[Yzandrl]

So, with a little extra dramatization tossed in, like any good story, this started with the best intentions, and ended in misery.

I began the journey by performing a health check code pull that I do every few months to stay ahead of failing components, not thinking I could end up in the mess I am in now…and a rear pad/rotor job (that part was fine, btw).

2006 e90 330i n52, 105K miles

What I pulled was:

A114 CAS Line fault, el. Steering lock – been clearing this for 8+ years, no issues.
A554 KOMBI: No message (telephone) - been clearing this for the life of the car
903D FRM: Clearance Light, right, faulty but fault not present ( I understand this means failing bulb, no problem), so basically a clean bill of health!

Then digging in further, I noticed increased negative fuel trims -8% ish for multiplicative bank 1,2 which used to be -3% a while ago. In a misguided pursuit of perfection, I thought I’d dig into it read a bunch online books and posts about diagnosing fuel trim issues and their possible reasons. I didn’t want to just throw $$$ at parts and came across a wealth of reasons for this to occur.


• MAF overestimating o2 (likely, but the kg/h seems within legit range, and this would affect the whole system)
• O2 Sensor (unlikely, because both banks)
• Fuel pressure too high (maybe, but too low is more common)
• Leaking Injectors (unlikely, both banks again)
• Stuck evaporation system purge valve (maybe, but no gooey oil cap and or did I find gooey oil when I did valvetronic gasket replacement which had me disconnect the back of the valvetronic)
• Possible leak crankcase vent tubes (strange, as this should make it lean, but enough people said this goes into a rich condition on this model/design and I did touch this recently)
• Sleepy ignition system (105K coils, 35K plugs, maybe?)
• Bad coolant temp sensor (wierd but I read this, but it is giving temps in range of what would be normalish without having a voltage spec to compare it too)

So, the most interesting finding during my research was that some people were having fuel trim issues and after replacing several costly components (injectors, maf, o2s, pre/post cat, etc…) it turned out to just be ‘worn connectors’ losing continuity fixed by cleaning, bending pins, or replacing the MAF’s socket connector, (which is analog on this model) and it fixed their problems 100%.

So, before throwing parts I did some diagnostic digging myself.

I pulled plugs and coils, observed grey but evidence of carbon fouling (from running rich I assume) in plugs, cleaned and reinstalled. Cleaned coils and connectors with electronic contact cleaner, reinstalled

Cleaned every sensor connector that I’ve touched during the valve cover gasket replacement with contact cleaner. Did the same to MAF and slightly bent the contacts with a fine needle, and drowned the MAF in an ocean of MAF spray.

With the engine off, I did a smoke test to check for any slight vacuum leak. I did not find any evidence of any leaks.

Didn’t get to test fuel rail pressure test as planned as the kit I rented from AutoZone didn’t fit the non-domestic cars (Schrader valve) but INPA said it was basically functioning, so unsure if it was getting the PSI required, but had to move on without testing rail pressure with car on/off timed values either.

The O2 sensors I didn't check as the likelihood of both banks would be strange reading the same over reported o2 values (and the kit I had rented wasn’t metric either)

So, after cleaning all the sensor connectors and reassembled them all, I hopped in and cleared all adaptations. And that set off a Christmas tree’s worth of ‘oh-oh’ lights. Transmission, DSC, Check Engine, and maybe one more that I don’t remember. I didn’t even start the car at this point, and the cleaner had plenty of time to evaporate.

At this point (after minor non-life threatening medical event) I was thinking I fried something electronic and expensive. Finally, out of wishful thinking, I couldn’t see how cleaning adaptations would do that, so I guessed it was low voltage during adaptation reset, and failed to execute correctly. So I charged the battery, cleared the codes, cleared adaptations again (no Christmas tree this time), and started the car, checked everything, and it seemed fine with code checks and smooth idle with cylinder run time values (so the crankcase sensor worked too). Feeling much better, I took it for a test drive and it drove just normal for several miles, maybe 15 mins and then Christmas tree came again and barely made it home with the engine stopping only once at a stop light.

