Intermittent Control Fault Decision Tree
Why this matters
Intermittent faults are the ones that humble good techs. The equipment works fine while you are standing there, then quits the moment you leave, and the customer is convinced you did nothing. A fault that comes and goes is almost always a connection, a thermal effect, or a marginal component that only fails under a condition you have not reproduced. This tree teaches how to chase a problem that will not hold still, across control boards, sensors, and switching in any trade. The discipline is the same whether it is an HVAC board, a pool controller, or an appliance.
Start here: capture the conditions
You cannot fix what you cannot reproduce, so your first job is to learn when it fails.
- Ask exactly when it acts up. Time of day, after running a while, when it is hot or cold out, when other equipment cycles, after a vibration or a slam. The pattern is the diagnosis.
- Look for a trigger: temperature, time-in-operation, load, weather, or a specific sequence. Intermittents almost always have one.
- Get a fault history if the control logs one. A code that appears and self-clears tells you what the control saw at the moment of failure.
A truly random fault is rare. Find the condition and you can force the failure.
The number one suspect: connections
The vast majority of intermittent control faults are bad connections, not bad parts. A part either works or it does not; a marginal connection works until something disturbs it.
- Inspect and reseat every connection in the suspect circuit - plugs, terminals, wire nuts, spade connectors, ribbon cables. Loose, corroded, or oxidized contacts make and break with heat and vibration.
- Wiggle test. With the system running (safely), gently move connectors and harnesses. If the fault appears when you move a connection, you found it.
- Look for heat damage. Discolored, melted, or pitted terminals carry resistance that comes and goes with temperature. Repair the connection, do not just push it back together.
- Check the ground and common. An intermittent ground or common gives wildly erratic behavior that looks like a failing board.
If it fails when hot (after running a while)
Heat-related intermittents point to thermal expansion or marginal components.
- A connection that opens as it warms - metal expands and a marginal joint loses contact. Reseat or replace the connector.
- A component that drifts or fails hot - a capacitor, relay, or solid-state device near its limit works cold and quits hot. If the fault tracks temperature and the connections are clean, a heat-sensitive component is the candidate.
- A safety opening on real heat - confirm it is not a genuine overheat (airflow, flow, load) before blaming the control.
If it fails when cold or on weather change
- Moisture and condensation bridge or corrode contacts when temperature crosses the dew point. Common on outdoor and pool equipment. Dry, seal, and protect the connections.
- Cold-cracked solder or a cold-contracting connection can open at low temperature.
- A component that will not start cold behaves differently from one that quits hot - note which.
If it fails with vibration or when other equipment cycles
- Vibration intermittents are loose connections or cracked solder, full stop. Find what shakes and what loosens.
- Failing when other loads switch points to a voltage sag, a shared neutral problem, or electrical noise. Measure the supply at the control while the other load cycles.
Diagnostic summary
| Failure trigger | Most likely cause | Action |
|---|---|---|
| Random, with movement | Loose connection | Reseat, wiggle test, repair |
| After running (hot) | Thermal joint or marginal part | Reseat, then suspect heat-sensitive component |
| Cold or wet weather | Moisture, condensation | Dry, seal, protect contacts |
| Vibration / other loads cycle | Loose joint or supply sag | Secure connections, measure supply |
The discipline that wins
Do not throw parts at an intermittent. Replacing a board you have not proven bad is a coin flip, and when the fault returns you have spent money and lost trust. Work connections first, reproduce the condition, and confirm the fault on the failed part before you replace it. An intermittent solved by finding the actual trigger stays solved. One papered over with a parts swap comes back as a callback.
References
- Equipment manufacturer general practice on intermittent-fault and connection diagnosis
- OSHA electrical-safety practice for live-circuit testing
- See related: The Control Isn't Responding Decision Tree
- See related: The Part That Keeps Failing Decision Tree