W204 320 CDI zero boost

[Freeform]

Hello,

Regular reader but (one of) my first posts - as most questions I have already found the answer here.

Problem with my 2008 W204 C320 CDI (OM642, no DPF), 88000 miles.

This week, Monday morning, I encountered a new issue: on starting out for the commute (about 75 miles of mixed & motorway) I noticed I had little power - not enough to cause the transmission to kick down, and struggled to get up to speed on the on-ramp to M25.
I pulled off at services, keyed off, waited, keyed on and continued with power restored.
Nearer the office, I encountered traffic and, when that cleared and I accelerated down the on-ramp, the power had gone again. I arrived at work.
No fault indicated.

Lunchtime, went for a quick test drive and all the power was available (there was little traffic). Again, on the journey home, and both ways on Tuesday, there was no problem.

Today, Wednesday, I set off monitoring with the TorquePro android app; OK on starting off but there was slow traffic before joining M25 and accelerating down the on-ramp there was no power and boost registered zero. I pulled in to services, key-cycled and continued with boost and power restored. All OK until I hit the traffic nearer my office and lost boost. This time there was a fault and I had time to log: P0237 - Boost sensor A circuit Low.
I key-cycled once, boost returned and I arrived at office.
Lunchtime, performed fault scan with Carly for Mercedes app: Code 119E00 'no explanation available' and OBD00 P0237, as above. I am fairly sure P0237 is just a symptom of boost not matching the engine control module's expectation.

There appears a pattern developing; the lost boost condition occurs when the care is in traffic, at little or no load. On boost, I don't hear any boost leak.

as it has just occurred this week, I have not had the opportunity visit anyone for a STAR diagnostic session.

Help and advice appreciated. I can measure and understand engine parameters with TorquePro or Carly.

With thanks,
Andy

 

[kid-jensen]

Maybe check the turbo actuator is actually moving and not going sticky?

 

[Lit91]

Also think that it may be the actuator, common problem. May be worthwhile checking the MAF's as first port of call though.

 

[Stegel]

My 2006 CLS CDI, with 117k miles had the same issue - loss of power, no kickdown and occasional EML. It occurred with ever greater frequency over the last few weeks - initially it only happened after a warm re-start, but towards the end would kick in a little while after a cold start and turning off/on rarely resolved the issue.

I've just collected it from the Indy - STAR indicated a turbo fault, and I was advised to replace the turbo as a whole as actuator-only repairs frequently fail and require the same amount of labour. I have only done a few miles yet, but it feels far crisper and responsive than before, and most noticeably there is no cloud of smoke on acceleration.

 

[Freeform]

Thank you. I didn’t get a chance yet to examine the actuator function yet. It crossed my mind. The weather has been miserable, dark or both; misbehaviour has been so far unpredictable and gone are the days there was an accelerator cable to pull from the engine bay! I shall enlist a colleague tomorrow lunchtime, with luck… At least it can be seen on this car. Is this something that can be fixed without major work or expense, or does it lead to turbo replacement?
I am suspicious that, if it is the actuator, there is a reason the actuator is failing - and that suggests to me the vanes are seizing(?).

I didn’t reset the Diagnostic Trouble Code P0237 but it seemed to clear itself in the course of the subsequent drive & key cycles.

Though I have never seen an intermittent mass-flow meter, I added MAF to my list of monitored channels. As a plausibility check, I got a maximum mass-airflow of 360 grams/sec @4000 rpm @1.6 bar boost (things were happening too fast to make sure all these numbers occured at the same instance but let's say they did!). Quick calculation shows that to represent 90% volumetric efficiency so that suggests to me the MAF sensors are about OK. There's no lack of push when it's on boost!

Was a short drive this evening so no failure.

[An aside:
I have some experience of MAF sensors of the types currently used: I worked for the biggest automotive supplier and had access to one of its chief product experts, so I became aware of common failure modes, the likely causes, and symptoms. This was usually gradually reducing engine power through diminished sensitivity through contamination.
Some OEMs at the time were experiencing field failures and we visited to explain the design evolution, common failure modes, likely causes of failure (those we accepted as our own and those we considered could be caused by end user customer).
For our own side, we had through successive iterations optimised the air path past the thick film hybrid sensor element, with an exit path for debris and oil mist (heavier than the air so more inclined to follow a straight path) and improved the stability of the protective gel on the circuit board to prevent leaching on to the hybrid film.
For the customer side, we considered causes to be intake design, end-users fitting aftermarket air filters using filter oil and possible over oiling of the same, and ‘chip tuning’!
We didn’t recommend cleaning (we wouldn’t!).
From a diagnostic perspective, there was not always a fault registered. Confirmation of reduced sensitivity could be achieved by disconnecting the electrical connector; a fault would register, of course, but it forces the engine to default map based upon MAP and engine speed that gave reduced power but perhaps not as reduced as a failing MAF sensor.]

