How Circuit Boards Fail

Why this matters

A control board is the brain that reads inputs, runs logic, and switches outputs. It is usually the most expensive single part in a machine and the one customers most want to blame and most want replaced. That is exactly why you should be slow to condemn one. Most "bad board" calls are actually a bad input feeding the board, a power problem reaching the board, or a connection on the board, not the board itself. A board that gets replaced and fails again the same week was almost never the real fault.

What is actually on the board

A control board is inputs (sensor and switch signals coming in), logic (the chip that decides), outputs (relays and drivers that switch loads), and a power supply that feeds it all. Each section fails differently, and the board's behavior tells you which section. The single most common true board failure is in the power supply or the output relays, not the logic chip.

The failure modes

• Output relay or driver failure. The board commands an output but the load never engages, or it engages and never stops. The on-board relay or switching transistor that drives that load failed. This is the most common real board failure because outputs switch high current and take the most abuse.
• Power supply failure. The board cannot make the low voltage its logic needs. Dried filter capacitors, a failed regulator, or a blown on-board fuse leave the board dead or resetting. A board that reboots randomly is usually a power problem, not a logic problem.
• Cold solder joints and cracked traces. A connection on the board cracked from heat cycling or vibration. The board works when cold, fails when hot, or fails when tapped. Intermittent faults that come and go with temperature or a bump point straight here.
• Burned or scorched components. A visible char mark, a blown component, or a burnt smell. Usually caused by an external event: a shorted load, an over-voltage surge, or water. The burn is the symptom; find what caused it.
• Corrosion. Moisture, condensation, or a leak got onto the board. Green or white crust on traces and connectors. Corrosion creates leakage paths that cause erratic behavior, then eats the copper.
• Surge and lightning damage. A voltage spike punched through. Often the input or power section is hit. Multiple boards or components fried at once usually means a surge event, not a coincidence.
• Connector and pin failure. Not the board itself but its plugs. A backed-out pin, a corroded connector, or a loose ribbon mimics every board symptom there is.

How to test one (and avoid a wrong condemn)

The discipline: prove the board is bad rather than guess.

1. Verify board power. Confirm the board is getting its supply voltage in, and producing its low-voltage logic supply. No input power means the fault is upstream. No output of low voltage with good input means a board power-supply failure.
2. Check inputs before outputs. A board acts on what its sensors and switches tell it. If a safety switch is open or a sensor reads wrong, the board correctly refuses to run. Verify the inputs are sane before blaming the logic.
3. Test the output at the board terminals. When the board calls for a load, measure for the output signal or voltage right at the board. Present at the board but the load is dead means the wiring or load is the problem, not the board. Commanded but absent at the board means a failed output stage.
4. Look and smell. Pull the board and inspect both sides under good light. Scorch marks, bulged capacitors, corrosion, and cracked solder are visible. A burnt smell is a strong tell.
5. Reseat connectors. Unplug and firmly reseat every connector. A surprising number of "bad boards" are bad connections.

Warning signs in the field

• A function that is dead or stuck on while the board's inputs all check good.
• Random resets, freezes, or erratic behavior (power supply or corrosion).
• A fault that tracks with temperature or a physical tap (cracked joint).
• Visible scorch, bulged capacitors, corrosion, or a burnt smell.
• Multiple components dead at once (surge).

What to do about it

Do not replace a board until you have ruled out its inputs, its power feed, and its connectors. When a board truly is bad, find what killed it before you install the new one: a shorted load, water intrusion, or a surge will take out the replacement just as fast. Protect the new board from the cause. Match the replacement exactly; control boards are configured for a specific model and inputs, and a near-match often will not run right.

References

• Manufacturer control-board specifications, input/output mapping, and supply voltages
• NFPA 70 (National Electrical Code) for grounding and surge protection
• Trade-standard practice for input-before-output diagnosis and board condemnation
• See related: how capacitors fail (generic); how sensors fail (generic)