380SL Still Won't Idle

[ctaylor738]

In the spirit of "just the facts:"

Car ("Wayward Woman") is a 1981 380SL. Engine is from a 1984 euro 380SEC, but with US injection and emissions.

Problem: rough idle when warm. Not a miss, but a "cycling" between 400 - 600 RPM. Engine will start to die, then surge. Worse in Park than in gear.

New parts: plugs, wires, cap, rotor, fuel distributor, O2 sensor

Other info: Runs excellent otherwise. Mixture adjusted, lamda control cycling correctly. OVP relay working. Vacuum is 16" with idle stable at 750. "Gently pressing" on air intake plate does not smooth idle. All wires test OK, no corrosion found. No detectable vacuum leaks.

Tests done: Idle air valve is getting rapidly changing 3.76 - 4.8V when problem is occurring. Manual says it should be 5V +/- 1V. Bridging the contacts on the coolant switch (simulating temp < 42 degrees) brings the idle up and stabilizes it at 750 RPM. Temp switch shows 0 V output with engine warm. So, alles gute?

Dilemma: Is the idle controller defective and sending bogus signals to the air valve? Or is something else, like a vacuum leak, or a bad injector causing a feedback loop where the RPM drop and the idle controller "catches" it and overcompensates.

One way it's a $200 bet on a non-returnable electrical part. The other, it's a $200 bet on injectors and seals (4 of which are new).

Any additional tests or places to look?

TIA

[stevebfl]

First disconnect your O2 sensor when the symptoms occur. This is because you only want to deal with one feed back control at a time. Verify that the throttle switch is closed (electrically - the euro car had a different throttle switch arrangement). Be sure of your vacuum control to the distributor retard (it should be off when the motor is warm).

[ctaylor738]

1. Very slight improvement when O2 sensor disconnected.

2. Injection is US, not euro - will perform test tomorrow (had to look up procedure).

3. No vacuum is being applied to distributor at hot idle.

[joe p]

What about the throttle body and lower intake, a/f sensor and such? Was the entire intake swapped including the bottem section and throttle body?




Joe

[ctaylor738]

Yes. It is a US car that was believed to have bottom end damage. I replaced the engine with a euro block, found a set of euro cams. But I kept the entire intake manifold and injection off the original US engine.

[Thomaspin]

It ran so roughly it almost stalled.

Replacing the fuel injector holders/seals cured most of the problem - they were leaking. (125k and 20 years old). Inexpensive. 8 seals and holders are under $40 if you DIY. Job is similar to the 560 one on my site.

The local indy did a white smoke test (pumped into engine) and determined there were also intake manifold leaks. We still get some idle cycling, but I have yet to do the i.m. job.

[ctaylor738]

Points well taken.

Four of the holders and seals are new (right side) because of a head replacement due to a timing chain problem - read "jump." I have been holding off on the other four pending the decision on new injectors, which I don't want to make until I resolve the idle controller issue.

No vacuum leaks found hunting with carb cleaner, though.

[Thomaspin]

....and it did not disclose the cause of my leak. The white smoke test, or whatever it's called, did.

Could be that so many of the seals were leaking that spraying one at a time made little difference.

[ctaylor738]

I was able to test the throttle switch today. Works the way the book specified - pins 1 & 2 have no resistance at idle, and 2 & 3 have none at full throttle. No connectivity at other positions.

What's next?

Would it work to test the idle control by giving the valve a steady 4.5V, and see if it smoothes the idle? If it does, it's the relay, if it doesn't, it's on to the the injectors and seals.