2009 - W211 E220 Cdi - Rough Cold Start & Idle

Page may contain affiliate links. Please see terms for details.

Mintyfresh77

Member
Joined
Oct 17, 2021
Messages
43
Location
Southampton
Car
S211 E220 CDi OM646 Evo
Chaps,

Before I start it's best to get the obvious ones out of the way first when mentioning cold start issues. The car had 4 new OE glow plugs and Glow plug relay fitted early this year to resolve very bad cold starts and it has run beautifully since but slowly over the last few weeks I've been noticing a rough initial startup almost like not all cylinders are coming to the party and once started it idles but with a gentle surging that actually rocks the car. Once the engine is up to temperature it's no where near as noticeable but I can still feel it when sat at lights or at junctions. On the go it performs perfectly fine, smooth power delivery, no smoke, no hesitation. I cleaned PCV pipework and turbo intake yesterday with brake cleaner and rag but was only very slight oil residue anyway I also blasted the MAF with brake cleaner. I removed the EGR today and was surprised at how little cleaning was required on a 122k engine. I did a leak off test last week on a warm engine as the wife forgot I needed it cold and nothing untoward on leak off quantities but wonder if it would be a different story on a cold start leak off?...I tried unplugging the MAF while it was idling earlier and it seemed a little smoother but engine was warm (again perhaps something that needs to be done on 1st crank of the day). Rail pressure according to my icarsoft is a constant 235bar at idle but honestly I'm not sure if that's normal or horrifically low? I can see lots of data but i'm not ashamed to admit I don't know what I'm looking at in all honesty. Any ideas guys?

Not sure if it's worthy of mention ad it's always done this in the 6 months I've owned the car but when turning ignition on before crank there is a curious noise from the fuel lines area like a pressurising noise that creates a whirring or muted whistling noise. I figure it's the fuel system pressurising and perhaps purging any air in the lines but that is purely a guess. It lasts for about 20-30 seconds which seems overly long for a pre startup procedure.

See youtube link Strange Noise - OM646 pre startupto video.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
Mine make that noise. It sounded near to the EGR valve but may have been in the fuel filter. I guess if you dont start it within a time it shuts of the fuel pump ?
 
I have invariably found the rough idling and smooth running otherwise, is related to the pressure switch at the back of the rail. Take it out-very easy- and spray with releasing fluid.
The other cause I have found is the clutch seizing on the alternator pulley, though that should also cause the belt to flutter
 
Update...
Still has intermittent misfire between idle and about 1500 revs BUT the rough start has totally gone and overall there has been improvement. Some days its fine and others it shakes every few seconds while stationary. Its totally baffled me. No fault codes logged and when taken to an indy a for diagnostic they didn't find any problems, charged me for a software update and star session and said to monitor but when I went to pick it up at the end of the day it was chugging at idle. Bast@rd obviously behaved itself while being looked at. The tech who originally looked at the car came over, witnessed it and agreed something was definetly going on but said that he only noticed that injector 4 was close to tolerance when testing it earlier in the day. So £135 lighter and no wiser I went on my way.

I plan on keeping the car for a 3-4 years so took a punt on a very good deal for 4 brand new Delphi injectors, fitted myself (properly with injector seats cleaned, 7nm torque +90 + 90 turn) and the overall difference in the engine smoothness, responsiveness and pulling power is astonishing so happy with that outcome but alas the problem still remains...

So far the following parts have been thrown at it (new OE parts).

Fuel leak off lines with Venturi T piece
Inlet Metering Valve (Fuel Quantity Valve)
Fuel Rail Pressure Sensor
Fuel Rail Pressure Regulator

I am now that guy who has thrown parts at the problem after lots of research and a personal score to settle with this thing but still drawn a blank...

My ear maybe deceiving me but I am sure I am hearing noise (whirring and clicking) from the altenator pulley but not noticing the belt jumping about especially?

