It finally happened - my 85 190E 2.3 failed the emissions test - the readings were:

HC: 142.9 ppm (limit 189 - pass)
CO: 4.48% (limit 1.21% - FAIL)
CO2: 12.32%

I did all the usual stuff - take it for a spin on the freeway for 15 minutes, oil change, etc. before the test. Also, had a full tune up 10K miles ago (plugs, wires, rotor/cap, injectors, air/fuel filters, OVP relay replaced) - my car has 121K now.

What could be the reason? What should I look for? It also has a cold running problem - runs rough until the engine warms up.

I looked through the archives, but could find an answer to my specific problem.

thanks for your help,
Rags

If the O2 sensor is original, this may be the problem. Since you also have cold-running problems, - the ECU may not be functioning.

Steve

I usually tell people to change oil, add STP (helps thicken oil a bit) and run premium fuel before the test.

Also, if you haven't used a fuel injector cleaner recently, this can help a great deal. I get STP injector cleaner and put two of those bottles in the tank.

O2 sensor is a good suggestion too.

Ken

On the cars that need emmissions testing, I also drive them for about 45 minutes to an hour before testing to thouroughly heat the cat. If the car has to sit for a bit before they get to it, the cat should still be retaining some of it's internal heat.

__________________
Mike Tangas
'73 280SEL 4.5 (9/72)- RIP
Only 8,173 units built from 5/71 thru 11/72

'02 CLK320 Cabriolet - wifey's mid-life crisis

2012 VW Jetta Sportwagon TDI...at least its a diesel

Non illegitemae carborundum.

I concur about the O2 sensor...if it's not doing its job, then it can't tell the ECU to adjust the mixture to correct the situation.

I don't know if a check engine light comes on for the O2 sensor on the '85. Obviously, it's not sending any codes...

__________________
2009 ML350 (106K) - Family vehicle
2001 CLK430 Cabriolet (80K) - Wife's car
2005 BMW 645CI (138K) - My daily driver
2016 Mustang (32K) - Daughter's car

The oxygen sensor is probably original - I am sure its not been changed at least in the last 5 years. The O2 light has never come on. Could that still be dead?

Also, the HC reading was within limit, but the CO reading was way off - shouldn't the HC reading also be higher if the mixture is not correct?

Thanks for your responses so far!

Rags

The cat will clean up excess HC emissions - for awhile. Once it plugs up, you'll have to replace it too.

The O2 sensor can measure incorrectly, without failing to function altogether. If the ECU THINKS it's ok, it won't trigger a code. Of course, if the ECU isn't functioning, there won't be a code either. The non-loop control of the CIS injection will still allow fairly normal apparent operation.

Steve

If you can't get it to pass there in Ohio, I'm sure Dr. Kervorkian in Michigan would be interested....... what with all that carbon monoxide......


Boo, Hissssssssss !!!

I think my car is working in non-loop control, since even slight changes in the air-fuel misture (my guy had richened the mixture 1/8 of a turn some time back to reduce the rough idle when cold) seem to affect the car - if the O2 sensor was working properly, shouldn't the mixture screw adjustment be redundant?

I'm getting the O2 sensor changed anyway. However, I don't think that will solve my rough running on cold engine, which is the root cause of the high emissions (since the mixture had to be made rich because of that). I've already spent over a grand on the car on this problem, with no solution, so am very wary of another tech throwing parts at me.

A very noticeable symptom in my car is that when its cold, it hesitates, and has a "slingshot" behavior, i.e., it moves in spurts. During this time, I often hear a click sound from the bottom driver side, which I can even feel in the accelerator pedal - as if something is jammed, and suddenly opens up. This happens 5-6 times, and then when temp reaches 80, it suddenly smoothens out and everything is fine.

Anyone had similar experiences? I am pretty sure its a fuel delivery problem - could it be the air flow sensor? How can one diagnose it?

thanks.

Fact is, it should be pretty easy for a tech with the proper tools to distinguish whether the problem is or not the O2 sensor. If the connector to the O2 sensor is unplugged and the running characteristic does not change, the ECU may be bad - unless the O2 sensor failed 'open'. An 'open' symptom you would expect to trigger a code for a bad O2 sensor. If the running does change, and a code is set when you unplug the O2 sensor, then it may be bad and replacing it would be a worthwhile diagnostic.

Steve