 

[Stegel]

I didn't get a chance to check the actuator when the car was in fault, but certainly the actuator was still moving when the engine was revved, which suggests, as you speculate, the vanes were causing the issue and hence the complete turbo replacement recommendation.

 

[Freeform]

I had meant to post the Freeze Frame data for the P0237 fault occurrence:

Engine load 20%
Coolant temperature 87°C
Engine speed 596 rpm
Speed (OBD) 21 km/h
Intake air temperature 12°C
Mass air flow 13.19 g/s
Fuel rail pressure 27280 kpa

​

This appears typical of the instances of boost failure but is so far the only time the fault code was triggered.

​

 

[Freeform]

Subject to the issue returning and making me eat my words, the problem is solved!

As the logged fault was "... circuit low", I took it more seriously and temporarily substitute a not quite suitable boost sensor from a scrap OM651 test engine at work and no further incident occured.

The correct replacement sensor is now in and working and all is back to normal.

[Confession: I made a mistake in my MAF arithmetic, above. I realised later, the 360 g/s of air registered by the MAFs was a peak signal range check sweep at key-off; the actual peak value I saw live was 208 g/s, and the engine (being a 4 stroke!) takes two revolutions to swallow 3 litres x max boost, equalling 200 g/s!:failAt least, I think I am right, now...]
Thanks for the input.

 

[cws196]

I can't help but admire the way you've meticulously investigated the issue - I bet you could teach many MB dealer techs a thing or 3.

 

[Freeform]

Thanks, CWS196. I was a bit thrown at first by there being no fault recorded until the third day, so I initially treated it as a symptom - but "... circuit low" suggests a wiring or component fault.
"...Circuit Range/Performance" might have warranted more thought.

Something else that might be useful to others experiencing similar:

The last time boost failed before I swapped the sensor, I left the engine running. (On the V6, with the plastic engine cover removed, standing one foot on the ground and the other on the accelerator pedal, one can see the vane actuator arm under the back edge of the open bonnet.)
The arm was pointing straight up and did not move when the engine was revved.

I keyed off. The arm immediately swung to its lowest position, gave another kick and settled, as if performing a self test [Many functions self-test at key-off (part of the ECU 'after-run') - some other functions when driving, on the over-run.]

On starting the engine once more, boost control had returned, as previously related. On revving the engine, the turbo vane controller modulated the vanes. I got no impression of stickiness in the vane mechanism - which gave me confidence that both turbo vanes and controller were OK.
​

I have been using 'Carly for Mercedes' Android app. I find it very useful, interrogating every module in the car, not just the engine control module, and I have had good support from the Carly team. I have no experience yet with STAR.

 

[Freeform]

Well, this evening, I got another no-boost event! Not so smart after all.

A colleague has STAR so will try looking with that some day next week.

More later...

 

[Freeform]

For those still following:

Since Wednesday’s setback with the delayed return of the problem.

I discovered that, though the MIL does not always light, the P0237 ‘Boost sensor circuit low’ fault is now always present (I was a bit slow there and should have checked more often).
I thought, if ‘…circuit low’ means electrical fault, how is it I can read boost pressure at the OBD socket, surely it should not work at all?
The fault will not reset with the engine running. It will reset with the engine stopped (and key-on, of course).
Once reset, the fault will recur almost immediately after starting.
All this made me think perhaps I do have a boost leak?
I don’t hear any leak but I am not sure I would pick it out from all the other noises – intake, engine, exhaust etc.

I looked online for other reports of Mercedes with P0237. Some of them were traced to leaks.

​

The car is booked for an oil service on Monday, at my favoured garage that is not a Mercedes specialist but with a lot of experience including race and modified turbo cars. He has agreed, we will perform a boost leak test with pressure and smoke.

[Other thoughts:
If due to boost leak, why is the zero boost event only triggered in slow traffic or stationary driving? Suppose this occurs if, during the faster part of the drive, the boost error vs. MAF air in exceeds a certain threshold. The boost may not be cut immediately and the engine sent into ‘limp mode’, when the car is under load as this could be dangerous but the ECU may wait until the car is near rest so the limp mode may be safely triggered. (I am just surmising, here).
Why does the MIL not always light (only twice, so far)? The MIL need only light for an emissions relevant faults and only then if there is no countermeasure to keep emissions within OBD limits. (There is some discretion allowed here; an OEM may choose to alert the driver of any fault, or if the car can be kept within OBD limits may choose not to alert the driver and leave a less serious to be discovered at service.)
What about the other potential failures of turbo actuator or inlet flap motor? I don’t get either of the fault codes associated with these. I have bought a replacement turbo inlet seal, will get that area and the flap motor cleaned up and try to verify that the flap system operates correctly.]

