Worked on 300E, now it stalls with idle valve connected!
 

I did a bunch of work on our new 300E over the past couple days attempting to solve a cold-start problem.

Symptoms: Car is very hard to start when cold, but starts ok when warm, once started and let run a minute or two, it will never stall out. Idle is very low (500) at all times.

Those were the original issues.

I replaced the idle control hoses (NOT FUN), cleaned the idle control valve, checked all sensor connections, installed a brand new OVP relay (old one had blown fuse but they wrapped wire around the base! :eek:)

I get it all back together, begin cranking....nothing, nothing....battery weakening....so I hit the gas during cranking, it fires Right up and holds like 2,000 rpm with the pedal, I hold it around 1200 or so for about a minute, let off the pedal, it dies. The engine coasts to a stop, then the fuel pump shuts off about 2 seconds later. :eek: I restart, it will fire, rev, die. I start using pedal, starts right up, I can hold the idle as low as 700 using pedal....let off, dies. I'm ready to set it on fire and get a diesel....ok maybe not yet. In desperation I unplug the idle control valve, turn key, fires right up! Idles at 750ish, smoother and more steady than ever before. In drive it lowers to 600. Check engine light comes on. I take it for a drive. It runs great, idles smooth, acceleration is strong.....0 issues. Come back, turn off for 10-15 mins......start it back up, fires right up and idles around 700.....I put it away for now, and we'll see tomorrow evening if it will cold start properly....(doubt it).

What did I do?? :o The engine wiring harness is Ultra-brittle....but everything seems to be plugged in ok (aside from idle valve). Why would unplugging the idle valve stop it from stalling?

Could my starting/idle issue be a weak coil? I don't want to throw parts at the problem.....that'd only dig the hole deeper. :o

With the valve disconnected your idle should be much higher...1500+ When not powered up the valve defaults to full open which should give much more than 700 RPM. I'd suspect a general running problem that is exacerbated at idle. Hard to diagnose what without being under the hood but I'd start by getting a steady idle by whatever means and listening closely for a vacuum leak.

So if it is only idling at 800ish with the valve disconnected then their is another issue at hand? What could that be? The car litterally drives perfectly with the valve disconnected....better than it used to WITH it connected. :dizzy2: Not sure on the cold starts as I haven't tried yet....obviously its not operating as intended.

This is really odd. You may try with the air cleaner off pushing down on the fuel door and seeing if the idle goes up. This would indicate that the fuel mixture is too lean.

I had same problem which was cured when I changed the the fuel computer.

The fuel pump is kept running by a signal from the ignition, if the engine dies the fuel pump shuts off. That is normal.

I believe that this would indicate the engine was running too rich, an easier way to check would be to unplug a manifold vacuum line to introduce more air and see if it runs better = too rich. Running worse with extra air indicates too lean. Gasoline engines are sensitive to air leaks unlike your diesel, and to temperature-dependent mixture settings.

It is difficult to diagnose completely from symptoms, often a sensor out-of-spec will cause mixture issues, would need to be diagnosed with an ohmmeter/DMM vs temperature using your FSM. Can be the fuel computer also, they do go bad, I've had two replaced under warranty from driveability issues.

It is unlikely IMO that it is the coil based on your symptoms.

If you can retreive the fuel computer PN, I will look to see if my spare is the same, which would be a 5minute swap, and one way to eliminate that from the mix.

I don't know what years the wire-harness was bad, my '91 was only 5y/o when I sold it, wiring was fine, others here very likely can tell you if yours is a suspect year.

Hot-wired fuses/relays are never a good thing, did the fuse remain good after replacing it? Only the fuse was bypassed, not the relay?

Check also the wiring to the O2 sensor, an unplugged O2 sensor will cause the engine to default to mechanical-injection settings. I don't think it'll affect idle though.

How do the engine grounds look? When it doubt, clean it. Weak grounds can cause you to chase symptoms forever, a couple of millivolt leak to a sensor ground can signal a whole different parameter. Ditto with the fuel computer's ground.

The grounds and wiring connectors all look ok, the car is a desert car, so there's no corosion on anything...

Only the fuse was bypassed....I haven't checked to see if the new relay-fuse has blown. The relay itself looked old/original.

