Plot measured lambda on the Y axis and engine rpm on the X axis. Sweep from idle to 3000 rpm and back. Different faults produce visually distinct curves on that simple graph — hence the nickname.
Healthy baseline
A well-running engine shows lambda tracking close to 1.0 (±0.02) across the entire rpm range. Small oscillation from closed-loop cycling is normal.
Vacuum leak fingerprint
Lean at idle (λ > 1.3), improving steadily as rpm rises. By 2500 rpm lambda is back near 1.0. Reason: leak area becomes negligible relative to commanded airflow.
Exhaust leak fingerprint
Opposite of vacuum leak. Near normal at idle (low exhaust pressure, little ingress), progressively lean at higher rpm as exhaust pulses pull more air in. Lambda worsens as rpm rises.
MAF under-report fingerprint
Roughly flat lean offset across the whole range, with LTFT positive throughout. Unlike vacuum leak, it doesn't normalise at high rpm.
Fuel delivery failure fingerprint
Normal at idle, lean at high rpm as injectors can't keep up with demand. The curve breaks sharply upward past a certain rpm/load point.
Lazy O₂ sensor
Measured lambda lags calculated lambda during rpm sweeps. Plot both on the same graph — calculated reacts fast, measured trails by hundreds of milliseconds.
// run it both ways
Sweep up and sweep down. Some faults are asymmetric (exhaust leaks are often worse on decel because the scavenging pulse is stronger). Asymmetry itself is diagnostic.