Skip to content

When a Swap Test Changes the Evidence

Interpret substitution tests without confusing a working replacement with proof about the removed part, and preserve evidence when connections or operating conditions change.

Published
Reading time
6 min
  • Fault Isolation
  • Field Service

A working replacement answers only part of the question

A technician replaces a suspect module. The equipment passes its functional check. Was the removed module faulty?

That may be the right conclusion, but the successful check alone does not establish it in every case.

Replacement can change more than the module. Connectors are disconnected and reconnected. Power is cycled. A configuration may be restored, an adjustment changed or the equipment allowed to cool. Any of those changes can affect the symptom.

A swap test is an intervention, not simply another reading of an unchanged system. Its diagnostic value depends on knowing what changed and what the before-and-after observations actually demonstrate.

This article develops the substitution-test caution introduced in Reducing MTTR Without Replacing Good Parts. Its focus is interpreting the evidence, not deciding that replacement should always be the first or last test.

Separate the system result from the removed item's condition

Two questions need separate answers:

  1. Does the equipment now meet the tested requirement?
  2. What evidence establishes the condition of the removed item?

A successful functional check with a replacement installed supports a conclusion about the tested system in its new state. Under a well-controlled substitution procedure, it can also support identifying the original item as bad. That second inference depends on the procedure's assumptions.

The replacement must be suitable and its relevant function established. Required configuration, inputs, load and operating conditions must be accounted for. Other changes must not provide an equally plausible explanation of the improvement.

Likewise, an unsuccessful swap does not automatically establish that the original item was good. The replacement may be incompatible, another fault may remain, or the test may not have reproduced the reported condition.

Observation Supported conclusion Conclusion requiring more evidence
The equipment passes after replacement. The tested requirement is met with the replacement under the observed conditions. The removed item alone caused the earlier failure.
The symptom persists after replacement. This intervention did not restore the tested requirement. The removed item was healthy in every relevant respect.
The removed item passes an approved bench test. It meets that bench test's requirements under its conditions. It could not have failed under the installed load, temperature or intermittent condition.

These are engineering interpretation limits, not claims that iTech automatically identifies every change made during maintenance.

Worked example: a module and its connector

Illustrative example only—not a customer installation, an equipment procedure or a recorded iTech session. A control module intermittently fails to deliver its required response. The initial investigation leaves both the module and its connection path plausible.

An authorized technician installs a suitable replacement. Doing so requires disconnecting and reconnecting the module's plug. The subsequent functional test passes.

At least two explanations fit that sequence:

  • The original module had a fault that the replacement removed.
  • The connection was unreliable, and disturbing it temporarily restored the path.

The passing result is useful: it establishes the tested response in the new state. But it does not, by itself, separate those explanations.

If the original module later passes a bench test, that result still needs context. The fixture may not reproduce the installed load or the conditions associated with the intermittent symptom. “No fault found under this test” is a more defensible statement than “the original module was definitely good.”

The next action should follow the approved investigation and return-to-service procedure. It might require an independent assessment of the removed item, inspection of the connection or further controlled observation. Do not improvise repeated reconnections, reinstall a potentially unsafe part or recreate a hazardous fault merely to strengthen the diagnosis.

Decide what the swap is intended to prove

Before an approved substitution, define its purpose.

It may be a diagnostic test intended to distinguish candidates. It may instead be an authorized assembly-level repair after other evidence has already isolated the assembly. Those are different uses of the same physical action.

For a diagnostic substitution, establish:

  • The hypothesis: which item or functional boundary is being assessed?
  • The replacement: is it compatible, correctly configured and suitable for this test?
  • The baseline: what failed, under which operating conditions and against which acceptance criteria?
  • The intervention: which connections, settings, resets or adjustments will change?
  • The interpretation: what will a pass or fail establish, and which alternatives will remain?

The iTech test-procedures guide explains that good or bad results are assessed against specified tolerances and performance limits. Apply that discipline to the post-swap check: “the equipment started” is not necessarily equivalent to “the required function passed.”

The appropriate boundary also matters. A justified assembly-level conclusion does not automatically identify a particular internal component. See Choosing the Right Diagnostic Boundary.

Preserve the distinction between pre-swap and post-swap evidence

An observation belongs to the equipment state in which it was collected.

Keep enough information in the service record to distinguish the original symptom from the post-intervention result. Useful information includes the equipment configuration, tested function, operating conditions, original and replacement item identities, other work performed and verification outcome.

This is a recommended service-record practice, not a claim about automatic iTech logging.

A pre-swap failure remains evidence that the earlier arrangement failed its test. It should not be silently reinterpreted as a failure of the replacement. Equally, a post-swap pass should not be entered or described as proof that the removed original item passed.

When modeling the procedure, make its interpretation explicit. Does the reported result describe the installed system's response, or the inferred condition of the original component? Reversing those meanings can reverse the diagnostic conclusion.

The iTech modeling-tests guide illustrates the broader principle: the physical observation and the modeled conclusion are not always identical. Its indicator example relies on an indication as evidence of an output. The relationship must be justified; a substitute observation does not justify an unlimited conclusion.

Keep bench tests and installed functional tests distinct

The iTech test-methods guide describes in-circuit testing as observing a component's output while operational, allowing a passing result to support conclusions through the modeled dependency path.

A removed-item test has a different scope. It may establish a property or function on a fixture without exercising the original installed connections or supporting equipment.

Neither method is inherently sufficient for every fault. Choose the evidence required by the hypothesis, and retain uncertainty where the available test does not reproduce the relevant conditions.

For how a pass can support upstream conclusions without certifying every function, read What a Passing Test Really Tells You.

Validate the interpretation before field use

The iTech model-validation guide calls for expert review, simulated troubleshooting and real-system validation where possible. Include the meaning of substitution results in that review.

Check cases where replacement restores the function, where the symptom persists and where the available evidence cannot distinguish the original item from its disturbed connection. Confirm that the modeled result refers to the intended item and does not claim greater certainty than the procedure supports.

Repair and return to service are separate obligations. The repair-procedures guide includes isolation, removal and replacement, adjustment or calibration, and reset. Follow the equipment's approved requirements; this article does not authorize live replacement, interlock bypassing or unsafe fault reproduction.

Use a swap result as evidence about a defined change under defined conditions—not as automatic proof that the removed part caused the fault. A useful diagnostic conclusion states both what the intervention established and what remains unresolved.

See model-based isolation on your system

Walk through a diagnostic model with an iTech engineer.

Request a Demo