[Ted Toal]

In my 2006 330i E90 N52B30 non-turbo, my check engine light came on and I read the codes: P0172 and P0175 - both banks running rich. It seemed to be getting bad mileage too, and smelled of gas when first started, though it ran fine.

(Note: at the very end of this thread, I describe what finally fixed the problem: using a $310 Continental MAF rather than any of the four cheaper MAFs I tried. Also, my battery was weak, and this can trigger low voltage MAF and MAP DTC’s without triggering a battery DTC. Replacing a BMW battery REQUIRES telling the computer some things).

I brought it to Midas, they replaced the MAF, cleared the codes, and said if the light came on again it might need O2 sensors. It continued to run excellently and gas mileage returned to normal, but the light DID come on again within 100 miles and I decided to try to diagnose it myself. I started out replacing the upstream O2 sensors but could not reset the codes, then realized it would not likely be O2s because both probably wouldn’t go out at the same time. I bought a new reader that could show live data (I now realize there are many different types of codes and reader interfaces and mine only connects with KWP-2000 format). The live data showed long term fuel tuning in both banks to be around -20 to -25, which accounts for the two codes. Short term fuel tuning WAS near zero. Upstream O2 sensors were flickering around 0 and downstreams were around 0.6 I think. I now know that the upstream sensors are different types than the downstream sensors, and the expected output levels are different. My best understanding is that the upstreams being around zero is correct and expected if the computer is accomplishing its goal of getting the mixture right. At one point I also got a MAF code (one time only), and I disconnected the MAF and worked on the connector pins to be sure they were making contact, and then the MAF code was gone. I tested fuel pressure and it was normal. I thought perhaps the injectors were worn and replaced them, and then the short term fuel tuning went to -15. I put the old ones back and it remained around -15 and since then has crept up to -30. I have a feeling the problem is unrelated to the injectors, but it’s just getting worse. Upon first starting, at idle, it smells strongly of gas and it runs very rough, but I see that it’s running open loop then, and as soon as it switches to closed loop (30 seconds), the idle smooths out, hinting that the strong negative fuel tuning returns the engine to the proper mixture. It continues to run great on the road. I got a couple of cylinder misfire codes (#2 and 4), which happened when I was starting it over and over and looking at live data, and it seems the problem is worst at idle and the repeated restarts may have made the cylinders misfire. I removed the plugs for those two cylinders, and one was wet with fuel and the other was somewhat sooty, which I take as indicators that it truly is running rich at idle before closed loop starts, and that this is not simply a sensor problem saying that it’s running rich. I’m at a loss as to what could be causing this based on what I read about the causes of running rich (on all cylinders apparently, or maybe just those two cylinders, one was left bank and one was right). I keep suspecting that the new MAF is bad so I replaced it and also retweaked the connector pins but that made no difference. I’m wondering about the MAF live data, which shows 18 g/s at idle. That seems high from what I read, but I haven’t been able to find a spec on this particular engine. I looked for bubbles in the coolant, and there were bubbles, fine foamy bubbles coming up. I can’t tell from what I read whether I would expect large bubbles if there’s a gasket leak, or if small foamy bubbles could indicate it. I’ve had to add coolant a couple times recently, so I think there IS a coolant leak, but I noticed a lot of gunk and wetness around the coolant temperature sensor, so it looks like it might be leaking there a little. What should I do next? Vacuum test? Compression test? Pressure test on coolant? Replace EGR? It really seems like the thing is getting excess fuel from somewhere but I can’t imagine where that would be. The 0172/0175 codes are stable, they don’t ever go away, and NO other codes are happening. I would think I would get other codes indicating something else that would help me isolate the problem. I’m also wondering about cheap Chinese MAFs. I don’t know what kind Midas put on, but the one I bought was about $25 so it’s probably a cheap one and maybe worthless?

Late addition: freeze frame and live data.

