Exhaust Emission Test

[johnwallace]

I have a Mercedes C180 (fuel Injection), 1995, which has just failed its emissions test. The car passed on the natural/normal idle test but failed on the fast idle test. The fast idle Lambda reading was 1.04 (the legal range is 0.97 to 1.03) and the CO level was 0.77% (the legal maximum is 0.3%). I replaced the CAT last year. The mechanic who tested the car advised that try I using a good fuel injector for 100 miles prior to a retest as this may normalise the readings. The car has been doing a number of short journeys over the past year with an occasional long trip. I will follow the advice but would welcome any other views. In particular, is it likely that the Lamda unit could be faulty. I should add that the car is running perfectly and I have no other symptoms.
Thanking you in anticipation.
John

[manny]

Your lambda reading of 1.04 suggests a slightly lean mixture, yet, you have 0.7 % of CO, saying slightly rich.
Could be a lazy ( on it's way out ) O2 sensor.
Ckeck- in on another topic here ( by " 95e320" ), in which Arthur Dalton gives a very brief, but practical hint on how to determine the activity, or lack of, an O2 sesnsor.

[johnwallace]

Manny,
Thanks for the response. I have cleaned the fuel injectors using Redex and changed the Lambda probe. Unfortunately, the car still failed the emissions test today. I have arranged for a diagnostic test on 24/4/03 but would be grateful for any advice as to what else could explain the high CO and Lamda readings that are reported. I should add that the Hydrocarbon readings were well within the acceptable range and the car does not have any other symptoms which could help the diagnosis. The CAT was replaced last year...could it be faulty?...there are no strange noises from the CAT or any other signs such as blockage.
With thanks
John

[johnwallace]

For completeness I provide all the readings of the emissions test in case these can be informative to the experienced eye!

Normal Idle Fast Idle

CO 0.19 (Max 0.5 PASS) 0.77 (Max 0.3 FAIL)

HC 120 (Max 200 PASS)

Lambda 1.04 (0.97-1.03 FAIl)

Lamda Probe Replaced today (pre second test)!
CAT replaced last year
No oil loss

No other symptoms

Please Help
John

[manny]

Here are, in generic terms, some items that can cause elevated CO readings:

1. dirty or restricted airfilter
2. contaminated oil
3. saturated charcoal canister
4. non-functioning PCV system
5. dripping fuel injector (s).

I am still puzzled, how you can have a high lambda reading ( indicating a lean fuel mixture ), and yet have high CO's.

When was the last time this exhaust gas analyzer was calibrated ?

[johnwallace]

Manny,
Thank you very much for your response. The filter is new and the oil is fully synthetic and was changed with the filter 2 months ago. I haven't recently replaced the charcoal canister but I suppose my frustration lies in not understanding the logic of the readings. As you say the readings appear to give a mixed message (I have checked the official printout a number of times!).
If the system was running rich then I would logically work through and replace as many of the associated components as I could prior to the need for the specialist MB equipment. The fact that the Lambda indicates a lean mixture but the CO is high (the Hydrocarbons are well within the limits) is confusing.

Has anyone else any experience of this type of emissions profile?

And in particular, what might underpin the readings?

Any suggestions would be gratefully received

John

[johnwallace]

For Information. I spoke to an exhaust centre today who suggested that I could have two problems. Firstly, a fuel problem which was causing the high CO level, and secondly, a leak after the CAT which could allow air to be sucked in by the exhaust analyser. The problem that I have with this explanation is that I would have expected the additional air being sucked in to have weakened the CO levels. Any advice for a frustrated DIYer.

[manny]

John

I am at a bit of a loss here.

As long as your O2 sensor keeps saying " lean ", as indicated by a lambda reading of 1.04, you would be getting more fuel than required.

Yes, your HC's are looking great, so there is no issues with oil consumption or ignition problems.

Too bad you are so far away, or I could possibly help you out with my infrared 5-gas exhaust analyzer.

Let me know what the " verdict " is after your diagnostics check on the 24.

[95E320cab]

I hear not all these guys operating this equipment are totally competent. Also, can't the equipment get out of calibration? Before I threw a bunch of time and effort at it, I might take it to another diagnostic center and have it retested. I also believe in the Italian tune-upp, especially for a city car.

