Edited with a better summary of what's wrong, and more tests i've done. Any CIS experts please chime in!


I'm pretty stumped with this problem i'm having with my 1992 N/A 2.6 190e. In the battle of trying to figure out why i can't pass emissions, i found something quite weird with my car while looking at the mixture at the X11 Diagnostic port.


According to BOSCH KE3-JETRONIC MIXTURE ADJUSTMENT, the static duty cycle should be 70%, with ignition on engine off. which means the voltage between pin 2 and 3 on the X11 connector should average around 4v.


1) Yet I'm averaging around 8v, which is around 33% duty cycle? What could be causing this?


If I open up the throttle, the reading would increase to 10.5, and drop back down to 9.33 at WOT. This is mostly normal.


2) However when i deflect the air flow plate with ignition on, engine off, the duty cycle does not change at all. What could be causing this?

- The airflow meter tests fine, with smooth rise of voltage to 5v, all within spec.
- The connection between the AFM and the ECU is also good
- The throttle valve switch is also good, both at idle and full contact
- The connection between the throttle valve switch and ECU is also good


Other symptoms:

When I open up the throttle, i'm suppose to see momentary enrichment, the duty cycle should temporarily drop (voltage goes up), however it actually does the opposite, and leans out.

With the engine warm and idling, my duty cycle is around 8v or 30% (rich). However when i rev the engine to 2500rpm, it drops down to 50%.

If i adjust the duty cycle to 50%, or 6.8v at idling, when i rev up the engine to 2500, the duty cycle will lean out to 60%. And if i rev it hard, the duty cycle will lean out to over 90% briefly!

My emissions report also reflect this. I failed on NOx (lean/hot combustion), and you can see it spike everytime the car accelerates.

Any help is appreciated! Sorry for the long post.

bump: updated with more info

Look for a vacuum leak. The hoses from the rubber boot under the air plate that go to the cold start injector can get hard and leak at their connections I had to replace the rubber boot itself. You can also have bad injector seals or o rings . My lambda would go lean when ran the rpm up toward 2000

Yeah I'll look into that, but that doesn't explain why the static duty ratio with ignition on engine off is still 30%, and that it doesn't change with AFM movement...

I'm really hoping it's not the ECU at this point

oooo new development!

Ok so nowhere has i ever read that the static test should be done with the engine hot... but I took a long drive and tested it, and it's giving me different readings!

I've done tests after idling for 15min, but I guess that didn't get the coolant hot enough, even though the temp gauge was in operating range already.

So now, with the engine fully hot, I'm getting 6v (50%) static, and the AFM is changing the duty cycle, however only up to 7 volt (40%), not to 10% like spec.

this just gets weirder and weirder...

I always thought it seemed wrong that the procedure says to do the static check of moving the airflow sensor plate to get 10 percent reading. If not check throttle valve switch. Then open throttle should show 20percent if not check airflow sensor potentiometer. It seems as though these two are mixed up. I would think that the airflow plate check should coincide with check the airflow sensor potentiometer and the throttle check should go with the throttle valve switch check. I however am not even close to being an expert on this . Just seems strange. I also found that if I set lambda to get a 0 center point for eha current I do not get 50 percent lambda. I think it was around 60percent. I set it slightly rich from there because idle was a little rough and I would occasionally get a code 17(oxygen sensor signal problem). My car has 270000 miles on it so my tolerance for perfect running is not as high as when it had 80000 miles on it. My car is a 91 300e

Rob I agree, I thought those 2 are mixed up as well. I actually contacted Dan from that site. He said he actually got the info from someone else, so he's not 100% sure himself.

that's a lot of miles! good to hear it's still going. I've only got 220k kms on mine, that's around 140k miles.

Do you get a check engine light. I got code 17 which is o2 sensor signal short to pos or ground. Sensor was fine I think it was seeing a lean mix on higher rpm that could not be adjusted for. I remembered another place I once found a vacuum leak it was at the canister(carbon canister I think) in the area behind the plastic fender liner behind the drivers side front wheel. It connects to the connector you are supposed to disconnect and plug when checking lambda