The freeze frame:
DTCFRZF: P0172
FUELSYSA: CL
FUELSYSB: CL
LOAD_PCT: 67.5
ECT(°C): 90
SHRTFT1(%): -25.8
LONGFT1(%): -10.9
SHRTFT2(%): -19.5
LONGFT2(%): -11.7
RPM: 1984
VSS(km/h): 22
SPARKADV(°): 17.0
IAT(°C): 20
MAF(g/s): 38.67
TP(%): 27.1

And now live data at 60mph, 45mph, idle:
DTCCNT: 2,2,2
FUELSYSA: CL,CL,CL
FUELSYSB: CL,CL,CL
LOAD_PCT: 45.1,34.5,46.7
ECT(°C): 97,87,92
SHRTFT1(%): -26.6,-20.3,-20.3
LONGFT1(%): -13.3,-14.8,-15.6
SHRTFT2(%): -25.8,-14.1,-14.8
LONGFT2(%): -11.7,-12.5,-12.5
RPM: 2055,1446,730
VSS(km/h): 92,65,0
SPARKADV(°): 23.0,24.0,-6.0
IAT(°C): 13,15,16
MAF(g/s): 27.05,12.97,7.49
TP(%): 23.5,18.4,15.7
O2S12(v): 0.810,0.895,0.905
O2S22(v): 0.815,0.910,0.905
LAMBDA11: 1.024,0.940,0.941
O2S11(mA): 0.047, -0.285, -0.277
LAMBDA21: 0.980,0.983,0.947
O2S21(mA): -0.094, -0.082,-0.250

1. Things that seem strange to me:
Idle load 46.7% ???
2. Short FT more negative in bank 1 than bank 2 at 45 mph and idle
3. Idle rpm 730? Tach shows about 500.
4. Throttle position 15.7% at idle??

Another thing of note: I was concerned that my OBD tool would not reset the codes, but then learned that the codes are not regular DTCs but pd DTCs, meaning they cannot be reset by the tool, but instead are reset by fixing the problem triggering the code AND THEN doing a Drive Cycle.

I’m wondering if maybe the injectors WERE a problem and I was misled by strange highly negative short and long fuel tuning parameters after such a change, and maybe I should have done a drive cycle before dismissing them as not the problem?? Maybe that would have cleaned up the fuel tuning values??

[PhaseP]

If you have not already seen this thread I had started already sometime back:
https://www.e90post.com/forums/showthread.php?t=1304563

Couple of years after I had done that connector replacement, I had the symptoms happen one more time. I cleaned the connector pins one more time and fixed that again. Bu the pins need to be deeply cleaned, and the pins at the cable end connector, not just the MAF sensor's pins.

That 2nd time I had also noticed that oil filter housing gasket was starting to leak (again.. it just does with these engines). After that I am suspecting the hot mixture oil and coolant that is leaking out the oil filter housing gasket is condensing and depositing nearby area, including the MAF connection.

2006 E90 MAF's are analog, not digital and I think due to this are more sensitive to connection resistance. MAF's connector pins are very thin too, the female side pins do not bite with high force.

[Ted Toal]

