Only runs correct when no MAF is plugged in?!

[fishhouse4]

I have been dealing with a rough idle and hesitation for about a year. My symptoms have progressed from minor to really bad currently.

In a nut shell here's what I'm experiencing:

• Really bad hesitation (like a shudder) for a few seconds only when applying light gas (maybe around 2,500 rpm and usually under load). Most of the time it is ok if you drive super slow with feather light acceleration and ok if driving it hard at WOT.

• This comes also with a very round idle. Minor sometimes...VERY bad at times - everything shakes and rattles. Sounds like it is going to stall out.

• Additionally, CEL lights with many codes. "Multiple misfires" (all cylinders), and "VANOS" codes. Clear codes and CEL stays off a few miles and then at one random time you give it gas and it hesitates they will all come back - in random order each time.

• Occasionally on cold mornings it will hardly accelerate at all. It just sputters and is barely drivable (is that limp mode?). I have to let it warm up, pull over and clear the codes and then it will run as described above - a hesitation only at specific times. I've learned to drive "around" it, so to speak.

Ok, now here's where it gets interesting! I've tried several things over the last 6 months to troubleshoot and finally gave up and took it to my BMW tech. One of the first things he tried was unplugging the mass air flow sensor. It was replaced last year, so I never even thought to look at it.

Magically, it runs near perfect with the MAF unplugged. I have since learned that it defaults to stored parameters if unplugged (it stumbles for a few seconds and then purrs like a kitten).

Great = needs a new MAF, right? Wrong!! I've now bought and installed 3 new MAFs and as soon as you plug one in = back to rough idle/hesitation/misfire codes. Unplug them = runs near perfect.

Ok, so wth is wrong when it does this when ANY mass air flow sensor is connected but runs fine without one???


FYI, I have done the following:

• Cleared codes.
• Smoke test = no air leaks.
• Die test = no oil leaks.
• Removed my Dinan stage 1 tune and restored stock.
• Removed foam air filter and replaced with stock paper filter.
• My mechanic reprogrammed it with a fresh copy from his INPA.
• Checked plugs and coilpacks (both new last year). VANOS solenoids new last year as well. MAF new last year.
• Cleaned fuel injectors (direct through the rail with cleaner).
• Replaced MAF (with 3 different new ones).
• Replaced integrated supply module (IVM).



FINALLY SOLVED!!! UPDATE INFO HERE: https://xoutpost.com/1236219-post18.html

[crystalworks]

What codes did it throw?

[oldskewel]

Could it be a problem with the MAF connector or wiring?

[Overboost]

What MAF sensor brand did you replace it with? There are a lot of horror stories out there with cheap MAF sensors. When you unplug the sensor, the DME reverts to Alpha N mapping (throttle vs. RPM). Plugged in it uses the MAF values. If the MAF was Genuine BMW or OEM then I would check connections at both ends (MAF and DME) and look for any physical damage to hoses or orings.

[fishhouse4]

Originally Posted by Overboost

What MAF sensor brand did you replace it with? There are a lot of horror stories out there with cheap MAF sensors. When you unplug the sensor, the DME reverts to Alpha N mapping (throttle vs. RPM). Plugged in it uses the MAF values. If the MAF was Genuine BMW or OEM then I would check connections at both ends (MAF and DME) and look for any physical damage to hoses or orings.

All Bosch. That would be OEM, correct?. I've been through 4 of them now - and all from different sources (eBay, Advance Auto, Napa Auto, etc). It would be insane to believe I have had that many duds.

[fishhouse4]

Originally Posted by Overboost

What MAF sensor brand did you replace it with? There are a lot of horror stories out there with cheap MAF sensors. When you unplug the sensor, the DME reverts to Alpha N mapping (throttle vs. RPM). Plugged in it uses the MAF values. If the MAF was Genuine BMW or OEM then I would check connections at both ends (MAF and DME) and look for any physical damage to hoses or orings.

I will check the connections on both ends.

The only two ideas my tech had was a new engine wiring harness (which would address those connections you just mentioned) OR a new DME. Both are stupid expensive...

Ok, help me here...this is what I don't get about the wiring/connection possibility:

It's just a wire, correct? It simply transmits the signal from point A (MAF) to point B (DME). Like... all or nothing, I would assume. So if the wiring or connections were damaged, then the signal wouldn't go to the DME. Which would be the exact same as having the MAF unplugged. So I would think in that case you would not get MAF signals to the DME regardless of if it was plugged in or not.

Or are you suggesting that the wiring/connections could be damaged in a way that could somehow degrade or corrupt the signal from the MAF??

[Overboost]

Originally Posted by fishhouse4

I will check the connections on both ends.

