Cold Start Gremlin!

[ac427]

My cold start problem is gone after fuel pump and cleaning injectors, but my LTFT after the car is warmed up is very high / lean. (+15 to +20%)

smoke test didn't show any leakage, this is really puzzling

I didn't have time to replace all the 3 lines connecting to the airbox, but very likely they are having tiny holes or broken hose since they have 190k km on them.

Thanks. Which 3 lines are you thinking of?

I know of the following pipes:

• Air shut off valve from the top of the airbox.
• OIl separator pipe from the top of the engine.
• ICV feed from the bottom of the airbox.
• Crankcase vent pipe to the bottom of the airbox.

Did you mean the brake servo/booster lines and air shutoff valve setup?

Are you measuring your LTFT through INPA ?

[y2k_o__o]

I replaced the oil/coolant sensor (the brass screw in one) becuase i had one and i also had a slight leak from the heater matrix return hose.


Is your car still fixed or have you had some cold start issues?

I'm not sure how i would seal my CSL airbox inlet.

How did you go about using the smoke tester?

My cold start problem is gone after fuel pump and cleaning injectors, but my LTFT after the car is warmed up is very high / lean. (+15 to +20%)

smoke test didn't show any leakage, this is really puzzling

I didn't have time to replace all the 3 lines connecting to the airbox, but very likely they are having tiny holes or broken hose since they have 190k km on them.

[ac427]

After reading through the first few posts in regards to cold start I wonder if having a cold start "delete" isn't the best idea.

I have my SAP removed and cold start deleted, I always figured that the cold start raised revs only to warm up the cats, but it would seem there is more to it than that.

Also, I don't know what all is "deleted" when you check the box for cold start delete on the binary modification tool program.

I understand your point but i want to know what is actually wrong rather than mask the issue.

My SAP has been deleted for years but i have not removed the actual hardware.

[Cubieman]

After reading through the first few posts in regards to cold start I wonder if having a cold start "delete" isn't the best idea.

I have my SAP removed and cold start deleted, I always figured that the cold start raised revs only to warm up the cats, but it would seem there is more to it than that.

Also, I don't know what all is "deleted" when you check the box for cold start delete on the binary modification tool program.

[ac427]

are your LTFT value from both banks normal when the car is warmed up?

I replaced the oil/coolant sensor (the brass screw in one) becuase i had one and i also had a slight leak from the heater matrix return hose.


Is your car still fixed or have you had some cold start issues?

I'm not sure how i would seal my CSL airbox inlet.

How did you go about using the smoke tester?

[ac427]

are your LTFT value from both banks normal when the car is warmed up?

The engine temp was around 54°C. I can warm the engine up more and get some Analog readings when it is hotter. I thought at this temp it would be in closed loop mode, so any sensor issue would show up. Do you think a higher temperature is required?

[y2k_o__o]

Thanks for the reply. I wish mine was just a water temp sensor as in the the posts above. I just checked INPA after leaving the car on the drive without starting it for 2 days. The both the coolant (brass one) and radiator return (Plastic one) were the same value. I watched the temperature rise on the the block mounted sensor the outlet lt stayed the same.

Like you i swapped or replaced quite a few parts:

Coil Packs
Spark Plugs
Injectors - Made a load of diffrence to performance but the start up idle was still rough.
Pre cat O2 sensors

I even swapped the ECU to no avail.

I checked the fuel tank split by looking at dash test. 007330 which means most of the fuel is in the RH tank.

If i stab at the throttle whilst in neutral, the acceleration response form the engine is good. As it returns to idle, there seems to be a very slight bit of hunting to acheive idle. Perhaps the ICV is the issue or at least contributing to it.

are your LTFT value from both banks normal when the car is warmed up?

[sapote]

Thanks for the reply. I wish mine was just a water temp sensor as in the the posts above. I just checked INPA after leaving the car on the drive without starting it for 2 days. The both the coolant (brass one) and radiator return (Plastic one) were the same value. I watched the temperature rise on the the block mounted sensor the outlet lt stayed the same.