 

[Stegel]

My car (see above) would go into the "flat" mode, no codes, no EML, once every 4 or 5 months, usually after reasonably hard acceleration and after a warm restart, and always cleared when restarted, for about 2 years - I put up with it as it was a mild nuisance. In the few weeks before I had the turbo replaced it would happen 2 days out of 3 when idling in traffic or gentle driving, and would put the EML light on about 50% of the time. It only occasionally cleared by restarting (the rest of the time it cleared itself overnight). Reading codes with an OBD reader (an Autel, not STAR) gave a number of codes relating to O2 sensors and wastegate solenoid - I think simple readers do not give codes necessarily relating to the real issue. Replacing the turbo has removed all issues - I'm not familiar with your OBD reader but it may be the codes are not necessarily indicative of the real issue.

 

[Freeform]

Strictly speaking, a generic off the shelf OBD scanner need only show emissions relevent failures covered by the OBD-II/EOBD legislation and nothing more. They're not much use to people like us.

A STAR license and correct hardware interface must surely be beyond us enthusiasts, though some here might have a pirated version with a knock-off hardware inderface (as does my colleague).

I have a full license for 'Carly for Mercedes' smartphone app, as stated above. It does read all codes, and interrogates all other control modules fitted (24 controllers are scanned in my C320). What it lacks is funtional tests, enabling of optional features (e.g. cruise control) and secuity functions (e.g. key matching). The full license is €40, IIRC - to cover all OBD-II/EOBD compliant Mercedes (those produced since the late 90s).

I also use the TorquePro app to monitor some engine operating parameters on a virtual dashboard. It seems to catch most, probably all engine DTCs, is good for a quick scan and reset but does not interogate other controllers.

 

[Freeform]

Following another week without recurrence, and without any fault logged, it has failed a couple of times each day since Wednesday. Once the P0237 fault appears (as pending fault, usually), zero boost is likely to occur at some point in the journey.
The fault can be reset with engine off and it may stay clear but very often reappears soon or immediately and next start.
The car will then likely drive normally until conditions for an internal plausibility check are met. Most predictably, this is when descending a hill coasting, without accelerator. I suppose this is a known no-boost situation (normally) and that the ECU SW compares the pressure value seen with the expected value. When significant mismatch is seen over this period, next time one accelerates there will be no boost, system disabled.

Anyway, whilst this may be interesting, I need my car fixed! I have it booked in with the Mercedes specialist local to me, on Friday next, 2nd Dec.

(Though it's frustration not to get to the conclusion myself, or with the colleague and his STAR kit but still, I will have a hand in the solution by opening my wallet!)

Thanks for all inputs so far.

 

[Freeform]

To update:

Recurrence of this issue has reduced to sporadic. Immediately after, the associated fault, P0237, can be seen as pending but will usually reset itself after the next drive cycle and will not be showing by the time I get to anyone with a STAR kit.

I have only made two changes since last post:

I realised this saga started after I switched to Tesco Diesel so I am am not using that anymore. I don't say it was the cause but it is easy to use something else.

I ran a couple of bottles of Forte turbo cleaner through it. I am not a great believer but my trade friends passed the stuff to me so I thought nothing to lose.

I shall wait until the frequency of occurence becomes more frequent, predicable or repeatable - or heals up by itself!

 

[wu56Shoozz]

Millers Diesel is the best OR Diesel Rhino.. add to the tank and watch ..

How mechanical are you? Good with a spanner, got a Torx Socket set?

Have a look at this - it may improve things but not necessarily the cure..

https://www.youtube.com/watch?v=oy1eP_RwlNY

I have the same vehicle incidentally with 70k miles on it......

 

[Freeform]

At least additives are easy to try, though the often accompanying advice to give the car a thrashing is not so practical with 270 bhp & 590 N-m!

EGR had crossed my mind; if the valve did not close properly, I can imagine that causing a signal range problem - but I get no undue smoke.

The valve is at least accessible, I have the tools and experience [I mispent many hours in friends' garages to get ramp time for my own cars]. I might have a go over the 'holiday'!

[Currently 92000 miles, up from 53000 when I bought it 16 months ago.]

 

[wu56Shoozz]

Its hard not to enjoy the comfort & performance..Lols!

 

[Freeform]

Those of us who enjoy motoring most are condemned to do too much!