Bad MAF, vacuum leak or leaky/burnt valves

[rpmihail]

Hello, I ran into a dead end troubleshooting my car. I searched forums for weeks, still can't figure it out, so I thought I'd ask here. Background: 2003 325i, replaced the block after overheating caused head warpage, stripped block head bolt threads. I tested valves by pouring gasoline into ports, both exhaust and intake, very minimal leakage, one could see traces of gas over a few minutes if you looked carefully. I replaced all gaskets (intake, throttle body, IAC, new intake boot, new DISA, etc). Car idles rough and lean codes are set for both banks. I checked, double checked, tripple checked for vacuum leaks everywhere I know. I'm also getting code for timing over advanced or performance on bank b, I think that means the exhaust side for VANOS related codes.

I have a graphing capable obd2 scanner and here's what I observe: with MAF plugged in, car idles rough and pre-cat O2 sensor voltage doesn't oscillate normally, is mostly around 0.02-0.1V (lean) with spikes up to 0.9 when you accelerate, and short term trim is always maxed out to 27.3%, long term 0%. This behaviour is the same parked and while driving (in load).

When I unplug MAF sensor, engine idles smoothly, pre-cat O2 sensors oscillate "normally" between 0.1 and 0.9 and short term trims also oscillate negative to positive, I believe in response to o2 sensors to keep cats happy.

My question to the experts of this board: what is more likely: bad MAF, some hidden vacuum leak that I failed to find for weeks now, valves leaking (mechanical issue), or timing set incorrectly (maybe even failed VANOS, although I rebuilt it).

Other notes: with MAF plugged in, opening oil cap lowers RPM, degrades idle quality. With AC on, MAF plugged in, idles much better than AC off MAF in. I cleaned the MAF sensor many times. MAF unplugged, idles/runs better but shift points are off, shifting (auto) is jerky.

[MarcoZandrini]

Unplugging the MAF causes the DME to go into the open loop mode. This mode is used when the engine is cold. It helps to get the cats up to temperature quickly. It does thus by running an overly rich fuel/air mixture. In your case, unplugging the MAF masks vacuum leaks.
The only way to find ALL of the vacuum leaks is to smoke test the intake.

[rpmihail]

Thank you for the reply. When the engine is cold, it idles rough with MAF plugged in. This rough running happens during the open loop mode when rich mixture should be seen. Does the startup "routine" to get cats hot use air volume readings? When I unplug the MAF it runs smooth cold as well.

I don't have the means to do a (good) smoke test, but here's something else I've tried to detect a vac leak: take air box with filter out, leaving MAF sensor housing last. I then put my hand on it, blocking the air intake, and look at how long it takes the engine to stall. For such a major leak that o2 and stft report, I should be able to almost entirely cover intake up, but it stalls really easily. While doing that, I also listen carefully for amplified hissing or suction noises somewhere else (where supposed vac leak would be) but can't find any.

[MIKYZZ4]

Until a smoke test is performed with a professional smoke machine, you will be chasing leaks blindly.
You can make an attempt at checking the usual suspects. You missed your chance at easily replacing CCV,
as that system is completely made out of plastic and rubber that cracks easily with age, when the head was off.
Remove all rubber intake boots, hoses, carefully check for cracks by squeezing and bending them to expose tears.
Check secondary air pump hoses in the same manner. Btw-have you tested fuel pressure at the rail?(45~51)psi.
A cylinder leakdown test can reveal if valves or headgasket are sealing correctly. You seem to be mechanically savvy,
so I'm just throwing out options for you, that can help reveal hidden leak locations.
Have you scanned for error codes?, as they can help provide clues, and you did not post any.

[rpmihail]

I have replaced all plastics, including CCV related gaskets. Selling the car is more likely than a smoke test... it has 128k miles.

Please help me understand how, assuming MAF is reading correctly, the short trims are maxed out, but precat o2 reading lean. Without MAF, trims hover around 0 and o2 oscillates normally between 0.1 and 0.9.

It seems to me that a vac leak would cause lean mixture even in openloop, or whatever the mode without MAF is called. Any insght is very much appreciated. You both seem to think it's a leak rather than bad MAF.

- - - Updated - - -

Ah, and the only code thrown is p0014, system overadvanced. Can poorly timed VANOS lean the mixture? I also replaced VANOS internal gaskets.

- - - Updated - - -

Fuel pressure is at around 50psi. Compression is 120ish on all cylinders.

[MarcoZandrini]

The DME ignores the data from the O2 sensors during open loop mode. It relies on the data from the MAF as long as the MAF is plugged in to its harness.

[MIKYZZ4]

Is there a possibility o2 sensors connections were X'd?
That is the only option I can suggest,
Besides a $100 smoke test that can reliably confirm and locate all intake leaks.
P-codes are notoriously unreliable and may be erroneous,
Perform a scan with BMW specific scan tool/software.
OBD2 Codes are not BMW Codes.

[White94RX]

It's not timed properly. You said in the original post you've got timing fault codes. That will cause everything you described.

[BimmerBreaker]

It's not timed properly. You said in the original post you've got timing fault codes. That will cause everything you described.

He also never mentioned having the head decked or anything else that should have been done after the major overheat... sounds like it was basically band-aided back together and is falling apart at the seams again now...

[rpmihail]

All, thank you for your valuable input. The issue is now resolved. There were two simultaneous issues. The secondary timing chain sprockets were one tooth off sync. I have no idea how I didn't catch that,... but it ran, poorly. So if anyone ever runs into low compression, uniformly over all cylinders, the timing may be off or, more tragically, valves bent. I was lucky, valves are fine. Second issue was a faulty MAF sensor. It reported less air than reality, which created the lean condition, lack of power and surging/loping for a minute or two after start-up, I guess before the ECU goes into some default mode to stabilize idle.

BimmerBreaker, I did resurface head and lapped all the valve seats, among other things, but the assumption was fair given my poor initial description.

Thanks again and I hope this is helpful to others who tackle this job.

[White94RX]

Glad you got it running.