Help - Emissions!

Just looked through the emissions sheet again and I missed this:-

Natural Idle Test
CO % VOL: should be 0.50, actual 0.90.

Not so bad at idle. Hmmm

Does it have a after market ss exhaust or a standard one as there was talk a while ago about having to cover one of the exhaust out lets to get the emissions down maybe David can post the thread

It has a Milltek stainless system fitted. It seems to breath through both pipes.

I could try blocking one off. :lol:

Ask the MOD's for the link for exhaust emissions when they fail the MOT, i have a feeling you had to block the left hand pipe off with a rag but not sure on a Milltek if this is the case

I replaced the manifold, Lambda, Front pipe, Cat (with a good one) silencer from a 2004 TF, NGK platinum plugs, distributor Cap and rotor arm, Magnecor leads and new coil. Headgasket has been done and filled with Castrol magnatec oil. the emissions on the Mot were so low the tester was surprised.

Mine smelt over fueled before I set fire to the old cat.

Thanks for the reply Bryan. I should have mentioned that I've already replaced the dizzy and rotor arm and checked the plugs which looked in reasonable nick - might change them though as a precaution.

Thinking about it overnight, I think my first action is to run a multimeter across the lambda and inlet temp sensor to check that they are reading right. I should also check the vacuum hoses to make sure there is no leaks anywhere.

Just need it to stop raining now.

Lambda probes are VERY high impedance devices, you can check continuity across the heater circuit but not an awful lot else with a multimeter unfortunately. As they work by degrading a reference potential a "dead-test" as such can't prove much, also as they are switching devices they'll switch far to fast for even the best multimeter in the world.

If you are near Bedfordshire I can have a check of your probe with a 'scope if that's any help to you.

Sounds to me like the loop isn't closing, and the engine is running without any feedback. This will unfortunately foul your catalyst if this is allowed to continue, if it hasn't already. A lambda ratio value of 0.88 indicates that the sensor is either failed not providing any reliable feedback to the ECU or the engine is not looking for the value yet believing itself to be still in "open loop" mode.

An engine stuck in "open loop" points in the majority of cases to a failed or otherwise inaccurate coolant temperature sensor - don't worry about the air-intake temperature sensor, this is a bit of a red-herring it only provides a tiny fuel trim; a lot of cars have lived and died with a failed one and nobody has ever noticed.

Before we get in to the nitty-gritty of diagnosing this, do you have any holes in your exhaust system before the catalyst and what is your oil consumption like?

Regards,
Andy

Ah right, so nothing short of an oscilloscope will be able to test that. Unfortunately I'm in Leicester so cant get to you but thank you for the kind offer.

In answer to your questions, oil consumption I am unsure about as I have only really driven the car a total of 30 miles since buying it.

The exhaust system should be ok. I did think that the recommended torque ratings for the manifold bolts were fairly light but used them anyway. I shall double check for leaks.

Can the coolant temp sensor be tested?

EDIT/ just looked and for the £14 that these cost I'll fit one. I take it the GTR206/brown sensor is the correct one?

Are you losing oil? Have you looked at the plugs recently?

Discolouration or fouling of the plugs can tell you a lot.

Was the cat a brand new one and was it a good make or a cheapy off eBay?

My k series Freelander was failing on emissions badly like yours and it was cured with a new cat. I only fitted one at the last MOT but it had gone. There are reports of some of the eBay cheapies failing straight away.

A failed catalyst wouldn't upset the lambda values so much though, it'd increase CO and NOx but it HCs and lambda control is carried out further up the chain. It appears that whatever is reaching the catalyst is full of partially combusted fuel and the engine itself is running rich, this value would be determined by the lambda probe path.

Another over obvious thing that may be being overlooked is the wiring for the lambda probe, this you can check with a multimeter if you dig around a bit.

Was the engine definitely very hot when it was tested, did they dip the oil during the MOT with a sensor, or will this be another "engine temperature indicated by cooling fan" report?

The most basic way to test the sensor would be to firstly check for continuity across the sensor (in series with wiring, to rule out wiring faults) then unplug the sensor and check continuity across the pins. I'm digging out my magic book now with the specs' in it.

Regards,
Andy

CJJ; Cant say for certain about the oil - I haven't driven the car much since I bought it but it looks clean.

I guess your asking about the oil because of bore wash?

Plugs seemed ok when I checked them. I might throw a new set in as a matter of caution.

The Cat is an eBay special. I'm pretty sure that it's not the issue though as it seems that the emissions issue is due to over fuelling rather anything else.

Cheers Andy.

I shall check continuity of the wiring. Possibly something to mention is that I am using a later type of Lambda sensor (the one with the wider 16mm head) with the standard manifold adapted to take it (drilled out and new boss welded on to take the extra length). From what I could find though, both new and old type of sensors had the same working parameters.

The engine was hot - I had it idling on the drive for 45 minutes before I left so I could bleed the coolant system, then I drove it fairly hard to the testing station which is 8 miles away.

The test says: Eng Oil Temp - Cooling fan cut in so I assume this is how the engineer checked it.

He did say that he ran the car for nearly half an hour to try and get it through the emissions so it should have been plenty hot enough.