The outlet temp won't change until coolant is hot enough to open the Tstat and more delay time for the radiator to heat up.

[ac427]

I think I did, but I haven't driven long enough to said fully resolved as currently the interior is gutted for refreshing and cleaning

I smoke tested it, turn out the intake elbow is causing a very very tiny leak (high positive LTFT +14%, but not great enough to trigger CEL). I thought I clamped it at the right location, but no... I ended up rotating the clamp and wiggle the elbow a little and tighten it. It seems to bring the LTFT down a bit which is a good sign. However, prior to re-tightening the intake eblow. I replaced pre-cat O2 sensors, fuel pumps (Hella) and injectors that fixed the CEL or P030x misfire at start up, even cold start, so far. I threw alot of money to catch up the maintenance as it was neglected by the previous owner.

Here's my journal of my E46 m3, I tackled all hectic oil leak and CEL one by one since I bought the car 3 years ago.
https://www.e46fanatics.com/threads/my-02-m3-steel-grey-oem-build-journal-finally-back-to-the-road-after-1-5-year-of-diy-fix.1300910/page-2?post_id=19241887&nested_view=1&sortby=oldest#pos t-19241887

Thanks for the reply. I wish mine was just a water temp sensor as in the the posts above. I just checked INPA after leaving the car on the drive without starting it for 2 days. The coolant (brass one) and radiator return (Plastic one) were the same value. I watched the temperature rise on the the block mounted sensor the outlet it stayed the same.

Like you i swapped or replaced quite a few parts:

Coil Packs
Spark Plugs
Injectors - Made a load of difference to performance but the start up idle was still rough.
Pre cat O2 sensors

I even swapped the ECU to no avail.

I checked the fuel tank split by looking at dash test. 007330 which means most of the fuel is in the RH tank.

If i stab at the throttle whilst in neutral, the acceleration response form the engine is good. As it returns to idle, there seems to be a very slight bit of hunting to acheive idle. Perhaps the ICV is the issue or at least contributing to it.

[y2k_o__o]

Did you get to the bottom of the issue?

I think I did, but I haven't driven long enough to said fully resolved as currently the interior is gutted for refreshing and cleaning

I smoke tested it, turn out the intake elbow is causing a very very tiny leak (high positive LTFT +14%, but not great enough to trigger CEL). I thought I clamped it at the right location, but no... I ended up rotating the clamp and wiggle the elbow a little and tighten it. It seems to bring the LTFT down a bit which is a good sign. However, prior to re-tightening the intake eblow. I replaced pre-cat O2 sensors, fuel pumps (Hella) and injectors that fixed the CEL or P030x misfire at start up, even cold start, so far. I threw alot of money to catch up the maintenance as it was neglected by the previous owner.

Here's my journal of my E46 m3, I tackled all hectic oil leak and CEL one by one since I bought the car 3 years ago.
https://www.e46fanatics.com/threads/my-02-m3-steel-grey-oem-build-journal-finally-back-to-the-road-after-1-5-year-of-diy-fix.1300910/page-2?post_id=19241887&nested_view=1&sortby=oldest#pos t-19241887

[ac427]

sorry to bring this back, I have the exact same symptomes as yours and I am still chasing what it is

Did you get to the bottom of the issue?

[y2k_o__o]

Still chasing this down. To me, it all adds up to a vacuum leak but I can't find anything with starter fluid or a smoke test.

Getting an EGT code and injector 4 code. But I believe that injector code is a symptom not a root cause.

Maf has been replaced.

Car runs brilliant once the o2s warm up, and if I clear the light, it won't come back on until the next day's cold start

Sent from my moto e5 play using Tapatalk

sorry to bring this back, I have the exact same symptomes as yours and I am still chasing what it is

[JeremyJames]

Time for the peanut gallery 😁 to chime in - cars in general (gasoline ones that is...) when they are stone cold, they run in an open loop, meaning without the use of O2 sensors to govern the air-fuel mixture. The ECU relies on a temperature sensor to default to closed or open loop. Old 911s use a sensor in the block, as with the family of VW/Audi/Porsche products that use early Bosch Jetronic CIS systems, which include K, KE, etc. Our modern M’s use the water temp sensor, or this below:

Take this as information rather than a diagnosis or an instruction/suggestion to spend $50 bucks on a sensor. Like others stated, do some testing to determine your faulty part(s), then proceed to buy. Good luck squashing your gremlin.