So, the new codes:

DME

2A94 DME: Crankshaft Sensor, Signal
No Signal
Fault = Currently present
SAE P0335
1st atMileage/KM = 169360
1st atEngine speed = 0.000 rpm
1st atEngine status = engine off
1st atStatus, engine management = engine at standstill
1st atVoltage term. 87 = 14.320 V
Last atMileage/KM = 169368 km
Last atEngine speed = 0.000 rpm
Last atEngine status = engine off
Last atStatus, engine management = engine at standstill
Last atVoltage term. 87 = 10.766 V

I checked the DME fuse pack and it was fine. I then snaked my arm in there below the starter and made sure the connector was at least pushed in - it was (though there was no way I could remove, clean, or inspect without getting serious about lift bay or manifold removal) And my activities didn’t go anywhere near this sensor at all as I had to contort myself to get in there.

FZD Function Centre, roof
A670 FZD: Fogging Sensor
Short circuit to UBatt
Fault currently present
Frequency 2
1st AtMileage 169352
Last AtMileage 169352

This one seems a symptom of something else, it wasn’t broken before, so why now? I disconnected it to remove it from the ‘short’ path in entirety; but it didn’t change the code. I disconnected the rain sensor too, just generated another error. When the error was present, it was giving humidity readings 32%, when I pulled it, it just thought it was 127% humidity when I tried to pull a value. So it functions, not sure about short to UBatt, but this seems a significant finding to correlate to the crankshaft sensor failure.

EGS transmission

51A7 Transmission control: no message (engine speed) from engine management
Implausible signal or value
Fault currently present
Last atMileage = same
Last atEngine speed = 2016.00 rpm
Last atBattery Voltage 12.640 V
Last atStandard fuse Protection ACC = O.K.
Last atStandard fuse Protection CAS = O.K.
Last atStandard fuse Protection DME3 = O.K.
Last atStandard fuse Protection DME2 = O.K.
Last atStandard fuse Protection DME1 = O.K.
Last atTemperature 104.000 C

This seems downstream to the crankcase sensor being wonky.

DSC
5E1A DSC: Engine Management interface (ok, not much info there)

This also seems downstream to the crankcase sensor being wonky.

This is where I need the expertise here to help save the day. I thought I was careful but perhaps I did half-fry a component that isn’t registering codes? Or, perhaps I’ve impacted the electrical cables in some area that would reach this crankcase sensor, perhaps correlated with the condensation sensor on the windshield? Lastly, did the smoke test pressurizing the crankcase slightly push oil into a possibly failing crank case sensor and short it out? That one seems a bit off I admit with the other unexplained short to UBATT.

At this point I'm without a direction and any help would be very much appreciated!

- Yzandrl

[Emilime75]

You sure your battery's good? Very often, when people get a myriad of codes, it turns out to be a weak battery.

[Phil325i]

Originally Posted by Emilime75

You sure your battery's good? Very often, when people get a myriad of codes, it turns out to be a weak battery.

+1

[hassmaschine]

8% isn't much to worry about IMO. Pretty sure the max is 25%, and it will throw a code if it's too far out of tolerance.

[Yzandrl]

Thanks for the help...I'm pretty sure the battery itself is fine. I didn't find any of the precursors that come with an aging battery before this. And, when I fully charge it and keep it on a charger when doing testing the voltage is 14V+ and it will still register code for 'no' signal.

However, if that short to UBATT reference means I have a short somewhere, then that would drain the battery or prevent re-charging, leading to what you're talking about. I'll monitor voltages for a while by checking pre drive, and post drive V at the terminals. If I find it's 13.9 V at start and 10 V at end, then we're onto something.

I still think I've banged up some wire(s) or connector in the harness and have zero clue what and where to check. I can't even get access to the sensors themselves to 'follow the wire' because i'd have to pull out the intake manifold and the other is in the rearview mirror.

And your right about the 25%, but i wanted to make it more efficient and figured it wasn't going to be that hard to improve if I just had an air leak or pressure regulator failure.

[E92William]

Have you tried disconnecting the battery for a bit and putting it back on after clearing everything. I'd double check your work with the connectors you cleaned maybe you mixed some up or they aren't fully closed who knows.
My fuel trims are usually between 8 and flat out 0 mostly on the lower end I wouldn't mess with it. These cars use maf for just emissions purpose so it won't be perfect with it.

