Hello all,

 

I have problems with my engine BKY and mixture, ECU is Magneti Marelli 4TV, manual gearbox, without EGR.

Symptoms: 
- high CO emissions, high HC (both idle and rised) and smells bad (previous owner has mounted emulator, so was without faults codes)
- problems with rich mixture - wideband lambda 1 permanently between 0,42-0,45V
- jumping short time fuel trim - in 10 sec idle even between -10% to 4% (all others logs I have seen are more stable)
- empty long time fuel trims - checked many times in month - always empty... 
- sometimes "kangoo effect" when starts off - fast move, lost some power for 1 sec and next normally run
- strange fuel consumption, once 8l for 100km, once 10l for 100km at those same roads.
- 1 lambda probe period is 1.36s (block 034). I have read that should be less than 1s, but passes test in VCDS 

Facts:
- 1 lambda is universal substitute NTK (cheaper than dedicated) changed 2 years ago because old had heater fault
- no faults codes, checked also most ECU blocks with spec
- engine load at idle shoud be 10-20%, in my car 22,5% 

Until now done:
- replaced cat - universal substitute (not tested emission, but without errors and smells "good"), read 2 lambda is quite good, but not perfect (but cat shouldnt causing my problems)
- cleaned and adapted throttle body
- replaced EVAP, old had leaks in vaccum
- tested vaccum leaks, changed most seals
- tested piston compression (11,2 | 10,2 | 10,8 | 12 bar)
- cleaned oil separator
- basic injector test - no leaks, all have similar stream

In plan:
- testing fuel pressure
 

All things I done in last 3 months, and still have problem with rich mixture, empty long time fuel trims, strange fuel consumption and sometimes "kangoo effect" - its dangerous expecially at juctions... 
I'm afraid that too rich mixture will kill my new cat and still is problem with emission.

Why ECU dont have long time trims? Why short trims are extremely jumping at stable idle? Because of universal lambda probe? 
What should I also check or replace? Which logs can by helpful?
I thought about exhaust leaks, but cant seen them (I can replace seal) and about engine tightness test (but compression is ok?).  

Please help me and give me any advices 

PS. Sorry for long story and my english in some parts, I'm not native.

Hi, I have exactly this engine in a Polo, so I should be able to make some comparison measurements. The car is not road legal at the moment, but I can run it in the garden. 

 

I guess there's nothing obvious like a coolant temperature sensor reading -40°C at ambient temperature or anything? Have you checked the PCV valve diaphragm by feeling for suction at the pressure equalising port?  

 

8 hours ago, Wino said:

Hi, I have exactly this engine in a Polo, so I should be able to make some comparison measurements. The car is not road legal at the moment, but I can run it in the garden. 

Thanks for your investment, if you can make comparison, It will be great! 

 

8 hours ago, Wino said:

I guess there's nothing obvious like a coolant temperature sensor reading -40°C at ambient temperature or anything? Have you checked the PCV valve diaphragm by feeling for suction at the pressure equalising port?  

 

All readings in VCDS, except strange 1 lambda and short & long fuel trims and little increased engine load look normally 
Tested with workshop-manuals

 

I cleaned oil separator and PCV valve last month, also tested other one used and no difference. Even tested pipes and blocked up entry in airbox, still this same lambda readings.

Ignition plugs and coils also new from 2020... Problems started in 2018 or 2019 with little "kangoo effect", but at beginning I didn't have time, tools and knowledge to overwhelm them. In this year was problem with cat and I bought cable for VCDS and want to diagnose my problem.

I did some logging just now, will look at and share results tomorrow.

I will need to do some more, longer duration logging/driving I think, I was in a hurry and didn't log all the best channels.

Here's what I logged. Forgot about the fuel trims on channel 32, but had a quick look after that log and both were at zero.  

Not sure whether it's cos the engine had only run for the time of the log since the battery was last disconnected, or maybe the car has to actually move off the spot to trigger the 'learning'. This was all done parked, but with a bit of playing with the accelerator pedal now and then, as you'll see from the rpm column in gp34

Ambient temperature probably just below 20C.

 

You'll see that after 50 seconds the wideband sensor stabilises and maintains 0.45V after that. It's only after 415 seconds that the dynamic factor starts reading, but it is reliably below 1 after that.

I have no idea what the rightmost column is, in group 36.

 

 

Lambda stuff 9N3 040821.xlsx

Look closer at the wide band lambda. That fixed response looks suspicious. 

 

If its idling, steady state, give the throttle a good stab, Rev the engine. You should see the lambda go rich for a short while. If it doesn’t move, then your lambda sensor is duff.

Is that aimed at the OP, or me?

 

1 minute ago, Wino said:

Is that aimed at the OP, or me?

 

Aimed at the OP.

ive seen these sensors fail in this way before.