Swap a Known-Good Part In to Confirm or Rule Out (Decision Tree)

Why this matters

Before you pull a part and put a substitute in, there is a decision to make that most techs skip: should this be a substitution test at all, right now, on this component? Reaching for the swap out of habit, when a five-minute measurement would answer the same question with less handling risk, wastes time and exposes a customer's working equipment to unnecessary disassembly for no gain. Skipping the swap when it is genuinely the fastest and most conclusive path costs you a slower, less certain diagnosis. This tree is the decision you make before the part comes out, not the decision about what a result means after.

Start here: can you get a confident answer without moving any hardware?

If a direct measurement (a meter reading, a pressure test, a visual inspection) can confidently confirm or rule out this component, do that first. Substitution testing is a tool for when measurement cannot get you certainty, not a default first move. Reaching for a swap when a thirty-second reading would settle it adds handling risk for no benefit.

If measurement is inconclusive, borderline, only valid under a load condition you cannot easily simulate on the bench, or the component's failure mode does not show up on a static test, continue to the next check. This is the actual use case for substitution testing: components that test "fine" on paper but still misbehave installed and running.

Do you have a genuinely known-good part to test with?

If you have a part you can prove is currently working (freshly tested, or pulled from a system you have confirmed is operating correctly), continue.

If the only available substitute is untested, new-in-box without verification, or of uncertain condition, stop. Swapping in an unverified part and getting a result tells you nothing, because either outcome could be explained by the substitute's own condition. See the related decision tree for what to do when no known-good spare exists.

Is the swap safe to perform on both units?

Do this check before anything else touches a tool.

• If either unit cannot be safely de-energized and verified dead, or safely depressurized, before the swap, stop. Do not disconnect or connect a live electrical component or a pressurized line or fitting.
• If the donor unit is a life-safety system that cannot be taken offline even briefly, stop. Find a different known-good reference or a different diagnostic method.
• If both checks are clear, continue.

Are the two parts genuinely equivalent?

If the substitute matches the suspect part's exact rating, specification, and configuration, with no modification required to fit, the comparison is valid.

If the parts differ meaningfully in rating, age, wear condition, or require any modification to install, the test result will not be reliable evidence either way. A part that is close but not exact is not a known-good reference; it is a second unknown.

Will the fault actually show up in your available test window?

If the fault is present continuously, or reliably reproducible on demand (a specific load, a specific button press, a specific run duration you can trigger), the swap will give you a clean answer in a reasonable time.

If the fault is intermittent and you cannot force the triggering condition, a clean swap in a short test window may produce a false negative (looks fixed, actually just has not recurred yet) as easily as a true one. Decide upfront whether you have time to observe over a longer window, or whether you should set that expectation with the customer now rather than after a rushed, ambiguous test.

Could a second, unaddressed cause make this result misleading?

If you have reason to believe more than one component could plausibly explain this exact symptom, and you have not yet isolated which, a single-part swap only clears you on that one part. A persisting fault will not rule this part out if a second failing component is compounding with it or masking the result. See the related reference on when substitution testing risks masking the real fault.

If every check passes, proceed

You have: a case where measurement alone was not conclusive, a genuinely known-good reference part, a safe procedure on both units, true compatibility, a realistic test window for the fault's behavior, and no known compounding second cause. This is the point where the swap is the right move, not a shortcut around a slower but more honest path. Follow the known-good swap test method for the mechanics: change one thing, restore, energize, observe, and reinstall the original if the swap clears the suspect.

Decision recap

1. Try measurement first; only reach for a swap when measurement cannot get you certainty.
2. Confirm you have a genuinely known-good reference part, not just an available one.
3. Confirm the swap is safe on both units before touching anything.
4. Confirm true compatibility between the two parts.
5. Confirm the fault will actually show up in your test window, or set expectations if it is intermittent.
6. Rule out a second compounding cause that would make the result ambiguous.
7. Only then proceed with the swap itself.

References

• Manufacturer documentation on component compatibility and safe removal procedures
• OSHA lockout/tagout guidance for de-energizing equipment before component removal
• See related: The Known-Good Swap Test Method
• See related: When a Borrowed Swap Test Isn't Safe or Valid
• See related: No Known-Good Spare Available to Test With (decision tree)