Dear All,

I have a 1992 W124 300e with CAT and oxygen sensor, EU version and it's produce an annoying issue which I cannot figure out.

So my problem is when I start the car cold the duty cycle sits around 50% as expected because it is in open loop. Once the engine is at operation temperature the duty cycle starts increasing and stops around 92% there no really any fluctuation. It takes a few seconds to reach the 90%. Once I rev the engine to 1500 RPM to duty cycle changes to a 50% without any fluctuation so I am assuming it goes to open loop again. Once I release the pedal it goes back to around 92% within few seconds.

I can smell the mixture is rich and my fuel consumption is around 23 liter / 100 km (~10mpg) and there is a lot of smoke from the exhaust. When the car is cold there is almost no smoke from the exhaust.

Anyway the car starts nicely, little bit rough idle but nothing serious (probably because the rich mixture). Maybe I can feel a little hesitation when I accelerate strongly but again I am suspecting the rich mixture.

What I checked so far:

- Duty cycle when the ignition on - 71%
- Duty cycle when the ignition on and air plate pressed - 11%
- Duty cycle when the ignition on and gas pedal fully pressed - 21%
- Duty cycle when the engine running cold - 50% (no fluctuation)
- Duty cycle when the engine is warm - 92% (no fluctuation)
- Duty cycle when the engine is warm RPM 1500 - 50% (no fluctuation)

- Fuel cut off switch works signal reaches the KE control unit
- Throttle position switch works (both idle and acceleration), signal reaches the KE control unit
- EHA currents seemed ok

- Tested with other KE control unit (used)
- Tested with other EZL ignition unit (used)
- Tested with other EHA (used)
- Tested without the cold start injector

- Tried to adjust the mixture but nothing changed

A lot of things were replaced in the last ~5000km
- Ignition cables
- Ignition cap / rotor brand new
- Ignition coil brand new
- Spark plugs
- O2 sensor replaced multiple times,
- Injectors
- Over protection relay (after market)

Is vacuum leak can cause this kind of symptoms?

I would appreciate any suggestion because at this point I am lost.

It's not clear to me exactly how you did the measurements, but look at the following Web page. I used the procedure 20 years ago to check the duty cycle on my California model 190E 2.6 and I recall it worked. The duty cycle was within spec and needed no adjustment.

https://www.landiss.com/mixture.htm

Duke

I believe a vacuum leak could cause this and the suspect areas are the boot under the fuel distributor, the IAC hoses and the 90° elbow from the intake to the valve cover. As good measure replace the hose from the top of the air cleaner to the valve cover and the oil dipstick O-rings unless they are excellent. None are expensive but can be tough to install.

IIRC Duke has dug deeply into these systems.

Good luck!!!

__________________
"Rudeness is a weak man's imitation of strength" - Eric Hoffer

Duke: I don't have a multimeter which can measure duty cycle so I measure voltage on the X11 socket and calculated the duty cycle from the voltage.

Sugar Bear: All those hoses are looks good to my eyes. The o-ring around the dipstick was replaced so should be good. There are some joints around the thermovalve which are bad so I ordered all of the small joints around the KE jetronic vacuum and will replace them once arrived.

Today I checked what happens if I unplug the O2 sensor and absolutely nothing changes. So the duty cycle slowly goes up to 92%. In this case the duty cycle should be at fixed 50%.

I will try to figure out why this is happens, but I would appreciate any idea.

Consider that the input signal(s) that change slowly correspond to change in engine, and/or air, temperature. Bridging the temperature sensors will, at the least, fix their values. This is a bit of a long shot, but may indicate something.

I had a 300ce do something like that ,i had a lambda meter hooked up and saw it going rich , it was the ignition coil , when the coil got hot the spark was weak , black smoke coming out the tailpipe, replace the coil and fixed the problem, the fuel wasnt able to burn due to weak spark when coil got hot

Frank Reiner: In my case the slowly means 5-10 seconds. Yesterday I also measured the CTS, the values are within specs. If I unplug it I get static 30% duty cycle which is correct.

nulu: The ignition coil was already replaced twice, pretty much the whole ignition system is new.

I am still suspecting some kind of electrical issue, because unplugging the O2 sensor should give static 50% duty cycle. But again pretty much all of the electrical parts are checked.

Frank Reiner: In my case the slowly means couple of seconds (5-10 seconds). Additionally I measured both the CTS and IAT sensors and both are in spec. If I unplug the CTS I get static 30% duty cycle which is again correct.

nulu: The ignition coil already replaced twice, because it was suspected. So probably I have 2 good working ignition coil.

However I made same progress today:

I am not sure if you are aware but the O2 Sensor cable is shielded after the connector under the passenger seat (the single barrel plug with green wire). This is because the distance between the lambda and the KE control unit fairly big and the lambda V is fairly low (between 0 and 1 V). To not pick up any noise this cable must be shielded.
The point is, within the harness somewhere the cable shield is separated and goes into the KE control unit pin 7 (the O2 sensor signal goes to the pin 8). The pin 7 actually goes to the ground. And in my case between the pin 7 and pin 8 there is a short circuit.

I have a spare KE control unit so I took it apart and cut the pin 7 (in this case the shield is gone but the the O2 signal should be "intact"). I warmed up the car with the original KE control unit and once I could see the annoying behavior I swapped out the KE control units. I didn't got the result what I expected but at least I could see different duty cycles and the car run horribly.

Once I have time I will try to adjust the mixture to see if it sees the lambda signal now, but at the end it was dark and cold so I gave up for today.

My prime suspect right now is the engine wiring harness. It will be a challenge to replace it or even to find a new one.

The signals that have the largest influence on fuel flow are from the:
- Air flow sensor (mechanical)
- Air flow potentiometer (electrical)
- O2 sensor (electrical)
- Manifold absolute pressure (MAP) sensor (electrical)

If the MAP sensor is sending a signal that corresponds to WOT (high MAP), the ECU will call for high fuel flow. The MAP is in the EZL, connected to the manifold with a flexible pipe. Is there a break in the pipe or the rubber connectors?

The first step is to confirm fuel pressures are in spec, along with fuel volume. While the gauges are attached and the engine is at warm idle, rev the engine- the pressure must not dip very very much. This tests that the fuel pressure regulator is doing it'd job , keeping constant pressure. ( you could have a blockage in the return line too)

Check that the throttle switch is closing.
Check the fuel distributor plunger isn't leaking.

I would also pull out the fuse in the OVP and see how it runs. This will test to see if its a electrical or mechanical problem.

You can mechanically set the mixture by reading the O2 sensor voltage.
Unplug the O2 sensor and read DC volts. Adjust the mixture screw so that the O2 sensor reads ~.45 volts dc fluctuating. between .1-1 volt.

At cold start does the engine rev ~1100 rpm then after 28 seconds settle in at 750?

__________________
1993 2.6 Sportline 5 speed
1992 2.3 Canadian delivery

The suggestion by DK (above) that there may be a blockage of the fuel return line is worth investigating. The time frame of 5-10 seconds to reach 90% duty corresponds with the fuel pressure rise due to a blocked return. It also would correspond to a fuel pressure regulator that is not opening, or is only partially opening.