The Intermittent Open: The Hardest Fault Decision Tree

Why this matters

An intermittent open is the fault that works fine while you are standing there and fails the moment you leave. The connection is good sometimes and broken other times - a wire that breaks contact when it gets hot, a terminal that lets go under vibration, a crack that opens when a panel is closed. It is the hardest fault to catch because the usual test, "does it have continuity right now," reads good during the window it is behaving. The skill is forcing the fault to show itself instead of waiting for luck.

Safety: plan your tests around the system state. Dead-circuit continuity checks are verified dead first; live monitoring is done with proper PPE and a proven meter.

Start here: catch it in the act

You cannot fix what you cannot see fail. The whole game is reproducing the failure while you are watching. Before deep diagnosis, ask what condition the fault correlates with - heat, vibration, time, position, load, weather - because that is your lever to make it fail on command.

Step 1: Find the trigger

Intermittents almost always have a trigger. Identify it and you can summon the fault.

• Temperature. Fails after it warms up, or only when cold. Metals expand and contract; a marginal joint opens at one end of the range. A heat gun or freeze spray on suspect connections can trip it deliberately.
• Vibration or movement. Fails when the equipment runs, a door slams, or a panel is moved. Wiggle-test connectors and harnesses by hand while monitoring.
• Position. Fails only with a cover on, a wire bundled a certain way, or a part in a certain spot. The act of buttoning it up flexes the bad spot.
• Load. Fails only under full current. A connection with high resistance passes a tiny meter current fine but opens or overheats under real load.
• Moisture/weather. Fails when damp. Corrosion bridges or breaks depending on humidity.

Pin the trigger from the customer's story and your own observation, then use it.

Step 2: Monitor instead of spot-check

A single continuity reading is useless on an intermittent because it samples one instant. Watch the path over time while you provoke the trigger.

1. Put a meter across the suspect section in a mode that will flag a momentary break (min/max capture or audible continuity that breaks when the path opens).
2. Apply the trigger - heat it, wiggle it, load it, run it.
3. Watch for the dropout. The instant the reading jumps to open or the continuity tone stops, you have caught it and narrowed the location.

If you have only a basic meter, a helper watching the load while you flex sections by hand works too. The point is continuous observation during provocation, not a one-shot reading.

Step 3: Wiggle and isolate by section

Once it drops out under provocation, localize it.

• Halve the run. Move your test points to the middle of the suspect section and re-provoke. The fault is in the half that still drops out. Repeat to close in.
• Wiggle test connections one at a time while monitoring. A connection that toggles the fault when you flex it is the suspect. Terminals, splices, plug connectors, and crimps are the usual offenders.
• Inspect what you find. Look for discolored or loose terminals, green corrosion, a wire that flexes too easily near a crimp (broken strands inside intact insulation), or a melted connector. A loose lug that arcs under load shows heat damage.

Step 4: Suspect the usual offenders

Intermittent opens cluster at predictable places. When the section is narrowed, check these first.

• Crimps and terminals that were never fully seated or have corroded.
• Plug-in connectors with spread or burnt pins, especially in vibration or heat.
• Stranded wire near a stress point where strands fatigue and break inside the insulation - looks perfect, reads open when flexed.
• Spade and screw terminals loosened by thermal cycling.
• A component with an internal intermittent (a switch, relay contact, or board joint) - rarer, but real. A cold solder joint on a board flexes open with temperature.

Step 5: Fix it properly and prove it

A reseated connection that you cannot make fail again is the candidate, but verify before you call it.

1. Remake the connection right: fresh crimp or terminal, clean contact, correct torque. Do not just nudge it.
2. Re-run the provocation that caused the dropout - the same heat, wiggle, or load. If you cannot make it fail again after the repair, you have it. If it still drops out, you have not found the only spot; keep isolating.
3. Cycle the system through a full operating run before you leave, ideally through the condition that triggered the original complaint.

The ordered path

1. Identify the trigger (heat, vibration, position, load, moisture).
2. Monitor the path continuously while applying that trigger.
3. Halve and wiggle to localize the dropout.
4. Inspect the usual offenders in the narrowed section.
5. Remake it properly and re-provoke to confirm it is truly fixed.

Do not certify an intermittent by a single good reading. Certify it by failing to make it fail.

References

• NFPA 70 (National Electrical Code) on connections and terminations
• Manufacturer wiring documentation and connector specifications
• OSHA electrical safety and lockout/tagout standards
• Trade-standard intermittent-fault and connection-diagnostic practice
• See related: Open vs Short vs Ground (The Three Faults), Friction as a Diagnostic Clue