Emission failure - Lambda too high and HC too high on idle

[Sam Russell]

My Passat Estate 1.8 manual (non-turbo) year=2000 with 56000 miles if failing the test on emissions.
Test Results
Low idle CO 0.05% (Allowed 0.5%)
HC 9ppm

High idle Lambda 1.11 (Allowed 1.00+- 0.03)
CO 0.01% (Allowed 0.30%)
HC 9ppm (Allowed 200ppm)

Got the car checked and new Lambda sensor fitted which gave the following result.

Low idle (860rpm)
CO 3.15% (Allowed .5%)

High Idle
Lambda 1.029 (Allowed 1.03)

This was after a small leak in the exhaust was fixed.

New Lambda sensor (another one)

Two new test runs:

Low idle CO 3.32%
High Idle Lambda 1.090

2nd run
Low idle CO 3.23
High idle Lambda 1.120

All other figures OK.

Mechanic says he can do no more - take it to someone else.

Help.

I have owned the car from new so mileage is correct.

[martin1810]

Has the car been checked for any stored fault codes. The ecu may still be getting a faulty lambda reading because of a wiring fault. This will then muck up the fuelling and further upset the lambda readings. How clean is the air filter. The catalyst may have failed. A proper vcds code read is the best place to start.

[Sam Russell]

There are no fault codes as far as I know. The plugs were changed - no difference. The car drives normally. The tester said that the car was missing at the high idle ("you can hear it missing"), but the mechanic said it was fine. High lambda in the exhaust can be due to a leak in the exhaust and a small leak was fixed, but the problem persists.

The difficulty I have is I have no reliable way of testing the emissions myself.

Would cleaning the MAF do anything?
Would changing the airfilter do anything?

The difficulty is that the mechanics are faced with working on a car whose value is very low, and very much lower without this problem fixed.

[Sam Russell]

Has anyone any ideas on this? I need to solve this.

[martin1810]

A VCDS read will tell what the lambda probe is reporting or get a garage that know what they are doing. Any decent garage should be able to test the lamba probe. It needs to be checked for heater function and input/output signal. These things are not difficult si it's hard to see why the garage are having problems. You should have fault codes stored as a result of the lambda reading being wrong.

[Sam Russell]

Thanks for your reply.

The lambda sensor was changed for a new one, and the car failed the test again. It was then sent for a test again where the exhaust was found to have a small leak, and it was failing on the low idle CO, and just making the lambda (1.029 - limit 1.03). The mechanic changed the lambda again, but the problem was worse, with the lambda way out.

The lambda sensor is doing its job (cycling on and off), but still not passing the test. What else causes a high lambda reading and high CO on low idle?

[martin1810]

Still think you need a live data read with VCDS or similar. Generally high Lambda means excess oxygen and high CO means not enough oxygen. !!!!!!
I would ignore the lambda and concentrate on the High CO which is usually the result of poor air supply or more likely a fuel problem. This could be a misfire, egr fault, fuel vapour control fault, crankcase ventilation fault etc. You will save money by getting a good VCDS read. It is odd that changing a lambda sensor on one occasion, made things worse. That suggests a possible wiring fault in the lambda circuit but it could have been just a bad earth connection. Finall the CAT may have failed and if it hasn't but you are overfuelling, it will.

[Sam Russell]

Thanks martin1810, I will go that way. I posted the various readings I got in my post above. I assume VCDS is the Vag-Com software.

I will see if I can find someone with it.