Cold-Start Faults vs Warm-Start Faults: The Real Difference

Why this matters

"It won't start" means two very different things depending on whether it fails from a dead-cold sit or after it has already been running and warmed up. Techs who do not separate these two questions test the wrong condition, get a clean reading on a unit that is not currently in its failure state, and leave having "found nothing" on a fault that will be back the moment the customer needs it again. Knowing which category you are in before you touch a meter is the single fastest way to stop wasting a visit.

The two families, defined

• A cold-start fault fails, or fails to start at all, from a dead-cold condition, and once it does get going (sometimes on the second or third attempt, sometimes not at all until it warms from another source), it runs fine. The failure lives in the cold state.
• A warm-start fault starts and runs fine cold, then fails, will not restart, or misbehaves once it has run for a while and built up heat, either from its own operation or from ambient conditions. The failure lives in the warm state.

They sound like mirror images and in one sense they are, but the physical causes behind each are largely different, which is why lumping them into one "intermittent" bucket wastes time.

Why cold-start faults happen

Cold conditions change material properties in predictable ways that can prevent a start:

• Fluids and lubricants thicken when cold, increasing resistance to motion right at the moment a motor or mechanism needs the least resistance to get moving.
• Electrical components can need more current to start cold than they need to run once moving (this is normal for many motors), and a marginal component that can supply enough current to sustain a warm run may not supply enough to break static resistance cold.
• Contraction can open a marginal connection. Metal contracts as it cools; a connection that is fractionally loose can lose contact entirely in the cold and reseat once things warm from operation or ambient heat.
• Condensation from a cold-to-warm transition can affect components that are otherwise fine, particularly right after a cold night.

Why warm-start faults happen

Heat changes different properties, and the failure mode is different in kind, not just degree:

• Metal expands when hot, and a marginally loose connection can be tight enough to conduct cold, then expand open and lose contact once warm. This is the "hot intermittent" pattern.
• Capacitors and some electrical components lose capacity as they heat. A weak component can hold enough charge or resistance cold to function, then fade past the working threshold as it warms.
• Bearings, seals, and clearances can close up with thermal expansion, binding a mechanism that turned freely cold.
• A genuine overheating condition (restricted airflow, fouled cooling, an overload) causes a real high-limit or safety to trip after a run, which is not a component fault at all, it is the protection working. Do not replace a safety device that is correctly protecting an overheating unit; find the heat source instead.

The one test that tells you which family you are in

Do not guess from the customer's description alone; confirm the pattern before you commit to either branch. At or below normal ambient starting conditions, does it fail immediately, and does it clear once warmed? That is a cold-start fault. Does it start and run fine cold, then fail only after a run or once genuinely warm? That is warm-start. If it fails in both states, or fails randomly regardless of temperature, you are probably not looking at a thermal fault at all, and should look at load, a marginal component that is simply failing regardless of temperature, or a control fault. See related: Fails Cold, Runs Fine Once Warm (Decision Tree) and Fails Warm, Runs Fine Cold (Decision Tree) for the branch-by-branch walk once you have confirmed the pattern.

The shared discipline: test in the failure window

Whichever family you are in, the same rule applies: you cannot diagnose a state you are not currently in. Take baseline readings in the working state, then get the unit into its actual failure condition (let it sit cold, or let it run to the point of failure) and take readings again immediately. The component whose reading changed between the two states is your fault. Testing only in the state where it works tells you almost nothing.

References

• NFPA 70B, recommended practice for electrical equipment maintenance
• Manufacturer nameplate values for capacitor microfarads and rated current
• See related: Fails Cold, Runs Fine Once Warm (Decision Tree); Fails Warm, Runs Fine Cold (Decision Tree); Why Thermal Expansion Explains More Faults Than Techs Expect