Dropping duty cycle
 

I admit that I’m in way over my head. I’ve been chasing a rich fuel/air mixture for weeks now and have yet to get a solution. This is the current status as it stands. The car is a 91 300e 2.6 from California. I have been checking duty cycles using voltage readings off the x11 port. In the past the readings after the ecu kicked in would peg at 12.7 volts and not drop at idle or 2500 rpms. I have changed out the ovp, O2 sensor, fuel distributor and throttle position switch. I seem to be getting correct readings from the air flow sensor potentiometer and EHA (but replaced as well). Finally, after doing all this and not getting any change, I went ahead and adjusted the lamba to lean out the mixture. This did lower the numbers to 6.7 to 7 volts, but now instead of holding at these levels, it slowly drops to .09 volts and holds. It will jump up to the higher numbers when rpms are raised, but again drop even as the higher idle is held. O2 sensor gives similar numbers, starting at .84 volts and dropping to .09v. Sometimes, at the leaner settings, the idle will surge and duty cycle will fluctuate back and forth till it settles and again starts to drop. When it does this surging, if I pull the connection to the idle control valve, the surging stops and idle smoothes a bit. With these setting, the car runs but has no power and will bog down at higher rpms.

Does this mean the ecu is bad? And if so, does anyone have an extra one I can borrow to swap out for checking?

In addition, the entire ignition group has been changed, new plugs, cap, rotor, wires and coil.

Perhaps you leaned it out too much? Try adjusting the the screw back the other direction just slightly.

BOSCH KE3-JETRONIC MIXTURE ADJUSTMENT

That's one of the guides I've been following along with the manual. I'll try bumping it up again and see what that does. Just can't figure out the steady drop.

My 190E did the same thing when it was at idle until I adjusted the screw again. I managed to get it settled around 40-45%.

Are you testing with the purge valve hose connected OR disconnected?

Heavy fuel vapors in the charcoal canister can cause RICH readings.

Purge line is plugged, both sides to be sure. One thing does bother me, its a California car and the manual states it has to be in diagnostic mode. Following the directions, I push #2 and get the one flash all clear. Pushing it a second time I am supposed to get a constant light. My does nothing, no blink, no constant light.

Tried to enrich it to see if that stopped the dropping readings. It did, but now pegs at 12.4v. If I lean it out, it goes back to the .09v drop. In both cases, if you rev the engine, it starts out at 6.8v and either climbs or drops.

Have been working on the moving duty cycle now for over a week. I just can't seem to find that sweet spot. It either drops to .09v or raises to 12.7v. The difference is no more that a click of a turn in the adjustment screw. I got it to hold between 6.9 and 8.5v for about a minute. While it was like that, I checked the O2 sensor. It seemed to be functioning, sweeping up and down in voltage. But after a little more than a minute, voltage dropped as before, setting at .09v on both the lamba reading and O2 sensor. After adjusting it clockwise a click, the readings eventually went back to 12.7v on lamba and .84v at the O2.

Going back to make sure I hadn't missed something, I rechecked the throttle position switch. The reading between post 1 and 3 seem correct, but the reading for post 2 and 3, show open no matter what the throttle is at. I get this same reading from both the original switch and the replacement one. That seems wrong to me. The readings for the air flow sensor seem to be within spec. I also checked the idle control valve just to be sure, and it seemed to be working.

Last week, I had to take the car to get gas. On the way, after about a mile, the check engine light came on. Checking codes it showed a 17 indicating O2 sensor. I assume this is because of fouling from running so rich. My question to this is, does this mean I need to replace this O2 sensor again or does it just reset so to speak? The readings from today makes me feel like its working.

To me, this has got to be a sensor or something. It all came on too fast to just be a matter of an adjustment. In open loop, it runs pretty good. Once ECU kicks in, it goes to pot. I haven't checked the coolant temp sensor yet, that's next on my list. Can anyone tell me what the correct readings for this are?

Check below the air flow plate to see if the fuel distributor is leaking into the air horn. If it is it can cause a wonky fuel mixture reading that will be impossible to dial in.

I'll take a look, but I did change out the fuel distributor already.

Took a look and didn't really see any fuel in the throttle body, but did smell raw fuel. Used a swab and it came out wet. Not a lot, but wet enough to tell it was fuel. How much can be expected.

I was able to get a replacement fuel distributor that came off a working 300e. Installed, and adjusted the duty cycle and it seemed much better. Held at about 50%. Let my son take the car to work, about 5 blocks from home. This morning, I checked on the car and found the CEL on and the code 17 for O2 sensor. Cleared the code and hooked up the dmm and found that once again the duty cycle is going rich or lean but not holding. Played around with it, got it to hold at about 50% again, but with higher revs, it goes to rich and pegs at about 12.8v. I jumped the O2 sensor and took the car out for a test. At idle, the sensor is swaying as it should between .86v and .01. But with acceleration, it goes to .95v and stays till the revs are dropped.

I don't understand why each time I make a change, it seems to run good for a while, and then goes back to this rich mixture. It also would seem to be quite a coincidence that both replacement fuel distributors would be bad to the same degree. Is the O2 sensor still good once it sets off the cel and I reset it. Could it be the EHA?

I am not going to declare this a fix, but I can say that measurable progress has been made. The problem all stemmed from a failed intank fuel filter. I got the idea from an old post from ps2cho. The thing had completely collapsed and was allowing sediment to enter the system. I had a spare to replace it, cleaned the tank as best I could. Put on a new Bosch fuel filter and started it up. At first the it went lean and then rich again, but after a minute, it steadied out and has held for almost a week. Still had a little hesitation, but that cleared up when I reinstalled the original OVP for the cheap Uro one I had picked up. I may still need to replace the fuel pumps, but it now runs and is actually drivable.