The Known-Good Swap Test Method

Why this matters

Sometimes you cannot measure your way to certainty. A part is borderline, your meter says "maybe," and the bench test does not load it the way the field does. The known-good swap test cuts through that. You substitute a part you trust for the part you suspect and watch what happens. If the fault clears, you found it. If it does not, you just eliminated a suspect for the price of a few minutes. Used with discipline it is one of the fastest tools you have. Used carelessly it sends a healthy part into a circuit that will destroy it.

What "known-good" actually means

A known-good part is one you have proof is working, not one you assume is fine because it is new or because it came out of a runner. New parts fail out of the box. The part you swap in must be either freshly tested or pulled from a system you have confirmed is operating correctly. If you swap in an untested part and the fault persists, you have learned nothing, because the substitute could be bad too.

Step 1: Confirm the swap is safe before you touch anything

This is the step people skip and regret. Before you put a good part into a suspect system, make sure the system will not kill it.

• Find out why the original part failed. If the failure was caused by something upstream (a short, an overvoltage, a seized load, a blockage), that upstream fault will destroy your known-good part the instant you energize it.
• Check that the substitute is correctly rated and compatible. A part that is close but undersized will fail or mislead you.
• De-energize, lock out, and verify zero energy before you open anything.

If you cannot rule out an upstream killer, do the swap test last, after you have measured the supply feeding that part.

Step 2: Change exactly one thing

Swap only the suspect part. Do not also reseat three connectors, clear a filter, and adjust a setting in the same move. If you change several things and the fault clears, you do not know which change fixed it. One variable, one swap, one result.

Step 3: Restore, energize, and observe

Put the system back to a normal operating state, energize, and watch the exact symptom you started with.

• Fault gone? The original part was bad. Confirm by reinstalling the original and verifying the fault returns, when it is safe and practical to do so. That reinstall step is what turns "probably" into "proven."
• Fault still there? The part you swapped is not the cause. Reinstall the original, move to the next suspect.

Step 4: Decide whether to reinstall the original

If the swap proved the original was bad, the customer part goes in and the original comes out, no question. If the swap exonerated the original, put it back rather than leaving a borrowed or stock part in place. Leaving the wrong part installed creates a second mystery for the next tech.

Where the swap test shines and where it bites

It shines on parts that are hard to test under load: a capacitor whose value drifts only when hot, a control board with an intermittent output, a sensor that reads fine on the bench but lies in the system, a valve coil that opens cold and sticks warm.

It bites when:

• The substitute is not truly known-good, so a null result is worthless.
• An upstream fault smokes the good part the moment it is energized.
• More than one thing got changed, so the result is ambiguous.
• The swapped part is expensive or single-use, where measurement is the smarter first move.

A clean swap, every trade

The method is identical whether you are subbing a run capacitor on a condenser, a thermocouple on a water heater, a pressure switch on a pump, a heating element in an appliance, or a flow sensor on a pool heater. Confirm good, confirm safe, change one thing, observe, prove with a reinstall. That sequence keeps the swap test honest.

References

• OSHA lockout/tagout requirements for de-energizing equipment before component replacement.
• Manufacturer documentation for correct part ratings and compatibility before substitution.
• See related: The One-Variable-at-a-Time Rule; Working From the Symptom Backward; Eliminating the Easy Stuff First.