I don't have a FSM for the W124....I have been going by what it says for the 1990 300SE W126.....I'd assume the engine/mechanicals are the same, it looks to be. I have a good multimeter/ohm meter. I guess the next step is to verify all sensors to be in spec....

I am pretty sure the cold start injector is working, as there was some fuel residue on its tip when I removed it. After the engine has been running/warming up for about 3-4ish minutes, there is a "click" sound and then it seems to idle a little slower (only slightly) but its noticable. What is occuring when it does this?

I will try to get the part number off the fuel computer, that is the larger unit in the back right? Does the fuel computer control the idle?

I'll also try disconnecting a vacuum hose with it running to see what happens.

Also, when I finally got it started the exhaust was smokey and smelled like unburned gas. Running way too rich I'd assume??

When just sitting there idling with the valve unplugged the exhaust smells normal and clean.....and it runs/accelerates great. If it starts good cold then maybe I will just leave the valve unplugged. :eek: :D

Sounds like the cold-start injector is working. You can verify by putting it in a jar and watching it when you cold-crank the engine.

I'd also guess that the SE engine controls are identical to yours.

I just hope that by the time I get this all sorted out I am not in need of a new starter! :eek:

Linkage could be stiff. Mine looked ok before I removed it and lubricated the joints. It's worth the 30 minutes work before buying more parts.

What did the CEL code indicate?

How do you read the codes? I know where the diagnostic plug is and the little button + red LED, but I don't know how to use it.

The linkage seems to move ok, I also added lube when I was working on it.

Search engine is your friend, got this from a post:

The engine needs to be off, but the key must be in the "on" position.

Hold down the button for 3 - 4 seconds and count the blinks. The number of blinks is the code. Repeat the process again. If you get the same number of blinks, you have just one code. If you get a different number of blinks, then you have more than one code.

When finished, turn the key off, then back on, and hold down the button for 6-8 seconds. When you start the car back up, the CE light will be off.

The only codes that will cause a check engine light to illuminate are codes relating to smog equipment. Your ECU will hold a lot of codes that will not cause that light to come on.

91-93 190E and 300 series (2.6L and 3.0L)
1 No faults in system.
2 Throttle valve switch(full throttle contact).
3 Coolant temp sensor.
4 Airflow sensor potentiometer.
5 Oxygen sensor.
7 TNA (RPM) signal.
8 Altitude pressure signal from EZL Ignition control unit.
9 Current to Electro-hydraulic actuator.
10 Throttle valve switch(idle contact).
11 Air injection system.
12 Absolute pressure valves from ELZ ignition control unit.
13 Intake air temp signal.
14 Road speed signal at CIS-E control unit.
16 EGR.
17 Oxygen sensor signal.
18 Current to idle speed air valve.
22 Oxygen sensor heating current.
23 Short to positive in regeneration switch over valve circuit.
25 Short to positive in start valve circuit.
26 Short to positive in shift point retard circuit.
27 Data exchange fault between CIS-E control unit and EZL ignition control unit.
28 Loose contact in coolant temp sensor circuit.
29 Difference in coolant temps between CIS-Eunit and EZL unit.
31 Loose contact in intake air temp sensor circuit.
34 Faulty coolant temp sensor signal from EZL unit.

The code will immediately or eventually come back if you don't resolve it.

My linkage looked fine too but complete removal, cleaning, and lubing the pins, joints, and moving parts resolved my idle issues. Merely spraying it with WD40 did not.

My car is a 1990, manufactured in Nov 1989....will that make a difference?

I verified that my linkage moves smoothly....and clicks the microswitch properly......so thats probably not the issue.....

Thanks for the code info....I will check that out today....

If your idle control valve is anything like the one on my BMW this might work. Remove the valve and spray into it with carb cleaner. mine was stuck in a partially opened position due to gunk. I had to spray it about three times before it loosened up.

Also try putting it back on and then disconnecting the O2 sensor. If the engine runs better, the sensor is shot.

I tested the idle valve when I had it off, and it operated properly when I applied 12V to it.....doesn't seem to be stuck, I cleaned it out a few times with carb/throttle cleaner too....

Ok get this, I went and started it tonight, outside temp is about 20F or so.....I cranked, and in under 2 seconds it fired off, with no throttle applied! It was running around 550rpm, but was missing a bit (shaking slightly from time to time) but evened out in maybe 30 seconds to a minute.....then ran fine. Idled at 650-700 or so, and I took it on a longer drive....ran great the whole way.....and the exhaust doesn't smell gassy like it did before.

