4 replies to this topic

[tjspragg]

I'm looking for some advice on an ongoing (2 years and counting) fault with my 1998 MPI with a MEMS 2J ECU. I've been trying to diagnose it methodically, but I'm now running out of ideas.

 

The problem:

• Fails emissions test (in Sweden)

• Slight but noticeable stutter around 2500-3000 rpm, both under load and when holding the engine at a steady RPM.

• Otherwise drives reasonably well.

• Idle is generally stable, although a little rough.

• Slight petrol smell in the cabin.
   

I've been using the RoverMEMS online tool to monitor live data in the car.

The logs consistently show:

• ECU enters closed loop normally.

• Coolant temperature appears correct.

• MAP readings appear sensible.

• Primary and secondary sync remain present.

• At 2500-3000 rpm, the fuel feedback/adaptive correction repeatedly climbs to 119-120%, suggesting the ECU is trying to add as much fuel as possible.

• Lambda voltage spends much of its time indicating a lean mixture.

• No significant fault codes

Work already carried out

• New battery.

• New coil pack.

• New HT leads (all tested at approximately 5 kΩ and consistent).

• New coolant temperature sensor.

• New NGK BPR6ES spark plugs, all gapped to 0.80 mm.

• Spark plugs all show very similar, healthy colouring with no evidence of one cylinder running differently.

• HT lead routing checked and confirmed correct.

• Exhaust inspected by a mechanic with no leaks found.

• Cleaned ECU/engine earth connection.

• Cleaned battery terminals.

• Found and repaired a cracked vacuum hose to the fuel pressure regulator.

Engine specification

The car has a genuine John Cooper S Works Plus conversion, including:

• S Works cylinder head

• 1.5:1 roller rockers

• High-pressure fuel boost valve

• K&N filter and S Works induction pipe

• Maniflow LCB manifold

• Maniflow exhaust system

The conversion has been on the car for many years and it has previously passed emissions without issue.

 

Some questions:

Has anyone seen an MPI behave like this?

What could cause the ECU to consistently reach 120% fuel feedback?

Would you suspect fuel pressure, the regulator, lambda sensor, or something else?

Are there any common MPI faults that fit these symptoms which I haven't considered?

 

Any advice would be much appreciated - I'm starting to pull what little hair I have left out over this!

[68+86auto]

Can you give us the actual values rather than the summary you have given (which I suspect is AI?)

 

For example, "MAP readings appear sensible." cannot be confirmed without knowing the values.

Anyway, with the fact it happens at a certain rpm range regardless of the load, I'd test all the sensor signals with an oscilloscope.

Edited by 68+86auto, Yesterday, 12:18 PM.

[tjspragg]

Thanks for the reply! Using an oscilloscope may be beyond my budget right now, or at least beyond my skills - but I will look into it.

You asked for data, this is what I have (and much more) from MEMS 2J ECU readings:

 

Ignition on, engine off:

MAP: approximately 101 kPa

 

Warm idle:

Engine speed: approximately 817 rpm

Coolant temperature: approximately 91°C

MAP: approximately 42 kPa

ECU supply voltage: approximately 13.1 V

Fuel feedback: elevated, approaching the upper correction limit

Closed loop: active

Primary sync: present

Secondary sync: present

 

Held at approximately 2500 rpm:

Fuel feedback: rose to 120%

Lambda reading: predominantly indicated lean

Primary sync: present

Secondary sync: present

 

Held at approximately 3000 rpm:

Fuel feedback: reached and remained around 120%

Lambda reading: predominantly indicated lean

Primary sync: present

Secondary sync: present

Noticeable stutter at this engine speed

 

Fault codes:

No relevant fault codes except oil-temperature and outside-air-temperature faults, but these are maybe ghost faults on the Mini MEMS 2J installation?

[mab01uk]

These PDF files in link below are also worth a read if you have not already seen them.

 

Mini Mpi Diagnostics - Car Mechanics Magazine:-

https://www.theminif...anics-magazine/

 

[NLinPEN]

These two sentences from your opening post:

• At 2500-3000 rpm, the fuel feedback/adaptive correction repeatedly climbs to 119-120%, suggesting the ECU is trying to add as much fuel as possible.

• Lambda voltage spends much of its time indicating a lean mixture.

give me a clear impression of the engine being too lean. Either not sufficient fuel is delivered to the engine. Or there is an air leak into the engine, thus messing with the air-to-fuel ratio.

Looking at the fuel delivery side: how old is your fuel pump? Are you certain that it pumps sufficient fuel? Your fuel filter isn't (partially) blocked?

Looking at the air leak side: have you verified the air intake, all the way from the air filter until the engine block? No leakage there? No hissing sounds? Another trick to look for air leaks is to spray some brake cleaner (or similar) to those pipes and hoses and see whether that influences the engine temporarily. If it does, you have a leak at that location.

 

Out of curiosity: what kind of emission test did your car fail? What parameter was tested, and what were the results?

Edited by NLinPEN, Today, 08:15 AM.