Strange Idle problem...

[jamiholzbach]

I have a strange Idle problem. When I start the car cold, it idles fine. However, Once it warms up the Idle stays up between 1000 & 1200 RPM. Any ideas? I've checked my throttle cable (& lubed it), TPS, and did an operational check on the idle control valve(appears to be working correctly) I've yet to do a resistance test on the O2 sensor, but when the car first starts to warm up(needle on the temp gauge starts to register) i notice that the idle smooths out, so I think the O2 sensor is working. Possibly the Temp sensor?

[jlevie]

Did you do an electrical check on the idle switch in the TPS? And is the TPS correctly adjusted? The idle switch should close when the throttle is .020-.060" off the throttle stop screw.

If all of that is correct and no one has tampered with the bypass adjustment in the AFM or the throttle stop, I'd suspect a sticking ICV, bad temp sensor, or intake leaks.

[jamiholzbach]

Sorry, been kinda busy...

I'm gonna start messing with it again today after work. I'll check stuff out cold first, then let it warm up and take readings again.

[Goya20]

spray the icv with air intaker cleaner while off the car and dry it up and put it back on it helped my car. Also remove the air intake boot and clean the inside of the manifold, play with the butterfly, make sure to put a rag at the bottom of the manifold because fluid willl drop out.

[jeff_b]

I would guess one of the temp sensors...

[llll1l1ll]

My two hoses connecting to a small box on the side of my intake manifold (right where the butterfly is) are leaking lots of coolant, leading to a higher idle and a bad smelling exhaust.

[jamiholzbach]

I believe its...

I went through checking everything involved with the idle, and I found that the resistance on my coolant temp sensor was way too low (1700ohms instead of 3000ohms when cold). Just for grins I did the STOMP test, and wouldn't you know it, code 1223. The coolant temp sensor! So, now to get a replacement. I'll let you know what happens when I swap it out.

[jamiholzbach]

Nope!...

Its not the coolant temp sensor. When I unplugged the ECU to check the wiring I reset the codes. after the next run it flashed 1444 (no codes) then after the next run it flashed 1222 (O2 sensor detecting rich/lean conditions). Again this is after it warms up and I drive it around the block, the Idle goes up to 1200 RPM's (Oh yea, by the way I realised that I misread the Bentley manual, 3000 ohms is what it should read cold in Alaska).

[jamiholzbach]

I think its...

I think its the O2 sensor. I ran the engine up to temperature and of course the warmer it got, the higher the idle. Once it reached operating temp, I unplugged my O2 sensor to take some readings with my voltmeter. When I statred it again wit the O2 senor unplugged, it idled just fine. With my voltmeter I checked pin 2 on the sensor side and it registered .068 volts. when I rev'd it up, the reading went up to .0756. But when I backed off the revs, it stayed there, and did not go back down. So, I think its the O2 sensor. Now to just get one...

[jamiholzbach]

Not The O2 sensor!!!

I replaced my O2 sensor and once the engine warmed up, it was 1200 rpms all over again! Since my "87 was running good, and other than the O2 sensor connectors being different, I decided to try swapping out parts.
So, first up was the ICV. Now I had already tested the ICV previously, and it had checked good according to the Bentley manual(buzzed when ignition was turned on, resistance & voltage checks were good), but when I swapped ICV's between the cars, the problem followed from the '89 to the '87. so I visually checked them (just plugged the electrical connector on and set the valve fully open and set it where I could watch what happened inside when I turned the key) and the ICV from the '87 closed up completely. When I did the same thing with the ICV from the '89, it left an 1/8" gap. So I tried cleaning it out with brake cleaner but that didn't work either. So I went to the pic-n-pull and got lucky. One 325i in the lot, with an ICV with the same part number. I got to test it on the car before paying $35 for it. Now the '89 idles fine! Next up, the clutch;)

[Adrian_Visser]

nice work, I've always doubted that the tests in the bentely completely rule out the icv as being bad

[jlevie]

nice work, I've always doubted that the tests in the bentely completely rule out the icv as being bad

You are looking at those tests wrong. They aren't meant to prove that the ICV is good, but rather to tell when it has failed. If an ICV fails those tests it is definitely bad. But it can pass those tests and still not function correctly.