The Required Flow Contradicts Your Experience (Decision Tree)

Why this matters

You are running the manufacturer's required flow and it points somewhere your experience says is wrong. It calls a reading normal that you think is off, or it routes you away from the fault you are sure of, or its logic just does not match what the unit is doing. Now you are stuck between a procedure you must follow to get paid and a gut you have earned over years. Handle it wrong in either direction and you lose: fabricate the flow's answer and you install the wrong fix, or ignore the flow and forfeit the claim. This tree is how you resolve the contradiction without doing either.

Safety comes first

If the flow's next step would create a hazard - re-energizing a faulted circuit, re-pressurizing a component you believe has failed, running a combustion appliance you suspect is cracked - do not perform it to keep the sequence moving. Make the situation safe, document why you stopped, and call the manufacturer for direction before you go further. A contradiction about diagnosis never justifies an unsafe step.

Start here: assume the flow might be right

The hardest move is checking yourself first, and it is the correct one. The flow encodes the maker's failure data across far more units than you will ever see, so it may know a failure mode, a look-alike symptom, or a common misdiagnosis that you do not. Anchoring on the first plausible cause is one of the most reliable ways good techs go wrong. Before you decide the flow is broken, treat the possibility that you are the one who is off.

Verify your reading before you trust it over the procedure

Most "the flow is wrong" moments are a measurement problem, not a flow problem.

• Confirm your instrument. Check the meter against a known reference or a second tool before you build a case on its reading.
• Confirm your conditions. Are you testing at the same point, in the same state, that the flow specifies? A reading taken cold when the flow wants it hot, or upstream when it wants downstream, will contradict for no real reason.
• Confirm you are measuring the same thing the flow means. Terminology and test points vary.

If verifying makes the contradiction disappear, the flow was right and your test was off. That is a good outcome, not a defeat.

If you have verified and it still contradicts

You have re-checked your instrument, your conditions, and your method, and the flow's premise still does not match reality. Do two things, and neither of them is "quietly do it my way."

1. Do not fabricate the result the flow wants. Writing down a value you did not get corrupts the record and can void the claim worse than a deviation would.
2. Call manufacturer tech support. Describe your readings and the contradiction. They will either correct your understanding with something you did not know, or authorize a deviation from the published flow. Get that deviation noted in the claim record with a reference number.

Tech support exists partly for this exact moment, and a documented call is what turns "I ignored the flow" into "the manufacturer authorized a variance."

If the fight is about the remedy

There is a specific version of this where the flow's conclusion tells you to replace a part you have tested and believe is good. That has its own decision tree, because the question there is about swapping a known-good component, not about a contradicted reading. Hand off to it rather than resolving it here.

Document what you actually found

Whatever the outcome, record your real readings verbatim, not the ones that would have made the flow tidy. If the claim is later questioned because you did not reach the flow's conclusion, your documented readings plus the tech-support reference are your appeal. Truth in the record is the only position that survives a reviewer.

When the flow wins

If you cannot verify your position, or tech support explains what you were missing, follow the flow. Humility here is cheap; being confidently wrong on someone else's equipment is not. The goal was never to prove the flow wrong, it was to install the right fix and get the claim paid, and sometimes that means the procedure knew something you did not.

The order to run it

1. If the next step is unsafe, make it safe first and call for direction.
2. Assume the flow might be right before you assume it is wrong.
3. Verify your instrument, conditions, and method.
4. If it still contradicts, do not fabricate; call tech support and get any deviation noted.
5. Hand off the known-good-part-swap case to its own tree.
6. Document your actual readings, and follow the flow if you cannot verify otherwise.

References

• Manufacturer service documentation and technical support escalation procedures
• Trade-standard practice for diagnostic verification and avoiding anchoring bias
• OSHA general guidance on halting unsafe procedures
• See related: The Flow Tells You to Replace a Part You Know Is Good (decision tree); You Know the Answer but the Warranty Wants the Steps (decision tree); Verifying Your Test Equipment Before You Trust It