Diagnose an O2 sensor before you blame it

Incorrect: An exhaust restriction cuts power at higher rpm; it does not cause a lean trim that is worst at idle.

Incorrect: A misreading mass airflow sensor usually skews trims across the airflow range; trims that fall from +23% to +5% as airflow rises point to a fixed leak instead.

Incorrect: Low fuel pump pressure shows most when fuel demand is highest, yet here the lean correction fades as load rises.

Correct: A large positive trim at idle that nearly vanishes at cruise is the signature of unmetered air; at idle a small leak is a big share of total airflow, and a smoke test finds it.

Incorrect: The upstream O2 sensor is reporting a real lean condition that the trims correct; a new one would see the same leak.

Incorrect: A restricted exhaust costs power at higher rpm and does not cause a high idle.

Incorrect: A MAP sensor that under-reports would show a low voltage, but this one reads high at 1.8 V.

Incorrect: A shorted coolant temperature circuit reads hot and trims idle enrichment; it does not explain high MAP with the idle air control nearly closed.

Correct: Unmetered air leaking into the intake raises manifold pressure, so MAP reads high, and the PCM pulls the idle air control nearly shut trying to bring rpm down; the leak keeps idle high anyway.

Incorrect: Carbon in the throttle bore chokes idle air, so the PCM opens the idle air control wider and idle tends to sag, not climb.

Incorrect: A real lean mixture from a vacuum leak would also read lean at the downstream sensor, not a steady 0.8 V.

Incorrect: Low delivery from a restricted fuel filter makes the exhaust truly lean, which the downstream sensor would confirm instead of reading rich.

Incorrect: An under-reading MAF sensor leaves the mixture truly lean, so the rear sensor would not sit rich.

Incorrect: Outside air from an exhaust leak would also reach the downstream sensor, so it would not read a steady rich 0.8 V.

Correct: The upstream sensor sits at 0.05 V and never moves while the downstream sensor reads rich at 0.8 V; the trims are adding fuel to a mixture that is already rich because the front sensor is wrong.

Incorrect: A faulty mass airflow sensor would hold trims off by a steady amount, not walk them back toward zero trip after trip.

Incorrect: An oxygen sensor stuck lean would keep driving trims positive, but STFT sits near zero and LTFT keeps dropping.

Correct: LTFT drops from +14% to +3% over three trips while STFT stays near zero: the adaptive memory is unlearning the correction it built for the old vacuum leak.

Incorrect: A second intake leak would hold LTFT high; this trim keeps falling.

Incorrect: A fuel pressure regulator failing open lowers fuel pressure, which would push trims up over time, not down.

Incorrect: A weak fuel pump would lean out both banks, mostly under load, not one bank at idle.

Incorrect: A torn intake boot ahead of the throttle lets in unmetered air that feeds both banks, so Bank 2 would read lean too.

Correct: Only Bank 1 reads lean, and only at idle; a Bank 1 manifold leak draws in outside air between exhaust pulses at idle, fooling that sensor, and matters less as exhaust flow rises.

Incorrect: A restricted fuel filter starves both banks and gets worse with load, yet Bank 2 is normal and the lean reading fades at 2,500 rpm.

Incorrect: A mass airflow sensor that reads low shortchanges fuel to both banks, yet Bank 2 trims are normal.

Incorrect: A MAF sensor that under-reads would not pull manifold pressure 3 inHg below BARO at WOT; the MAP reading confirms airflow really is restricted upstream.

Incorrect: An exhaust restriction also cuts airflow, but with the throttle wide open manifold pressure stays close to BARO instead of dropping 3 inHg.

Correct: Both readings say less air is getting in: MAF is short of spec and manifold pressure sits 3 inHg below the 30.0 inHg BARO reading at WOT, the pressure drop of a clogged air filter.

Incorrect: At WOT there is almost no vacuum to pull air through a leak, and a vacuum leak would raise manifold pressure, not lower it.

Incorrect: One faulty MAP sensor cannot also make the MAF read low; two readings that agree point to a real restriction.

Incorrect: Clogged injectors would also raise pulse width, but they do not pull rail pressure down to 30 psi.

Incorrect: An exhaust leak before the upstream oxygen sensor gives a false lean reading that raises trims without touching fuel pressure.

