Plato

W221 m273. I only notice this problem on idle speed. Throttle and MAFS was replaced, plugs/compression/coils are fine, also no errors. There are times when it shows up a lot, there are times when there is almost no defect noted. And I emphasise again it's only at idle. I'm attaching a video (better listen in headphones). I have my own Star Diagnosis, I can take all the necessary readings.

BOTUS

when driving or when stationary ?

if when moving from rest TC lock up clutch ?

Plato

Only on P position (stationary) and never noted it after a journey. Only during warmup. I'm gonna test engine with smoke machine for air leak. Just ordered device - will take some time to deliver.

BOTUS

its won't run funny without some clues showing up in diagnostic fault codes

coil packs and crank position senors are normal - and if you are running on idle to warm up - that will help clog up the cat sensors - 2k RMP ought to be considered min sensible speed of operation on almost all petrol cars - and yes well aware the v8s and worse the v12 like to do lots lower

Plato

There are no errors on Xentry...
Do you mean 2k RPM on idle would be good for engines?? I never seen vehicles with idle speed more then 750 rpm

chassis

Cracked intake duct or manifold. Sounds like unmetered air, but should also be accompanied by misfire and cel.

Replace plugs and coils (yes all coils) with OEM parts purchased from a dealer if this has not been done, regardless of visual inspection results.

BOTUS

for the longevity of the car's engine, and much shorter term the longevity of the CATS below 2000 RPM is terrible - you could try some posh injector and CAT cleaner and give it a good old fashioned Italian tune up

mileage and ave use would help - any test ideas in the dealer tools - they can hold back a spark and test each cyl and then say how much each pot is bringing to the party, they had this in the 80's

Plato

Yeah sounds like that. I have ordered fog machine and will test the system for unmetered air. Will post results here.

Was done a year ago when I removed LPG from the car from previous owner.

Plato

The mileage on this baby is about 210k km. I have dealer tools (Star diagnosis with all software). If you could tell me what and how should I test it will be much appreciated.

Plato

In my case it most likely unmeasured air. I have tested already the system with smoke machine and found a huge air leak. Awaiting for part to fix it and continue further testing. Will share report here one it's done.

BOTUS

I just re read the post - Torque converter lock up clutch ? its normal these play up - not sure why I didn't put that up as an idea first time round....

this should show up as misbehaving in diagnostics though ?

Plato

There are zero errors on dealer's Star diagnosis. Just nothing. When this "defect" goes more stronger there could be missfires recorded but in very rare cases. Like once at 3 months.

So do you mean that you hear is normal?? It doesn't sound like normal especially when it goes too deep down.

BOTUS

just listened to the video - if the AC is having to work a bit and or maybe low on gas or the alternator is pulling a bit more current in strange ways the speed would fluctuate a bit more that you'd want...

- what is it we are supposed to hear ? - sounds pretty normal - nasty rattley PAS, water pump, alternator and drive belts (like some panzer tank having a bad day) - but that Mercedes for you !

Plato

Well... I have finished my research for unmeasured air. I used smoke machine to complete it. According to the results of the tests, I replaced the exhausts gas recirculation hose A2730180582 and two rubber gaskets - right after the MAFS and before the throttle, also changed the gasket between MAFS and air filters (A2720940080, A2731410780, A2721410780). I would suggest to anyone with 273 engine if you haven't changed these parts yet - you should change them asap due to age. The gaskets were hard and thin, and the rubber hose was all cracked and easily broke in half as if it were made of plastic.

Results

• Unmeasured air is not the main issue however it intensified the effect and when engine speed dropped the car would noticeably choke. Now the problem remains, but it is harder to hear. The search continues, the next step is to study the lambda correction.
  
• After replacing the hose A2730180582 the hose itself became very noisy, as if there is a huge hole in it, but it is completely sealed. The noise comes exactly from the intense air movement inside the hose. If the recirculation valve is switched off (plugged) - the noise remains. If you squeeze the hose in any place - the noise disappears. Who knows if it’s normal? I understand that the vacuum creates a lot of pressure and perhaps it should be so, but it is very unusual by ear. After I turn off the engine, the air noise is still audible for a while and at the end it makes a "spit" as a compressor usually does when bleeding air. You can hear it clearly on this short video. Please tell me if this is normal.

