[boostd92]

I'll try to cut this down to the meat. I have had a cold start misfire (that goes away after a block or two of driving) forever. The car always stores both banks lean air mixture faults, and then it will store cylinder #6 misfire (which I can feel as a singular misfire occasionally on cold starts), and if codes left long enough uncleared (a week or so), all cylinder misfire codes will join #6 and the lean mixture codes.

Here is a link to a 5 minute cold start datazap log - during this log, cyl 6 was misifiring constantly (worse than usual):

datazap.me | apcrews | E46 330xi cold start misfiring

datazap.me

For reference, after clearing the codes (so the DME will turn cyl 6 injector back on) and driving for 30 minutes, here is a few seconds of hot idle:

datazap.me | apcrews | E46 short hot idle

datazap.me

What I've done so far:
Replaced MAF with OEM BMW new MAF
Replaced DISA with OEM BMW new DISA
Replaced vacuum cap on back of cylinder head
Removed CCV (did the DIY bypass with a PCV valve)
Smoke tested the car 3 times, no leaks - paid a BMW master tech to do it again, absolutely NO INTAKE LEAKS.
Sent all 6 fuel injectors out for testing and cleaning - returned all now flow perfect, new o-rings, and none of them leak

When cylinder #6 code was stored and actively misfiring, I swapped the coil to #1 and the plug to #3 and retested - cylinder #6 misfire returned (this means the plug, coil, and fuel injector on #6 are ok).

What I haven't done:
Fuel pressure check (don't have right adapter)
Compression test for cyl #6.

Notes on these - the car does not lose fuel pressure overnight. It always fires up in less than a second - even after sitting for 3 days in 21F weather.
I don't believe the compression is bad on cyl 6 because it drives perfect after it's gone a block or two. When it's chugging hard and SES light on for cyl 6, I can clear the code while the engine is running and it smoothes out perfectly (an indication of the DME turning the fuel injector for 6 off, then back on when the code is cleared).

What I don't know - does the car reference the 02 sensors in open loop - and I'm not sure how to read the data from the streams above to see if the O2 sensors are reading correctly.

My suspicion is that I have a clogged fuel filter or bad fuel pressure regulator. Being a returnless system and cyl 6 being the last cylinder fed fuel, if pressure is good enough to start the car, but low at the back of the rail until I drive it for a while, that would explain things. But sometimes I'll get the lean air mixtures after the car has been driving for 20 minutes (if they were cleared previously) and the car is still running smoothly.

Any help would be appreciated. Thanks!

 

[BaliDawg]

Can you please share the raw CSV files somewhere? DropBox is free.
Graphs are ok, but need to see real data to spot issues.

Hot idle log observations:

• You are missing Long Term Fuel Trims for both banks.
• Fix the naming of the O2 sensor voltage PIDs, hard to tell which are pre-cat (sensor 1).
• Short Term Fuel Trims are quite high at +13% to +15%, indicating lean condition
• MAF value of 3.0 g/s is low for a 330 engine, should be at least 3.4.

Regarding the vacuum leaks, read our smoke test thread.
Test the power brake booster for vacuum leakage.
Check fuel pressure at rail, should be around 50 PSI.

 

[boostd92]

The .csv's are attached and thank you for making me do this, I realized I had the WRONG LINK for the hot idle above, I have since fixed the datazap link.

Here are the .csv's

5_min_coldstart.csv

drive.google.com

Hot_idle.csv

drive.google.com

 

[BaliDawg]

Updated my comments in post #2.
After playing with datazap a little, I like it!

 

[boostd92]

Can you please share the raw CSV files somewhere? DropBox is free.
Graphs are ok, but need to see real data to spot issues.

Hot idle log observations:

• You are missing Long Term Fuel Trims for both banks.
• Fix the naming of the O2 sensor voltage PIDs, hard to tell which are pre-cat (sensor 1).
• Short Term Fuel Trims are quite high at +13% to +15%, indicating lean condition
• MAF value of 3.0 g/s is low for a 330 engine, should be at least 3.4.

Regarding the vacuum leaks, read our smoke test thread.
Test the power brake booster for vacuum leakage.
Check fuel pressure at rail, should be around 50 PSI.