[Yzandrl]

I just rechecked all connections after your post, and I don't see how my actions could have done this. I wasn't pressing or tugging anything that should have this result. That harness has retention points on both side and is shielded by conduit. I only messed around the front sensors, fuel rail connectors, and MAF, nothing else was part of my 'gasket' replacement.

Anyway, the battery is at 12.66V.

So, thinking of that short to UBATT on FZD, I left the car alone without key, opened trunk, pushed in the latch in to trick it to think it was closed, pulled off the neg terminal, and then reconnected using multi meter in a series, it made a noise and showed a 5.73 A draw, that dropped to 2.xx A then down to -.80 mA (not sure why it has -) after twenty mins, then -.04 mA after 2 hours, which isn't 22mA as expected.

Anyway, I pulled the related FZD fuses, and no change...so it shouldn't be related as they are completely separate circuits.

I think my options are:

Damaged a major control component
a short to ground in harness
a short to supply/power in harness
a crankshaft engine sensor has failed, and has the power to impact FZD/humidity sensor circuit by sheer indomitable will (not sure why they happened at same time, but they shouldn't be remotely related DME+FZD)
Or two separate unrelated issues...most likely

What is short to ubatt anyway? I can't find a translation, it didn't say short to earth. Does it isolate ground/supply or component? Any ideas? I hear all the time that tracking electronic gremlins are a horror show and I'm not sure how to apply a methodical method to test. I guess as a start I need to know what is normal draw on a healthy e90 in sleep mode. Does anyone know that or where I can find it? (edited: found it and tested, states 22mA in fully loaded vehicle after 1 hour with TCU) Also, could reset adaptations resulted in computer damage?

[Yzandrl]

Okay, I am not sure if this needs to move to electronics, but it is engine related so here it goes.

I've taken the time to map every wire/pin layouts from the B+ potential distributor to the final end point of all Fuse #3 components. Using my Bentley e90 book and https://www.newtis.info/tisv2/a/en/e90-330i-lim/uT8YBGK starting at Fuse#3. (These are pretty neat resources, btw)

The issue I have now is testing. I need some help from experienced techs to develop a safe test procedure to further isolate a short or broken wires without having to yank the harness or further harm the older harness (introducing new problems). The best I can come up with on my own is pull negative battery terminal, wait 10 mins, then pull DME connector X60005 out and begin testing resistance X60005, and the sensor pin end points. Or, do I leave power on and test each endpoint connector for their DC supply voltage (which I actually can't find yet). - What should I do for the most accurate/safe? Academically, I am also curious what healthy resistance is across the load pin and the ground pin on some of these sensors are and could I can test their 'health' by doing this as well...or would that damage them...I'm curious if there is a resource out there on this level of detail.

Lastly, is my meter not right? I can't understand how I'm getting a negative reading on mA when the vehicle is sleeping (closed circuit current draw).

[Canuck335]

Saw your post over in the Audio section.

Just an idea-- maybe it's a bad ground, someplace where all the control units are grounded or whatever the common return path is.
Maybe X173, which is driver's side footwell.
https://www.newtis.info/tisv2/a/en/e...-module/q-X173


Side note:I think the slight negative current indicates a capacitor discharging into the battery.

[Emilime75]

I don't think the negative reading on the meter is anything, just an anomaly within the meter, itself...or, since we're talking about DC, reversing the + and - probes will indicate a negative reading, i.e. if you connect the positive probe to the negative terminal of the battery, and the negative probe to the positive terminal, it will read a negative voltage. Maybe that's a clue if you're sure you tested positive to positive and negative to negative.

Not likely to be your issue, but worth checking - the braided ground strap between the engine block and chassis on the left side of the engine compartment. These tend to corrode and cause issues. Check it, make sure it isn't frayed or corroded. Whitish/greyish dust like residue, or a green tint, indicated corrosion.

[Yzandrl]

I'll go and check that ground; including the engine compartment one.