The only two ideas my tech had was a new engine wiring harness (which would address those connections you just mentioned) OR a new DME. Both are stupid expensive...

Ok, help me here...this is what I don't get about the wiring/connection possibility:

It's just a wire, correct? It simply transmits the signal from point A (MAF) to point B (DME). Like... all or nothing, I would assume. So if the wiring or connections were damaged, then the signal wouldn't go to the DME. Which would be the exact same as having the MAF unplugged. So I would think in that case you would not get MAF signals to the DME regardless of if it was plugged in or not.

Or are you suggesting that the wiring/connections could be damaged in a way that could somehow degrade or corrupt the signal from the MAF??

The terminals can be dirty, corroded or even loose. Just visually inspect the terminal contacts and make sure they are not spread making a soft or poor contact on the MAF connector and on the DME connector. If you have an ohmmeter, check the continuity of the wires from one end to the other. It doesn't cost anything but a few minutes of your time to rule that out or validate an issue.

[oldskewel]

Originally Posted by fishhouse4

I will check the connections on both ends.

The only two ideas my tech had was a new engine wiring harness (which would address those connections you just mentioned) OR a new DME. Both are stupid expensive...

Ok, help me here...this is what I don't get about the wiring/connection possibility:

It's just a wire, correct? It simply transmits the signal from point A (MAF) to point B (DME). Like... all or nothing, I would assume. So if the wiring or connections were damaged, then the signal wouldn't go to the DME. Which would be the exact same as having the MAF unplugged. So I would think in that case you would not get MAF signals to the DME regardless of if it was plugged in or not.

Or are you suggesting that the wiring/connections could be damaged in a way that could somehow degrade or corrupt the signal from the MAF??

First, I think the fact that 4 different MAFs all give the same result means they're all fine. More likely for that to be true than for all to be bad.

When the MAF is unplugged, the control system figures out how to manage things (engine and AT) based on what sensors and system model it has. Many different ways to do something like this, not sure on specifics. But the fact that it runs just fine with the MAF(s) unplugged suggests the rest of the car is good. Engine, even the ECU.

Just a thought, maybe when it switches out of MAF closed loop mode or whatever it is, it also switches out some other sensors (e.g., O2 sensors) while it runs open loop. If you can get good info on how the open and closed loop circuits work, that may be helpful in deciding where else to look for the root cause.

On the connector / cable problem I suggested, I expect they are just wires. But for example, what if one of the (two?) pins with the actual MAF signal is bent / wire broken internally / etc. but other connections within the same connector / harness / cable are still good. Could it be that the ECU gets enough to think the MAF is there and working, yet it's actually getting garbage or no data on that line? Smarter fault detection methods could help in this, but maybe they kept it simple (for once, LOL).

So I would start with a careful visual inspection of the connector, making sure all the contacts look good. Then the easiest thing to do to check for a wire issue is to get the car running (with MAF connected, so running badly) and wiggle the wire as far as you can see it, hoping you intermittently restore something to make it run differently. Think about where an internal wire failure is likely to happen - at the connector, bends, any place it may have been pinched, etc.

After that, it's trying to find the other end of the MAF connector wires, and testing continuity through the whole wire.

[upallnight]

Originally Posted by fishhouse4

I will check the connections on both ends.

The only two ideas my tech had was a new engine wiring harness (which would address those connections you just mentioned) OR a new DME. Both are stupid expensive...

Ok, help me here...this is what I don't get about the wiring/connection possibility:

It's just a wire, correct? It simply transmits the signal from point A (MAF) to point B (DME). Like... all or nothing, I would assume. So if the wiring or connections were damaged, then the signal wouldn't go to the DME. Which would be the exact same as having the MAF unplugged. So I would think in that case you would not get MAF signals to the DME regardless of if it was plugged in or not.

Or are you suggesting that the wiring/connections could be damaged in a way that could somehow degrade or corrupt the signal from the MAF??

Nope. Mass air flow sensor measure the amount of air going through it. How does it do that? It's a wire (or a film) that's heated by electricity and the air passing through it will cool the wire down.

The DME can setermine how much air is passing through the MAF base on the temp of the wire after it has cooled down.

Unplugging the MAF and the DME default to a preselected mixture base on the engine coolant temp and the incoming air temp.

I know on a 3.0 M54 engine the MAF is a Siemens unit, and not a BOSCH unit. Cheap MAFs on eBay or Amazon are known to cause more problem than what you save on buying them.

[andrewwynn]

Is that M62 or N62? i can't remember which is the 4.8; sounds a lot like the intermediate lever issues with the N62 motor.

what are the realtime voltage and g/sec readings from your MAF? get some control readings from other folk out here.