Both Banks Lean, but Only at Idle
A 2006 V6 with 98,000 miles sets P0171 and P0174. Freeze frame at idle: 620 rpm, MAF 3.1 g/s, long-term fuel trim +24 percent on bank 1 and +22 percent on bank 2, short-term trim +8 and +7, ECT 190 F, IAT 90 F. Raised to 2,500 rpm in park, long-term trim falls to +6 and +5 with short-term near zero. Fuel pressure is 58 psi against a 55 to 62 spec and holds for 10 minutes after shutdown. There are no visible cracks, and the engine idles smoothly.
1. What does the idle-versus-2,500-rpm trim pattern tell you, and which of the three candidates does it eliminate?
Read the worked answer
The trims correct as airflow increases, which separates the candidates:
- Unmetered air (vacuum leak) downstream of the MAF: leaks flow a roughly fixed volume. At idle that volume is a large share of a small total air mass, so trims go very positive. At 2,500 rpm it is a small share of a much larger air mass, so trims fall. This matches the data.
- Under-reporting MAF: the error is proportional to airflow, so trims would stay high at 2,500 rpm. Eliminated. The 3.1 g/s idle reading is also plausible for this displacement.
- Fuel delivery weakness (pump, filter, restricted line): the engine needs the most fuel at high load and high rpm, so trims would worsen when airflow rises. Eliminated, and the 58 psi pressure with a clean hold supports that.
The pattern is classic unmetered air at idle. A leak large enough for +24 percent on both banks should also raise idle speed slightly, so look at PCV, brake booster, intake gaskets, and the intake tube between MAF and throttle body.
2. A technician adds propane at the throttle body. Long-term trim drops from +24/+22 to -6/-5. What does that prove, and what does it not prove?
Read the worked answer
It proves the control loop and the oxygen sensors are working: when extra combustible gas is supplied, the PCM sees the richer exhaust and pulls fuel, and it has enough authority to cross from positive to negative. So the lean reading is real, not a biased sensor or a maxed-out trim table.
It does not locate the leak. Propane added at the throttle body is drawn in through the same path as the leak, so it enriches the mixture whether the extra air is entering through a booster hose or a gasket. It also does not distinguish a vacuum leak from a small MAF error by itself; the idle-versus-load pattern from the freeze frame is what made that distinction.
Use the propane result only to confirm that the system can correct, then move to locating the source.
3. List the pinpoint tests you would run, in order, and what result would make you stop and look elsewhere.
Read the worked answer
- Smoke test into the intake with the throttle closed. Watch for smoke at the booster, PCV, intake runners, and throttle body gasket.
- Pinch or plug the brake booster hose at idle and watch trims. A 10-plus point drop confirms the booster or its check valve.
- PCV system: pinch the fresh-air side or temporarily plug the valve. A stuck-open PCV is a common idle-only leak.
- Spray test around intake runners with a combustible spray, watching short-term trim for a negative jump.
Stop and look elsewhere if the smoke test and pinch tests show nothing and trims remain +20 or more at idle. At that point re-examine the MAF signal at idle against a known-good vehicle, check for a misreporting MAF at low flow, and verify the throttle body is not carboned enough to be bypassing air past the plate in a way the PCM does not model.