That's brilliant about the negative draw, I assumed it was the connection of the meter, or the connection from neg clamp to neg terminal. I'll leave it overnight and see if it is the same; i read that a capacitor, left alone could hold power for months, but this is draining into the battery possibly, so I can't imagine that 'load' would be long. I guess I'll find out tomorrow.

[Emilime75]

Yes, a capacitor can hold power for a long time. The bigger the capacitance, whether it is 1 large, or multiple small caps in a circuit, determines how much voltage they can store/discharge.

[juld0zer]

Originally Posted by hassmaschine

8% isn't much to worry about IMO. Pretty sure the max is 25%, and it will throw a code if it's too far out of tolerance.

I agree. I've had mine up to -15% before and it didn't log any too rich faults.
I'm not sure what happened recently but all the trims are within -4% now. They used to hover around -8-9%. I almost threw a MAF sensor at it too. I do seem to remember scraping the middle of the exhaust when I had a full car over a big speed hump prior to this change, so maybe it fixed a tiny exhaust leak? There's no obvious signs of a leak though.

Ubatt is battery positive. Some fault descriptions on some scanners are badly translated. You will also find BMW wiring diagrams are littered with abbreviations. For sensors you will find M (ground), U (B+), P (signal). Just when you thought Bosch terminal designations was already a mouthful...
Don't be fooled. A short to positive fault can be caused by an out of nominal range signal or an open circuit. It just depends what the control unit is expecting on the signal line.
For your stalling problem, everything is pointing to the crank position sensor. It's important to not cloud your thought process with irrelevant faults, ie faults in other areas of the car unlikely to cause stalling. A windscreen condensation sensor would only impact the HVAC.

N54 engines use the same sensor and I've replaced several before. None on N52 yet but my experience is only a small sample. Perhaps you could punt on a budget aftermarket sensor before spending on a genuine one? The bolt is aluminium also, FYI.

Your multimeter reading is because you have it set to auto-range.

Resetting adaptations (I assume you used INPA) always does that. INPA doesn't tell you what ISTA does. Essentially you can't start the engine immediately after waking the car after resetting (all) adaptations. You have to turn the ignition on and wait 30s or so.

[Yzandrl]

Originally Posted by juld0zer

Don't be fooled. A short to positive fault can be caused by an out of nominal range signal or an open circuit. It just depends what the control unit is expecting on the signal line.

Interesting. I understand there is a fuse, a relay/switch and load on each separate part.

So, do I understand this? The relay/switch could be stuck in the on position, a positive wire touches ground after the relay/switch, or another B+ is completing a circuit it behind load? Or, finally, when it is running power, the device is just feeding data out of nominal, and triggers the same code?

I will tackle this issue separately. Until then I will pull fuses 5,12, 28 and quarantine those circuits.

Adjusting the auto range, after overnight, it was drawing .011 A; so 11mA. Well below the 22mA 'fully loaded' technical spec.

Next, I managed to get the sensor out after lots of fun. I couldn't disconnect it safely with the connector retention tab holding strong, so I just pulled the sensor out and with the extra slack safely removed it.

Since I had it in hand and in interest of not pulling the DME and causing further wear on the wires. I tested continuity on the sensor itself on pins 1 & 3. I expected A) resistance B) 000.0 registering a 'short' or C) open circuit/infinity. And the winner was C. Just incase I was off in my map I tested 1,2 and 2,3 as well, all open.

It has a magnet in it and I placed it near metal and on wood since I didn't know if it needed something inside to pull...and tested..same

So, that means my sensor is straight dead if I am not mistaken.


EDIT: I read that this is a "Hall effect" sensor. I read it can only be tested when 12V is applied and sample the signal wire (by opening it) and grounding the negative lead and moving metal across it will register change in volts in a range of 0-5v. I also read it should still have some resistance when moving metal across it while measuring continuity, but would validate its connected. It didn't register anything at all. (I didn't do the voltage one as I didn't want to strip open the conduit covered harness wire to test it)

So next question. It says Siemens VDO, stamped with BMW. What is the quality difference from a OEM VDO, or the BMW variant. Are they just the exact same quality, only for a large order, or tested better, or slightly different design with patented tech, or have better materials, ie, platinum vs something cheaper for example (I read that about cheap MAF sensors)?

