@HD
At the moment, my memory of that system is less than perfect
That said, I’ve come across a few “irregularities” in the Mercedes-Benz factory literature.
To be fair, I’ve also seen the other manufacturers make mistakes in their literature
I have some news:
While I was driving and stopping at traffic lights today, I realised a further decrease of shakiness at idle in “D” compared to yesterday. I wondered what could possibly make the difference between today and yesterday, and I thought - besides adding 1 liter of engine oil last night – there was no change.
And then I remembered a comment which ok4450 posted at the beginning of this thread - he said:
… slightly rough idle simply because whoever checked the engine oil level did not fully seat the oil dipstick in the tube …
I stopped to check the dipstick. It was sitting properly. But I took it out anyway to check the O-ring which seals it in the tube. It’s as hard as hard plastic and it has shrunk, so that it can’t seal anymore. That’s obviously what age and changing temperatures can do to rubber.
I think you’re onto something here ok4450 !
It seams that after adding oil the higher level of oil with it’s higher pressure at the lower end of the dipstick tube causes less (or even no) air to get into the crankcase and thus under the valve cover. These areas are under vacuum, because - besides the breather hose between the valve cover and the air filter housing - the intake air ducts around the injectors are connected to the valve cover through a separate breather hose.
Now I think I know why idle was less shaky at times. I can’t remember in detail, but probably that was each time after I added engine oil (1 liter (0.26 gallon).
So I guess, false air has a finger in the pie here after all !
Maybe the fact that idle is more shaky when the engine is hot has (also) to do with the decrease of the oil’s viscosity with it’s temperature rising. With the oil becoming “thinner” - friction for penetrating air becomes lower.
Tomorrow I will renew the O-ring on the dipstick.
I haven’t installed the new injectors or a new o2-sensor yet.
But we are narrowing down the (combination of) problems step by step and at least I am learning something here.
There is still a very small rest of shakiness left at idle. If a new dipstick O-ring does not eliminate that, my next suspects would be those two injectors which don’t atomize the fuel nicely at idle anymore and / or the old o2-sensor.
I’m sure the idle speed fluctuation in “N” or “P” is caused by the old o2-sensor.
And I think I will install the new injectors anyway, because I suspect them to be responsible for at least the other two problems:
- warm engine start (starts and runs for 1 second – stalls – start again – runs)
- engine sometimes stalls at idle in “D” and has to sit for about 5-10 minutes before it starts again
P.S.: The reason why I have to add engine oil from time to time is not the engine burning it. It leaks out about 1 liter on a little less than 1000 miles. It’s the most common leak on an M103 engine – I think ok4690 knows which one I’m talking about. I may take you up on that later ok4690, if you don’t mind.
In a complete bout of laziness, Im throwing something out there without reading all the responses so this may be a repeat BUT,
A friend recently inherited a 190e and had rough running problems. After throwing parts at it, it turned out to be from a “sock” like filter inside the gas tank that was plugged up and closing from suction from the fuel pumps. Removal of that and washing out the gas tank cured it all. Apologies if this has already been suggested.
@HD
Are you referring to the infamous 103 engine head gasket “seepage” . . . ?
If so, yes I know exactly what you’re referring to, as I’ve replaced my fair share of headgaskets on 103 engines, and 104 engines also
If that is indeed what you’re referring to, I think you’re quite capable of replacing the head gasket by yourself.
I’ve done so many of them, I even have the torque specs and angles memorized. Even though I haven’t done one in a few years.
The 300CE gas tank is also equipped with that filter.
The fuel pressure measurements were all in order and the car doesn’t show any problem or lack of power at full acceleration up to higher revs. So there doesn’t seem to be any filter problem neither with that filter nor with the one on the pressure side of the fuel pump and also not with the tiny filter at the entrance to the fuel distributor.
Thank you for mentioning that! … And don’t hesitate to throw in suggestions! … Maybe you saw what I experienced with the engine oil level and the not sealing O-ring of the dipstick. Who knows if I would ever have found out about that, if ok4450 wouldn’t have talked about his experience with oil dipsticks and false air on CIS cars.
It sounds like you’ve gone through the whole fuel system and ensured thats all happy, and the ignition is all good as well, how about the cooling system? Are the fans coming on, new coolant, new radiator hoses? Sounds like heat is making the problem worse (sluggishness or rough idle).
Outside of that then Id guess sensors of some sort. If you can go to a junkyard and grab some spare sensors or relays to try. Its beyond me really but if you’ve ruled out all the major components of the mechanical systems then it seems to be some computer issue. I have read on here the crank position sensor or cam sensor to be faulty when hot and cause rough running or hard starts and then when its cool the car magically starts no problem.
