Floobydust

I was helping a friend diagnose a high idle problem with his M102 KE Jetronic car the other day. Since I have an M104 with HFM fuel injection, I did some searching of the forums for info on IAC troubleshooting. I found a really mixed bag of sometimes contradictory information on how to troubleshoot and diagnose problems with the IAC. The problem with his car turned out to be a relatively straight forward sticking IAC valve, but I got curious about why there was so much differing information on the IAC.

So I took his old IAC, cleaned it up, a decided to see how the valve functions. What I found was very interesting and may help explain some of the information that's out there.

The IAC valve itself is a rotary valve with a galvanometric movement that provides rotary motion in response to an input control voltage. The valved allows metered air to flow from the FI plenum to the intake manifold to provide idle air. Since the air is "metered" by the air flap, fuel is added to this air by the injectors in proportion to the amount of air flowing. Note that any air that leaks into the IAC circuit from cracked hoses or bad connections is not "metered" and will only lean out the mixture and cause a poor/rough idle.

Operation of the IAC would be very straight forward if zero volts corresponded to zero air flow through the IAC, and if air flow increased in proportion to the control voltage. But it turns out it doesn't! Air flow through the IAC is, in fact, non-linear, with about 20% flow at 0 volts, no flow around 4 volts, and flow increasing with voltage beyond that. This characteristic means that with no control voltage (e.g., the IAC is electrically disconnected), the engine will receive some idle air - enough in some cases to cause a high idle. A bad OVP relay will also cut power to the IAC and cause a high idle. Conversely, an excessively high control voltage will also cause a high idle situation. It appears that during normal, warm engine idle the IAC control voltage should be somewhere around 4 - 6 volts and will probably vary depending on the condition of the engine. Cold engine high idle should be around 6 - 8 volts. The DC resistance of the IAC coil should be about 8 ohms.

I have attached the airflow (derived from cross sectional area) versus control voltage characteristics that I measured along with photos of the valve position versus control voltage. I haven't figured out why Bosch designed the valve this way, but perhaps it may be that the no voltage air provides for better starting characteristics. Whatever the reason, the valve characteristics make diagnosis a little more challenging.

I hope this information is helpful for someone.

- FD

idontlikepeople

very nice, thank you

PhilDoc

Thank you so very much! I've been searching the web for info on this for weeks, to no avail. My car has no registering fault codes; the Throttle Valve Microswitch, Coolant Temperature Sensor, and Air Mass Potentiometer all tested normal. The warm-up idle is normal at 1100-1200 RPM. And it runs great... except for a 1600-1800 RPM idle when warm.

Disconnecting the electrical connector to the Idle Air Control causes the car to immediately idle normally at about 650 RPM. But then, of course, it gives a #18 DTC Fault Code of "Illogical value for Idle Valve".

I removed the Idle Air Control thinking that it must be sticking and was surprised to find that it was partially open, about 1 1/2 mm, just as your picture shows. I assumed this was where it sits when I disconnect the electrical connector. But when I measure the voltage at fast idle I find 8 volts - which, according to your picture is almost wide open.

Now, at least, I understand what is going on. All I have to do now - and I don't expect that this will be easy - is figure out why the ECU is sending such a high voltage. I'm guessing it's not a faulty ECU and more likely a faulty sensor, possibly the Oxygen Sensor which I have not yet checked.

Still, my ECU sends no fault codes to the diagnostic module, so I don't expect to prove the Oxygen Sensor faulty.

Thanks again for your help! I should have had the good sense to acquire a variable voltage apparatus to test the thing myself.

Floobydust

I'm not really an M103 guy, but I would suspect that the airflow pot may be miss adjusted. Here is a link to an article on how to adjust it: http://www.sg-motorsports.com/AFS%20Pot%20write-up.htm.

Good luck.

- FD

Vincenzio

Sorry about digging up an old thread and not first introducing myself, but I just stumbled upon this thread searching for a solution to some idle problems with my engine, using the same type of Bosch ICV valve.

Firstly I want to share some knowledge I do have, which does answer a question posted above. The reason for the strange opening curve is not that mysterious, it provides a base setting so that if the ICV fails it will position itself in a fail safe aperture that still provides a reasonable idle speed.

