Hello there and welcome to the BMW Car Club of America.

If you are a BMW CCA member, please log in and introduce yourself in our Member Introductions section.

Idle Wild Problem - 92 3181

Discussion in 'E36 (1992-1999)' started by suva76, Sep 2, 2008.

    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild Problem - 92 318i

    Aloha, Forum Members!

    A while back, I began having a wildly-fluctuating idle problem (well, my car did, that is). Manual trans/110k miles, mine since Jan 93.

    I lived with it for a while, eventually took off the idle control valve (ICV) at the rear of the engine and sprayed it out w/brake-cleaner. Seemed to help a little, but not for too long. Lived w/it for a while longer! Eventually figured the electrical part of the valve must be "bad" so bought new part and installed it (yes, correctly, I am sure - very difficult to mess it up, although I DID see an on-line reference saying that the arrow on the part has to point to the block, whereas in MY case, this is not possible, so I cannot but conclude that reference was not correct).

    Problem remains. What else did I do? Glad you asked! Checked all vacuum-related and intake hoses and boots (some I had replaced not too long ago), no obvious leaks. Checked the throttle position sensor w/multimeter and it seemed to be w/in specs. The old ICV appeared to work OK, insofar as, when I removed the electrical connector, engine revs changed, and when connector replaced, revs changed again, but I replaced it because I could not find anything else that seemed to be a culprit.

    I cleaned the air-filter and the plugs. I have also been experiencing some oil on the plugs and black fouling (no pitting and otherwise, plugs look fine). I was suspecting a leaking valve-cover gasket. Replaced the (original: 100+k on car now) O2 sensor not long ago during this problem, due to check engine light being on, Peake code-read on O2 sensor, and fact that it WAS the original sensor! Still getting check-engine light, sometimes stays on for a while, or comes back on when at stop and idle goes way down (ocasionally, engine has stalled at idle). Was checking the vacuum hoses under the manifold, where there is a control valve (forget what it's called right now, but a tiny "flying saucer shape) and a nearby "one-way" or non-return valve, a tiny item with rubber hoses entering (or leaving) from bottom and exiting (or entering!) from top. These small hoses WERE a little aged and loose. I temporarily tightened the connection from the top with a twist-tie (all I had!). Perhaps the hoses and that valve need replacement, but not sure these are critical points in ths system, and I saw no improvement in the idle at any rate.

    A neighbor threw this at me: he has "seen many idle probs before" (appears to be a reasonably decent shade-tree mechanic - and a pro welder) and my spark-plug problem, and more often than not, he finds that due to "bad fuel" (his claim), the catalytic converter gets "junk" in it, which in turn can cause all the other symptoms I have w/my car. he takes the cat off and shakes, beats it or whatever, getting all kinds of crud out, and after replacement, problem is cleared-up.

    My exhaust system is the original, so there is no doubt that the cat COULD be at the end of it's useful life. And problems in the cat would give me O2 sensor codes, even w/a new sensor, I suspect. I DID replace the o2 sensor myself, a pretty simple job and I do not think it is possible to mess up its electrical connection. And when I did that, the exhaust system looked pretty good (of course, what's inside couldn't be seen). The guy suggested a compression test and fault-code analysis, and I am leaning toward that, since there seems to be little else to do right now.

    One other real oddity: a week or so ago, I was driving on a freeway, maybe 10 minutes or less after starting the car, when I heard a distinct "thud" seeming to come from the right side or maybe under the car. I had not noticed anything in the road that I may have hit, tho it was dark, so who knows. It almost sounded like a soccer-ball or something had maybe hit the lower right side - but this was freeway, with no playing areas or residents adjacent. I saw no dents or anything. Yesterday, after moving the car only 100 yards to an area with more daylight, in order to continue my troubleshooting (e.g., the car was virtually completely "cold", and I ran it one other time for a minute or two to check things under the hood, then off again), out of nowhere, I heard the "thud" again! Seemingly same area (right side, maybe underneath). I was on left side of engine bay. Car was not running or moving when this happened, unlike the first time.

