Any BMW experts? N55 rough idle only with MAF connected.

[JD!]

So this 2012 X3 runs awesome with the MAF disconnected, perfectly smooth idle. Has a small-scale rough idle with it hooked up. Obvious symptom would be an air leak somewhere, but I've smoke tested the intake 3 times and can't find any leaks. It held a lot of pressure before discontinuing the test.

I also replaced the maf with a new BMW part thinking that would fix it. No change.

No codes of any sort. I took it on a 600 mile road trip last month and nothing came up.

I've attached some live data at idle RPM below. Looks like the MAF reading is pretty unstable - any thoughts on this?

 

[shortyb]

What are your fuel trim readings like? Short term going up? Where are you introducing smoke?

 

[GMBoy]

Every N55 I ever worked on with your condition was due to 1. an oil leak at the valve cover (vac leak) or 2. a bad PCV (which is in the valve cover). Smoke test should catch those leaks and the oil leak would be obvious as well. There was also one time on an N54 that the B1S1 oxygen sensor was near its limits but still in spec enough not to set a code - replaced it and all was well again.

 

[Trav]

Is the with DI?

 

[JD!]

So at idle with the MAF plugged the LTFT remains at about -15-16, STFT remains very steady within +/-1 whereas with the MAF plugged in it was +/-5.

I've searched for oil leaks and only found one, the turbo return line which I replaced. Front main seal looks good, maybe the rear main is weeping but I can't really tell due to oil everywhere from the turbo line. Regardless I don't seem to lose any oil on the e-dipstick.

Valve cover looks relatively new, no leaks there. I don't see or hear the usual PCV issues like excessive crankcase vacuum, squealing seals, or oil consumption. Possible that there's just a slight crack in the diaphragm?

 

[JD!]

What are your fuel trim readings like? Short term going up? Where are you introducing smoke?

I'm smoking it just about a foot downstream of the MAF sensor.

 

[iCurious]

Your long term fuel trim is negative 16%.

You should plug the MAF sensor in and post up your STFT and LTFT both banks. One at idle and one at 3000 rpms.

 

[shortyb]

I suspect you have a leak in the intake manifold area, like a bad gasket between it and the cylinder head, loose bolt(s) etc. If possible smoke the manifold, ie; find a fitting on the manifold and introduce smoke there. I've done the intake just downstream of the MAF on my VW CCTA, but got nothing. Filled the manifold at the breather hose inlet and found the leak instantly. Also smoke the valve cover through the breather and check the breather diaphragm.

 

[Marcozi]

What is happening is the engine computer (DME in BMWspeak) defaults to baseline fuel/air mix when you disconnect the MAP. You need to have the intake professionally smoke tested to find all of the vacuum leaks. Try www.bimrs.org.

 

[iCurious]

I suspect you have a leak in the intake manifold area, like a bad gasket between it and the cylinder head, loose bolt(s) etc. If possible smoke the manifold, ie; find a fitting on the manifold and introduce smoke there. I've done the intake just downstream of the MAF on my VW CCTA, but got nothing. Filled the manifold at the breather hose inlet and found the leak instantly. Also smoke the valve cover through the breather and check the breather diaphragm.

But he has a rich condition not lean. Vacuum leak would cause fuel trims to be positive.

 

[shortyb]

But he has a rich condition not lean. Vacuum leak would cause fuel trims to be positive.

Didn't catch the negative trim, that changes the ballgame. Leaky injectors? Bad O2sensor? How old is the air filter?

 

[JD!]

I'm certain the O2 sensors and injectors are at least "good enough." The car runs beautifully with the MAF sensor unplugged, the AFR readout is very stable in this state around 14.7:1.

It doesn't just revert to a baseline in the MAFless failsafe mode - I can still see trimming going on and the car still runs closed loop targeting stoich the whole time. The N54 which this engine is an evolution of did not have a MAF.

Last night I held a flame to the nipple on the atmospheric pressure side of the PCV diaphragm and saw it was being sucked in more than pushed away, not a regular pulsation like it probably should be. I am going to replace the valve cover tomorrow and see what happens.

 

[iCurious]

I'm certain the O2 sensors and injectors are at least "good enough." The car runs beautifully with the MAF sensor unplugged, the AFR readout is very stable in this state around 14.7:1.

It doesn't just revert to a baseline in the MAFless failsafe mode - I can still see trimming going on and the car still runs closed loop targeting stoich the whole time. The N54 which this engine is an evolution of did not have a MAF.

Last night I held a flame to the nipple on the atmospheric pressure side of the PCV diaphragm and saw it was being sucked in more than pushed away, not a regular pulsation like it probably should be. I am going to replace the valve cover tomorrow and see what happens.

If the PCV is sucking more than that would be a lean condition not rich condition. Changing the valve cover is not going to solve the problem. Again, you are chasing a rich condition, not a lean condition.

 

[JD!]

If the PCV is sucking more than that would be a lean condition not rich condition. Changing the valve cover is not going to solve the problem. Again, you are chasing a rich condition, not a lean condition.

As I see it, the problem is some sort of instability. The car runs fantastic with a -15 fuel trim so long as the MAF sensor isn't plugged in. The problem starts when the MAF is plugged in and the STFTs start fluctuating wildly, seems like a sort of feedback loop.

At this point I just have 0 ideas. No reason to suspect my wideband or my injectors are bad, no intake leaks, MAF is good, no fault codes. PCV is the only thing left that I can think of.

 

[JD!]

I should also mention that with the MAF plugged in the car drives great as soon as it gets moving, it's purely at idle where the shakes come in. Makes sense to me as a PCV issue as these fully open the throttle and let valvetronic take over off idle.

 

[shortyb]

I should also mention that with the MAF plugged in the car drives great as soon as it gets moving, it's purely at idle where the shakes come in. Makes sense to me as a PCV issue as these fully open the throttle and let valvetronic take over off idle.

Yeah, Valvetronic can be a b!tch when they start acting up. Common issue I've heard are the eccentric shaft sensors fail and cause issues like rough idling etc. Could be worth a check.

 

[Trav]

This as far as I can tell a DI engine, I have seen these get lots of fuel in the oil from leaking injectors it will start burning the fumes coming through the pcv system causing an uncontrolled rich mixture.
Check the oil for fuel smell, remove the pcv pipe to intake and plug the intake, if it runs perfect change the oil before diagnosing it further.

 

[JD!]

Not too shabby for 225k miles.

 

[JD!]

Back to the drawing board... ugh.

Valve cover didn't help. Smoke tested the intake again, noticed a slight leak at the turbo inlet clamp, tightened it further and no more smoke. Still, car runs poorly at idle.

Does anyone know if disconnecting the MAF sensor disables the Valvetronic system? Because if it does, that's likely my issue. If it doesn't, I'm sure the Valvetronic is fine.

 

[CleanSump]

Does anyone know if disconnecting the MAF sensor disables the Valvetronic system? Because if it does, that's likely my issue. If it doesn't, I'm sure the Valvetronic is fine.

Disconnecting the MAF throws the DME ('puter) into Safe mode. Then it uses the throttle to control engine load and the Valvetronic is essentially off. You'll have a code and RPM limited (I think to 6500 RPM). So it may seem to drive OK, but......
Are you sure your Valvetronic motor is good and installed correctly?