First, 2006 R52 "Justa" 83K on Odometer, 120k on engine (2005) and 85K on R53 6sp trans (2006). Complete engine and transmission installed last fall. Only 900ish miles since. Nothing was reused from the original transmission, and only the passenger engine mount bracket was reused from the engine. Original engine wire harness was not used.

Next, I've read through the starts then stalls thread that is stickiyed. It wasn't exactly helpful

Symptoms:
Car runs great when cold and for a short while after it warms up.
After the car is warm, it feels as if it is down on power. Most noticeable from a dead stop. No codes stored.
Intermittent hard start when cold. Hot or cold ambient.
Hard start when warm, when car starts it revs to 1k, then stalls/dies. Noticed when ambient >about 60. Not experienced when ambient temperatures are in the 30's. No codes stored.
About 5-10 mins after the first restart attempt. the car will start and idle. 5-10 mins key on or key off, I've seen no correlation.
Holding the gas pedal to the floor will allow the car to stat on an extended crank (15-20 sec) car will idle and can then be driven and will not stall. Repeatable. Since I have discovered this, it has worked every time I have tried.
Intermittently, P0300 when cruising at temp. Car runs rough, if it is babied it will run rough with a miss and stall/die. If I can throttle out and accelerate (freeway on ramp), it clears up and stores the P0300 code.
Have seen one TB code, forget which. unable to repeat and have only seen once.
No disappearing oil or coolant.
Fuel MPG went from 28 to 23 on a tank with the same driving, and same route to and from work. Almost even split hwy and city with stop and go mixed in at a toll bridge. Not sure this is related, but it is the main reason that I record MPG in an app.

Things done/troubleshooting.
New NGK plugs and wires when engine was installed.
Fresh oil change and filter with engine install.
Various gaskets and seals replaced with engine install.
New Radiator. Original was damaged by engine shrapnel.
All coolant system hoses and plastic fittings replaced because it is still a BMW..
Car dosed liberally with injector cleaner.
Premium fuel from various sources no ethanol.
New battery after the first couple of times this happened.
All engine and body grounds verified re-installed good electrical contact from various points to battery neg terminal. (PITA to verify)
Long and short term fuel trims were very near 0. As in 0.XX. ECU reset after engine installation and after checking trims.
Cold compression is in the 140's and all cylinders are within 10%.
Cold fuel pressure ~57 PSIG.
Warm Fuel pressure 55 PSIG but questionable due to auto zone fuel pressure destroying schraeder core. Probably an operator error, but that's never happened to me before.
New fuel filter.
New OEM coil. Some corrosion on original and cracked @ #4. Haven't driven much since I changed the coil, but maybe that was the P0300 cause.
Verified gaps on plugs. Plugs have consistent color and appearance.
I believe that I verified ECU coolant temp with gauge. Results were consistent with each other and with ambient if cold or 190-200f when warm.


I need to buy a fuel pressure gauge and recheck warm fuel pressure. If that is not it, I am at a loss as to where I should look next. Is there something that I have overlooked? I use Autoenginuity for diagnostic software, but I do not have their BMW add-on.

 

Rechecked fuel pressure. <1/8th tank of fuel. Fuel pressure pec is 43 +/-3 psig.
Engine off and cold, gage vented, cycled the key to key to let the fuel system primed. Reading 55 psig.
Engine idling, cold. fuel pressure 43 psig.
After about 45 mins of idling, engine warm, fuel pressure 40 psig.
Rev the engine a few times, fuel pressure 38-39 psig.
Turn the key off, engine starts and dies, fuel pressure 50 psig.

I cleaned the throttle body and IAT sensor from the original engine and reinstalled them. Still hard to start and stalls on restart when warm, but I haven't gotten CELs or misfires in a few weeks.

 

Been some time since I updated this.

I've had the car in a specialist shop, where it was identified that I had sourced the wrong NGK plugs from the local parts house. Also, I've done a bit of reading and replaced a few more parts and began putting a ton of miles on this car (6k since Nov 1). The changes made in the last post, actually did not correct the misfires.

The cold hard start / hot start-stall seem to have been cured by a fuel pressure regulator. A crank position sensor and cam position sensor seemed to help, but were probably not necessary. I also saw about a 1.5 MPG average improvement per tank.

The misfire began to get more common, especially in the ~3200 RPM range but not limited to those RPMs. I recently moved and began adding at least 100 miles per day while commuting, and it got so that it was not uncommon to get several misfire incidents per trip. Cycle the key and it would clear up and run great, for a while. Occasionally the CEL would go out. Always it was at least a P0300, if I was less than, I think, a 1/4 tank, I would get a P0314. Seeing as how the car sat for a few years, I replaced the fuel pump shortly after replacing the fuel filter.

Finally last week, I experienced an unrelated failure. The muffler mount broke loose from the muffler. I believe that it was a failed weld. The rubber isolators were still in serviceable shape. I didn't give it much thought, but got it properly sorted before it became necessary to replace the flex joint or other parts as well. This morning I realized that i've put almost 1k miles on the car with out having a misfire incident, and the muffler is the only thing that I've done immediately prior to this. The last thing I did before the muffler mount broke was only replaced the serpentine belt and tensioner pulley.

Could a failing muffler mount cause enough noise or interference to cause the ECU to think misfiring cylinder? Or am I fooling myself?

 

Misfires can be flagged because the engine controller detects one or more cylinders are not contributing the expected acceleration to the flywheel during the cylinder's power stroke. If too many of these occur in succession a misfire code is logged and the CEL is turned on. If the misfire is believed to be a rich misfire -- not sure how the engine controller detects this -- the CEL is flashed to indicate risk to the converter.

A P0300 just indicates more than one cylinder is misfiring. So I would think there should be P0301 to P0304 codes (maybe not all of them) and the freeze frame data might identify a specific cylinder and its misfire code.

Hard to imagine a loose exhaust pipe would affect cylinder performance.

Now it might generate a vibration that that fools the knock sensor into signaling the engine is experiencing detonation and thus causes the engine controller to back off on the timing. If the timing is backed off too much this can trigger a misfire.

Guess you'll just have to see if the misfire error code stays away.

 

I almost wonder if there is an injector leaking down and draining the fuel pressure, necessitating repressurization of the fuel system. Holding the pedal to the floor when cranking triggers "clear flood mode" in most cars, and shuts down the injectors or fuel pump. the loss of fuel economy also makes me think there is an injector that just isn't closing. I would pull the plugs and see what they look like. They should be light brown to gray. If you find one that is black, that cylinder is running rich and likely the source of the problem.

 

Have you checked the fuel pressure regulator for leaks? If it leaks it will fill the vac line and that fuel will get sucked into the intake manifold. Been down this road twice now..

 

Thanks for the suggestions, I have already replaced the Fuel pressure regulator and that seems to have sorted the hard starts (cold and warm).

The misfire stayed away only long enough for me to think it was gone. Came back yesterday. with a p0300 and a p0304. I have a start I guess. It came in several more times with only a p0300. Have to work on the wifey's car this weekend, but maybe I'll have time to dig into the mini some more.

Would INPAS or ncexpert help me gain a better insight into this? The standard freeze frame data hasn't told us much.