change in idle D/N

Hey Guenter, still looking for that?

It is normal, that the engine changes sound when shifting from load to none, but the RPM should be more or less the same. A difference of

100-200 rpm is definately not normal. But you sometimes hear a mechanic claim a problem as "normal", if he can't find the fault.

It is a closed loop, where the ECU will control the idle speed air valve for a constant RPM, just as you could do yourself with the accellerator (the idle speed air valve is simply a small electically controlled throttle in parallel with the ordinary throttle).

If it changes speed, it indicates that the ECU (or rather the loop) is not functioning correctly (as if the driver controlling the accellerator fell asleep).

OVP relay is providing power to the ECU, and if that is failing, the ECU will not get power and will of course not function then. But there could be many other causes.

If ECU is not operating, you will probably also exerience a mixture problem with hot or cold engine, although if you are lucky, the preset mixture will provide a good compromise and allow the engine to operate reasonable.

Try to measure the voltage across the idle speed air valve. In normal idle, it should fluctuate rapidly between 0 and 12 volts (constantly adjusting idle speed higher and lower).

Guenter Scholz skrev:

Jens, thanks for that. I wouldn't have guessed that it was an 'active' servo. Will put a scope on it tomorrow and see what it's doing

cheers, guenter

ps I know there are many with this difficulty who simply live with it, maybe a 'fix' or specific source for the problem can be discovered eventually

OK, please tell the result.

So you are a man with an oscilloscope and knowing about servo loops?

/Jens

Guenter Scholz skrev:

Yep, normal.

tiger, what's normal? the decrease in idle rpm by 200 rpm or so when shifting from Park or Neutral into Drive? Local dealer service manager who I think is very knowledgable thinks so.... some on this board don't seem to think so. You?

cheers, guenter

In park or neutral, the RPM is around 700 RPM... In R or D, it should be around 500 RPM. With AC on, it should maintain similar speed of 500 RPM.

..... I dabble. Finally warm enough today to look at the ICV behaviour.

- the duty cycle of the ICV appears to be 10 mS. With the ignition ON, but the engine not started there is 13V across ICV with a duty cycle of 9 mS and 0V for 1 mS. After about 10 seconds or so, the 13V disappears and there is only 0V applied.

- when engine is started the 13V/0V duty cycle is about 5 mS each.

- Then when engine is seriously beginning to stumble at idle the 13V duty cycle increases to about 6 mS (4 mS OFF) and idle smoothens out again and duty cycle returns to 50/50 ON/OFF.

.... so it sems everything is working as it should. Was a bit suprised that the ECU shut off the ICV after a little while when engine is not started .. smart I guess :-)

comments Jens or anyone?

cheers, guenter

I never measured it with oscilloscope (I don't have one anymore), so thanks for the results.

- ICV is controlled by 100 Hz with variable duty cycle

- Higher duty cycle gives higher "throttle"

- Normal idle is approx. 50% duty cycle (in your case)

- When you have the problem, it increases "throttle" to 60% to keep up RPM

So, the loop works perfectly.

But I'm not sure what to conclude further on. Too lean mixture could cause it, and that could be caused either by wrong setting of the mixture adjustment or by vacuum leak.

A vacuum leak allows the engine to draw in air, which is not measured by the air mass meter, and consequently too little fuel is provided for the actual air mass drawn (air through the air mass meter plus the leak). Engine would then tend do die, if it was not because the idle control loop tried to keep up the RPM.

Guenter Scholz skrev:

thanks for that Tiger, there was the notion that the ICV should keep the idle at 700 rpm if everything is working correctly. Judging by your comments as well as a senior MB service rep that notion is incorrect. Still, the car 'stumbles' and recovers .... it's not the plugs, so it might be distributor/rotor. Is there a reason they should be good at high rpm but not at idle?

- cheers, guenter

If you are talking about a lumpy idle, then that is not normal. As with older cars, things tends to wear out and cause problems.

First I would do is to take a carburator cleaner and spray any possible gaskets or rubber/vacuum hoses to detect any leakage. When there is a leak, you can detect change in idling characteristics. Fix these first.

Second what I would do is do a compression test to check the condition of the internal engine condition... the reading should be fairly consistent from one to another. I would pull out the fuel pump relay to do this test... and pull out all spark plugs... test one cylinder at a time... you want the reading at the third compression... third jump.

Thirdly, I would run synthetic oil.. My old 380SE was suffering severe idle problem and I was baffled... I knew my valves are the problem so I ran synthetic oil in the car and bam... after a trip around town, my idle was silky smooth. I think the hydraulic valve compensator/lifter were sticking due to all the gunks and residue from dino oil... or that my oil tubes to the lifters were coked (typical on V8)...

Then report back to me.