When the RPM PID Lies to You
The complaint
“Scan tool says RPM is steady, but the engine is clearly stumbling. Or it reads zero during a crank that's obviously turning over.”
The RPM PID is calculated from the crank sensor signal — interpreted, filtered, and reported. Several layers between the wire and the screen can lie. A dropout signal can show as steady RPM, a noisy signal can show as smooth, and an intermittent sensor can show zero on a cranking engine. Trusting the PID without verifying the source can put you on the wrong diagnostic path for hours.
The assumptions that burn techs
- Scan tools sample slowly. A dropout faster than the sample rate is invisible.
- Many platforms filter and smooth the RPM PID for display. The number is averaged — not raw.
- A failing crank sensor can produce a partial signal that the PCM interprets as 'cranking but no fire' even while the scan tool shows healthy RPM.
- A reluctor wheel with a cracked tooth gives the same averaged RPM but a misfire pattern that nothing on the scan tool will reveal.
Inputs, commands, and expected results
Inputs — what to read
- Raw crank sensor signal on a scopeThe source of truth — not the PID.
- Cam sensor signal alongside crankSync errors hide in the relationship.
- RPM PID at the same momentCompare PID behavior to raw waveform.
- Misfire counters during the eventSensor dropouts often flag as misfires.
- Stored DTCs (P0335, P0336, P0340)Often the only direct hint.
Commands — what to do
- Scope crank + cam during a crank attemptConfirms the signal the PCM is actually seeing.
- Wiggle the harness with the scope runningReproduces intermittent dropouts the PID smooths over.
- Heat-cycle the sensor with a heat gunSurfaces heat-failing sensors before a tow truck does.
Expected results — what good looks like
- Crank sensor waveformClean, consistent amplitude — no missing pulses except the sync gap.
- Cam-crank correlationStable relationship cycle to cycle.
- RPM PID at idleSmooth, within ±25 RPM of commanded.
- RPM PID during crank150–250 RPM, stable from first revolution.
What sends techs down the wrong path
The questions a real diagnostician asks
This is the difference between a parts changer and a diagnostician — not what you test, but the order you think about it.
- 1
Does the PID match what the engine is actually doing?
If the engine is stumbling but the PID is glass-smooth, the PID is filtered or sampled too slow. Don't trust it — scope the sensor.
- 2
What does the raw signal look like?
Crank and cam on a scope, simultaneously. Missing pulses, amplitude collapse, or sync slip will jump out — none of which show on a scan tool.
- 3
Is the failure heat-, vibration-, or time-related?
Heat-soak failures need a heat gun reproduction. Vibration failures need a wiggle test. Time-related failures need a longer recording.
- 4
Are misfire counters or sync codes pointing where I'm not looking?
A 'random misfire' on a stable RPM PID with a P0336 in history is almost always a reluctor or sensor problem the PID hid from you.
- 5
Did I prove the PID by comparing it to the source?
PID and waveform side by side. If they disagree, the PID lied — and so does any diagnostic based on it.
Stop guessing. Start thinking.
DiagCoach helps technicians follow structured diagnostic logic using real-world test results — the same way the best techs in the bay actually work.
Start a guided diagnostic →Frequently asked questions
Why does my scan tool show RPM when the engine isn't actually starting?
The crank signal exists and the PCM converts it to RPM, but injector or spark logic has another condition not being met (sync, immobilizer, etc.). RPM PID confirms only that the crank signal is there.
Can a bad crank sensor pass cranking and fail at idle?
Yes. The signal envelope at cranking speed is forgiving; at idle and load the PCM needs cleaner edges and can lose sync intermittently.
What's the fastest way to prove the RPM PID is lying?
Scope the crank sensor and the PID at the same moment. Any disagreement is your answer.