Note: This is all now that I have Unplugged the Idle control valve!! This makes no sense! Warmed up and having been driven a while it idles around 750-800 in park/neutral with no loads on the engine. Idles around 600-650 with A/C comp or heavy electrical load (or both) And idles around 600-650 in drive or reverse. When driving it is responsive and feels perfectly normal and smooth.

How is this possible?!? :dizzy2:

It still needs a coolant flush with MB coolant + a new thermostat.....and I might just replace the distributor cap/rotor/coil for the heck of it.

IF you would like to be educated on your cars fuel injection i can send you documents on it.

Jsut curious did you check your fuel pressure and for internal leaktightness?

That would be much appreciated. Which documents do you have? How would I check the fuel pressure? I have none of the required tools/equipment for that.....

Still doesn't explain why it starts and runs ok with the valve unplugged.....but stalls with it plugged in. Why?

the fuel injection and idle the 300E is controlled by the PCM not by a control valve and I can prove that.

To check fuel pressure you need a fuel pressure tester

http://www.sears.com/shc/s/p_10153_12605_00902169000P

Whatever you disconnected to make the problem go away was not the idle control valve since the system doesnt have one. The round part with 2 prongs is the idle speed adjuster.

Also check this:

test coolant temperature sensor
test starting device of KE fuel injection system

I need your email for those docs btw

My 300SE had the same exact symptoms after my coolant temp sensor broke. Seriously, the same EXACT symptoms except I never tried disconnecting the idle circuit.

But I would have to hold the gas when starting, make it rev to 2K for a couple minutes, then it ran fine.

Worth a check, very easy to test. I just posted how to test this in another thread.

-tp

Coolant temp sensor has been replaced. No change. Trying to start with the valve plugged in results in it struggling to start, but does if you push the throttle a bit, but it won't idle. Letting off throttle results in the engine shutting down. Disconnect valve, starts and idles fine....with a sliiiiiight miss every now and then.

New development. Temp sensor does do something, connecting and disconnecting it while car is running makes the switchover valve trigger/not trigger.....doesn't affect how the engine runs, but it does do something. :D

I wondered about the intake air temp sensor.....so while it was idling cold (and missing a bit) I unplugged that too, voila....idle went up by 100-150 or so (to about 650) and smoothed out, no missing at all. Smooth as silk. :eek: Whats the deal here?! This engine seems to like less and less electronics to be active! As a result of these two things unplugged the car runs smoooth and nice when driving, idles great, starts great....everything great. Very annoying. Check engine light on. Diagnostic light gave code 18 (idle valve illogical signal) and now code 13 too (intake air temp value illogical) :D Makes sense sure, but with them plugged in it runs poorly or not at all.

Plugging in the idle valve while the car is running does not affect it....this may be because it has already disabled the connection due to no signal though....so it may be ignoring it.

Plugging in and unplugging the air intake temp sensor results in lowering idle and a bit of missing (when plugged in), to smooth and slightly higher idle when its unplugged.

Help?! I'm probably going to just leave it as is......I'm doing a coolant flush and thermostat replacement this weekend and possibly a trans fluid replacement, but other than that I may just leave that stuff unplugged and leave it be for a while. Perhaps later on I'll have the dealer diagnose it. :o

Did you read Lyle's docs, they explain much of each components functions?

No.....where might I find these?

I have the W124 bible, and understand how the system works....but that still doesn't explain why when everything is plugged in it doesn't start/idle correctly. (With no computer codes by the way)

Initially it would start with about 10 seconds of cranking with everything plugged in, then I did some work and replaced some tubing and such, and then it would not start unless I unplugged the idle control valve. It runs so nicely with that and the intake air temp sensor disconnected that I may just leave it that way, and have the dealer diagnose it sometime in the future when I have more $$ to throw at it. :D

The short version is that the CIS-E is basically the mechanical CIS (Constant Injection System), with electronics added to control emissions via a frequency-valve (can't remember what it is called here) on the side of the fuel distributor. With a fault in the system, such as un-plugging the O2 sensor, it will revert to the mechanical system where the air-flow sensor plate moves a metering pin and directly changes the injection flow rate (all six cylinders, all of the time, at the same rate).