    Anybody have ANY idea/theory/experience as to what the heck THAT was/is? I thought, OK, something exhaust-related? Cat converter, maybe? But what could cause this "thud"?! Some sort of pressure build-up that randomly releases?
    • Member
    • Technical Service Advisor

    mooseheadm5

    Post Count: 1,880
    Likes Received:15
    Wow.
    Anyhow, there are few things on that car that can cause vacuum leaks and/or erratic idle. I can tell you right now that just because they are not obvious, does not mean they are not there, they can be tricky. Take it somewhere that is familiar with BMWs and has a smoke machine so they can test it for vacuum leaks. There are hoses and gaskets that can leak, potential injector problems, AFM problems, problems with water infiltrating the computer box, etc. Can't help you with the thunk. That is the kind of thing that you need to hear in person, unfortunately. One of the most common problems with the check engine light is that often people see the O2 sensor code and immediately think it is the sensor. Usually it is just telling you that the mixture is off for whatever reason, which is often a vacuum leak.
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Thanks for the response, Paul!

    I also realized that I can (duh) read my car's stored codes myself, using the flashing-lights in instrument panel technique (and explanation of codes from Bentley Manual), so I will do that as soon as I get time. But something else happened last evening that was oddly coincidental and curious! At our chapter monthly meeting, held at the local BMW Center, I asked one of the service writers to pull the "campaigns" on my car and up popped up a recall dealing with erratic idle - for my OTHER '92 Bimmer (325is) - but NOT on my 318!!!

    Weird! I have not had the problem with the 325, and in all the years I have had the 318 (15+) this is the first time this has occurred (well, as far as I recall, anyway, but 15 yrs is a long time! At least the prob has never been like it is now, I can say!). So, I have the 325 scheduled to go in and get the fix (just a EPROM chip or something) at no cost, but I can't get the 318 fixed for free because the recall does not cover it! However, I WILL try to talk to the Service Mgr about that, to see what they will do for me.

    Good tip on the water in the computer box! I remember this being mentioned more than a few times in Roundel letters and elsewhere. Again, as far as I know, 'taint happened with my car, but there is always the first time.
    • Member

    Bimmerdan

    Post Count: 422
    Likes Received:4
    OK, here's a random thought on the "thud" sound you've been hearing...

    I get a similar sound on one of my E36's after I shut the car off and it sits for just a moment (it also happens now and then a couple minutes after start-up). I finally tracked it down to one of the vacuum operated HVAC flaps under the dash on the passenger side (the one that re-directs air between the different vents). I think one of the vacuum actuators has a leak so after I shut the car off, the vacuum slowly leaks out and the flap slaps shut with a distinct 'thud'. On start-up, it takes a minute for the vacuum to build in this particular actuator so it's the same thing, it eventually goes 'thud' when it opens (at least, that's my theory...). In my case, it has no effect on driveability so I just live with it.

    It's not impossible that the sound and the rough idle are related to a vacuum leak like Moosehead said. If you can find a shop with a smoke machine...do it! I did that on one of my cars that had a bad idle and I was sure I had checked for leaks everywhere! We hooked up the smoke machine and in 10 seconds I found exactly where the leak was!!

    Good luck and let us know what you find!
    • Member
    • Technical Service Advisor

    mooseheadm5

    Post Count: 1,880
    Likes Received:15
    The actuators are stepper motors, not vacuum, but it can still happen like you said because they do move slowly after the car is turned off.
    • Member

    Bimmerdan

    Post Count: 422
    Likes Received:4
    And it was such a GOOD theory...oh well, another one shot to heck...
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    Bimmerdan, you take those cars back to where you found them...right now, Mister!

    Thanks, guys, for the responses. Yeah, I get the "slammin' flap" now and then, but that has always been right in the dash, and not nearly as loud or "dense", if I may use that descriptor, as the one I have heard twice now.

