Turbo Underboost — Requested vs Actual Low, Charge-Air Leak Found

Vehicle / system context

Educational case: 2.0 turbocharged petrol (TFSI-style). Synthetic/demo scenario for boost-fault methodology training.

Complaint

Low power under motorway acceleration, occasional limp mode, MIL on with P0299.

Initial evidence

VCDS: Engine — P0299 (Turbocharger Underboost). Freeze frame: 82% load, requested boost 1.8 bar, actual 0.9 bar. N75 duty at max. Gateway: all modules communicating.

Working differential

Ranked by evidence (actual below requested under load):

1. Charge-air leak — high probability, fastest to find with a boost-leak test
2. Wastegate stuck open / actuator not holding
3. VGT / control (if applicable)
4. MAP sensor bias (correlate against gauge)
5. Exhaust restriction limiting turbine drive
6. Turbocharger wear

Tests performed

1. MAP key-on plausibility: MAP read barometric — sensor plausible.
2. MAP vs mechanical gauge (road test): gauge also read low — sensor accurate, boost genuinely low.
3. Boost-leak test: pressurised the intake from the turbo outlet — audible leak at the intercooler end-tank.
4. N75 / wastegate actuator: actuator moved and held vacuum — control functional.
5. Requested vs actual log (road test): requested rose, actual stalled at 0.9 bar; N75 duty at max — control trying, physical loss confirmed.

Evidence updates

Boost-leak test found a split intercooler end-tank before any turbo or control testing. MAP correlated against a gauge ruled out sensor bias. N75/actuator test ruled out control. The evidence pointed to a physical charge-air loss — confirmed by the leak test.

Confirmed root cause

Split intercooler end-tank losing charge air under load. Turbo and boost control were functional; the leak dropped actual boost below commanded pressure.

Repair summary

Replaced the split intercooler. Requested vs actual within tolerance on full-load road test.

Before / after

Before: requested 1.8 bar / actual 0.9 bar at 82% load; P0299; limp mode. After: actual tracks requested within tolerance; P0299 cleared; full power restored.

Verification

Road test at 82% load: requested vs actual within tolerance. No P0299 return. Limp mode did not recur.

Lessons for technicians

1. Requested vs actual under load is the core boost evidence.
2. A boost-leak test finds the cause in minutes — before the turbo is condemned.
3. Correlate the MAP against a mechanical gauge to rule out sensor bias.
4. Confirm the wastegate/actuator moves and holds before condemning control.
5. The turbo was never replaced — a split intercooler was the root cause.