Field Guide · For Techs

Reading Live Data
without getting fooled

A scan tool hands you a screen full of numbers. Numbers without context lie — a value can look perfect while the thing it describes is broken. Here's how to read PIDs like they actually mean something, and how to spot the ones playing you.

Start here

A PID is a claim, not a fact

  • Some values are measured — a sensor voltage the ECU actually reads (MAF g/s, MAP, TPS).
  • Some values are calculated — the ECU infers them from other inputs (calculated load, calculated airflow). These can look flawless while the real signal is garbage.
  • Some values are corrections — fuel trims aren't a reading, they're the ECU compensating. A trim is a symptom, never the root cause.
Rule #1: before you trust a number, know where it comes from. Measured, calculated, or commanded — each one lies in a different way.
Step 1Sort every PID into three buckets
  • MeasuredMAF, MAP, TPS, O2 voltage — a real signal; trust it, then verify it against a scope
  • CalculatedCalc load, calc airflow, timing advance — derived; fine until an upstream input is wrong
  • CorrectionSTFT / LTFT, idle adaptation — the ECU covering for a problem, not the problem itself
  • CommandedCommanded EGR, VVT, boost, AFR — what the ECU wants
  • ActualActual EGR, actual VVT, actual boost — what it got; the gap is the whole story
Step 2The PIDs that tell the truth — if you read them right

Fuel trims (STFT + LTFT)

The confession
  • Big positive = the ECU adding fuel → lean (unmetered air or low fuel delivery). Big negative = pulling fuel → rich.
  • Read at idle vs cruise to localize: lean at idle but fine under load points at a vacuum leak; lean everywhere points at fuel delivery.

MAF (g/s)

Reality check
  • Rough rule at wide-open: grams/sec should land near the engine's liters × some, not half of it.
  • Low at WOT = restriction or a lazy/contaminated MAF. Compare to calculated airflow — a split between them is the tell.

O2 — upstream vs downstream

Loop & cat
  • Upstream should swing (the ECU hunting in closed loop). A flatliner = a dead sensor or stuck-open loop.
  • Downstream should be lazy and flat. If it swings like the upstream, the cat isn't working.

Misfire counters (per cylinder)

Aim before you scope
  • Per-cylinder counts point you at the offender before you drag out the scope.
  • One cylinder climbing = that circuit/mechanical. All of them = a shared cause (fuel, timing, a lean bank).

Commanded vs Actual

The gap is the fault
  • EGR, VVT, boost, fuel pressure — put commanded next to actual and watch the gap.
  • Commanded moves, actual doesn't = a stuck actuator, restriction, or mechanical fault. That gap is the diagnosis.
How live data fools you

The traps that send good techs the wrong way

  • Idle-only data hides load faults. Watch it under the condition that actually fails — drive it, load it, heat-soak it.
  • A calculated PID looks fine while the sensor feeding it is wrong. Cross-check measured against calculated.
  • Default / substituted values — the ECU dropped to a fallback table and it reads like a real number. Suspiciously "perfect" is a flag.
  • No baseline. "Looks about right" isn't a spec. Know the good number for that engine before you judge.
  • Trims that "look OK" because they maxed out — clipped at their limit (±25%). Pegged is not the same as happy.
HabitsWhat makes you fast with a scan tool
  1. Watch it under the condition that failsIdle data proves almost nothing. Recreate the fault — load, RPM, temp — then read.
  2. Graph it, don't stare at one numberTrends and correlation tell the story. A single frozen value hides the moment it went wrong.
  3. Put commanded next to actualHalf the diagnosis is just watching what the ECU wanted vs what it got.
  4. Always compare to a known-goodSame as the scope: the fault jumps out against a healthy reference. Build the library.
The unfair advantage

Data pointing two directions at once?

When the trims say lean, the MAF says fine, and the O2 says something else — that's the moment to talk it through with Kate. Give her the actual numbers and the conditions you pulled them under, and she'll hold the whole picture and tell you which reading to trust. No guessing, no "consult a professional." Just the next test.