I think I'm getting close now at least. Just need to order a part.

EDIT: I just ordered the BMW one as it was 30$ more than the OEM VDO (piece of mind I guess). Still interested in learning if OEM is the same.

[juld0zer]

The FZD is one of those unique components in an E90 which has access to constant battery power (T30). It needs this for various functions like cabin lighting, alarm functions, etc. It's not officially documented anywhere but I would expect the input from the condensation sensor to be only required while the IHKA is operating, not something that is woken up for a status update while the car is asleep.

Do you have a current draw problem or just trying to understand the faults logged?

Siemens/VDO is the OEM for the CKP sensor, CMP sensors and MAF.

[Yzandrl]

Originally Posted by juld0zer

Siemens/VDO is the OEM for the CKP sensor, CMP sensors and MAF.

Does that mean 1:1 and maybe some quality control testing for a big customer, or could they be physically different materials? Its academic at this point, and I'm just curious...i guess its part specific perhaps.

As for the FZD, what you write makes perfect sense. I don't know if I have a short/draw while in different wake states but I'm interested in learning the systems. I'm at that point where I have a 14 year car and the only can keep it going cost wise if I do the work myself, and isolate the correct items needing repair instead of throwing $$$ worth of parts at it trying to guess, so some sort of proficiency is required and I'm trying to build it as I discover issues.

[desertman123]

Originally Posted by Yzandrl

Does that mean 1:1 and maybe some quality control testing for a big customer, or could they be physically different materials? Its academic at this point, and I'm just curious...i guess its part specific perhaps.

Usually it's the exact same piece, but they took it off the factory line before they went into the BMW logo stamping machine

Check out random parts on the car. You'll see the OEM brand name - like Continental for the belt, or Siemans/VDO/Bosch for sensors.

You should see the OEM TRW brand M3 control arms. They literally grind the ///M logo off. It's an extra step in manufacturing and yet they still charge less!

[Yzandrl]

Well, the part came in, installed it, and voila. Fixed. I don't understand why the sensor failed at the same time I was cleaning contacts and bending pins for analog sensors (far, far away from this digital one) and reset adaptations, but it likely was caused by something I did. I am glad I found the from under the car repair on youtube instead of the 'through the manifold' method..the idea of moving all those vacuum tubes makes me think air leaks are quite possible when finished...and this was quite easy to do (minus mapping the pins from the DME to the endpoint and testing the entire harness path for breaks or shorts, but that was me trying to be precise instead of just buying a part and finding it was a bad wire).

So, here are some notes for replacing the bmw N52 engine crankshaft sensor who might find this thread during a search. It was my first attempt to really isolate engine electrical without having to throw parts at it, and while it consumed a good bit of time, I learned skills I can apply elsewhere in the future... so I'm happy with that.

notes:

- feel around carefully until you can visualize it after studying a picture/diagram. It takes a while to get it.
- it was pretty dirty around that sensor and was much harder to remove by wiggling it out, but it was super easy to install, no friction that made me think I had the wrong gasket actually. I'd suppose directing some plastic safe evaporative cleaner in there to flush it out would reduce the chance of pushing the grime back in, I just pulled it out with q-tips, overkill, but I had the time and was concerned it would push dirt into the oil sump area.
- I had to remove the sensor before disconnecting it, its too hard to release the tab with a pick.
- i tied dental floss to the screw, and put it in with fingers and it went easy. (dental floss has saved me a few times with aluminum screws that could go somewhere I might not recover... I dropped one in this engine once, only to fish it out with the floss, so I always do it now on anything aluminum).
- i made a bench test using a battery and setting up the condition I saw on the internet to validate if it was functional in the range, but quite simply for this situation, that was a waste of time. A simple continuity test on the new sensor on pins 1,3 told me it was connected with a resistance, and while the bench test told me it worked perfectly. The old one's basic open circuit made all that unnecessary...my old one was straight broken, not 'inaccurate'.

So, a good result and I want to thank all of you for your pointers and and assistance in helping me figure this out.

- yzandrl