Thanks, yes, I did see your post! As a matter of fact, just yesterday I decided to clean my MAF connector for a third time and this time I took it apart and very carefully inspected it, and did a more thorough job. However, I did not replace the pins because it seemed like it was more likely to cause problems because I don’t have the right tool to put new pins on, and actually, the old pins did not look corroded at all. However, I did find by sticking a solid wire into them, that one of them was kind of loose, and so I bent back a small tab on top of each pin so I could gain access to the interior, where there were 2 spring loaded tabs that press against the MAF pin, when it is inserted. I was able to press down on each of these pair of spring loaded tabs and bend them down so they would have much more force for closing the connection. I then cleaned it multiple times very thoroughly with an electrical connection cleaner. I put it back together and at first it seemed like things were running better because the short term fuel trim was zero or positive most of the time, but as I drove it around for a while, I finally decided it actually wasn’t really any better and it wasn’t getting better, the long term fuel trim remains at about -20 and the short term fuel trim sometimes went far into the negative range.
I’m curious about what you mean by the connector at the other end? I tried to figure out where the other end is and I see that the cable enters the junction box under the intake manifold. I tried to find out if there are connectors inside of that, but web search is indicated nothing so I suspect that junction box is just a wire junction center with no connectors. So I’m thinking that the connector you were talking about must be located inside the electrical box with all the connectors that go to the engine control unit? Wait a minute, maybe you are simply referring to the connector pins on the MAF sensor itself? The thing is, mine is brand new, so I don’t need to clean those connectors. And I see no sign of oxidation or rust or corrosion or dirt or oil or anything on any of the pins or connector.
Besides working on that connector, I also replace the MAP and that made no difference. I just now finished doing a compression test and found something unusual. Cylinders 1 through 5 are 230 psi, which, as I understand it, is way too high, they should be around 170. cylinder 6 is in fact 170. I read that high compression can be caused by carbon deposits, and if it’s running rich, maybe there are a lot of carbon deposits. Yet I don’t think it’s really running rich because it runs too damn good, AND as far as I can tell the downstream O2 sensors are reflecting a good mixture. Lambdas are around one, which is I understand it also means the mixture is good. It really seems more like the MAF is likely the problem.
Many people say that cheap MAF’s absolutely do not work. Furthermore, I see no cheap MAFs sold on all the BMW parts websites that sell not just official BMW but other brands also. So, I’m thinking I really ought to break down and spend $200 on a new MAF, my third one.

I just looked for the connector on the other end of that MAF cable, in the computer connection box, and although I think I see the wires, I really can’t see where they go to because they dive down into this hole.

Let me add something else. I’ve been monitoring the MAF output very carefully with my OBD tool live data mode. Earlier it was suspicious because it seemed to go up to about 20 at idle a lot and that seemed way too high. However, after the work I did working on the connector yesterday it doesn’t seem to be doing that. That makes me think that it really did help to work on the connector. However, despite that the fuel trims are nowhere near wanting to go back to normal, as far as I can tell. My understanding is that they both should be around zero, that is all four, short and long term for both bank one and bank two. Perhaps the long term could be somewhat negative but it shouldn’t be -20, probably not even -15. Maybe -10 at the most from what I read. The thing is, I am unclear as to how the algorithm computes the long term fuel trims. I would think they would be sort of a low pass filter on the short term trims, but something also happens with regard to engine load and rpm. I think maybe there are fuel trim CURVES, not single values as the live data would suggest. That makes it real hard to tell what’s actually happening and whether things are getting better or not. Maybe I actually fixed the problem by doing the connector work and maybe 50 miles of driving just wasn’t enough to reset the light yet? Well, I should never see long term fuel trim going to -30 should I? This happens sometimes at idle.

[PhaseP]

I’m curious about what you mean by the connector at the other end?

I just meant not the pins on the sensor itself, but the connector that goes into the sensor. Sorry not being clear on that.

I put it back together and at first it seemed like things were running better because the short term fuel trim was zero or positive most of the time, but as I drove it around for a while, I finally decided it actually wasn’t really any better

When you first tried did the idle settle on 650 rpm after warming up? From my experience when everything is good idle settles on 650 rpm. If it is going below that, and staying at 500 rpm, or keeps fluctuating something is wrong. This is assuming AC is off. AC raises the rpm a little I believe.

After you drove around, the connector might have moved. It is not only that pins get dirty, it is also that the connector housing doesn't fit tightly in the sensor. The replacement blue housing sits tight. When I was troubleshooting my issue, many times I made the engine run good by just reseating that connector to the sensor. A quick check for this was, if engine was idling at 500 rpm, I would shut off the engine. Disconnect and reconnect MAF connector, start the engine and then engine would settle on 650 rpm idle.

So, I’m thinking I really ought to break down and spend $200 on a new MAF, my third one.