You are indicating through your experimentation that the basic CIS including mechanical parts such as the fuel pressure regulator, and the fuel pump, the injectors and injection distributor, seem to be operating correctly, and likely (not certainly) within spec. When the electronic brain is put back into the loop (plugging in all components), it fails, which indicates that somewhere in a sensor, the computer, the wiring and grounds, there is something out-of-spec. There are restistance vs temperature values for the temperature sensors, and I believe values for the O2 sensor available and the sensors can be measured with a good DMM/VOM. They can also be bypassed using a resistance-substitution box and a few BOSCH connectors cut from junkyard cars to see if the proper resistance will create the proper running condition.

It is my further opinion, that the ignition system is in adequate condition based on your experimentation and previous postings.

Check the sensors, a couple of sensors out of spec can really make a car run poorly. I'm betting someone here knows the correct resistance for that intak-air sensor for example, you'll need the temperature at which you measured it BTW.

He offered to email them since they are large documents.

If you are happy the way it runs, you're done. If you want it to run how it was designed then you have more work. I didn't like that the prior mechanic unplugged my ICV and added a helper spring...

pawoSD, I would resolve the idle control issue before tackling other maintenance items on your 300E; otherwise you're liable to complicate the situation further.

The only other maintenance I am doing is fluids/thermostat (runs too cool).....with the electronics taken out of the loop it runs fine, and you can't even get it to stall if you try. (even when cold, bumping the throttle, putting a big load on it (electrical, A/C, ect...) it still won't stall)

I'll keep checking sensors....I'm guessing the air intake one is bad...its not in very good shape physically....however the engine temp sensor is brand new. What other sensors should I check? Or, what others are there that could cause that problem? The air intake sensor is cheap, I'll pick one up sometime in the future and try that....perhaps the computer itself has issues....I'll figure it out eventually....

" I need your email for those docs btw"

"Did you read Lyle's docs, they explain much of each components functions?"

Never got his email addy so I figured he didnt want them anymore.

But your stalling problem is the reaction of the pcm taking the inputs of all the sensors and devices and taking that information and making adjustments to the A/F mixture. its apparent that one of your sensors or part of your fuel injection is malfunctioning.

"thermostat (runs too cool)" is heading in heading in the right direction because if the thermostat is not working properly it will the pcm to run rich and stay in an open loop when the system should be in a closed loop.

Did you pull your plugs and see the color of them?

If you suspect the intake air temp sensor, and don't feel like continuing to order/pay/wait for each test, look up the resistance value for that sensor at (for example) 0C/32F, buy a reisistor at Radio Shack of that resistance value, plug it into the wire harness and see how it runs. I'm not a big supporter of buying parts until it runs correctly, that's what Betten does for me at ~$100/hour.

I figured its a sensor issue, but I am not sure what all sensors it is looking at (how many are there that play a role?).....one is brand new, and the other is known bad...I'll replace it sometime in the near future. It runs sliiightly rich, but not real bad. Exhaust looks clean but has a slight rich smell to it....much better than it used to be though. Spark plugs have been replaced with the correct non-resistor Bosch. Old plugs were junky champion resistor style, they were a bit oily and dirty (as was expected)....but the engine does not burn/lose any oil. I've driven it nearly 1000 miles since its oil change (M1 0w40) and it has not lost any.

The thermostat could explain the slightly rich running, but not the cold start issue.....as its cold anyways. At running temp right now it hits about 67-69C or so.....tomorrow I'll fix that.

As temp sensors go bad, the resistance tends to drift upwards, which is indicative of a "colder" temperature.

If this is what's happening (likely), the computer thinks that the car is getting more air intake than it is, since cold air is denser than warm air.

If the computer thinks that it's getting nice thick cold air, it will increase the amount of fuel delivered to keep the mixture correct. But, if it's just a faulty reading then you wind up with a rich condition.


-tp

Still doesn't explain why it would have problems cold starting with everything plugged in. It starts warm fine and runs fine when warm.

Doesn't matter, it runs fine mechanically. :D I'm not going to worry about it for now.....any car that can pull 107mph in 3rd gear at 6000rpm counts as running GOOD in my book. :D :eek:

Swapped in a new intake temp sensor and idle control valve (both from the junkyard), problem solved! It fires right up now and idles smooooth! Not bad for $10!

I did throw about $250 of parts at it prior to this though. :D :D :D