Revs dip - sometimes stalls - STILL NO CODES!

[chrisben40]

Last December I posted a thread about the rough idle on my 328is. I never did find the problem. Even after replacing a dozen or more parts including a few sensors.

It is getting a bit more serious now. Lately my car has been stalling when I slow down to stop. It happens when I back off the accelerator and disengage the clutch. At that point the tacho does a quick dip to about 500 rpm them (usually) recovers to just over 700. Oddly it only does this after being under load. If I just blip the throttle in neutral I can't get it to happen. The idle it settles to, although a bit low is fairly smooth. I still get no codes from my Peake Research reader or Casrsoft 6.5. For a full rundown of what I've tried see this thread: http://www.bimmerfest.com/forums/showthread.php?t=331995

The only thing I do know about the stalling from live data is that there is a big negative fuel trim value of about -4.8 or more when this happens. The other thing is, this was happening badly yesterday after some heavy rain, making me think it is something electrical.

Does my car have to fail completely before I will see a fault code? :dunno:

[drivinfaster]

hmmm, maybe being at the bottom of the world makes the fuel settle at the 'top' of the tank away from where the pump is???:dunno:


i'm kidding....:p

the things that would have a direct input on idle would be the tps, throttle plate, maf, icv, intake boot, injectors, and dme, as well as the wires that tie them all together.* have you checked for a bad gound on the tps or icv?? have you tried replacing the icv with a known good one?? a bad ground can be checked by doing a voltage drop test (google it, or search, i think i did a link somewhere but i cannot remember...) and fixed by splicing in a similar wire to a clean ground. electrical gremlins can be a pain, especially as a vehicle ages.


*this is just a sophisticated wild guess here, as i am unable to get to the bottom of the world to check it out. i'm not sure that i would be able to cope driving on the wrong side of the car, though.


hope this helps.


df

[the_brouhaha]

hmmm, maybe being at the bottom of the world makes the fuel settle at the 'top' of the tank away from where the pump is???:dunno:


i'm kidding....:p

the things that would have a direct input on idle would be the tps, throttle plate, maf, icv, intake boot, injectors, and dme, as well as the wires that tie them all together.* have you checked for a bad gound on the tps or icv?? have you tried replacing the icv with a known good one?? a bad ground can be checked by doing a voltage drop test (google it, or search, i think i did a link somewhere but i cannot remember...) and fixed by splicing in a similar wire to a clean ground. electrical gremlins can be a pain, especially as a vehicle ages.


*this is just a sophisticated wild guess here, as i am unable to get to the bottom of the world to check it out. i'm not sure that i would be able to cope driving on the wrong side of the car, though.


hope this helps.


df

+1 on the MAF, maybe even a vac leak.

[chrisben40]

Thanks guys.

drivinfaster, you were a great help last time around. Your new suggestions are definitely worth a look. I have given the icv a good clean, as I have the throttle plate and maf. (don't worry, I used proper cleaner). The flow figures seem feasible and I had a second opinion from a diagnostics guy down the road. The tps readings also seem to check out. I have done exaustive vacuum tests and the reading is good. The icv earth sounds like a good possibility though. Especially since it seemed to be worse after some heavy rain.

I'll let you know how I go with your suggestions. Unfortunately, I have no access to substitute parts so I would have to pay our over the top Australian prices to do that.

It is hard, at first to get used to driving upside down and on the wrong side of the car but I used to practice a lot on the rollercoaster at the local theme park.

[nismosx08]

I actually had this symptom once immediately following a cooling system problem (radiator hose blew off, etc). It turned out that the reason the idle was fluctuating right afterward till i had it all repaired was that the coolant level was too low to provide a reading for the Coolant temp sensor. If the ECU is missing signal from this sensor, it can't properly manage the idle. Once the coolant level was high enough to allow the sensor to function, idle was back to normal.

[chrisben40]

That's very strange. I just replied to your post and it has disappeared. Maybe I dreamt it.

I did suspect the coolant sensor so I replaced it a couple of months ago. I haven't let the coolant get low in the two years I have had my car. Maybe I should test the wiring to the sensor though.

I have a theory it might be leaky injectors. I think the idle is basically o.k. but it is running rich in certain conditions. I was thinking that, because this rev dip happens after engine braking, it could be fuel leaking past one or more injectors under high manifold vacuum. That would make it run rich until next time you gave it some throttle and burn off the extra fuel. The injectors hold pressure when the car is not running because it starts fine after standing for a few days. I have tried a fair amount of injector cleaner but that probably isn't enough. If the suggestions on this post don't show up anything I might pull out the injectors and have them tested. I was resisting this because injector testing is quite expensive.

What do you think of that theory df?

[drivinfaster]

What do you think of that theory df?


forget the theory. i want to know how you got your e36 on the rollercoaster...oh wait, i musta misunderstood...:tsk::D


to confirm a possible leaky injector, you need to put a guage on the fuel line and take a key on engine off (koeo) reading. since bmw didn't put a fitting (that *i* have noticed on mine, or any other that i have had to do this to) i usually have hooked into the fuel line elsewhere to get this reading. (yes, it's a pain in the butt)

once you are inline and not leaking anymore, run the engine to full operating temps. note the runnig fuel pressure, and then shut the engine off. see what the 'resting' fuel pressure reads. it shouldn't drop much. then, koeo for the 2 second prime and clamp the feed and return hoses (safely with the proper tools as not to cut or damage the lines) and watch for the fuel pressure to drop. it shouldn't since the fuel has nowhere to go except through the injector(s). be preppared to have fuel in the engine oil, so do a fluid check (grab a tiny sample from the oil filter reservior) to compare with the oil after the test has been completed. this way you will have something to compare to when you are finished. (or just change the oil...)

now if there is no pressure drop, undo the clamps and perform a performance balance test. (yes, this is more specialized test equipment that will probably cost less than 50 usd at places like pep boys). the tester (i bought from aforementioned automotive establishment) will/should have 2 modes, pulse and single. you want to use pulse mode for this test. what it does is operate the injector a set amount of time (consistently) in order to check how much fuel is actually passing through the injector. a weak or faulty injector will flow differently than than one that is operating properly. this will show via different pressure drops on the guage. (think of it almost like a compression test for the cylinders) this test should be done both hot and cold (hint, do cold first...) and should be done 2 times for consistency (like a compression test...)

do you have a power graphing meter or pgm mode on your scan tool?? looking at voltages will not always indicate a poor ground (hashy signal), and if you are looking at 'values' such as temperature and degrees they're all processed. stick with the actual voltages, and if you can, scope them with a power graphing meter to look for drop outs, and high resistance as indicated by a hashy 'live' (vehicle running) signal. if no pgm, use your dvom and do a voltage drop test on the ground wire of the tps and icv, s/b >0.1v rule of thumb.

hope this helps.


df

[chrisben40]

We only use rollercoasters for driver training here in Australia. We also have to use the trains that come with them. Not a lot of people know this but a standard E36 won't actually fit on the rails without modifying its track. Most people down here are lazy and just use their cars on the roads. BMW Australia also specs larger fuel pumps, running at higher pressure, on account of all the gravity we have here.

Wow, that is a lot of complex information you have given me. I think it is getting a bit over my head. I will take my car and your suggestions to the diagnostics guy I know down the road. The good thing is, he also has an injector cleaner if that's what it needs.

Thanks.