The Apprentice Reports a Reading That Can't Be Right: Decision Tree

Why this matters

Your helper on site reads you a number that is physically impossible: an open circuit that somehow shows full voltage, a compressor pulling ten times its rated current, a pool reading a chlorine level no shock dose could reach. That number is a claim, not a fact, and a green hand's claim breaks far more often than the equipment does. Chase the impossible reading as if it were real and you order the wrong part, condemn a good component, or burn a return trip tomorrow. This tree is the order to run when a reading cannot be right, so you catch the real error before you act on a phantom.

Start here: does the reading, if real, mean a hazard

Before you assume it is a mistake, check what it would mean if it were true.

• If a face-value reading would signal danger - gas present, live voltage where they expect dead, pressure past a relief setting, a combustible level - have the helper treat the area as hazardous and stand clear first, then work the reading. A wrong number is common, but you disprove the hazard before you dismiss it, not after.
• If the number is merely odd, not dangerous, move straight to validation below.

Step 1: read it back raw, with units and scale

The single most common cause of an impossible number is transcription. "Forty-one" on the wrong scale is off by a thousand.

• Have them read the display literally: every digit, the decimal point, and the unit symbol shown. Not "it's forty," but "four one point zero, and it says amps."
• A number an order of magnitude off is almost always a scale, a decimal, or a units slip, not a dead machine. Confirm the units survive the phone before you believe the value.

Step 2: confirm the tool is on the right function

A meter on the wrong mode gives a confident, garbage answer.

• Volts versus resistance, AC versus DC, amps versus millivolts, a manometer on the wrong range - each reads a real number that means nothing for the check you wanted.
• Have them describe the dial position or menu setting in words, and match it to the measurement you actually asked for.

Step 3: confirm where the leads or probe actually sit

The right value across the wrong two points is still wrong.

• Have them describe each lead or probe placement anchored to something physical: "the left terminal with the red wire," "across the run capacitor, not start."
• Watch for a parallel path or a backfeed giving voltage on a leg that should read zero, or a reading taken on the load side of an open device.

Step 4: confirm technique and machine state

• Wrong energized state: a part read live when it should be dead, or dead when it should be running, flips the whole picture. Confirm what was on and what was off at the moment of the reading.
• Poor contact: a probe not biting through paint, scale, or corrosion reads open or erratic. Have them re-seat and re-read.
• The tool itself: a blown fused lead, a dead battery, or an un-zeroed gauge produces impossible numbers on every check, not just this one. A quick known-good reference reading exposes it.

Step 5: cross-check against a second observable that must agree

A reading you cannot trust gets confirmed by a companion fact.

• If a winding reads open, the unit also draws no current. If a pressure is impossibly high, a relief should be lifting or a safety tripping. If a temperature split is impossible, the airflow or water flow tells the same story.
• If the two observables disagree, the reading is the liar - re-take it. If they agree, the surprising number may be real.

Step 6: consider that the reading is real and your expectation is wrong

Sometimes the number is right and the "can't be right" lives in your head.

• A momentary inrush is not the run current. A cold reading is not the hot reading. A nameplate you are quoting from memory may be a different rating than the one on the unit.
• Have them read the actual nameplate before you decide the value is impossible against a spec you assumed.

Recap

1. If a true reading would be dangerous, secure the area first, then validate.
2. Read it back raw with units and scale; suspect transcription first.
3. Confirm the tool's function, then the test points, then technique and state.
4. Cross-check a second observable that must agree.
5. Only after all that, and after reading the real nameplate, act on the number.

References

• Trade-standard practice for meter and instrument use in field diagnosis
• OSHA guidance on qualified persons for electrical and hazardous work
• See related: Confirming a Remote Reading's Validity Before You Trust It; You're Guiding an Apprentice by Phone Through a Diagnosis