Also if I were to notice it happening more when fuel tank is say 1/4 full than when it's full would that point to fuel pump or tank related issue? I am going to test this theory as it was behaving itself the other week when tank was between half and full, today it's under a 1/4 and being a dog at idle so will fill up and monitor.

Come on Mercedes experts, throw the dog a bone here?!

Possible Vacuum leak, where to look?
Intake manifold gunked up?
High Pressure Fuel Pump?
Altenator pulley/clutch?
A sensor that's bad but not throwing codes?
Bad Catalytic converter?

Remember, it's intermittent and otherwise pulls like a train and runs smoothley except when this mysterious miss starts happening and then it's chugging with a little smoke and the car rocks abd shakes....and then fine again for a few minutes, hours or days...

Help!!!
 
Last edited:
One to consider is the main crank pully / harmonic balancer.

Sounds like the injectors are happy, but did you do the zero calibrations?
 
One to consider is the main crank pully / harmonic balancer.

Sounds like the injectors are happy, but did you do the zero calibrations?
Hi M80,

Thanks for your input.

I coded the injectors to ECU with compensation values, is that the same thing as zero calibrations?

Harmonic balancer sounds worryingly like open heart surgery!

Cheers
 
Hi M80,

Thanks for your input.

I coded the injectors to ECU with compensation values, is that the same thing as zero calibrations?
No,
What you will have done, I assume, is to type in the new injector codes so that the ECU can do whatever with that info.

Star can do zero calibrations at tickover, approx 15 minutes of continual adjustments.
There is also a running zero calibration that takes approx 22 minutes.
The running cal's need a min input air temp of 30 degrees, I couldn't attain that in winter so ran it again last week. I ran the car up to temp on a warmish day and then left it ticking over for the input temperature to rise.There is a noticeable difference to smoothness, and possibly consumption.
Harmonic balancer sounds worryingly like open heart surgery!

Cheers
Not so involved.
Remove starter and lock the fly wheel with a toothed tool.
Remove aux belt, remove main lower crank pulley, and replace.
Basically the pulley is 2 parts, bonded with a damping rubber. It that has deteriorated and the bonding is less secure (I guess) there can be vibes.
I'm hoping that is my problem anyway.
I'm not looking forward to removing my knuckles while removing the pulley tbh.
 
No,
What you will have done, I assume, is to type in the new injector codes so that the ECU can do whatever with that info.

Star can do zero calibrations at tickover, approx 15 minutes of continual adjustments.
There is also a running zero calibration that takes approx 22 minutes.
The running cal's need a min input air temp of 30 degrees, I couldn't attain that in winter so ran it again last week. I ran the car up to temp on a warmish day and then left it ticking over for the input temperature to rise.There is a noticeable difference to smoothness, and possibly consumption.

Not so involved.
Remove starter and lock the fly wheel with a toothed tool.
Remove aux belt, remove main lower crank pulley, and replace.
Basically the pulley is 2 parts, bonded with a damping rubber. It that has deteriorated and the bonding is less secure (I guess) there can be vibes.
I'm hoping that is my problem anyway.
I'm not looking forward to removing my knuckles while removing the pulley tbh.

Thanks for explaining the zero calibrations, it's interesting food for thought and I had no idea about the Star continual adjustments so def an avenue I'm sure will be taken by the next indy I choose to take on this mystery as I am fast approaching my DIY capability threshold. I was absolutely shitting myself doing the injectors due to all the horror stories of injector clamp bolts shearing off either through undoing or tightening up! I was so careful to clean everything properly including the bolt bore/threads. I was so convinced it was a set of tired injectors and that I'd nailed it when the engine fired up sounding quieter than before and upon taking it for a test drive was chuffed how it relentlessly pulled on boost but sitting on the driveway idling my old friend the slight miss could be felt and heard again. B@stard....

On the subject of thr harmonic balancer I have an interesting (but probably incorrect) theory. I was advised by an indy doing a quick health check 6 months ago that the engine mounts were heavily worn, I have both mounts and the trans mount in the garage waiting to be installed when i get time. Could it be that the combination of a worn balancer pulley and the mounts creates a bit of a Mexican wave from time to time and i say this because when i select reverse or drive to come off the driveway it really has a good wobble sometimes.
 