    In other related news....BMW of Honolulu left me a message that "Oh, sorry, that campaign is really for specific conditions of cold temps (like, 30F)", etc., so it did not apply. That's OK, 'cause it was not for the 318, anyway. However, once I return on Monday from Maui, I will pursue gettin' some kind of (hopefully) free or (OK, I'll probably take it) discounted assistance on this prob.

    I SURE wish that BMW NA would let us have the 4-cyl diesel !!! I would take the 4-cyl Z4!!

    John in Honolulu
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    Back with some further observations......

    I got around to reading my "codes" via the blinking "check engine" light method and the only code that came up was the same one that my friend's Peake tool had: lambda (O2) sensor.

    And I had changed the sensor a couple of months ago (the wild idle problem preceded this by a long time). The sensor-change had no apparent impact on the fluctuating idle.

    But here's another odd thing I have noticed: with the AC on, the wild idling is pretty-much eliminated! Soon as I turn the AC off, the idle starts bouncing again.

    OK, does that mean anything? Well, doesn't the AC "tell" the computer to bump-up or otherwise adjust the idle when the AC goes on? Maybe this IS all a computer problem, related or similar to the problem that led to the campaign for the 325s (although this ain't a winter climate here). Should not be the idle-reg valve, because I replaced that, but COULD be the computer input to the valve. Not sure how/if I can determine that without a visit to the dealer or other mechanic.

    Still, why is this only causing a code readout on the O2 sensor!? Grrrrrrr!!!

    Oh, no further mysterious "bumps in the night".form under the car..at least so far.
    • Member
    • Technical Service Advisor

    mooseheadm5

    Post Count: 1,880
    Likes Received:15
    Oxygen sensor codes are rarely caused by the O2 sensor, but rather from the mixture being so far off that the O2 sensor's allowed adjustment of the mixture is reached and it throws the light. The AC tells the car to bump up the idle speed. This is likely the reason why the idle is smoother- there is a vacuum leak and the car cannot deal with the extra unmetered air unless the idle speed is turned up. As I said, find someone with a smoke machine and find the vacuum leak. Otherwise you are shooting in the dark. I am fairly certain this is your problem.
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    OK, Moosehead, I will do that - it may be a relatively "cheap fix" if some hose is leaking! They LOOK fine, but I guess it doesn't take much. BTW, you have some unusual machines there in your collection! Wow!

    Whoops! Meant to type "from" (vice "form") under the car, in last email!!

    P.S. I HAVE considered that the problem could be within the so-called "DISA" (dual resonance intake system) components - there is a small, 1-way check-valve with weak, small-diameter rubber hoses fitting from the bottom and top (not super-easy to get at with manifold in place) which slip off very easily and ARE a little old and worn. Not sure if a vacuum leak there (which could be quite possible) would be enough to cause the problem. There are also a solenoid valve, butterfly valve, and vacuum reservoir involved in the DISA.

    291979 guest

    Post Count: 1
    Likes Received:0
    fuel pressure regulator

    I had a similar prolem with my 325i, turned out to be the fuel pressure regulator.
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    Thanks, 291979 !

    Well, I took the car to an "indie" BMW shop on Saturday and it turns out it was the mass air-flow sensor, a part that sells for anywhere between $800 to $179!! (I found the latter on a parts-store website, and it does not even appear to require a "core" return. Others in the range of $300 - re-conditioned - + core, so if $179 seems too good to be true....). Have not ordered yet. The guys at the shop I took it to had one from another car and they put it in to check out and it calmed things down. They need the part back but let me "borrow" it under my hood for a couple of weeks. I am going to check with another indie shop where the owner is pretty good with electronics and see if he can repair my old one. I managed to pry off the top, exposing the electronics inside and nothing "jumps out" as broken. I suppose that it could be what appears to be the tiny sensor itself, a pin-head-sized amber-colored "ball" suspended by two wires, that hangs down into the air-flow. Did I read something, somewhere on-line about being able to clean that thing somehow (spray with carb-cleaner or somethingf?!
    • Member
    • Technical Service Advisor

    mooseheadm5

    Post Count: 1,880
    Likes Received:15
    The air flow meter is expensive for sure. You can get a used one. I would not trust those remanufactured units too much. We have not had great luck with them (usually Python or Fuel Injection Corp reman units.) Sometimes the flapper door in it is a little warped and sticks near idle causing an intermittent lean or rich condition.
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    Thanks, Paul,

    I wonder........if maybe the stuck-door could be the prob with mine? However, it appeared to move freely and easily when I reached inside and pushed it.

