Do you see smoke from the tailpipe when this happens or when it finally catches?
Yes. That is the common problem I was thinking about. Another common cause of a long crank is the camshaft sensors and/or the crankshaft sensor. If the camshaft sensors do not see the correct signal it will crank longer and then use the crank sensor as a reference for starting. Do you have a scanner that can read the adaptation values -multiplicative and addaptive of the mass air flow sensor?
Hmm. You should try to find the location of the water leak and repair it. Water can damage a wiring harness and can very expensive to repair the damage. I recommend using a friend with a light and spraying down the area with a hose (friend in trunk with flashlight....I guess it should be a good friend) and look for how the water is entering. Is it the trunk seal or throught the taillamp itslef. Once determined you should do this first. You should then repair as much of the damaged harness as possible. Including replacing the pins in the relay holder(harness side) and cut back as much corrosion as possible. It is always best to start with the obvious. Perhaps once the water intrusion is corrected it will stop this problem.
As for the electric t-stat. I have seen some strange things with this problem as well. You should have the t-stat replaced and check the electrical plug to the t-stat as it often times melts from the faulted t-stat.....I have seen a melted t-stat connector cause no start situations and intermittent no start situations and it is a direct input to the engine controller so if it pops the fuse in the fuse carrier in the E-box the engine might not start until this is traced down. One final thing I have seen....and it was only once...the electric t-stat melted at the connector and coolant was being sent up through the wiring harness and all the way to the DME (engine control module) and caused the module to short out from coolant in the DME. It took me a bit to trace down this problem. That is what I suggest at this point. I hope this leads to finding the source of your intermittent crank problem.
Hello again Crispin!
Wow! Sounds like things have gone bad. I will try to help you as best as I can.
The thing that stands out by your description is the oil in the spark plug holes. This is caused by a brittle valve cover gasket that then seeps into the spark plug holes over time. This causes the boots on the coils to deteriorate and sometimes fail. I would re-remove them and see if they are oily. I would them clean them very well with brake clean and spray some inside the boot hole and them blow out with compressed air to get any oil that might have gotten into the coil could be a contributing factor for the misfiring.
The next things is that you replaced the spark plugs. I imagine the oil that was trapped from the spark plug then goes into the cylinder. While this is not overly bad(it should burn off) it will produce smoke for longer than one might think(bluish smoke) as it gets into the honey comb of the cat and needs to burn off. Once in you can't get it out except by time and yes this is a damaging factor that could cause a cat to fail.(sorry I like to be completely honest and not sugar coat problems.) However, most of the time no failure results unless it is a lot of oil like an oil hydrolock problem. I would remove the plugs and clean them as well and then reinstall. Make sure everything is dry and clean before reinstalling both the plugs and the coils.
Also, did you use a BMW mass air flow sensor or and aftermarket mass air flow sensor? I have seen many problems with non-BMW mass air flow sensors. You could also try to install the original mass air flow sensor and see what happens. I hope this helps you Crispin!
I am glad I was right about oil on the boots. (been there :) )As for leapfrogging to
perfection....I wish problems like this were that easy my friend. I have been thinking
about your symptoms. A long crank can be caused by many things. A failing crank sensor, failing cam sensors, poor fuel pressure or volume.
I like to look at the big picture with problems like this. Because you stated you have
the problem switching form AC to no AC it sounds like it could be a failing Idle Air
Control valve where it is getting mechanically stuck and can not adjust itself fast
Also, you have done a lot of work relating to air and fuel. The dme has a stored value
called adaptaion values. They should be cleared to start the DME at zero so it can now
relearn how to control the air fuel mixture. I am not sure if that is something that can
be done with a normal scanner.