Not sure if you had a chance to check my link to the original thread - the car has no vacuum leaks and everything air related has been replaced (including brand new MAF).

I haven't ever smoked the booster - just disconnect it from the intake a smoke towards the booster?
I'll hopefully have fuel pressure tested today.

I took the long term fuel trims out of the parameters logged because I thought the 5 minute cold start with 22 parameters wouldn't give quick enough samples. I'll re-do it.

I did not name the O2's, just selected them from OBD Fusions PID choices, not sure how to rename them, or if they are accurate. Do the readings look normal for the sensors?

 

[BaliDawg]

Not sure if you had a chance to check my link to the original thread - the car has no vacuum leaks and everything air related has been replaced (including brand new MAF).

I haven't ever smoked the booster - just disconnect it from the intake a smoke towards the booster?
I'll hopefully have fuel pressure tested today.

I took the long term fuel trims out of the parameters logged because I thought the 5 minute cold start with 22 parameters wouldn't give quick enough samples. I'll re-do it.

I did not name the O2's, just selected them from OBD Fusions PID choices, not sure how to rename them, or if they are accurate. Do the readings look normal for the sensors?

What original thread? Maybe add a link to that in post #1?
For completeness, add the year and model E46, and any other mods to it.

Cannot smoke test the brake booster because of one-way valve. But there are indirect ways of checking if it is leaking vacuum. Do that.

Add back in LTFTs, as those are super important. Remove these non-essential PIDs:

• post-cat O2 voltage (sensor 2)
• ignition timing advance
• intake air temperature
• commanded secondary air status

You can edit the CSV file to fix the O2 sensor voltage heading names.

 

[boostd92]

Appreciate the response. I'll amend PID's logged and run another cold start log when I leave work, and a hot idle lot when I get home and upload it. Thank you in advance for your time.

Sorry about the "original thread" reference - I have 2 threads going and this IS the original thread.

The other thread was specifically about the "Fuel System 1 Status" PID. In the cold start log, it goes to "8" which I found was "open loop due to fuel system malfunction" or something like that.. I was looking for more info on that - it appears that happens on my car in the cold start log at 107 seconds.

 

[boostd92]

One more thing - I have the sample rate set to 1s. The OBD Fusion logging thread said 1000ms (which is 1s), but I noticed when I graph the logs it's very blocky. I'm using a VEEPAK OBDCheck BLE to log and I'm going to change it to 500ms to get better resolution unless you guys who are more experienced know it won't...

is the sample rate tied to the speed of the obd tool being used or the OBD Fusion software?

 

[BaliDawg]

Higher resolution sampling isn’t needed and just doubles the data size, which is a pain. Honestly, a 1-second sample time is fine.

The fuel system statuses go to ERROR (8) because the pre-cat O2 sensor voltages dive down (lean condition) and don’t start switching.

 

[boostd92]

Ooops, I logged before reading your reply above. The following logs are 500ms, hope that's not too much of a pain - file size is still very small.

I changed parameters to the ones you suggested. This time on cold start, I felt the single cylinder misfire again - no SES light on but when I scanned it it had P1353 which is Cylinder 6 misfire with fuel cutoff again. I cleared the code and the car started to instantly run smoother (as it should when the fuel injector is turned back on. Log here:

Cold-Start_500ms.csv

drive.google.com

The drove absolutely perfect after it was on the road for a minute or two. Never had an SES the whole way home but it did have both banks lean air mixture codes pending when I checked. Here is the highway cruise log, I was going about 68mph:

HighwayCruising68mph.csv

drive.google.com

Then came home and idled in the driveway for a minute or two. Idle was exceptionally smooth, still no SES light yet, but lean mixture codes pending:

HotIdle500ms.csv

drive.google.com

I was not able to check fuel pressure after work - we closed early due to NYE and they kicked everyone out (I work at a dealership). I also would like to check compression sooner or later.

If anything jumps out at you, please let me konw.

Thanks!!!!

 

[BaliDawg]

The cold start log exposes a serious problem, that the pre-cat O2 sensors are not responding within 1:30, so fuel management goes to ERROR state. They should start switching by 30 seconds.

The hot idle log shows the MAF value is low (might be under-reporting) and all four fuel trims are too high.