Also, by my M2.5 ECU the ICV is not controlled by a DC voltage, but is pulsed by PWM to control it's position. It is also controlled only in one direction, the other direction is controlled by a spring. Later models (3-wire) had an opening and closing PWM connection, which is rendered superior to the 2-wire valves and a lot more accurate.

I'm now having issues with decelerating and/or gear changes. When I let go of the throttle and decelerate the ICV is sent open for a split second after a second or two, which is quite unpleasant as it shakes the car for a bit. Also, when I change gears, or let go of the throttle and keep the clutch in the engine revs down to near stalling, then the ICV is sent open and makes it rev at almost 2000rpm for a second. Quite annoying! Probably it's the same phenomenon, with or without clutch, basically when I let go of the throttle it's opening the ICV for a blip after about two seconds. I'm not sure what the reason is, maybe the mixture is too rich and the lambda triggers the ICV to open position. I've pressurised the whole intake and there is no leak, all the vacume operated parts work fine.

I've thought about adjusting the grub screw at the side of the ICV, to enlarge the zero voltage gap, maybe that solves it.

Any ideas about this would be great!

PhilDoc

You must forgive me. I am not a mechanic and do not know much about these things. I can tell you, however, that I went to my Mercedes mechanic three times to have this problem fixed. Each time he replaced parts that did not need replacing, at great expense, without either diagnosing or fixing the problem. The last time alone cost me upwards of $800. He's a great guy, but absolutely no good at diagnosis.
Finally, realizing that I would either have to figure this out myself or get a new car, I went to the internet, downloaded a bunch of info on how the system worked, read it, learned it, bought a $40 multimeter, and figured out what was wrong within about 5 minutes of working on the car measuring voltages and resistances.
It turned out that I had a failed Throttle Position Sensor / Throttle Valve Switch - easily determined by using the multimeter. The place where the wire attaches to the switch had been bent so many times (presumably, by my mechanic working on my car, since I never so much as opened the hood to add oil) that it was coming apart.
There is much confusion in the literature about the Throttle Valve Switch (Throttle Position Sensor), which measures whether the throttle butterfly valve is actually open, and the Throttle Microswitch, which measures the Throttle linkage position; and the terms for the two completely different switches are often, incorrectly, used interchangeably. So reading the information available from do-it-yourselfers is not much help.
The bottom line was that the part is no longer available; my mechanic had one; I couldn't do the work anyhow; so I paid him to put the thing in. Now all works perfectly.
As with most things in life - and completely contrary to what physicians and most mechanics do - if you want a good outcome, the first thing to do is to UNDERSTAND THE PROBLEM. Treating a problem you don't understand is a HUGE WASTE OF TIME AND MONEY, and is not likely to help anyone but the receiver of the money.
Fortunately, the internet is a great resource that provides Official Mercedes documents on how to diagnose and repair just about everything. All you have to do is pull it up, READ IT, and make sure you understand it.

Vincenzio

Thought it would be a good time for an update as I've found out my engine works a LOT better with the valve set to beyond the closed point. Every other sensor, switch, cable etc. etc. has been replaced by brand new packed, or through the exchange program, all original Bosch material and though that did give some improvements, and cost a fortune, the idle quirks were still there every now and then.

With the valve trimmed to set a good 900 rpm idle (normal idle is at 1000 with my engine) to still leave some adjustment range for the ECU the idle speed never first drops to a near stall. It just goes down gradually to stabilise at the programmed speed of 1000. I can't understand at all why the valve should even have a fully closed point, neither why it should be open at zero volts, then first close. I've still got that small opening now when it is unplugged, but then on the other end of the valve, where it can only move to further open.

I've talked to a technician at the dealership and he too agreed it runs and drives so much better now, that he could only conclude the base setting made by Bosch is a mistake!

Hope this helps, and for those who want to try, beware not to damage the adjustment screw, you need to precision heat the thread locking compound, so the valve doesn't fully heat up. Then with a small screwdriver scrape out the compound until the allen key fits in nicely. Then off course adjust it when unplugged to a lower idle than normal, to leave some adjustment range for the ECU, and then use some nailpolish or whatever to lock the screw again.

Enjoy!