Ongoing problem with fuel cutoff and idle

[SRR2]

To recap: '89 535i manual. 150000mi. Changed battery and immediately ran into idle and fuel cutoff problem. Idle is often too high, usually in the range of 1000 rpm. When it's in the state of misbehavior characterized by the high idle, the fuel won't cut off above 1500rpm like it's supposed to. Take your foot off the accelerator and it continues to feed fuel at a rate that almost prevents deceleration. The ECU isn't throwing off any "stomp" codes. No CEL. I cleaned the ICV and got a small improvement (the idle was a little better controlled) and then replaced it with a new one. Problem persists. At the same time, I replaced some old rubber in the vicinity of intake and breather. I checked the fusible links. The O2 sensor seems to be working, as the problem persists virtually the same way with (no CEL) and without (CEL on) the sensor connected, and when the engine is cold (when the sensor isn't used). Water temp sensor is in spec. Throttle position sensor switch is functional and has voltage. I have reset the ECU too many times to count. I've tried the ECU calibration procedure (posted by Shogun) many times without improvement. Something is causing the ECU to supply fuel when it shouldn't, particularly above the 1500rpm closed-throttle cutoff point.

Now I'm a little suspicious of the AFM. When I had the rubber off to replace it, I tested the temperature sensor (in spec) and measured the resistance between pins 2-3 as the Bentley manual describes. Bentley is not particularly helpful in determining operation because they say something like "the resistance should vary". Well thanks a lot. Vary HOW? And here's why I'm suspicious: With the vane closed I get 550 ohms consistently. As the vane opens the resistance rises to almost 900 ohms in the center of its travel, and then as it approaches full open, the resistance drops again to around 400 ohms. That doesn't make a heck of a lot of sense to me. I'd expect it to increase or decrease as it traverses, but not rise and then fall again.

Does ANYONE have some spec on resistance v. travel?

And while I'm at it, can this be converted to a hot wire or other less restrictive MAF sensor? If I'm going to have to replace it, I'd like to use better technology.

One other observation... The bad behavior is predictable. On a warm start the engine is perfectly controlled at idle, while delivering power, and during deceleration when the fuel cuts off as it should. After about two to three minutes, the bizarre stuff starts. This happens every time. Then, when it's misbehaving, I can come to a stop and watch the idle sit at 1000 rpm. After a predictable two minutes, the idle drops to exactly 700 rpm where it will sit indefinitely. The 700 rpm idle is smooth. Plus, I can cycle the A/C compessor and it doesn't move at all.

This is driving me . I hope someone can help me out here. At this point I'm seriously considering taking it into the dealer, even at the risk of them engaging in an expensive Parts Replacement Guessing Festival.

[ryan roopnarine]

if you were to induce a massive leak at the icv connection point on my 1992 525i, and get it used to it, you could get it to get used to having the air leak, which would result in a 1100 rpm drive idle (stoplight), and a 3k+ rev at park, ie no cutoff. this has only happened once, but i know that it can do it. i'd check for vacuum leaks if you are at all mechanically inclined first.

To recap: '89 535i manual. 150000mi. Changed battery and immediately ran into idle and fuel cutoff problem. Idle is often too high, usually in the range of 1000 rpm. When it's in the state of misbehavior characterized by the high idle, the fuel won't cut off above 1500rpm like it's supposed to. Take your foot off the accelerator and it continues to feed fuel at a rate that almost prevents deceleration. The ECU isn't throwing off any "stomp" codes. No CEL. I cleaned the ICV and got a small improvement (the idle was a little better controlled) and then replaced it with a new one. Problem persists. At the same time, I replaced some old rubber in the vicinity of intake and breather. I checked the fusible links. The O2 sensor seems to be working, as the problem persists virtually the same way with (no CEL) and without (CEL on) the sensor connected, and when the engine is cold (when the sensor isn't used). Water temp sensor is in spec. Throttle position sensor switch is functional and has voltage. I have reset the ECU too many times to count. I've tried the ECU calibration procedure (posted by Shogun) many times without improvement. Something is causing the ECU to supply fuel when it shouldn't, particularly above the 1500rpm closed-throttle cutoff point.

