Turbo MAF sensor faults: symptoms, tests and cleaning

Symptoms owners blamed on the MAF

• Struggling to idle when the MAF is plugged in. With it unplugged the car runs, but lumpy and in a safe mode with no boost. One owner's scan showed P0100 (MAF circuit) plus a misfire code (P0300).
• A general "off" feel, misbehaving on part throttle and the recirculation valve not working properly at times ("turkeys"), and the ECU seeming to rein the power in even though boost readings didn't change.
• Going into limp mode or the turbo "going on strike", which cleared when the MAF was unplugged. One owner had this every year for two or three years.
• An airflow fault with a high airflow reading code, about 75% power, revs hunting at idle, and the car driving fine with the sensor unplugged but chugging and coughing with it plugged in.
• With the MAF unplugged the engine warning light comes on and the exhaust can pop more on the overrun, and the car runs rich (one owner).

Testing

1. Unplug it and see. Owners did this as a quick test. If the car runs better with the plug pulled, that points at the sensor, but one owner who tried it wanted another way to confirm before spending money. The warning light comes on, and one owner understood the ECU also stops taking other readings, such as the lambda, but that wasn't confirmed.
2. Measure the output. One owner says check the voltage between pins 3 and 5, which should be about 1.2 V at idle and rise to about 4 V at full throttle. Back-probe the connector rather than cutting the wires, and the plug has an earth, a 5 V and a 12 V feed on other pins (his recollection). Another owner who tried found 1.3 V at about 1,000 rpm and 2.4 V at about 4,000 rpm, but only revved the car stationary, so couldn't say it was low. A third owner quoted a workshop data source as saying the output is just over 4 V at full throttle and wondered if his reading meant the sensor was on its way out.
3. Use a scan tool. An OBD reader that shows the MAF reading in grams per second is the alternative. The reading should rise as you rev the engine, even in neutral. One owner says the MAF scale goes up to about 200 g/s.
4. Swap with a known good one. The cleanest test is a borrowed working sensor for a day. Bear in mind that if it is mapped for a particular MAF size, a borrowed one must be the same size.

Cleaning

• One owner advises petrol or brake cleaner (toluene), because oily crankcase vapours coat everything in the tube. Another says if the sensor is contaminated with oil, wash it with petrol or white spirit.
• One owner gave it a very gentle clean with a delicate alcohol-based contact cleaner, pulled the fuse to reset the ECU and the car felt much better for about a week, then the problem returned. He wondered if the reset alone helped by making the ECU forget bad readings, so a quick improvement may not be proof the clean worked.
• Another owner sprayed the sensor with a petrol-engine EGR cleaner in short bursts, with the MAF in the intake hose but unplugged, while the engine idled at about 1,500 rpm. It improved the car for about 48 hours, then failed again.
• One contributor says a sensor should last years. The owner with yearly failures ended up fixing his by cleaning the connectors behind the rear wheel arch. See Throttle failure and the connectors.

Air filters

• Several threads say MAF sensors fail with anything other than an original paper filter. An over-oiled cone filter is a suspect (one owner's view), and a sensor can fail even without an oiled filter.
• One owner likes a dry filter (an Apexi) because it won't foul the MAF, though it may flow less freely than others.
• An owner running a standard paper filter had the same trouble, so a filter isn't always the cause. A bigger turbo (a VXR K04) was ruled out as a cause by another contributor.

Sizes and the ECU map

• One owner's sensor, a Bosch 0280218211, was identified by another as the 80 mm Z20LEH (VXR) sensor, an upgrade rather than original. The owner measured the bore at 71 mm, so the thread didn't agree on the size.
• A different owner says an 80 mm MAF needs ECU calibration, so a car mapped for a 70 mm sensor may run badly. Check the size against the map. A replacement 80 mm unit that was fitted to a car whose map hadn't been changed was suggested as a possible cause in the airflow thread.

When it isn't the MAF

• Wiring. One owner had the same fault, and the wiring to the MAF was at fault. He advises continuity testing before spending money, since it could be a five minute soldering job. Another suggests bending the loom around with the old sensor connected to see if the fault changes, and cleaning the plug.
• The connectors behind the rear wheel arch. See above.
• Ignition. A failing coil pack and plugs were suggested more than once. See Spark plug replacement.
• Air leaks. Rev hunting at low rpm, especially at idle, is often a leak in the air path before or after the turbo (one owner). Also check for a split pipe. A collapsing intake pipe was another suggestion but ruled out where the pipe is alloy.
• Boost control. The Turbo has two boost solenoids, and an owner says one is more prone to fail. See MAP and lambda sensor faults.
• Throttle body. A sticking butterfly was suggested. The owner whose car drove fine with the sensor unplugged argued that doesn't explain it.
• A new MAF that didn't help. One owner bought a new MAF for about £100 and it made no difference. The cause turned out to be the lambda sensor.

One owner's opinion is that the only true cure for repeated MAF problems is to ditch the MAF and fit a different ECU. That is an aftermarket route, and others in the thread didn't comment.