Everything you mention, I have tested – including the crank position sensor, which – as you say - can cause problems with rising temperatures. Especially the – luckily very rare - stalling at idle and hard restarting when the engine was hot could also be caused by the crank position sensor.
As a matter of fact, I’m currently building a test apparatus with which I can check several things during driving. Monitoring the crank position sensor is one of those things.
As for the idle quality, as it is now - most people would call it good enough. But there is still a little shakiness left, which I expect to disappear after the installation of new injectors. But before I install them I would like to renew the o2-sensor, which I couldn’t do yet.
I renewed the O-ring on the engine oil dipstick today.
The old one’s cross section wasn’t even O-shaped anymore – it was more a rectangular shape with rounded edges.
But I can’t feel any difference in the idle quality since I added engine oil to the maximum level two days ago.
It seems that the higher oil level already makes it impossible for false air to get into the engine through the dipstick tube. However with the new O-ring idle quality shouldn’t get worse with the dropping oil level anymore.
This posting may be only remotely related to the titel of this thread, … but db4690 has a lot of experience working on engines like mine, and I would like to access that (or anybody else’s experience with the Mercedes M103 engine):
“…Are you referring to the infamous 103 engine head gasket “seepage” . . . ? …”
I hope it’s not the head gasket!
Oil is leaking at the front of the engine on the passenger side. It’s where the following parts meet: cylinder head, block, upper front cover and lower front cover (see attached picture).
My suspect is the U-shaped rubber seal between the upper and the lower front cover.
Because when I drive longer distances at higher revs (like on the Autobahn) the leakage is bigger. As you know, the upward moving timing chain, where it meets and follows the camshaft sprocket, centrifuges engine oil into the upper right corner of the upper front cover. And at higher revs more oil is scooped upward and centrifuged into that corner.
I intend to renew that rubber seal and the sealing compound between the front cover and the front face of the cylinder head. I think I will use heat-resistant silicone instead of Hylomar sealant .
I will also renew the camshaft seal then of course.
And I want to check the timing chain respectively the engine timing based on the alignment of the 2 test pins (the one on the first camshaft bearing cap and the one on the camshaft flange).
Or would you renew the timing chain anyway after 120.000 miles?
Since the valve cover will be off, I could check and retorque the cylinder head screws - at least the ones at the front. What do you think?
And I also want to check the valve drive, because after starting the cold engine you can hear it ticking. Sometimes more, sometimes less – I guess depending on which hydraulic lifters are pressed or released after the engine is switched off.
A few weeks ago with my ear close to the warm running engine I could also hear the ticking sound (only slightly). Then my wife gave me a time window of 10 seconds, while I typed on my smartphone’s notepad for each tick I heard. Counting the characters amounted to 330/min (half of the crankshaft’s idle speed). So it seems to be related to 1 valve or at the most 1 intake and 1 outlet valve which move simultaneously.
Then I used a big screwdriver as a stethoscope in the attempt to narrow the problem down. First I pressed it onto the cylinder head close to each spark plug. The sound was the same at each spot, and not very prominent.
Then I opened the oil filler cap (with the engine still running) and pressed the screwdriver onto the first camshaft bearing cap, in which – as you know - also the rocker arms are pivoted. There it was very prominent.
Finally I pressed the srewdriver onto the swinging intake valve rocker arm of the first cylinder (close to the camshaft, so that the up and down movements of the rocker arm wouldn’t punch my ear). There it was the most prominent.
The fact that the sound was much less audible on the cylinder head’s surface than on the camshaft bearing cap leads me to believe that it’s not the valve(s) but the lifter(s).
Please correct me if I’m wrong!
@HD
I believe you are correct on both counts
Most likely your cam cover seal is leaking. This is actually a little tricky to install without dislodging it from its groove. I used a slightly oiled microfilm to assist me.
Most likely the lifters are failing
If I remember correctly, the lifters can be replaced without messing with the chain. It’s been awhile.
If I were you, I wouldn’t replace the chain at this time, although it’s smart to replace the cam seal
Thanks for your advice db4690 !
Actually I can’t see any traces of oil coming down from the cam cover seal, but I will install a new one anyway.
“… a little tricky to install …”
I thought you just stick the seal onto the lower edge of the cam cover and then just place it onto the flat surface of the cylinder head!? (see picture)
“…If I remember correctly, the lifters can be replaced without messing with the chain …”
That’s right, the M103 engine has 6 camshaft bearing caps (1 for each cylinder) in which both rocker arms are pivoted. You can dismount them (one at a time) and change both lifters on the workbench. No need to mess with the timing chain.