    My other indie shop just said same thing that you said....not much good fortune with Python brand! I believe that was the brand I saw on-line for $179!
    • Member
    • Technical Service Advisor

    mooseheadm5

    Post Count: 1,880
    Likes Received:15
    Yeah, sometimes they seem OK, but the door may bind when hot. It is also possible that the little circuit board inside is worn out. There is a carbon trace that has a little arm that wipes across that can wear out. I won't hurt to pry the plastic cover off and spray electrical contact cleaner on it. It may not help, but why not try? When all is OK, you need to glue that cover back on with silicone (don't go crazy.)
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    heh, heh...I guess that I was just a little ahead of you on that, Moose: I had pried off the top (not easy with al the goo holding it down!) last weekend. It LOOKS OK in there, and I saw the sweep-arm you mentioned. Just yesterday, I bought a can of electronics-parts cleaner spray, to spray the litle filament-sensor and the "flap-door", but I wasn't sure if it was wise to spray the electronic components (tho I figured that that's what this spray stuff is for, and, as you say, what've I got to lose?). So I will try that. About the silicon, it appears to be there not to keep air out, since the air-flow doesn't go thru the electronics, but to keep dirt/water out, no? What I am getting at is would I need to seal it just to test it or can/will it work with just the cover snapped back on. Incidentally, I broke some small pieces off of that cover in prying it off. How critical is the cover and seal to the operation of the sensor? And would the part get rejected as a "core return" if I have to buy a new one?
    • Member
    • Technical Service Advisor

    mooseheadm5

    Post Count: 1,880
    Likes Received:15
    It is just to keep crap out of the electronics. Not sure about the core return issues, but since they have to pry it off anyway, it should not be an issue if you make it look convincingly intact. Chances are they will have no idea if it is good or not until they go to remanufacture it anyway. At least you will get a warranty. If your mechanic buys his parts from Worldpac, he should get a one year warranty from them.
    Run it with the cover taped on for testing.
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    Right on, Paul. Will do that this weekend.

    Was looking into ordering the part on-line (if I need it). Apparently, "Python Injection" is the main re-conditioner of these sensors. A couple local mechanics here have said that they get some "bad" units from Python, so it is hit or miss. On-line sources appear to vary a bit in price, but the cheapest couple are near $235, incl the delivery (plenty to Hawaii) and core charge, so I guess that's better than the $600-800 new price!
    • Member

    suva76

    Post Count: 23
    Likes Received:0
    Idle Wild

    Well, upon close-examination of the mass air flow sensor, the "flap door" DOES appear to be a little warped! I grabbed it with some needle-nosed pliers and tried to bend it a little, but it's a pretty strong piece of metal - I think I bent it a slight amount in the "right" direction. Still, it tended to bind a little, not closing all the way after poking it open and letting it flap back. I cleaned the "air-flow tunnel" inside (a little dirty, but not terrible, especially considering it has over 110K on it), and sprayed generously with silicon lubricant. Seemed to help! I put it back together and so far it is working OK. Maybe not for long, but for now it is. And, at least I know what the prob is!

    BTW, I can see where the use of oil-soaked air cleaners might tend to gum-up the meter. I have only used paper filters.

    just1318is guest

    Post Count: 23
    Likes Received:0
    had a similar problem on my 94 318is. i bought a can of afm cleaner (under 10 bucks) and that has worked great for me so far. also while when the problem first surfaced it turns out my alternator was bad. it made the most difference to the idle when i replaced the alternator and then cleaned the afm. just an idea!

Share This Page