Rough Idle (blah, blah, blah...)

[rps1072]

I have an intermittent rough idle problem on my 94 525iT (87k). It starts well, but when it tries to go back down to idle speed, it usually dies. If I can get it to idle, it does so at around 600 rpm (or less) and runs very rough. Every once in a while it will go away. The last time this happened, I ran completely out of gas. When I refilled the tank—fixed as if by magic.
It runs and drives very well—just the idle sucks.

Here is the litany of things I have tried:

New fuel pump and sender
New fuel filter
New fuel pressure regulator
New fuel injectors
New idle control valve
New intake temp sensor (the one in the bottom of the manifold)
New spark plugs and connectors and tested coils
Tried a different air mass meter
New oxygen sensor
Propane test for vacuum leaks
Carbon buildup removal (in the shop)
Truckload of fuel system cleaners
Tested throttle position switch: resistance varies from 0.80 Ohms to 4.75 Ohms without interruption.

What else is left to do?

I have seen several posts that state that the resistance for the TPS should vary from ~1 Ohm to ~4 Ohm. Are my values of 0.8 to 4.75 within the "~" range? How would too low resistance affect the car at idle?

[winfred]

with a dead crank sensor it won't start, and on some cars a dead cam sensor doesn't really effect the running quality just the way it injects fuel, instead of one injector at a time it goes to gang fire like older cars and pulses all of them

[rps1072]

Is there anything between working and dead?

[winfred]

not usually, theres a few tests for it using a multi meter, theres 3 wires 2 signal and 1 shield, the center and one of the outers are what you want to hook to, you will get ac voltage while cranking, you will get a volt or two off the crank depending on cranking speed and it's hard to say on the cam as it only triggers once every other crank revolution, you will see something. another test is resistance, bentley should have the figure, it's on the board if you don't have a bentley

[rps1072]

Can I get to the plug for the crank sensor without removing the manifold?

[winfred]

yes it's just behind the oil filter housing, follow the wires

[rps1072]

I'll give it a try. Thanks again.

[SRR2]

Your TPS is fine. Most multimeters have quite a bit of error when measuring low resistances. The TPS input to the DME is high impedance anyway, so actually it wouldn't matter if the switch produced 30 or 40 ohms, it would still work. (That said, if the switch actually did that, it ought to be replaced because obviously it's not working correctly and may be in the process of failing.) But the values you have will drive the DME properly. By the way, do not be surprised if your TPS does not have a WOT closure. I've looked at three different p/n's of these switches and none of them had a WOT switch installed in the package. If you manually turn the throttle to WOT and hear a defininte click sound in your switch however, then you may have one that actually does have the WOT function installed.

I don't know what Winfred is talking about if he's referring to the TPS, maybe he was actually referring to the CPS. The TPS switches only at closed throttle and has nothing to do with crank speed.

[winfred]

i was taking about the crank and cam sensors, the later small thumb sized tps's don't have a set switching position they vary a voltage signal from 0 to 5 volts


I don't know what Winfred is talking about if he's referring to the TPS, maybe he was actually referring to the CPS. The TPS switches only at closed throttle and has nothing to do with crank speed.

[SRR2]

I was thinking about his 1 ohm to 4 ohm reading. The pot type TPSs shouldn't ever produce a resistance that low, so that's why I was assuming (I know... I know...) that it was the switch type.

Bentley's procedures for testing any of the pot type transmitters/sensors SUCK with a capital S. They consistently tell you to watch for some vaguely specified resistance variation and deem it 'good' if it changes. I've seen a number of these things that were in fact bad, but given their miserable troubleshooting procedure, you'd have thought they were good. Their spec for the AFM is another example of it.

The right way to check a pot type sensor is to power up the element with 5V and then watch the wiper voltage vary as the device is exercised. You can see right away if it's right or not.

[rps1072]

There was one place I had not checked for a vacuum leak. I took off the fuel tank breather valve and squirted some cleaner in it. That improved the idle a lot. So, maybe there is a vacuum leak in that valve. I am going to try to replace that valve--hell, it's cheaper than any of the sensors.

I think the growl was caused by every single ball joint in the front suspension being worn out--they are all original. I replaced the sway bar links, and now everything else is clunking with greater purpose...but that is another thread.

[M-what]

with a dead crank sensor it won't start, and on some cars a dead cam sensor doesn't really effect the running quality just the way it injects fuel, instead of one injector at a time it goes to gang fire like older cars and pulses all of them

I too have the "blap blap blap" idle and have high HC (unburned fuel) emission readings during idle on my 90 535i. But it clears up fine off idle. I just put in new plugs, caps and rotors, cleaned the ICV & AFM and check for vacume leaks. Its a lot better than when I first got the car but still idles rough. So if the cam sensor is bad and it injects fuel all at once, is this only felt/seen during idle-too much fuel being dumped in to burn complete or should they be apparent through the whole rev range? It takes off and revs up without bogging for spewing out dark exhaust gas. How do I check the cam sensor? (bently is on the way). Steve

[winfred]

m20 and m30s have a inductive pick up, the bentley should have a resistance reading for it, you'd need a scope to tell more


How do I check the cam sensor? (bently is on the way). Steve