Tinker, et.al,

I've seen a few items that mess the minds of people who work on these fussy systems.

1. The injectors can look ok when tested exactly as described in the test procedures and still be worthless. Are you using an exhaust gas analyzer in conjunction with the mA readings at the EHA? Have you ever tried a comparison of the idle speed change caused by individual cylinder shorting? IF all compression numbers are relatively equal, the combination of these things can lead to at least an indication of whether the problem is related to the whole system, or the system is trying to compensate for one (or more) weak spots. For example, any injector that isn't getting the fuel atomized correctly leaves unburned fuel in its related cylinder, which affects the lambda, causing it to lean out the system.
Are you getting the same quantity of fuel through each injector? IF you make asetup like I have, that uses a known good fuel distributor, EHA, and pressure regulator as a test system, you can check both output pattern and quantity, something the CD doesn't give you. Run all 4 injectors at once, letting them spray into individual plastic ( clear is very helpful ) containers. Do a 20 - 30 second run with approximately idle deflection and measure the fuel. Then do the same at a mid and full throttle setting. THey should be within 10% of each other at any setting. IF not, try changing them among the 4 output lines to verify the problem is not upstream. Using this method lets you observe the injector pattern as part of the system, not just on a test stand.

2. Wiring issues: Check both input and output values at both ends of the wiring harness. If you're electrically inclined, you may be able to convert the various values from volts to amps and vice versa. Between these you may find, as I did, that corrosion and age had reduced the number of connected strands from the normal 8 -15 to 2 or 3: enough to get a reading, but not enough to carry the load.

3. Low cold idle generally means not enough fuel or not enough air, assuming mechanical and electrical issues are resolved. Make a fabricated idle air bypass, plugging the inlet hole and connecting the fabricated unit to the output. With a regulated air source ( compressed air with a control valve ) start the engine and vary the amount of air you let into the system. IF idle increases with more air, the fuel side is probably ok, and your problem is in the air inlet and management of to intake system. IF it decreases, the fuel side is likely the problem. IF there is no effect, take the adapter that gets the air into the intake manifold apart and clean it so all orifices are open and try again.

4. Uncalibrated timing light, or incorrect selection of timing marks or timing mark indicator. Consumer reports did a check a few years ago and found timing lights varied by up to 10 degrees. What would a change of 10 degrees mean to your system?

5. Leaky high voltage wires. Cold, wet engines are most susceptible to high voltage leaks. Big leaks can be seen in full darkness. Smaller leaks are harder to find.

6. One last thing: did you ever get a load of bad fuel, or do you know if that ever happened? Though the filters are pretty efficient, you can get enough stuff through to mess up the distributor, even plugging the littel filter cones.

7. Finally, try purposely enriching the mixture, so the mA readings at the EHA indicate the system trying to lean out the mixture. Start with about a 4mA change and check the effect.

8. If you're brave, and can visually examine the computer circuit board, use a 3x or more magnifier and look for cold or fractured solder joints. These can be repaired.

Richard,

I appreciate the insight.

1. The injectors and seals are new, but this doesn`t rule out a problem with the fuel distributor. I might have to invest in some graduated flasks to measure individual (cc) outputs. If the fuel distributor is bad though, the performance would be lousy, especially at higher rpm. This is not the case

2. I had the same thoughts and have started down this road. I have already checked continuity throughout the harness to the brain which was ok. But it could be a problem with the brain itself. I plan on checking readings again from the brain connector to the various components both cold and hot.

3. Cold start is not a problem. Its only after its at 80+ C degrees that the stalling occurs. I might have to rig up some manually controlled idle valve.

4. Timing is completely controled by the computer. You can`t even see the timing marks on the crank, there is no room.

5. It does need a set of new wires. If I pull off the wires one at a time while the car is running, there is noticeably less arc present on cylinder two. But there is no visible arcing or miss while the car is running. Its another $180 investment I need to make.

6. Fuel filter was replaced. I also removed the smaller screens from the fuel distributor. They looked pretty clean.

7. I tried playing with the mixture, on both ends of the acceptable range with no difference.

8. I did open up the CIS computer and look at the curcuit boards, but saw nothing out of the ordinary. They are suppose to be fairly robust.

If I drive it while hot, it will attempt to stall at lights. But, if I stop and keep my foot on the gas to keep it idling and then put it into neutral and slowly release my foot it will generally idle a nice steady 900 rpm. If I rev it at this point, it generally will not stall.

I appreciate your time and thoughts. I am going to give it a few more hours of my FREE time and if I fail, seek new professional help.

Tinker

OK, so some things are likely fine from my list. Let's look at # 3 on your reply. Stalling over 80C engine temp. Put an ammeter in the EHA circuit and watch what happens during this event, and compare that to behavior below say, 60C. Are you leaning out purposely ( lower and lower current readings )? After obtaining the readings, put a resistor in the coolant temp sensor line, say 250ohms, and try again. Did the problem go away? If so, the temp sensor or its related wiringis not giving a true picture to the computer.

Second part of the check: Vacuum readings at idle, cold and hot. The engine needs to draw enough air to keep the sensor plate deflected or it shuts off the flow of fuel. IF you have a temperature related vacuum leak, such as a cracked casting, or a valve with too little clearance, heat will affect its performance.

Idle switch, A/C switch, etc. Make certain you're actually activating the supplemental idle control switches. IF the AC cycles on without giving a signal to the computer, you'll lose 5hp, and about 100 - 200 rpm. IS it dead yet?

Your # 4 response: If you can't get to the factory timing marks when running ( it's difficult on a 2 valve, but can be done), you should still be able to see them with the engine stopped. Get to TDC1 and put a mark on the cover and pulley wherre you can get at them with a light. Also mark off a second line on the pulley at 10 BTDC by measuring the distance with a caliper and transferring it to your new marks location. Then you can perform a visual check of timing location. Timing delayed by 5 - 7 deg is enough to give the system problems in meeting idle emissions, so it will keep leaning the system out to try to get to a good lambda reading.

Your #1 Fuel distributor failures come in all variants. A little cross injector leakage, or a marginal bleed volume to the pressure regulator could result in idle loss. Are you really dead perfect on the two set heights (zero and stop )?

Your # 5: This is part of what I was talking about with an idle test. The best way to do this is with an electronic, digital tach and an electronic cylinder shorting system like many diagnostic systems have. You're looking for say, 150 rpm drop at each cyl. If one drops 10 and the rest 160, you've narrowed the search to one suspect.

IF you go to open mode ( bypass the O2 sensor ) what happens?
Have you done the two primary O2 sensor tests?

Richard and all curious others,

FIGURED IT OUT!

It only took about six months!

I reset the air flow meter`s resting height. It was too low causing constant pressure on the fuel distributor`s plunger.

I also reset the throttle body resting stop. I believe that these DO wear causing air to bleed past resulting in a high idle. My problems initially started after I cleaned it.

So far (tonight) I have beat the heck out of it and can`t make it stall. The wife is amazed. Starts on the first crank of the key with no throttle.

I`ll keep you posted if it acts up again.

Thanks for all the thoughts.

Tinker

I have been following this thread for a while...thanks for keeping us up to date. I haven't heard anyone else with idle problem do this but this would make sense.

Please do post again if it starts acting strange...but for now good luck and enjoy your car.

__________________
'87 MBenz 190E 2,3 8vlv

Tinker,

Isn't it amazing that things we thought we checked once turn out to be in need of adjustment. Congratulations.

Richard

Also, after congratulations, note that I referrred to that very problem on 10/31