ground system; verify continuity end to end.
- Re-test with fall-of-potential or clamp-on and record the value.
Grounding matters here beyond code compliance: a well-grounded system gives surge current a path and reduces the chance that a chafed conductor energizes the messenger.
3. The overhead head measures 14 ft above the crown. What is the requirement, and what is the remedy?
Read the worked answer
Overhead signal faces must be mounted with the bottom of the housing at least 15 ft above the crown of the roadway at the center of the controlled lane (and generally not more than about 19 ft). 14 ft is deficient - it reduces clearance for trucks and tall loads, and it places the head where a high vehicle can strike it.
Remedy:
- Re-tension or raise the messenger to bring the head above 15 ft, checking that the change does not push other heads above the 19-ft maximum or alter their lateral placement.
- If the span cannot be raised, relocate the head to a different attachment point or a pole-mounted arrangement that meets height and lateral placement requirements.
- Re-measure and document every head on the span, not just the one that failed - installation errors usually repeat.
- Confirm the change did not compromise messenger loading or the conductor at the thimble, which is the same area already under repair.
Acceptance should be held until both the electrical fault and the clearance deficiency are corrected and re-inspected.