What do you think about retorquing the cylinder head bolts when the cam cover is off?
@HD
“What do you think about retorquing the cylinder head bolts when the cam cover is off?”
Bad idea, in my opinion
You said the headgasket isn’t leaking. Personally, I find that a little unusual, but that’s besides the point. I find it unusual, because Mercedes-Benz inline 6 engines are known to leak from the head gaskets
Anyways, I believe no good can come of loosening and retorquing head bolts, because you’re dealing with a 26 year old gasket.
You’re showing me a picture of the valve cover
“…You said the headgasket isn’t leaking. …” I said I HOPE it’s not the head gasket!
Oil is visible within the small elliptical area (see picture) - mainly on the block’s rectangular machining fixturing surface in the right half of the elliptical area, and of course from there on downward.
And the closest oil drain duct is about 4 inches further back. The small duct about 1.5 inches from the front face of the block seems to be a coolant duct.
Other then that spot at the front right edge there are no traces of oil around the head gasket.
That’s why I hope that - instead of the head gasket - it’s the rubber seal between the upper and the lower front cover which fits into that U-shaped groove on top of the lower front cover. I heard that’s an infamous M103 seepage too – some German guys on the internet call it the “Sollsiffstelle” (predetermined dribbling point).
“…no good can come of loosening and retorquing head bolts …” You’re right!
I was considering to check the tightness according to a quoted final torque. If they would move a little – okay, if they wouldn’t - I would leave them as they are. But now I see that there is no statement of the final torque in the Mercedes factury literature, but a description of this 3-step tightening procedure: tighten with 55 Nm, then tighten further by 90° rotation angle, finally by another 90°. Okay – my mistake!
I attached the picture of the valve cover to my last posting, because judging from the picture it doesn’t seem to look tricky to install. You said: “… This is actually a little tricky to install without dislodging it from its groove …”
I thought after sticking the seal onto the lower edge of the cam cover I just place it onto the flat surface of the cylinder head.
@HD
I said the camshaft cover seal is tricky to install . . . that u-shaped seal. You put the seal in that groove. Then you put sealant on the camshaft cover (not the valve cover, which you showed me last time). Then you install the camshaft cover. At which point, the seal is very often dislodged from its groove.
The problem, if I recall correctly . . . you push the cam cover horizontally, against the head. And that horizontal pushing is what often dislodges the seal.
Previously, you showed me a picture of a valve cover. The valve cover gasket is comparatively easy to install.
Okay db4690 – now I know what happened here – a misunderstanding due to different terminology:
To me the terms “valve cover” and “cam cover” have always meant the same thing, i.e. the hood on top of the cylinder head.
When we talk about the cover which is mounted against the front face of the cylinder head - you call it “cam cover” or “camshaft cover” – and I call it “upper front cover”.
Now I see what you mean when you say that’s a little tricky to install.
And now I also know what you meant by: “… I used a slightly oiled microfilm to assist me …”.
Maybe tilting it a tiny bit - so that the lower part of the cover is slightly leading - while pushing it horizontally against the front face of the cylinder head, would help a little to keep that rubber seal in it’s groove.
Unless you disagree – besides the surface which is mounted against the head, I want to put sealant (heat-resistant silicone instead of the usually used compound) around both ends of that U-shaped rubber seal.
Today I renewed the o2-sensor (genuine Bosch from the local Bosch dealer), with the following result:
-
idle speed fluctuation in “P” or “N” reduced by 50 %.
-
closed loop voltage: 60 – 770 mV (before: 300 – 600 mV, sporadically: 180 or 700 mV)
-
cycle time / frequency:
- at 2500 rpm: 1.1 sec / 0.91 Hz (before: 1.2 sec / 0.83 Hz)
- at 675 rpm: 4.2 sec / 0.24 Hz (before: 4.5 sec / 0.22 Hz
The new sensor’s amplitudes are much better, but it’s only a little quicker than the old one. It seems to me that the old sensor’s frequency was nothing to worry about.
After a test drive with the new sensor I checked the duty cycle, generated by the ECU:
- at 2500 rpm: 49 % +/- 1.5 % (before: 47.5 % +/- 2 %)
- at 675 rpm: 46.5 % +/- 1.5 % (before: 46 % +/- 2 %)
So, with the new sensor the engine runs a little leaner and with less mixture fluctuation, which means that the fluctuation of the current through the coil of the EHA must be less now too, which causes it’s valve plate to swing less wide between rich and lean.
