The Fault Behind the Fault Decision Tree

Why this matters

Some parts do not die of old age; they get murdered. A fuse blows because something shorted. A motor burns because a bearing seized. A board fries because a surge came through. If you replace the dead part and leave the killer in place, the new part dies too, and now it is your callback, on your dime. The skill that separates a parts-changer from a diagnostician is the habit of asking, on every failed component, "what killed this, and is it still here?" This tree walks you from the symptom you can see to the root cause hiding behind it.

Safety first

A fault that destroys parts is often dumping energy somewhere it should not go. If the failed component is in a gas, electrical, stored-energy, or pressurized system and shows scorching, arcing, a blown rupture, or heat, treat the upstream cause as live until proven otherwise. De-energize and verify dead, relieve pressure, or shut off the fuel before you put hands on it. The thing that killed the part can kill the next part, and it can hurt you on the way.

Start here: is this part a victim or the cause

The first call is whether the dead part failed on its own or was taken out by something else. The failure signature tells you.

• A protective device tripped or opened (fuse, breaker, overload, safety, fusible link) - it almost never fails on its own. It did its job; something made it act. Find that something.
• A part shows damage out of proportion to its age - a young component that is burned, melted, or seized was driven past its limits by a stressor, not by wear.
• A part shows clean wear consistent with its age - this may be a genuine end-of-life failure with no killer behind it. Confirm, but do not invent a phantom root cause.

If a protective device opened

Protective devices are the loudest "fault behind the fault" tell in the trades. They open on purpose. Replacing them without finding the cause is the classic mistake.

• Fuse or breaker tripped - there is an overcurrent or short downstream. Check for a shorted load, a pinched wire, a failed winding, a grounded conductor before you reset or replace. Verify the circuit before re-energizing.
• Thermal overload or high-limit opened - something overheated. Look for restricted airflow, a stuck load, low flow, a blocked filter, a dry bearing. The device saved the equipment from the real fault.
• Pressure relief or safety valve lifted - pressure went too high. Find the overpressure source (a stuck regulator, a blocked outlet, a failed control) before you reset or replace the relief.

In every case, the device is the messenger. Read the message, fix the source, then replace the device.

If the part died out of proportion to its age

Branch A: mechanical kill. A motor that burned, a pump that seized, a coupling that sheared. Look upstream and downstream for the load that did it: a seized bearing, a jammed mechanism, a blockage that stalled it, a misalignment that wore it. Spin the load by hand if you safely can; binding you can feel is binding that killed the part.

Branch B: electrical kill. A board, a contactor, or a winding that fried. Suspect a supply problem (over- or under-voltage, a surge, a lost phase or neutral), a short, or a stuck load drawing too much current. Read the supply and the load draw before you trust the replacement.

Branch C: environmental kill. A part destroyed by water, heat, corrosion, or debris. Find the source of the water intrusion, the heat, or the contamination, because it is still feeding the same failure.

Branch D: upstream-component kill. One failed part took out the next part in the chain. A failed capacitor that cooked a motor, a failed valve that deadheaded a pump. Trace the chain back to the first domino.

If the part shows clean age-appropriate wear

This may genuinely be an end-of-life failure with nothing behind it. Confirm by checking the things that would have stressed it: load, supply, airflow, alignment, contamination. If all read normal and the part's age matches its expected life, treat it as a straight wear replacement. Do not manufacture a root cause the evidence does not support; a phantom fix you cannot point to is not a fix.

Confirming the diagnosis

You have found the real root cause when you can name the mechanism that killed the part and show the evidence: the short on the meter, the seized bearing in your hand, the blocked airflow, the over-voltage reading. Correct the root cause, install the new part, and run a full cycle. If the new part survives a full duty cycle under real load, the killer is gone. If it stresses again, you treated a symptom, not the cause.

Next steps

Repair the root cause first, then the failed part, in that order. Note the chain on the work order: "overload tripped, found blocked condensate drain stalling the pump, cleared drain and replaced overload." That record protects you on the warranty and tells the next tech the real story. When you cannot find a killer after honest checking, say so and treat it as a wear failure, but set the customer's expectation that you watched for an underlying cause and found none.

References

• ISO 13379-1 (Condition monitoring and diagnostics of machines; failure cause analysis).
• NFPA 70B (Standard for Electrical Equipment Maintenance; overcurrent and protective-device root-cause practice).
• Manufacturer documentation on protective-device function and acceptable operating limits.
• See related: The Brand-New Part That's Also Bad Decision Tree; Reading Rust and Corrosion Patterns.