Last edited:
Engine mounts are the 1st call. Oil filled on the 211 apparently, they can lose some of that.
I would consider doing the easier and cheaper gearbox mount at the same time, if for nowt more than piece of mind.
 
Engine mounts are the 1st call. Oil filled on the 211 apparently, they can lose some of that.
I would consider doing the easier and cheaper gearbox mount at the same time, if for nowt more than piece of mind.
I've been putting them off for a while now being dead cert that it was an air/fueling issue of some kind because of the way it sounded and felt but thinking I should make that my next job. 2 engine mounts and trans mount are sat in my garage waiting to be fitted. Boy oh boy will I feel a plum if after fitting those the problem goes away.....

I'll let you know!
 
  • Like
Reactions: m80
M80,

I think you maybe right about the Harmonic balancer. I removed the v belt, started the engine and sure enough the constant belt chirp type of sound from that area remains although the crank doesn't appear to wobble to the naked eye when engine is running or have any noticeable play to the touch but I guess that it would have to be getting pretty terminal before that becomes glaringly obvious. I still think I need to investigate/clean the intake plumbing for the intermittent rough idle and miss because there is definitely an occasional miss when lightly feathering the throttle up to 1500 rpm but I am now thinking the idle rocking motion is not related to that, I think you are bang on the money with the Harmonic balancer. So 1 problem could actually be 2 issues combined. I still need to fit the engine and trans mount which I will do on Monday and recheck but feel like I starting to zero on the culprit and things will only start to get clearer after engine mounts are installed, intake cleaned and by me and HB diagnosed and fitted by indy.
 
If you have been told that the engine mounts are worn then you will never get a really smooth idle especially when the engine is cold.

Don't forget the transmission mount as well! 👍👍
 
No,
What you will have done, I assume, is to type in the new injector codes so that the ECU can do whatever with that info.

Star can do zero calibrations at tickover, approx 15 minutes of continual adjustments.
There is also a running zero calibration that takes approx 22 minutes.
The running cal's need a min input air temp of 30 degrees, I couldn't attain that in winter so ran it again last week. I ran the car up to temp on a warmish day and then left it ticking over for the input temperature to rise.There is a noticeable difference to smoothness, and possibly consumption.

Those injector characterisation codes on top the injector are determined at end of line. The delivered quanties at varying energising times and rail pressures is compared to a reference. The offset required to bring them back to that ref at key points( usually fuelling points that correspond to pilot, idle, mid load and full load) is calculated and encoded. These offsets are always used by the ECU when it calculates the energising time for a fuel request.

The zero fuel corrections are adaptions that are designed to cater for long term drift. They are called zero fuel as it's a measure of the energising time required to open the control valve and just about to open the nozzle, but not quite. The learning happens during overrun, one cylinder at a time, when conditions are correct. So not every overrun.
It works by slowly increasing injector energising time on the one cylinder and monitoring when a very small combustion has occurred, then just backing off a bit until it stops. Older Delphi systems used a knock sensor for this. Later they copied the Bosch systems that the monitor the crank sensor for small changes.

In theory you shouldn't have to force learn them, but given that the injector characterisation codes do not cover this point, speeding up the process can help.
 
If you have been told that the engine mounts are worn then you will never get a really smooth idle especially when the engine is cold.

Don't forget the transmission mount as well! 👍👍
Thanks DSM, So frustrated at the moment.....thought I'd attempt the mounts today and start on driver side but can't actually see a way of getting old mount out and the the new mount in without removing the DPF/turbo etc and I'm really low on time and confidence in myself at the moment. Passenger and trans mount are perfectly accessible but drivers side looks an absolute pig 😭
 
Thanks DSM, So frustrated at the moment.....thought I'd attempt the mounts today and start on driver side but can't actually see a way of getting old mount out and the the new mount in without removing the DPF/turbo etc and I'm really low on time and confidence in myself at the moment. Passenger and trans mount are perfectly accessible but drivers side looks an absolute pig 😭
There will be a way, perhaps raising the engine slightly once the mounts are disconnected. I’ve managed to do a few in the past, just take your time and you’ll beat it!
 