Lambda correction

After replacing all parts and resetting all adaptations, I was happy for a moment because the engine ran smoothly. Let me remind you that we are talking only about idling. But then floating (barely noticeable) engine ran returned. And after a while the work was normalised again for a moment and then again went floating. What should I check next? I don’t believe changing lambda sensors is a good idea as both were changed not so long ago (2-3 years) and it is unlikely that they failed at the same time and behave equally wrong.

Fig 1. This is how the lambda correction should look like on a good engine and this is how my engine works perfectly.
UPDATE: Wrong! See correct graph in my last post.



Fig 2. This is the picture when the engine start to float. As you can see the long term regulation starts to float evenly from 0.9 to 1.10.



Fig 3 This is the picture when the rpm was floating, then became OK, then floated again.

BOTUS

can we get a picture of the part you changed gaskets on please ?

spark plugs and coils packs ?

had mine melt down and popped another coil pack on cyl 2 two weeks ago - I swapped with a secondhand one I had, that measured same resistance on the only two pins (of four) you can get any reading from - it was like all the other good ones on the car - then checked the dead one and it was measuring numbers like dead ones do - misfire went, did a 70 mile drive and suddenly same pot threw teddies again !!!

and so of my 3 spare coil packs I always have with me when I drive this thing, I'm down to my last one - at least u can do any of them in 10 mins start to finish with just one tool

now contemplating aliexpress £15 each junk - as Genuine Merc ones are the worst in the world



.

Plato

Spark plugs is not a reason here. There is no missfires.

Plato

Interesting update. Today I was monitoring the lambda correction and decided to remove the plug from MAFS - the long term correction went up, but in general the floating went away. Threw in a new MAFS and for about 10 minutes I was happy... It worked smoothly and the lambda graph was perfect. But then those waves came back again... Looks like one of the "points" of adaptation is floating. But which one? Where to start?

BOTUS

one of my bikes was in and out of stupid for the last 3 years - then this year a CAT sensor died - its been less stupid ever since I fitted a new one

Plato

I also noticed that the lambda graph BEFORE the catalytic converter starts to float only when the lambda graph AFTER the catalytic converter "gets into its rhythm" and starts to give a stable sinusoid. Immediately after the start while the second ones have not warmed up yet and have not switched into their rhythm on the graphs everything is ok. But this is on a warmed up engine.

jerellorenzo

just curious if you think it's best to just replace all the cam position sensors when the first one goes? I replaced one - it had oil coming out the wire harness and the code was a short so seems to me that was the issue. Car was great for a month. Did the sputter and back on with the light. Xentry was doa, seems my power supply is bad. Cascade of BS... Anyway - presumably i now have another sensor short - seems advisable to do it all after one fails. Also read that swapping out the magnet sensors when you do the cam position ones is recommended. I asked the dealer and they said no - so presumably the correct answer is yes, and they will just do anything to get you in the garage... God willing my Xentry works with the new supply. Never easy.

Plato

these sensor are good on my car. no leak or smth.

still looking for help on forum

jerellorenzo

Wish i could offer up anything useful. Mine idols rough at start for 10-15 seconds but then always starts purring. Frustrating these rouge issues. Carbon build up or something crazy? I had a p03402 and p03401 after a miss. replaced 1 sensor - worked great for about a month light is back on, Xentry charger is broken ( what luck ) so i used the cheap-o one and p03401 - came back, presumably it's one of the 7 of 8 cam sensors or magnets i've not yet changed. Time will tell. Hilarious when the diagnosis computer doesn't even work... Maybe it's like a broken motor mount or something. I don't see the link on here anymore so can't listen.

kevm14

I tried to follow the lambda sensor waveform theory but I got lost.

Plato

In my case there is no missfires and no errors

Plato

Not so complicated. But Im not a good guy who can explain. I just learned it few days ago a little bit but do understand now how short and long terms correction works and how it looks on graph. you can send you graph I will try to decode it for you.