Incorrect: An oxygen sensor fault could push trims up, but the low gauge reading shows a real fuel delivery problem.

Incorrect: An intake vacuum leak explains positive trim, but not fuel pressure sitting 15 psi below specification.

Correct: Fuel pressure is 15 psi under spec, so each injector flows less per millisecond; the PCM stretches pulse width and long-term trim to make up for the restricted filter.

Incorrect: The new sensor's heater element tested within specification at its connector, so the element is not the problem.

Incorrect: An open in the heater control circuit stops current completely; the meter would read zero, not 0.1 A.

Correct: The heater element itself checks good, so the tiny current means the wiring or connections are limiting it; high resistance lets only a little current through.

Incorrect: A short to ground lowers resistance and would make current spike or blow the fuse, not fall to 0.1 A.

Incorrect: A blown fuse cuts the circuit, so no current would flow at all.

Incorrect: The oxygen sensor is reporting the lean condition the trims confirm; nothing here shows it is faulty.

Incorrect: Trims at their adaptive limit with a lean signal still present are far outside normal fuel system operation.

Incorrect: A restricted catalytic converter costs power and does not by itself push trims to their lean limit.

Correct: LTFT pinned at +25% with STFT still adding fuel and the sensor still lean means the PCM has no correction left; the cause of the lean condition still has to be found.

Incorrect: A low-reading mass airflow sensor is one possible cause of the lean condition, but the trim numbers themselves only say the correction is used up.

Incorrect: A clogged fuel filter leans the mixture, but propane at the manifold would not pinpoint it, and MAF and MAP would read normally.

Incorrect: A biased MAP sensor would not make the MAF read low or make the trims react to propane at the manifold.

Correct: MAF reads below spec while MAP shows weak vacuum, and propane fed near the manifold swings the trims: air is getting in after the MAF and is never measured.

Incorrect: A MAF sensor that under-reads would leave manifold vacuum normal; the 68 kPa MAP reading shows extra air really is entering.

Incorrect: A restricted exhaust raises manifold pressure under load, but it does not cause lean trims that react to propane.

Incorrect: Extra resistance in the injector circuit slows the rise and lowers peak current, the opposite of 1.8 A.

Incorrect: Low fuel pressure affects how the pintle opens, not how much current the injector coil draws.

Incorrect: An open injector winding passes no current, so there would be no ramp at all.

Incorrect: A failing PCM driver usually limits or drops current; it would not push a normal coil past its peak.

Correct: A shorted injector winding has little resistance, so current rises fast and overshoots the 1.0 A spec with no hump from pintle movement.

Incorrect: A mass airflow sensor that under-reads would keep trims positive warm or cold.

Incorrect: A leaking fuel injector adds fuel, which pulls trims negative, and a constant leak would not clear up once warm.

Incorrect: A stuck-open thermostat would keep the engine from reaching 95 °C, and these trims show lean, not rich.

Correct: Trims are high only while cold and return to normal at 95 °C; a gasket that leaks until heat expands the metal and seals the joint fits that pattern.

Incorrect: Purge vapor from a stuck-open EVAP valve makes the mixture rich, driving trims negative, and would also act at operating temperature.

Incorrect: A biased oxygen sensor could fool the trims, but it cannot raise measured HC to 450 ppm on its bank.

Incorrect: A restricted return line raises rail pressure for the whole engine, so both banks would run rich.

Correct: High HC from Bank 1 alone shows real unburned fuel, and trims pulling fuel only on that bank point to an injector dripping there.

Incorrect: A vacuum leak leans the mixture, so trims would go positive, not to -18%.

Incorrect: A mass airflow sensor error changes fueling for both banks equally, yet Bank 2 is normal.

Incorrect: A binding pintle shows up as a missing or late pintle bump, yet this bump is normal.

Incorrect: A shorted injector winding draws current faster and higher, the opposite of a slow, low ramp.

Incorrect: An open circuit in the engine control module's driver would stop current completely, not just slow it.

Correct: Resistance in the feed circuit slows current build and caps the peak, while the injector itself opens and closes normally, as the clean pintle bump shows.

Incorrect: A weak return spring changes when the pintle closes, which would move the bump, and the bump here is normal.