ARC-TUNING — WOT Cohesion Dashboard

✦

Scarecrow Briefing — WOT Cohesion Dashboard

The crow's-nest of the tuner room. This page doesn't tune anything itself — it integrates the outputs of PE, Spark, MAF (or VE+IFR), and verifies they agree under WOT load. The "cohesion" check: does the airmass model (MAF/VE) agree with the fuel model (PE+IFR) under load? Does spark advance match MBT without knock? If yes, sign off. If no, the page tells you which upstream section needs another pass.

(1) Gated on prereqs. Requires PE complete + Spark complete + (MAF complete OR VE+IFR complete) per spec §4.1. If any prereq is incomplete, page refuses with a "go back and finish X" link. (2) Cross-correlation, not new tuning. Inspects log channels: AFR Actual vs AFR Cmd (PE), spark advance commanded vs actual + KR (Spark), MAF g/s vs VE-derived airmass (Cohesion). Flags discrepancies. (3) STFT does NOT apply at WOT (spec §5.4). The page refuses to use STFT histograms for WOT analysis. AFR ground-truth comes from wideband. (4) Final sign-off. If everything agrees within tolerance (AFR ±2%, spark ±1° vs commanded, KR < 1° sustained), Scarecrow marks the build "WOT-validated" and writes a session report.

KB: afr-target-ranges-by-load.md, knock-detection-and-prevention.md, ve-table-tuning-methodology.md. Spec §4.1.

6.1 Prereq Gate — PE / Spark / Airmass complete ⚠ Section gate per spec §4.1

Page refuses if any prereq isn't complete. Required: aggregateState.pe === 'complete' AND aggregateState.spark === 'complete' AND (aggregateState.maf === 'complete' OR (aggregateState.ve === 'complete' AND aggregateState.ifr === 'complete')). Each unmet prereq surfaces a "go back to X" link.

This page only runs after PE, Spark, and either MAF or VE+IFR are done. Otherwise it sends you back.

⛔ SCARECROW HALTS HERE IF…

PE INCOMPLETE WOT AFR target not yet locked. Return to PE.

SPARK INCOMPLETE Spark advance under load not yet finalized. Return to Spark.

AIRMASS INCOMPLETE Neither MAF nor VE+IFR is done. Return to whichever path your build is on.

6.2 AFR Actual vs Commanded (PE check) 📍 LOCATE

HPTWB Channel: AFR Actual

— Tolerance: ±2% across cells visited

Under WOT load, scan the log: at each (RPM, MAP) cell, AFR Actual should be within 2% of AFR Cmd. Bigger errors mean PE table needs revision. Source AFR from wideband, not narrowband, not STFT — STFT does not apply at WOT.

Compare what AFR you commanded to what you actually got under WOT. Off by more than 2% means PE needs another look.

6.3 Spark Advance — Commanded vs Actual + KR — Tolerance: KR < 1° sustained

Per cell visited: spark advance actual should equal commanded minus knock retard. If sustained KR > 1° in any cell, spark is too aggressive — return to Spark. Brief KR spikes (<0.5°, <100ms) on transients are normal noise; sustained pulls are real knock.

Knock retard should stay under 1° during a WOT pull. Anything more sustained means timing is too aggressive.

6.4 Airmass Cohesion — MAF g/s vs VE airmass — Cross-check airmass models

If both MAF and VE are tuned (recommended), they should agree on airmass under WOT. MAF g/s × intake density should match VE-derived airmass within 3%. Larger disagreement means one or both is off — usually the MAF table at high g/s (above what the OEM curve was tested at). Adjust the lower-confidence side.

If you tuned both MAF and VE, the two should agree on how much air is going in. If they disagree by more than 3% under WOT, one is wrong.

6.5 Final Sign-Off + Session Report — Generates report when all green

When 6.2, 6.3, and 6.4 are all within tolerance, Scarecrow marks the build "WOT-validated" and writes a session report. Report contents: build context (engine/PCM/mods/fuel/sensors), final cell values changed, log timestamps verified, KB references cited, confidence level per subsection. Saved to disk for the user's records.

Once all the WOT checks pass, this page writes a session report you can keep — what was changed, when, why.