To summarize, your suggestions to date have been great, but the long cranking (9/10 times) remains. The FC A8 error I buzzed out to the DME and discovered a near-break in the wire near the idle control valve, fixed it and THAT problem went away. Also, I bought a good fuel gauge, put it in line with the rail, and saw correct pressure spike, excellent cranking pressure and 40 psi on idle (high frequency flutter with fuel injectors), with the pressure holding even overnight when engine off. Interestingly, the fuel pressure is always solid and in specs even during a long-cranking failure- no hint of any fuel issues. Note that I have a new fuel pump, filter, mass sensor, idle control valve, spark plugs a month ago, coils a year ago, O2 and cats a year or so ago. I now have NO error codes, no check engine light, and the car runs superbly EXCEPT for the starting and especially when at temp. Again, if we focus on electrical I would note that often during the long crank the car is trying to run fast, i.e. about 50% crank and 50% running. Flooring the pedal before, during or after makes no difference. I read on line about ignition switch problems, but am loathe to go change it without better data. What do you think? The long cranking is so bad that I have to re-charge the battery at night and occasionally will not get it in the 12 or so cranks before battery depletion. Well, now that the cabernet sauvignon is taking effect, let me share a spooky observation with you. Nearly every time that I think that I have found the problem (e.g. fuel relay, bad fuel hose, bad connectors) and make a change, the cranking is fine for about a week during which I feel like I have the upper hand, and then, about a week later, it reverts back to its old ways.... As the scientist who insisted that the thermos flask was the greatest invention because it worked hot OR cold, I would repeat his words: "But how does it know?"
Since fuel pressure is good.....it might be time for an exorcism (at least it would run good for about a week).
What your found online is correct. This vehicle does have a lot of problems with ignition switch failures....Usually the complaint is that the radio cuts out at times, or the dash lights flicer, or various electrical problems....there is an older tech that would start the vehicle and lower the visor on the driver side to turn on the upper lamp. When he did this the electrical problems would happen. This would indicate a bad ignition switch...it is something to try. I do not know if this could be an ignition switch problem. Even though it is a common failure, it does not fit the symptom quite right. I will try to review some wiring diagrams to see if it makes sense.
When the long cranking occurs....is it after a rain or a wet day? Can you explain the 50% crank and 50% running in more detail? Does it spin fast with no compression sound? Stumble while cranking like it is misfiring? Do you have another key?---If the EWS is not getting the proper signal from the key it will crank and not start, also it could be an intermittent EWS problem in the module. EWS=electric car immobilization and prevents the dme from releasing injector fuel and spark. This is only a guess. I am out of time...off to Beamer land...but I will do some more thinking on this...
By the way...Great find with that wire at the throttle!
All these car troubles are occurring in dry southern california weather and not rain dependent. I have NO troubles or events with any other electrical systems. Regarding EWS, the Bentley manual I bought tells me that for my particular model and date, the engine will not even crank without the key signal (I tried the second key anyway with no change). Note the car runs beautifully now and never stalls or shows any issues - only the start! 70% of the time the long cranking is normal sounding with compression, but 30% of the time it cranks really fast, with a weak compression feel, trying to run on gas (kind of hyperactively) with a slight ring sound of the flywheel. In a sense, it is tripping on its own timing it seems, being just ahead or behind of the sweet spot. Also, it never starts if I let it long crank- rather it starts immediately (WHEN it starts) when I turn the key. So I turn the key and kill it if it doesn't start right away. Typically it now takes 9 or so efforts to get started. Is it a crime to kill a car?
I think it would only be a crime if you use a shotgun on broad daylight.
I checked the wiring diagram, you are correct, the ews is also the starter relay so if it was failing the vehicle would not crank. Reading your description makes me think the 30% with a ring of the flywheel is like the starter is not engaging the flywheel correctly. This could be a failing starter or a worn out flywheel (from starter engagement) The slight ring sound from the flywheel makes me think this. You could remove the starter and take a look at the gear to see if it looks worn and also check the flywheel.
Crispin, has the crank sensor been replaced? The DME is looking for the crank sensor signal to know when to fire the injectors. If you are getting a bad signal it could take a few cranks before the DME figures out the position. This does not always set a check engine light or a fault. It is just a sensor so it is not too expensive. I hate to have you throw parts at it but it is a possibility.
If I had your vehicle at the dealer, we could check the signal or scope out the pattern while cranking to see if the signal is getting to the DME. The other item I would like to do is to reprogram your DME. This might be something worth trying as well. At the dealer we can see a lot of inputs using our equiptment. The good thing is that it is reproducable...not for you though...but it helps us to see the problem so we can check the inputs to see if something is not right. I hate to say it but it might be worth a run with the dealer to take a look at what is going on when the problem occurs.