Try part number DL-3089 from autozone, cheaper but real deal if they didn't change. A few years back I had fluctuating idle problem. Just to rule the MAF sensor out, I bought the cheapest MAF sensor from local autozone, which was autozone part number DL-3089. The sensor that was inside DL-3089 box turned out to be Continental Made in Germany part. The BMW logo and part number had been erased from it. OEM is either Siemens VDO or Continental. I still have that MAF sensor in my car, and keeping the original as spare. The culprit had turned out to be oil drain line from the CCV to the engine oil pan had came lose and was leaking air into the intake. Different issue.

I also highly suggest you get an INPA Ediabas cable so that you can read the codes and other diagnostic data with BMW tools on a laptop, as BMW intended. For instance you can easily reset the fuel trim adaptations with that. Search for INPA BMW Tools, INPA cable.

[PhaseP]

Originally Posted by Ted Toal

The thing is, I am unclear as to how the algorithm computes the long term fuel trims

Here is a good explanation of modern BMW engines fuel trim workings:

https://bimmerprofs.com/stft-and-ltft/

Note that INPA is referenced in that article several times. The tool you have is translating the BMW values to you. If you get INPA working, you will get the BMW values themselves from INPA. It will be easier to do diagnostics because then you can compare your values and the results to the above article or others who have posted their INPA screen shots. Plus you can do a lot more as mentioned before like resetting adaptations, checking rough running on each cylinder and so on.

[PhaseP]

The user with 1 post count AnthoyColucci made a sneaky spam post above!! (By the time you read this admins will probably delete that post)

Will report this post above, but before reporting wanted to show others. First, the message is copy pasted from autozone's system too rich information page: https://www.autozone.com/diy/diagnos...-bank-1-bank-2

Second, a URL link is inserted in the message as a dot at the end of first sentence. Forum users are not likely to click on that, but search engines will index this forum page as making a link to that url. And then if too many sites and pages are linking to that site, it will come up higher in search engine results. The site linked from its name and url definitely looks like a spam site.

[Ted Toal]

Thanks, lots of good info there.

I’ve ordered an MAF connector rebuild kit. But I also, I bought the continental detector this morning for $311 (returnable), not expecting it to work, but it actually does seem to make a big difference, negative fuel trims are disappearing fast, I’m up from -20 to -3 on the long term fuel trims already. But I think I’m gonna wait until I rebuild the connector and try one of my old ones again and I will return that thing if that works.

Meanwhile, a compression test, showing way too high compression, and all cylinders lead me to treat the engine with carbon blowout treatment and following that, the running too rich codes went away. But I’m not convinced yet that that is what really did it, because in my everlasting saga of events, during that process, my key got stuck in the ignition and the car wouldn’t start, and I had to remove the key circuit relay. It was very hot. I let it cool and put it back in and the car started right up, but my OBD tool was indicating that it reset everything and all the codes went away. Then, P0172 came back. But after I drove it for a while, it went away again. Currently, I am in the process of replacing the vent valve and vent pipes. I’m thinking that they may be causing carbon buildup and high compression.

[Ted Toal]

Thanks, I was wondering what INPA was. I’ll check out your reference on bmw fuel tuning.

[Ted Toal]

PhaseP, can you tell me how you were able to attach the pins when you rebuilt the connector? Those are crimped in place and there must be a very special tool to do that. I think my pins are somewhat damaged and I should replace them but I’m wondering how I’m going to do that?

[PhaseP]

I had ended up not replacing the pins, only cleaning them. But I had replaced the housing. So didn't need to use a crimping tool.
You would need an open barrel crimping tool of correct die size:
https://iwiss.com/how-to-crimp-open-barrel-terminals/

[Ted Toal]

Thanks. I’m hoping the $14 one at Harbor Freight will work.

Here’s an interesting observation. The wires on a fairly new crankcase vent hose’s heater pulled right out of the connector, same kind of pins as the MAF connector. They had been crudely crimped with a pair of pliers it looked like, completely unsatisfactorily. Wish I could remember where I bought that hose.

[PhaseP]

Make sure Harbor Freight one is for open barrel pins.
Amazon has $17, $18 open barrel crimper tools when I search.

