I have a '91 190e 2.3L. I was recharging my AC sometime ago when suddenly the AC compressor disengaged and the idle shot up to 1500. Over the last few days the idle has climbed up to 3200. The car can accelerate up to 25mph on its own and is extremely (and dangerously) hard to stop without popping it in neutral.

So far I have ran injector cleaner and carb cleaner. I have replaced vacuum hoses going to the fuel regulator and between the ICV and the air box. I have cleaned out the ICV with carb cleaner and ran a 12 volt test on it, seems to work fine, however it is stuck open about 1/4 of the way (I think this is normal). I have opened up both the ECU and Overvoltage relay to look for damage (none) and resoldered some joints that looked questionable. I lubed up the throttle linkage. I also tried unplugging (by electric plug and fuel) the cold start valve while the car was running, which made no difference.

What am I missing? Could there be a hole in the intake manifold?

Thanks
Andrew

 

i think it's a malfunction on the idle speed actuator.

if the wiring is disconnected on this one, the idle speed will go very high. (emergency measure).

try removing and cleaning it with carb cleaner. it might have got stuck in full open position (spring loaded to open).

check if the wiring is properly connected. see if it got current (milliamperes).

 

forgot to include this....

[IMG]local://upfiles/1/9C2FBA9710944AD6BCBC5D5A0CB27A49.jpg[/IMG]

 

Might the actuator be what I am calling the ICV? I can't find anything similar to that other than the idle control valve which I have already cleaned out. Current to it is .60mA.

 

Ok sleepwalker, here's what I got.

I took the hose off of ICV/actuator and keep an eye on the valve while I turned the key to "on". When the car is turned to on, the valve opens all the way and then closes about 15sec later. I put a piece of paper inside the hose after the valve to restrict air flow. With air flow completely cut off the car doesn't start (good). When air flow was limited to a small amount, the car starts with no problems and the idles at 2000 instead of 3200. I think this means that there is indeed too much air going through the ICV. I believe the actuator is operating correctly, but maybe the signal to it is incorrect? Is the actuator naturally suppost to be slightly opened?

 

i agree that it's just a signal problem since it happens only after you turn off the a/c.

sometimes by cleaning the a/c relay and fuel pump relay contacts can resolve this.

if the idle speed valve is sticking or faulty, the idle speed will only go high for sometime, then will be back to normal... intermittent.

air leak on the intake side will also cause high idle speed. to test this, you may need to spray some water on suspected area, esp on injector seals.

you may need to check the idle speed switch on the throttle valve also.

the idle speed adjuster valve is getting signal directly from engine control unit. however, it gets power directly from terminal 6 of over voltage protection relay (terminal 2 on ICV). better check if there is power on it at ignition on. terminal 1 is signal from ecu.

we might be barking at the wrong tree here, since few other things can cause high idle speed. the ecu is suppose to get the rpm signal and if that...is also missing, we will have this problem too.

linkage on throttle can also cause this hi idle speed.

i really cannot pinpoint the source without checking the car myself.

i can only assist you on some leads.

 

Hi Andrew.
You appear to have shown that the ICV is allowing to much air to pass at idle since by blocking it off (partially) the revs drop. Can you alter the orifice size and show different idle speeds?

This suggests the ICV (or possibly throttle butterfly) is at fault but you state that BOTH are closing correctly.

It is possible that air could be getting in via another route e.g. manifold gasket as you state but..
the ICV ECU is usually designed to monitor engine revs and control to the required idle speed setting. If you have a sufficient air leak then the ICV could be effectively "electronically" shut yet sufficient air is getting in to run the engine at idle (and also at abnormal revs) BUT if this were the case blocking the ICV hose shouldn't make a difference since the valve is nominally shut. Maybe there is a "designed" in leak to allow for nominal idle airflow just to make matters more complicated?

My "guess" is that the ICV is not closing adequately either due to mechanical or electrical reasons but why did this occur whilst you were working on the A/C?

Concerned that in some systems the engine speed ECU monitors whether A/C is ON since different settings (fuel?) may apply to idle to allow for load of A/C pump rather than controlling entirely by engine speed. You don't state whether A/C is currently working correctly.
Does idle have the same problem whether A/C is ON or OFF?

Stuart.

 

To clear up on the AC. The compressor is pretty much brand new, previous owner had it replaced about 3 years ago and let the car sit in a garage before I got it. I have the reciept. Basically the compressor has not been used a whole lot. While I was recharging the AC, the compressor clutch disengaged and the idle shot up to 1500. After that the idle climbed and the compressor has refused to engage whether AC is on or off.

I did look into it a little more today. The throttle linkage does not appear to engage the idle speed switch, and I can't get it to adjust to do so. It's possible that at first glance the valve appears to be closed, but is slightly opened. I'll have to take it apart tomorrow as I have to go to work soon.

Another thing I found was a post on a different forum that shows how far open the ICV should be open at different volts. At 3V it is closed, 10+ V it is fully opened, 0V leaves it opened 1/4. Wished I had booked marked it. Either way, my ICV operates as above. However voltage to the ICV reads 13+ V when the car is on...

To answer the rest of the questions:
AC and fuel relay contacts are clean
ICV is not sticking at all
I sprayed water on the injector seals with no change in idle
Idle becomes a little unstable when pushing the AC button, but the idle usually stabilizes back to 3200 (compressor does not engage).
Leaving the AC relay unplugged changes nothing.

Also is there anyway of test the ECU and the overvoltage relay? I would also like to test the input signals to them as well.

Thanks, got to go,

Andrew

 

While I was at work I thought of something. Is it possible that I overcharged the AC causing maybe the AC pressure switch to shut off the ac clutch and increase the idle, or even break the switch all together?

Edit: I am going to bleed the system tomorrow... you never know...

Andrew

 

if pressure is more than 20 bar, the a/c relay will disengage the clutch in compressor.

if the compressor got speed signal, the relay will also cut it off in order to protect the belt so other components can still function. (for example...seized compressor).