There will be a way, perhaps raising the engine slightly once the mounts are disconnected. I’ve managed to do a few in the past, just take your time and you’ll beat it!
Thanks Colin, was that the E220 OM646 engine that you've had success with? I did wonder about lifting the engine but it all just looked so bloody packed in. I have a day off work tomorrow with no 5yr old to constantly harass me so may have another bash.
 



Thanks Colin, was that the E220 OM646 engine that you've had success with? I did wonder about lifting the engine but it all just looked so bloody packed in. I have a day off work tomorrow with no 5yr old to constantly harass me so may have another bash.
On one of the hottest days on record.

I salute you! 👍👍

Presumably you will have appropriate refreshments to hand 👍👍😎
 
On one of the hottest days on record.

I salute you! 👍👍

Presumably you will have appropriate refreshments to hand 👍👍😎
🤣.... I was thinking that myself. I may on the other hand just take an air conditioned drive over to a very highly rated indy who is in my area to have a chat about my woes and get it booked in. I think his name is Olly and his company is PCS. I need engine/trans mounts done, HP power steering pipe is on it's way out (weeping), mysterious noise needs diagnosing and of course the reason for the original thread, the odd intermittent miss. I drive a truck 50hrs a week and have a 5yr old boy so I get very little time to do this stuff without either disappointing him or annoying my wife so it's time to face facts and make some money decisions because I'm losing the race on this one. If I knew 100% it was possible to do the mounts on this W211 OM646 evo DPF without removing the Cat/exhaust or DPF/Turbo I'd do it but even with Colin's suggestion of jacking the lump up a few inches I can't see it coming out. I would be chuffed if anyone can tell me from a tech perspective.
 
Last edited:
🤣.... I was thinking that myself. I may on the other hand just take an air conditioned drive over to a very highly rated indy who is in my area to have a chat about my woes and get it booked in. I think his name is Olly and his company is PCS. I need engine/trans mounts done, HP power steering pipe is on it's way out (weeping), mysterious noise needs diagnosing and of course the reason for the original thread, the odd intermittent miss. I drive a truck 50hrs a week and have a 5yr old boy so I get vert little time to do this stuff without either disappointing him or annoying my wife so it's time to face facts and make some money decisions because I'm losing the race on this one. If I knew 100% it was possible to do the mounts on this W211 OM646 evo DPF without removing the Cat/exhaust or DPF/Turbo I'd do it but even with Colin's suggestion of jacking the lump up a few inches I can't see it coming out. I would be chuffed if anyone can tell me from a tech perspective.
Olly and his team at PCS are one of the best in the business. You will get high quality work carried out at a fair price 👌👌👍
 
Those injector characterisation codes on top the injector are determined at end of line. The delivered quanties at varying energising times and rail pressures is compared to a reference. The offset required to bring them back to that ref at key points( usually fuelling points that correspond to pilot, idle, mid load and full load) is calculated and encoded. These offsets are always used by the ECU when it calculates the energising time for a fuel request.

The zero fuel corrections are adaptions that are designed to cater for long term drift. They are called zero fuel as it's a measure of the energising time required to open the control valve and just about to open the nozzle, but not quite. The learning happens during overrun, one cylinder at a time, when conditions are correct. So not every overrun.
It works by slowly increasing injector energising time on the one cylinder and monitoring when a very small combustion has occurred, then just backing off a bit until it stops. Older Delphi systems used a knock sensor for this. Later they copied the Bosch systems that the monitor the crank sensor for small changes.

In theory you shouldn't have to force learn them, but given that the injector characterisation codes do not cover this point, speeding up the process can help.
A superb explanation, thank you kindly 👍
 

Users who are viewing this thread

Back
Top Bottom