[Ted Toal]

Well, I have it all fixed! Quick summary:

1. The main problem was that cheaper MAF sensors did not work. I tried 4 different ones ranging in price from $40-$180. But when I switched to a $310 continental one, it immediately worked, the negative fuel trims immediately started becoming normalized towards zero.

2. I did redo the MAF connector, including replacing the pins. This was a big job and I will say a bit more about this below. But it made no difference at all.

3. The rich fuel DTC’s went away soon after I installed the Continental MAF. However, while trying some of the cheaper MAFs, another DTC relating specifically to the MAF came on, and it did not go off when I put it in the Continental. I thought I needed to do a drive cycle to make it go off, but in the process of doing that another DTC came on relating to the MAP sensor. This threw me for a loop for quite a while until I eventually discovered another problem: my battery was weak. Believe it, or not, with all those trouble codes, that computer still never gave me a battery trouble code.

4. I replaced my battery with a new AGM battery, which I didn’t even know about AGM it until now. The old battery only lasted two years, and the mechanic at Walmart asked me if I had coded the car when I put it in. I said no. I have no idea I was supposed to do that. It turns out that on a BMW you can’t just change the battery, you also have to tell the computer you did so, and tell the computer the kind of battery you put in. That is not easy to do if you don’t have the right equipment. I was ultimately successful, but although it only took about half an hour to put the battery in, it took all day to get INPA installed and figure out how to use it to code the new battery. All the codes are now off, but I’m waiting for 100 miles to see if anything comes back on.


Something I have now learned is that OBD readers come in a great variety of styles, and there is a reason why some of them cost more. Some are only able to read stuff from the car but not do anything to the car. As you go up in price, the amount of things they can do to the car increases. My $50 device was not able to even tell the car that it had a new battery, let alone recode the type of battery. I heard about a more expensive one that will let you tell the car you have a new battery but you can’t tell it what type. One video mentioned a $120 OBD tool that could do both operations. If I were buying another one, I would look for something that was at least of that ability.

The INPA software is part of a suite of Windows PC software from BMW, written in German. The software is very old legacy software, which BMW has continuously updated as they come out with new cars, so this software is a mess. It’s a mess to install in particular. (I am familiar with legacy software messes, having maintained such software myself during my career). There are a lot of different versions out there and if you don’t have the right one, you’ve got problems. It’s hard to use because even the versions that have some translation into English don’t have 100% translation into English. There are no good manuals for it, and it is very very complex, with many different programs that you run to do different things. The cars themselves are far more complex than I had appreciated. They are full of literally thousands of options that can be set one way or another using the software. If you can read German that is. I would strongly advise people to just stay away from the INPA software unless you’re some kind of professional that can get a professional version of it installed. Just get a more expensive decent OBD tool that can do what you need to do and make sure it can do it before you buy it.

Replacing the MAF connectotor pins was a challenge and my job was not as good as the factory job but it was adequate. If you plan to do that, a key thing to look for is a crimper that crimps style SN-28B terminals. That is not exactly the kind of terminal we are talking about but it is close. That type of terminal is often used with electronics such as Arduino computers, so the crimpers are pretty easy to find. I’ve never used one before, and it took me a while to even realize that I needed to find the right size hole, or die, and that each die had two sides to it. You might be able to find one that has the kind of die for that particular connector we are talking about. The main difference with the SN-28B is that the collar only crimps the wire insulation and not the small rubber insulating grommet, so only crimp the wire and its insulation, and then slide the grommet down over it. When you buy the connector parts, you have to buy the individual components from BMW, the connector body, the connector collar, the connector rubber insulating band, the five pins, and the five grommets. It’s about $30. But like I said, the connector turned out to not be a problem for me.

After all is said and done, and everything was working, I put one of the cheaper MAFs back in, and immediately observed highly negative short term fuel trims. Unfortunately, when people say that you need to get the factory MAF, they are correct. And it costs a fortune.

[Serf27]

The protools app isn’t the cheapest scan tool but it is very user friendly and can do almost anything ISTA can, for future reference if you look for a scan tool.