I want to wait several days with readjusting the base setting until I will have installed the new injectors. Then I will also check the EHA current.
The new sensor did not make any difference to the rest of the shakyness at idle.
Hopefully the new injectors will take care of that.
I have installed the new injectors today.
Now problem No. 3 is definitely solved !
The warm engine fires immediately, I don’t have to start it twice anymore.
I assume that problem No. 2 (engine stalled at idle several times in the past) is also fixed with the new injectors.
But problem No. 1 (the most annoying one) is still not completely solved. The new injectors did not further reduce the idle shakyness.
The fuel seems to be atomized more thoroughly though, since with the new injectors the meter for the duty cycle (which represents the o2-sensor emission control) read:
- at 2500 rpm: 47.5 % +/- 1.5 % (before: 49 % +/- 1.5 %)
- at 675 rpm: 45 % +/- 1.5 % (before: 46.5 % +/- 1.5 %)
That means that the mixture became a little richer.
The unchanged rest of idle shakyness leads me to believe that there might be mechanical problems involved.
I tried to pin down that hydraulic lifter noise further with a big screwdriver as a stethoscope, and I think it’s coming from the intake valve of the first cylinder.
I took out all spark plugs (after 15 minutes of idle) and – besides that all of them didn’t look clean – I saw that the spark plug of the first cylinder was quite wet compared to the others. (see picture!)
Then I cleaned the spark plugs thoroughly, plus narrowed the gap back to target, and set the mixture a little leaner to:
- at 2500 rpm: 49,5 +/- 1,5 %
- at idle: 50,5 +/- 1,5 %
Just a 5° turn of the adjustment screw (counterclockwise) was enough. That reminded me again how delicate the mixture adjustment is!
(A duty cycle of 50 % shows that the o2-sensor is fluctuating precisely around a center voltage of 450 mV).
Now there is only a little idle fluctuation left.
So - maybe because of the failing lifter the intake valve of the first cylinder doesn’t open completely, so that the intake gas can not properly fill the combustion chamber, thus can not be ignited properly, which results in a wet spark plug and remaining idle shakyness !?
Or maybe the timing chain is streched a little, so that the camshaft timing is not accurate anymore!?
If any of you guys have any feedback about this idea, please let me know!
@HD
Congratulations on your continuing progress!
I know you’re losing a little oil, due to that leak you mentioned earlier
Are you also using a little oil?
Tank you db4690 for your continuing assistance!
Regarding possible oil consumption I did 2 things:
No. 1)
Two months ago I added 0.15 liter of LIQUI MOLY “valve clean” to about 50 liters of fuel before I went on a 120 miles long trip on the Autobahn.
Since that was very early in the morning when the Autobahn is not crowded yet, it was a good opportunity to let the car run under different load conditions at higher rpm’s and clean the valves.
Usually I drive that car very modestly. The gas pedal is rather caressed than pushed, almost never using the full torque or exceeding 4000 rpm. The gas pedal has touched the kick-down switch less then 10 times within the last 60.000 miles – and always for testing porposes only.
So with the warmed up engine after 15 minutes on the Autobahn I accelerated from 50 mph up to 115 mph with the pedal close to the floor and then I took my foot completely off the pedal, keeping an eye on the rear view mirror where I saw substantial fume (not too dark or dirty).
After the car had slowed down to 50 mph I repeated that, but now holding the speed of about 110 mph for a minute or so – again I saw smoke when I took my foot off the pedal, but not as much as at the first time.
Than I accelerated a third time, staying at high speed for several seconds – that time (after I released the pedal) there was only very little smoke visible. Within the next half hour I repeated that a few more times – but without any more smoke visible at the rear of the car.
So I guess the valves were only contaminated a little.
No. 2)
Three weeks ago I let the engine sit at idle speed for a little more than 10 minutes, and then I let my wife floor the gas pedal while I was standing behind the car.
There was not a trace of smoke coming from the exhaust pipe !
So I guess that the 26 years old valve shaft seals (and guides) are still okay, which really surprises me, since I heard that problems with these seals are not uncommon on M102- and M103-engines.
Obviously the piston rings can also not be in bad shape either.
Besides that I want to add, that as far as I can look under the valve cover through the oil filler opening, everything looks very clean in a golden brown color. And after rubbing my finger over the underside surface of the valve cover (through the oil filler opening) only a little black debris is visible on my finger.
The last engine oil and filter change was 4.800 miles ago – and the oil looks almost as clean as fresh engine oil.
Plus it has no traces of fuel smell.
I m jealous HD. it would be nice to live in an area where I could safely “clean the valves”