 

[BaliDawg]

is the sample rate tied to the speed of the obd tool being used or the OBD Fusion software?

It’s how often the OBD app/tool requests the data, on a timer in this case.

 

[boostd92]

The cold start log exposes a serious problem, that the pre-cat O2 sensors are not responding within 1:30, so fuel management goes to ERROR state. They should start switching by 30 seconds.

The hot idle log shows the MAF value is low (might be under-reporting) and all four fuel trims are too high.

Do you think this is because of old O2 sensors (they are original, car has 117k miles) or because the car is actively misfiring during cold start (cyl 6)? I'm up for replacing the O2's as preventative maintenance anyway, I guess.

The car does have a K&N drop in air filter if that matters, otherwise it's completely stock.

 

[Archbid]

O2 sensors are a maintenance item, and I would definitely replace. You have high trims at idle and cruise which indicates sensors and you have already replaced the MAF. BTW have @BaliDawg update your ecu version.

Booster can cause vacuum leaks, but a few percentage points at most, not double-digit - I wouldn’t put energy there.

I find it curious that you have high trims at both, but very high trims at idle which does point to a vacuum leak. Take a look at the ltft guide in my signature.

 

[BaliDawg]

This statement indicates the crankcase hoses have been altered, which is where vacuum leaks are likely.
Removed CCV (did the DIY bypass with a PCV valve)

 

[boostd92]

Yes I removed the CCV. I did it via the DIY on these forums, also did it to my wife’s X5 and she has been running the bypass for five years with no problems, so I know what I’m doing there. There are no leaks in the bypass system. The symptoms were there before and after the CCV delete.

 

[BaliDawg]

Yes I removed the CCV. I did it via the DIY on these forums, also did it to my wife’s X5 and she has been running the bypass for five years with no problems, so I know what I’m doing there. There are no leaks in the bypass system. The symptoms were there before and after the CCV delete.

Gotcha. I know you stated it was smoke tested thoroughly, but just looking for potential leak sources.
Any progress on a fuel pressure test?
I think the leading candidates so far are vacuum leak, low fuel pressure, under-reporting MAF sensor, exhaust leak.

 

[boostd92]

I’d like to think the MAF sensor is a non-issue since it was installed a few days ago purchased brand new from BMW dealer. The symptoms have remained the same through my old MAF sensor, a cheap one from Amazon, and now this brand new one from BMW. I will hopefully be able to do a fuel pressure test tomorrow after work. I will also re-smoke test the car just for grins.

 

[boostd92]

Another thought after looking at the last 500ms cold start log.. The car was actively chugging because cyl6 was misfiring. At 97 seconds, the car goes to Fuel System Status 8 - malfunction - could this be when the car shut the fuel injector off for misfiring? Just trying to get an idea before I go buy two new upstream O2 sensors.

You can also see where I cleared the faults (while it was running) at 268 seconds. The idle jumps up, fuel system status goes to 2, and you can see the RPM smooth out as it's no longer misfiring.

I'm going to buy the new O2's anyway, they are due to be changed as regular maintenance, just curious.

 

[boostd92]

Update:

Checked fuel pressure today - solid 50psi - was really hoping this was all it was but it's perfect.

I re-smoke tested the car again and found a leak! It appears that I have air leaking from either a crack in my valve cover around cylinder 3, or the circular spark plug well gasket is leaking. I'm going to change the VCG (even though it's dry as a bone as far as oil goes) and check for cracks. Here is a video of the test:

I drove home and was watching the short term fuel trims. They didn't get very high until I was idling in my driveway (after 30 minute drive). Was trying to figure a way to still get it through emissions -when I revved it, the STFTs came down, but they come right back. Was considering monitoring STFTs and revving when I have to, to prevent a code from storing so I can just get it through emissions so I can get a new plate sticker which just expired. Here's a log of that:

HotIdlePostFPCheck.csv

drive.google.com

Is it strange that the O2 values are not cycling? Is this normal for when the STFT's are at 27% or should the O2's still be cycling?

That, and the fact that I don't think a slight air leak around cyl 3 is causing my cylinder 6 misfire and injector cutoff.

Monday after work I'll finally do a compression and leakage test to get that off the table.