Heat-Soak Faults: Decision Tree

Why this matters

A heat-soak fault is the most frustrating kind: the thing works, runs for a while, then quits, and by the time you arrive it has cooled off and works fine again. "Heat soak" means heat that builds up and saturates a component or an enclosure over time, often after the equipment stops running and the cooling stops too, while the residual heat keeps climbing. The part tests good cold because it is good cold. The only way to catch it is to recognize the timing signature and force the fault to show itself at temperature.

Start here: confirm the timing signature

Heat-soak faults have a fingerprint. Before you chase anything, confirm the pattern.

• If the unit runs fine, then fails after a predictable warm-up or run interval, and recovers after it sits and cools, you are in heat-soak territory. Note the time-to-fault.
• If the fault is worst right after the unit shuts down and sits (not while running), that is a textbook heat-soak signature: with the equipment off, fans and flow stop, so trapped heat rises and saturates nearby parts.
• If the fault is present cold and does not change with heat, this is not heat soak. Look elsewhere.
• If the fault improves as things warm, suspect the opposite (a cold-tight or condensation problem), not heat soak.

Same time-to-fault every cycle is the strongest confirmation.

Branch 1: electronic and control components

Heat is the number-one killer of marginal electronics, and they fail in temperature-dependent ways.

• If a control board, sensor, or relay drops out only when hot and recovers cold, suspect a heat-sensitive component or a connection that opens with expansion. Carefully warm the suspect area (per safe practice) and watch for the fault to appear.
• If a thermal protective device (a limit, an overload, a thermal cutout) is opening, decide first whether it is doing its job. A real overheat condition will trip it correctly. Do not jumper or bypass a protective device to keep the unit running; that defeats the safety and can cause a fire or a burnout.
• If the surrounding enclosure is running far hotter than it should, the root cause may be lost cooling: a failed fan, a blocked vent, or packed-in debris letting heat soak build. Fix the cooling, not just the part that quit.

Branch 2: motors and windings

Motors generate their own heat and protect themselves thermally.

• If a motor runs, then cuts out after a period and restarts once cool, its internal thermal protector is tripping. Ask why it is overheating: an overload, a failing bearing adding drag, low voltage forcing high current, or lost ventilation.
• If the motor is hot to the touch well beyond normal, treat the overheat as real. Find the load or supply problem behind it. Repeated thermal trips cook the windings over time.

Branch 3: fluid and pressure side

Heat changes how fluids behave, and heat soak after shutdown can push a system past a limit.

• If pressure climbs after the unit stops (residual heat with no flow to carry it away), trapped-heat expansion is pressurizing a closed section. (See related: Thermal Expansion as a Fault Cause.)
• If a pump starts rattling once the fluid is hot, hot liquid cavitates more easily because its vapor pressure rose. That is heat soak showing up as cavitation. (See related: Cavitation and What It Tells You.)
• If a hot restart is hard or impossible but a cold start is fine, vapor lock or a heat-affected component in the start path is likely.

Forcing the fault to appear

You cannot diagnose what you cannot see, so reproduce the heat.

1. Run the equipment through a full cycle and let it reach the temperature where it normally fails. Stay with it.
2. If the fault appears, take your readings immediately, while it is hot, before it cools and "heals."
3. If it fails most after shutdown, monitor through the shutdown-and-soak window, not just while running.
4. Carefully and locally adding heat to a single suspect part (within safe limits) can isolate which component is heat-sensitive when full-cycle reproduction is too slow.
5. Let it cool, confirm the fault is gone cold, then repeat to prove the timing is consistent.

The discipline

The fatal mistake is testing cold, finding nothing, and swapping the most expensive part on a hunch. Heat-soak faults only confess at temperature. Make the equipment hot, keep your meter on it, and read the fault while it is happening. The branch you land in tells you whether the cure is a part, a connection, lost cooling, or a load problem feeding the heat.

References

• Manufacturer thermal-limit and operating-temperature specifications
• Trade-standard practice for diagnosing intermittent and temperature-dependent faults
• NFPA and OSHA guidance on not defeating thermal protective devices
• See related: Thermal Expansion as a Fault Cause: Decision Tree; Cavitation and What It Tells You