[johnwallace]

Thanks Manny and 95 E320cab for your responses. I will certainly advise the forum of the outcome. The MOT analyser supposedly has an inbuilt check system (I am not sure whether the inbuilt check system has a inbuilt check system!). The test was administered by two different mechanics on the initial and retest assessment...so I can exclude individual incompetence.
Suggestions that I have received from a generic CAR DIY site in the UK, suggest that the high Lambda reading is often indicative of an engine rather than CAT problem. Before the second retest I took the car for a 170 mile run with Injector cleaner in the tank and on the day of the test I took it for an 8 mile run prior to the test. Another respondent asked for the other gas details but the emission test in the UK only records the CO, HC, and Lamba readings. I have not replaced the charcoal canister and I probably have been guilty of occasional over fill but I don't know if this would explain the symptoms. It has also been suggested that I check the vacuum hoses. Is the fact that the car easily passed the idle CO test a significant factor in fault deduction?
I appreciate your help...it helps to maintain sanity.
Regards,
John in overcast England

[manny]

Originally posted by mpnye
I think you're at the point where you need PROFESSIONAL HELP ! ! !

That's what I do for a living & the reason I have a 5-gas infrared exhaust gas analyzer.

A vacuum leak would have a greater effect at idle, and a lesser effect on the high speed test.

As far as two different technicians doing the test, if they used the same ( suspect ) equipment, it would give the same results.

Out- of -date calibration gas can cause some pretty weird readings.

Maybe we'll just have to wait until the 24 th to get our answers.

[Richard Wooldridge]

Hi there,
I think you MAY have a bad EGR valve, causing the computer to try to compensate for the problem and raising the CO. If the EGR diapghram has a weak spring this condition can occur. Try replacing the EGR valve and see if the problem doesn't go away. What happens is the EGR valve opens when it shouldn't, and the computer richens the mix to try to compensate.
Not a costly item to try. Do you notice a slight surging whilst driving down the road? That's another indication that the EGR valve could be opening when it's not supposed to.

Good luck with it!
Richard Wooldridge
'82 300D/4.3L V6
'94 Taurus
'88 Astrovan
'77 Jag XJ6L
'77 280Z
'85 GMC S15 PU
'80 VW Cabriolet convertable
Etc...

[manny]

Richard

You are correct about the EGR causing a lean condition ( as far as the O2 and ECU is concerned ).

The next piece of the puzzle is, if the EGR is acting up, the HC would normally go quite high. A reading of 120 ppm seems rather low, although it was not specified if that number is at idle or high rpm.

I guess there is nothing like " being there ".

[johnwallace]

Thanks for all your help...at least with your inputs it is an educational as well as frustrating experience.
The HC reading was at Fast idle ( I should add that in the first fast idle test the HC reading was 88...after failing the first fast idle on CO and Lambda the tester has to do a second reading having pre-conditioned the car i.e run it at fast revs for 3 mins). The basic idle test result did not refer to the HC levels.
Some additional queries.
(1) Is it possible that the lambda reading at the sensor is different from that at the tail pipe? That is, the data being provided to the engine manmagement system may not be that found at the tail pipe. I am wondering whether the Cat could be having a differential effect on the gases. I suppose a gas probe at the Lambda sensor location would be revealing.
(2) How easy is it to detect an exhaust leak. The whole system looks fine to me so if there was a leak it would be at either the manifold joint, manifold- to- CAT joint or CAT- to- tail pipe joint. There are no obvious signs but I am relying on uncalibrated eyes and ears. Could a small leak have such an impact. If so how would the professionals detect it?
(3) How good is the OBD and the MB diagnostic. Should I feel confident that my money spent on the MB diagnostic will definitely identify the problem or confidently exclude many factors.
(4) Should I hire an exhaust gas analyser and experiment with 'solutions'?
Any responses to these queries would be appreciated.
Thanks,
John

[manny]

John

For question # 1, the way to test the efficiency of the catalytic converter, we do a pre-cat ( before converter ) & then a post-cat ( after the converter ) test.
That's why OBD II vehicles have two O2 sensors.

# 2 , to test for an exhaust leak, plug the tailpipe & see if you can hear any hissing sounds in front of the O2 sensor. If this is done on a cold engine ( first thing in the morning ), you would possibly see some condensation cloud.

# 3, I am not very familiar with all the M-B engine controls specifically, therefore I tend to talk in " generic " terms.

# 4, other than running your diagnostic check, which you have scheduled for the 24 th of this month, I would postpone chasing any more demons until then.