The Customer Already Replaced a Part: A Decision Tree

Why this matters

"I already put a new one in and it still does it" changes your whole approach. Either the part was not the problem, the new part is also bad or wrong, the install was botched, or the original fault destroyed the new part too. A self-replaced or previously-replaced part is a loaded clue: it tells you what someone already suspected, and it usually means the real root cause is still sitting there waiting to kill the next one. Skip the obvious-component check, because it is already new, and go hunting for what made it fail.

Start here: ask three questions about the swap

Before you test anything, get the story straight:

• Did it ever work after the swap? "It worked for a day" versus "it never worked" splits the tree immediately.
• Was it the right part? Wrong rating, wrong size, wrong direction, or a generic substitute that does not match the nameplate is extremely common with self-repairs.
• Who installed it and how? A part can be perfect and still fail from a backward install, a missed connection, an over-torqued fitting, or a skipped step.

Write down the answers. They tell you whether you are chasing a bad part, a bad install, or a root cause that eats parts.

If it never worked after the replacement

The new part is suspect, the install is suspect, or the diagnosis was wrong. Work in that order:

1. Verify the part is correct. Compare ratings to the nameplate: capacitance, voltage, current, size, flow direction, thread. A close-but-wrong part is a frequent self-repair failure.
2. Verify the install. Polarity, orientation, seating, torque, and every connection that had to come off. An arrow pointing the wrong way or a connector left loose looks exactly like a dead part.
3. Confirm the original diagnosis. If the part is right and installed right but the symptom remains, the part was never the fault. Go back to first principles and diagnose the actual failure, ignoring the assumption that led to the swap.

The trap here is trusting the customer's diagnosis. They replaced the part they could reach or the part a video told them to replace, not necessarily the part that failed.

If it worked, then failed again

This is the most important branch, because it usually means the root cause is still present and it destroyed the new part. A part that fails twice was not unlucky; something is killing it.

If an electrical part failed again, look for what overstresses it: a motor drawing too much current cooks capacitors and contacts; a loose connection arcs and burns terminals; a voltage problem destroys electronics; an overload trips and welds contacts. Find the overstress, not just the casualty.

If a mechanical part failed again, look for what wears it: misalignment chews bearings and couplings; vibration loosens and fatigues; a missing filter feeds grit into everything downstream; the wrong lubricant or none at all.

If a sealing part failed again, look for what attacks it: heat, chemical incompatibility, a surface that was never cleaned, over-tightening, or pressure beyond its rating. A seal that blows twice is telling you the surface, the pressure, or the material is wrong.

If two different parts have now been replaced

When a customer has thrown two or three parts at a problem, stop replacing and start measuring. Parts-cannon repair means nobody has diagnosed the system; they have been guessing. Reset to a clean, evidence-based diagnosis:

• Measure the actual fault (voltage, current, pressure, temperature, flow) instead of inferring it from which part "should" be bad.
• Look for the common cause that could explain every failed part at once. Two unrelated-looking failures often share one upstream root: a power problem, a flow restriction, a contamination source.
• Trust the readings over the history. The repair history tells you what was suspected, not what is true.

Protect the new part you install

Whatever you put in, do not let it die the same way. Before you button up:

1. Fix the root cause first. Replacing the casualty without removing the cause guarantees a callback.
2. Verify the conditions the new part will live in. Correct voltage, clean flow, proper alignment, right lubricant, in-spec pressure.
3. Document the root cause on the ticket. The next tech, and your warranty position, depends on knowing the part failed for a reason and that reason was corrected.

A part the customer already tried is not a dead end. It is a free hint that someone already looked there, so the real answer is usually one layer deeper.

References

• Trade-standard root-cause-analysis and repair-verification practice
• Manufacturer service documentation for part ratings and installation
• See related: Two Things Broke at Once: A Decision Tree; Why Electrical Connections Degrade
• See related: Why Seals Leak: The Physics