Persistent P0103 MAF high signal, even after ECU replacement

[andreidx]

Hi everyone,

I’m trying to diagnose a persistent P0103 – Mass Air Flow circuit high input fault on my 2015 Peugeot 508 RXH 2.0 BlueHDi 180, engine code AHW / DW10FC, with a Delphi DCM6.2A ECU.

The fault appears consistently about 2–3 seconds after starting the engine, even at idle and without touching the throttle.

The car originally had the wrong MAF fitted by a previous owner, a Hitachi AFH60-38. I believe the correct MAF for this engine is the Hitachi AFH70-88 / PSA 9681917980 / 1920.RL.

So far I have tested several MAF sensors:

original wrong AFH60-38
another AFH60-38
correct Hitachi AFH70-88
another correct AFH70-88
HELLA equivalent

All of them result in exactly the same P0103.

The interesting part is the live data behavior.

After clearing the fault and starting the engine, Delphi initially shows approximately 2000 mg/stroke for the MAF. After 2–3 seconds, P0103 is triggered and the displayed airflow changes to much more realistic values around 400–600 mg/stroke, which I strongly suspect are substitute/calculated values(and are a little bit higher than the requested by de ECU).

I also performed a test with a correct AFH70-88 connected electrically but physically outside the intake pipe, so absolutely no intake airflow was passing through the sensor. The result was exactly the same:

~2000 mg/stroke → P0103 after 2–3 seconds → substitute airflow values

Electrical checks so far:

MAF supply: approx. 12 V ignition on, 14 V engine running
ground appears good
IAT integrated in the MAF works correctly and displays realistic ambient temperature
MAF signal wire has continuity from the sensor connector to the ECU, approx. 0.3 ohm
IAT wire also approx. 0.3 ohm
no shorts between the checked wires
MAF signal pin is about 5 V with the sensor disconnected
with the MAF connected, the signal is around 2 V
MAP sensor reads approximately atmospheric pressure with engine off

I have also smoke-tested the intake system extensively and found no obvious leaks. The EGR appears mechanically closed when commanded closed and its actual position closely follows requested position. Turbo VGT actual position also follows requested position very closely.

The ECU software was checked by a Peugeot/Stellantis-capable workshop and was confirmed to be the latest original software.

Because the MAFs, wiring continuity and intake system all looked good, I suspected the MAF input stage inside the ECU. I therefore bought another identical DCM6.2A and cloned it.

Unfortunately, the cloned ECU behaves exactly the same. The same ~2000 mg/stroke appears immediately after startup, then P0103 after 2–3 seconds, then the ECU switches to substitute airflow values.

At this point I am thinking about:

signal wire problem that a simple continuity test does not reveal
bad terminal contact at the MAF or ECU connector
ground/reference issue under load
electrical noise/interference on the MAF signal
wrong pinout or previous wiring modification because the car previously had the wrong MAF fitted
some configuration/calibration issue that was also copied during the ECU clone

My next idea is to run a temporary direct signal wire from the MAF to the ECU and compare the signal with an oscilloscope both at the MAF connector and at the ECU during the first few seconds after startup.

Has anyone seen this exact behavior on a PSA Delphi DCM6.2A / DW10FC, especially the temporary ~2000 mg/stroke before P0103 is triggered?

Also, does anyone have reliable information about the way the MAF is wired in this specific model?
Any ideas would be greatly appreciated.

[mickthemaverick]

After replacing the MAF sensor did you reset the BSi according to the procedure in the link below?

viewtopic.php?p=601679#p601679