I hope I am helping....let me know you thoughts and what you feel your next step should be. I have been enjoying trying to help you. If you would like to drive out to CT I will scan it for you :) I hope I can get you past this last hurdle (without the use of firearms!)
I did change the crank sensor, but no change to the difficult start, but now have some important new info for us to enjoy (!) Knowing now that the fuel side was fine, I removed coil #1 and attached it to a new spark plug that I carefully grounded with Cu wire to the engine, enclosed the end in a cone of black paper for easier viewing and had someone do the crank thing. When it starts right away (now 1 in 5 times and getting less often it seems), there is a strong spark, but when it does its long-crank-no-start thing there is NO SPARK every time (which explains why it doesn't start!) Now, FYI, all the coils are new upgrades from Bimmerparts about a year ago, so I do believe that the lack of spark reflects the lack of low voltage input... Also, remember, I still HAVE NO ERROR CODES or engine light, and the car runs beautifully once started. However, today for the first time, it died on the freeway, but I successfully restarted it in neutral by tapping the break etc. SO, knowing it cannot be the security system, I'm thinking that maybe it is the ignition switch, but could that cause death during a drive? What sayest thou, O master of the BMW?
Good to hear from you. We WILL fix this car!
I took out my magic eightball to try and help me fix your 7 series. After a good shake it asked if you have replaced the DME main relay? Smart little eightball.
As for the ignition switch....all I could say is maybe. I have seen a bad ignition switch keep a vehicle from shutting off, so why not causing a shut off problem. It is plausible that if the incorrect signal is being sent it would be like turning off the key. I think it would be a good idea to do, since it is a common problem. If you have been getting any strange electrical problems like intermittent radio shutting down this would be a good guess.
By the way, nice find with the spark!
What a trip it has been. It made me feel good that I was able to help you figure this out, though I must admit you have a pretty good grasp on things yourself! I hope she treats you well from here on out. I get a laugh over the progression from wanting to shoot the car to "this old loveable car". It has been enjoyable for me assisting you. Maybe I will get to help you again in the future. Take care and I wish you well!
Although I hate these long goodbyes, I wanted you to know that I had yesterday added a $30 bonus for all your hard work (one hundredth the cost of a BMW dealer visit) so that you can enjoy a decent Cabernet Sauvignon (but I'm guessing you're a single malt Scotch drinker). I did also have one last question please since I am trying to pass a smog test. After you and I produced a fully-working car, I reconnected that battery, but the codes that normally accompany this action =6C (battery disconnect) and CC (EWS rolling codes) persist (no other issues) for the last 5 days and I am worried that they will cause an automatic smog test failure. Should I (i) keep driving since they will one day soon disappear or (ii) use my code tool now to reset all codes, then redrive for a few days for the OBD to relearn?? Will patience be a virtue or a curse in this case?
Thank you Crispin for you bonus, it is much apreciated. I was in a hurry to read your recent post to see what new problem has popped up now that everything else has been resolved (knock on wood.) I think it would be safe to say that all you went through would drive anyone to drink (a single malt Scotch does sound good!).
I have hear California is one of the strickest states for the smog test. I think the battery resets might be a problem. Your best bet would be to ask the testers about that part. If it does not matter you should be set to proceed. If you use the scan tool, you have to wait for all of the test in the vehicle to run. This could take quite a few drive cycles. If you have time this might be the best bet. If you do clear the memory, you will want to drive the vehicle as much as possible until you go for the smog test. Possibly two or three tanks of fuel worth. Many tests have to run including, catylitic converter efficiancy, purge valve test, secondary air test, LDP(leak diagnosis test)....I actually had a z3 that took 3 months due to aging 02 sensors for it to pass.
You have done quite a bit of work so maybe a clean slate and some time driving would be a good idea, but before you take that step a quick call to the local smog tester might shed some light on the battery resets and ews. I think the only thing they would be concerned about would be the battery resets as the ews does not effect smog. Up in CT, we get a two shot deal. If we fail then we get to go back for free for a second test later within a certain time. If it is the same for you, it might be worth it to try and see if it passes. At least you would know exactly what it failed for and can clear and drive for next time.
It has been a pleasure helping you. Is it wrong to hope you car breaks so I can read your posts back? (Just kidding about the car breaking thing! Your 7 certainly put us to the test!) If you need anything else, I will give it my best to help. Bye for now Crispin.