Surging idle when warm.
 

1987 190 2.3 16v 178,000 miles

Starts up fine when cold and warm. Idles perfect at 950 when cold. Then as the temp gets up to 70ish the idle starts to hunt between 700 and 1300.

Could a leaking cold start injector cause this surging?

Is a cold start injector a common failure?

Thanks,

Christian

I'd be curious how to make a homebrew bench tester with a air compressor.

You and me both brother. Couldnt be rocket science, and saving a couple hundred on a new-not needed-injector would be nice also.

Im gonna smoke test this tomorrow.....if no leaks....Ill call a Priest or something.

ffp - Bosch cold-start valve in K- and KE-Jetronic

It should actually be easier than the main injectors since it is not pressure actuated. I assume 12v?

950 rpm is not good idle. It should be around 650. It is likely that your linkage is crusty, and not activating the idle switch. Do some tests first before you start swapping parts.

Take the air cleaner off and manually push the throttle linkage back to the rest position. Look to the idle switch (grey roller switch) and see it engaged.

Good thought. I already replaced the micro switch and the throttle cable is adjusted properly. The micro switch clicks open the second the throttle cable is engaged.

I have a fellow 16v guy lending me a Cold start valve, coolant temp sensor and fuel pressure regulator to test on Thursday. Side note: I swapped the Cold start valve from my Testarossa today.....no change but it is a different part number. In fact the part number installed on the 16V is different from the MB part site.


IIRC 950 idle is target for the Cosworth with manual tranny.

Ill update later. I love and hate this car at the same time hahaha. Not giving up on it though.

You may be right, I assumed you had an m103 of some sort. Checking linkage is a good start anyway, and maybe unplug the idle air control valve while the symptoms are present.

As far as the cold start valve, try just removing it, and block off the hole. Leave the valve plugged in (fuel line) so fuel doesn't go everywhere. It only runs for a few seconds when the engine is ice cold, so try this test with the engine somewhat warm..

It's an expensive part that almost never goes bad.

Hunting idle for me ended up being the ICV. New Bosch part fixed it up. I also believe the airflow potentiometer was not helping the matter. I got a new one a week before adding the new ICV. Though, the ICV is ultimately what stopped the problem.

Thanks guys.

It has a new ICV, no change and it does 'click'. The old one was good as well.

I disconnected the line to the cold start valve and ran a line to my spare fuel can. The car wouldnt fire.

The loaner parts will be here on Friday. After that.....?

Valve adjustment?

Rotor off by a sprocket?

I did check the timing chain and its new and lines up with the timing marks at TDC. But in the process I removed the rotor and I think I might be off by a spline on the gear. Tomorrow Im going to smoke test the engine to look for vacuum leaks.

Check the o2 sensor, temp sensor, and airflow potentiometer if you haven't already. All are easy to test. Just got done with this same problem if you need explanations.

My Coolant temp sensor is off a little...at 90 degrees it Ohms out at 393. I believe target is 200 to 300 at that temp. Thats what I get for buying ebay no name sensors. New Bosch 02 sensor, new throttle plate and new fuel / air plate with potentiometer pre adjusted. Im going to test the pot switch just to make sure tomorrow. Loaner temp sensor, fuel pressure regulator and cold start valve on Friday.

I believe these are the correct ohms for the coolant temp sensor:

60c - 600
70c - 435
80c - 325
90c - 247 etc

Like I said mine is at 393.

From what I've seen in the forums, hunting idle always ends up being: ICV, airflow potentiometer, temp sensor, throttle linkage, micro switch, or fpr. Or the thread just dies of with no resolution. Definitely curious what you come up with.

Throttle linkage?
 

Just discovered that the return spring on the throttle valve (W124) had broken/disappeared. Motion is transferred from the accelerator cable to the throttle lever by a roller that resides inside a roughly kidney shaped slot and it's wider than the little roller that transfers the motion. Without the return spring the little roller is free to float backwards and forwards in the slot, allowing the engine to hunt. Same phenomenon could occur if the ball joints and support shafts of the linkage overall are sticky. Just another possibility.

Mega, Im going to hold the throttle lever in place to rule that out. In fact, im gonna hold some of the other linkages in place.....interesting thought. Ill clean and regrease them to make sure they work as advertised.

Liquid...I have replace the usual suspects you mentioned with MB parts or good working parts from a loaner 16v.

Today I swapped out the coolant temp sensor, coldstart valve and fuel pressure regulator from a known working car. Symptoms improved but still hunts from 900 to 1100 when warm and drops down to 500 briefly on a sudden throttle let off. Combine a throttle let off with a sudden turn, and it will stall..30% of the time. If I load up the motor (lights, heater blower and tuning the wheels) with a throttle let off, It dies 60% of the time. Not as bad as before the new loaner parts.

The timing is dead on, I rechecked everything. Im going to burp the coolant system.....thinkng that air bubbles might be throwing off the coolant temp sensor thus sending a bad input to the ECU. I might also swap out the 02 sensor with the one that came with the car....Im running a generic Bosch 3 wire unit.

After I swapped the parts today, It idled perfectly at temp until I blipped the throttle, then surging.

Im off for training next week but will attack it after the 15th. I might bite the bullet and take it to the local guru IE throwing in the towel. Man card revoked.

Edit: seeming how the surging is only when warm......failing cat?

http://www.austincc.edu/wkibbe/lambda.htm

I've been meaning to take a inline reading off my o2 sensor just to see what is going on. I haven't found a equation yet to convert from volts to lambda, but I've read .4V is in ideal range.

What is your duty cycle number?