Now I'm a little suspicious of the AFM. When I had the rubber off to replace it, I tested the temperature sensor (in spec) and measured the resistance between pins 2-3 as the Bentley manual describes. Bentley is not particularly helpful in determining operation because they say something like "the resistance should vary". Well thanks a lot. Vary HOW? And here's why I'm suspicious: With the vane closed I get 550 ohms consistently. As the vane opens the resistance rises to almost 900 ohms in the center of its travel, and then as it approaches full open, the resistance drops again to around 400 ohms. That doesn't make a heck of a lot of sense to me. I'd expect it to increase or decrease as it traverses, but not rise and then fall again.

Does ANYONE have some spec on resistance v. travel?

And while I'm at it, can this be converted to a hot wire or other less restrictive MAF sensor? If I'm going to have to replace it, I'd like to use better technology.

One other observation... The bad behavior is predictable. On a warm start the engine is perfectly controlled at idle, while delivering power, and during deceleration when the fuel cuts off as it should. After about two to three minutes, the bizarre stuff starts. This happens every time. Then, when it's misbehaving, I can come to a stop and watch the idle sit at 1000 rpm. After a predictable two minutes, the idle drops to exactly 700 rpm where it will sit indefinitely. The 700 rpm idle is smooth. Plus, I can cycle the A/C compessor and it doesn't move at all.

This is driving me . I hope someone can help me out here. At this point I'm seriously considering taking it into the dealer, even at the risk of them engaging in an expensive Parts Replacement Guessing Festival.

[SRR2]

if you were to induce a massive leak at the icv connection point on my 1992 525i, and get it used to it, you could get it to get used to having the air leak, which would result in a 1100 rpm drive idle (stoplight), and a 3k+ rev at park, ie no cutoff. this has only happened once, but i know that it can do it. i'd check for vacuum leaks if you are at all mechanically inclined first.

Yeah, I'd thought of that and replaced all the hoses/rubber in that area. They were all original, not in real bad shape, but it was an easy and cheap bit of preventative maint. I didn't expect any improvement and didn't get any. Reason I doubt that it's a vacuum leak is because everything works correctly for 2-3 minutes after a warm start. Then it acts up. Predictably. And when it's acting up I can park it and let it idle for about 2 minutes whereupon the idle returns to the right value and the fuel cutoff will work for three or four shifts. These times are repeatable. I just demonstrated it for one of my neighbors (who's a car guy) this morning who also allowed as to how strange he thinks it is.

Also this morning I pulled out the ECU and inspected it under magnification. It looks pristine. No evidence of any component failure or bad solder joints. There aren't any aluminum electrolytic caps in it, only solid tantalum, so that rules out a nasty souce of long-term failure.

[infinity5]

what happens when you rev the car in neutral, while its parked? could it be a problem with the throttle linkage or anyhting like that? have you checked the fuel pressure? or replaced the FPR? i'd replace everything i could think of associated with the fuel delivery system, as well as the AFM and maybe the TPS just for fun (how much is it?) and then take it too a good indy or the dealer.

it also seems like, from what other people have reported about various components, correct voltage and/or resistance values are sometimes given even when the components are bad. see if you can't find someone near you to trade out parts with.

i agree that seems like an odd thing to happen with the AFM resistance values.. maybe thats your problem! good luck

[SRR2]

Thanks for all the suggestions. The problem is solved as I detail in this http://www.bimmer.info/forum/showthread.php?t=10783 post. Short version: it was an intermittent, temperature-sensitive TPS.

[SRR2]

Thanks for all the suggestions. The problem is solved as I detail in this http://www.bimmer.info/forum/showthread.php?t=10783 post. Short version: it was an intermittent, temperature-sensitive TPS.

[SRR2]

Thanks for all the suggestions. The problem is solved as I detail in this http://www.bimmer.info/forum/showthread.php?t=10783 post. Short version: it was an intermittent, temperature-sensitive TPS.