Hi all,

My 2009 R56 base cooper reported 2773 (DME: Combustion misfires, cylinder 1) and 2783 (DME: Combustion misfires several cylinders) recently. I've spent time perusing others discussing this in the forums and elsewhere and the facts don't quite line up (of course) so i thought I'd post for some thoughts.

Some facts:
- I did a near complete in-engine-compartment rebuild of the engine over the summer. Head was completely refurbished so there is almost no carbon build up at this point
-- I did not replace injectors
-- I did not replace the rings (and no.. I'm not sure skipping that was a good idea either. I accept your somber disappointment)
-- Timing chain is brand new
-- VANOS solenoids are brand new
-- Plugs are brand new
- The car ran smooth after rebuild for a solid month before these codes appeared.
- I do not have engine knocking issues or any HPFP codes
- I do not have problem with cold-engine start, however I live in Southern California and it doesn't get that cold here in general, so maybe not a fair trial

My ideas:
- Pressure test the fuel rail
- Move coils around, clear codes and retry.
- Compression check cylinders despite how sad the results may make me feel
- Check JBE for water intrusion/corrosion

Any other "relatively" easy checks I should try to gather data and diagnose?

Thanks for your time and energy!

Mike

 

Define "complete refurbished cylinder head."

Were the valves replaced or ground?
Did the machine shop grind the valve seats or replace them?

If they performed the typical 3-angle grind and ground the valve seats, i can almost guarantee you the valve seats are loose now. The valve seats are only .022/.023" deep. Grinding them forces the stone into contact with the seat that is pressed into the head. The head is locked in the machine, but the seat can twist within the head, altering the press fit and causing it to loosen.

When the seat is loose, it drops down; at idle, the valve lift is only 180 microns,, and only the forward intake valve opens at idle. So if the seat drops, it is as if the valve doesn't open at all, so the fuel and air can't enter the cylinder....misfire.

Tell me, do the misfires happen when the engine is hot or cold? Does it misfire under full throttle, at idle, or under all power settings?

Does the misfire go away when you disconnect the inlet camshaft position sensor?

 

The head was removed, cleaned, refaced and pressure tested. The valve seats were probably ground, but I'll have to ask. The only replacement parts I provided were new valve seals.

My son tells me the misfire happened when the engine was hot, but not I was not there. But it might have been at idle when hot and waiting in traffic. I suppose experiments are in order.

So you are suggesting the following test:

• Clear the codes, leave the coils in place
• Start engine and let warm up and leave it idling for.. some amount of time to see if it misfires.
  • This would tell me that the problem is with the valve seat?
• If I don't get a code, take it for drive and give it some power variation?
• Do I disconnect the inlet camshaft position sensor before this test? Or is that a separate test? What would no code with the sensor disconnected tell me?

Also, I just noticed there is a 2845 DME: VANOS, exhaust actuator movement code as well. The VANOS solenoids are brand new. Does that error code help narrow things down?
I really appreciate the help.

Mike

 

Do you have a scan tool? If you do, bring up live data (misfire data).

With the car cold; start the engine. Does it misfire when it's cold? Make a note:

Next, drive the car until it reaches operating temperature; while driving, monitor your misfire counters (if possible). If you don't have a scan tool, can you detect any misfires? Make a note:

Lastly, park the car, and monitor your misfires counters again. If you don't have a scan tool, you'll notice the engine running roughly. If it's running roughly, disconnect the intake camshaft sensor connector from the sensor. Does the idle smooth out? Make a note: Let us know.

As for your 2845 code: Here's the most probably cause from Mitchell1.


However, you'll need to troubleshoot to isolate the cause of your fault.

 

Awesome stuff.
Yes, my scanner can give me live misfires. I just did it, only at idle, and it was all in cylinder one. I was not able to do a road test right then.
The odd part is the engine isn't running noticeably rough. It sounds fine and runs steady, even when the idle drops as it warms up.
Since last time I wrote I also did a compression test and all 4 cylinders are running around 190 psi.
Thank you for the awesome troubleshooting chart for the 2845; I will engage on that later.
I was thinking of swapping the coils and see if the misfire moved or not.