PS, Every bit as important as the sensor is, it is only good enough based on the connector’s or wiring’s condition. On a car that’s now 15 to 20 years old, the connector might have oxidized a bit which degrades the signal, or if the wiring is moving around, this may lead to slightly damaged conductors that lead to an intermittent connection. Moisture, being that this sensor is very near a water/coolant source, that wicks and works its way under the insulative casing will have a corrosive effect and degrade the conductor’s ability to carry a signal.

Monday fun-day where I chime in again, from ECS’s site, temp sensors can be had for the cost of an In N Out double-double combo meal Including animal fries w/extra grilled onions. That seems fairly inexpensive. The original image got deleted somehow, so I am just re-posting the same info.

Let’s talk about cold start some more as I think it might help the discussion. First, some definition of terms: Open Loop engine operation means the engine is running without O2 sensors engaged, hence “open loop”; fueling is predetermined by a set parameter by the ECU. Closed Loop means the engine is warm and the ECU now engages the O2 sensors to govern the AF mixture.

Cold Start - When the engine is stone cold, the ECU knows this from the temp sensor, because it is cold. When one turns the key, a cold start (open loop) program is used, which separate and not related to warm engine operation (closed loop). This cold start consists of injecting more fuel in the bore, because when it’s cold, the very finely atomized fuel droplets are attracted to the cold cylinder wall where some fuel droplets condense while others remain suspended just long enough to touch-off from the spark plug, leaving sufficient fuel for cold start combustion. Next, more fuel requires more air, which means the ISV opens up resulting in increased idle. Engine timing as part of the cold start is retarded to compensate for the condition of cold start - remember this, as this is key.

With that background, let’s run through a scenario of insufficient fuel at cold start. Like that stated, very finely atomized fuel will condense on a cold cylinder wall, hence the term “washing of the bore,” which means condensed fuel washes off the thin oil film on the bore, which leads to ring wear that everyone is familiar with. Anyways, insufficient fueling when it’s cold will lead to a hard start because much of that fuel is condensing on the bore wall, and a rich mixture is mandatory.

Next for this discussion, let’s say the injectors are still providing insufficient fueling for cold start - at some point when the engine is cranked continuously and long enough, enough fuel will be available for cold start but it’s a rough idle. What causes the rough idle, a combo of a lean mixture and retarded engine timing. Next, you’re late to work as usual so it’s damn the torpedoes and full speed ahead. You hit the gas, the mixture is lean, engine timing is advanced on an engine with retarded timing and bingo, the engine stumbles hard, and the car bucks like your girlfriend when... - well, you know.

Fueling is key to a cold start. Without enough fuel, some of you may like the feeling of your M3 bucking like your girlfriend, but it’s bad for the sled. Tackle your fueling problem, the hard cold starting, stumbling in cold operation and a bucking ride’m cowboy M3 will likely be resolved.

Best troubleshooting advice I have seen. I learned so much, more as a bi-catch researching something completely off of this topic. None the less I enjoyed the read. Thank you for the education!

[High.miles.big.smiles]

Still chasing this down. To me, it all adds up to a vacuum leak but I can't find anything with starter fluid or a smoke test.

Getting an EGT code and injector 4 code. But I believe that injector code is a symptom not a root cause.

Maf has been replaced.

Car runs brilliant once the o2s warm up, and if I clear the light, it won't come back on until the next day's cold start

Sent from my moto e5 play using Tapatalk

[Speed Monkey]

OP, perhaps there might be some good info in these videos that can shed some light on your problem... Even though Nathan’s sled is not an E46M, much of the info crosses over to the M.