I really appreciate your help and for introducing me to the word "Justa".

Mike

 

Ok Mike, you've gotta help me out here.

Did it misfire when it was cold?

The problem with the compression test is it won't find a loose valve seat.

The compression on my cylinder 2 was within specifications, but when I pulled the head off, look where the valve seat was.



The reason you're getting the misfire in this situation isn't from the lack of compression, it's the fact that with the seat dropped, the fuel/air mixture can't enter the cylinder.

Suggested reading:

https://www.mini2.com/threads/prince...e-imho.362271/

You can swap the #1 and #2 coils, but I can already tell you, the misfire is going to stay with cylinder 1.

I my case, my cylinder 2 misfire became my Cylinders 2, 3, and 4 misfires, after the seats dropped in the other cylinders.

Here's a pic of cyl 3.



And cyl 4.

 

Got it. Those do look terrible.
Sorry I wasn't clear. The engine was cold when it was started and the misfires occurred at idle.
Thank you for your time and I'll update the thread as I pursue resolution. I will ABSOLUTELY keep your diagnosis in mind but I'm unfortunately recovering from a busted leg at the moment and will need to wait to pull the head off again.

Mike

 

@mkov608
So here's the status
Replaced the fuel injectors and coils since I didn't see a downside with the mileage on the car.

On cold start, Cylinder 1 still misfires (as predicted). The rate is enough to trigger CEL but does not make the engine feel or sound rough. After the initial idle the idle speed drops a couple hundred RPM (still over 1K) and the misfires stop completely. The graph completely flatlines.

I start driving. Initially, on low power starts (like calmly pulling away from a stop sign in a residential neighborhood) I will get a distinct lag and surge that correlates with misfires on cylinder 1 (about 40 on the reader). Once I get past the surge, no more misfires are detected. Higher power starts (like trying to take a right onto a busy 50 mph street from a stop light) do not trigger misfires. Accelerating at higher speeds does not trigger misfires.

Once the car is warm, no misfires are detected and no surges happen despite mimicking the starts that caused it before. Upon returning home, I parked the car, stopped the engine, and immediately restarted. No misfires detected. It clearly does not seem to happen when the engine is warm.

At this point I think I'm ready to buy a borescope to see if the valve seats are off. The pics you show seem like I should be able to see it with a borescope without taking off the head. But they also seem like they should cause problems under almost all temperatures given the interference with the valve.

I ran the VANOS adaptations when I put the car together, but are there other ones I may have missed? This was my first mini engine rebuild.

Thanks again.

 

Smoke check the intake manifold; many times a cold engine misfire is caused by a leaking intake manifold gasket. If you have an intake manifold gasket leak, that cylinder with the leak will pull in un-metered air, and that will make the fuel air mixture lean enough to cause a misfire. As the engine warms up; things expand and all of the sudden you have proper sealing again, so the misfire goes away.

On some engines you can spray soapy water (lots of bubbles) around the intake manifold and watch/listen as the soapy water will get sucked up. However, MINIs don't really have any intake manifold vacuum at idle due to the variable valve lift, so the soapy bubbles trick doesn't work very well.

Another option would be to pull up your short term fuel trim (STFT) on a scan tool, and spray some carb cleaner around the intake manifold gasket for the suspect cylinder. If the gasket is leaking you'll see your STFT go negative. Again, make sure you do this test when the engine is cold.

 

I will give this a shot.

Can it be just a wonky sensor? Is there any value in attempting to replace the camshaft and crankshaft sensors even though I don't have a code for it? I only ask because I watched a video of a guy dealing with a mini with a very rough idle and eventually he realized the coolant temperatures just were not right so he replaced that sensor and the car ran smooth. There was no light for that sensor issue. I replaced my thermostat during the rebuild so I'm fairly certain that is not my problem, but I wonder if the concept of a wonky sensor with no code could be something to consider.

 

For a misfire? I doubt it. There is no value in replacing perfectly working camshaft and crankshaft sensors. You can have a diagnostic shop connect to the two camshaft and the crankshaft sensors and see if they are dropping out, but I doubt a randomly dropping out sensor would cause a misfire in a single cylinder.