I’ll certainly keep you all updated. I think there’s enough evidence to swap the sensor now but I’m partly regretting not bugging a new MAF as well incase I need it.
I quite like a challenge…
Have you checked all the manifold to head nuts are present and tight? Just that your readings even if they are slow are showing that the system is cycling rich to lean across 14.7 as you would expect. An unmetered steady state air leak would be expected to fool the sensor into giving you the extra fuel to compensate.
I asked the garage to swing through all the fittings especially after the centre section was replaced but I’ll check again before I get around to replacing the sensor.
This happened to me about a year after I slapped a turbo on my NC, it drove me crazy as I thought I was looking for an air leak. Eventually I bit the bullet and got a new sensor, when I got the old one out it was really coked up, I think the tuning process just sent it over the edge. Hope you find a fix.
It certainly looks as though the sensor is functioning, although it still might not be in spec. Reading between the lines I guess the “3000 rpm to idle test” is checking how the sensor reacts to overrun fuel cut when you lift off. If it’s meant to be done while moving in gear then it wouldn’t surprise me if the equivalent check done out of gear was too brief to give the true result.
Do you have the actual figures for the failed MOT emissions check?
If the problem was an air leak around the manifold, for example, I would expect to see high CO but near normal Lambda (the sensor would see the extra O2 in the air, so the ECU would add fuel until the sensor read stoich but the engine itself was running rich).
1 Like
I have recently purchased a Mk 3.5 that has a center section decat. It had just had its MOT so I wasn’t too worried about the emissions. After driving it for under 30 miles an EMS light came on and my heart sunk. Thankfully it turned out to just be an O2 sensor, which I put down to it being a decat. After further investigation and an emissions test at the local garage (which it failed pretty badly) I thought it was a case of having a sports cat put into it in the downpipe.
I took it to an exhaust specialist and he says that the downstream lambda sensor is after the first cat in the manifold, so not an error with the lack of cat in the downpipe? He thinks that its the Lambda sensor spitting an error which is then causing the car to over fuel, and therefore fail the emissions test.
Is the best way to see if is that to have a look at the lambda sensor reading and determine if it is working as it should be?
Any advice would be greatly appreciated.
Well I’ve finally got an MOT with emissions that are well under the limits and much lower than last year’s test. Sadly I committed the engineering sin of changing 4 things at once so I don’t know what the root cause of the failure was…
Since failing on high CO I replaced the centre cat and had it tested again but there was no improvement. After that second test I replaced the wide band lambda sensor, cleaned the MAF and ran some cataclean through the system. I also disconnected the battery for an hour in the hope that perhaps the ECU fuel trim had learned something in error and it would forget it (seemed tenuous but was unlikely to cause trouble).
After all that the initial start was much cleaner and some high rev driving down the local ring road may have helped subsequent starts too. Long term fuel trim does appear slightly lower at about -4% and i’m sure it was -2% when it failed the MOT. The wide band lambda current trend did appear to behave slightly differently as well but hard to confirm.
So that was £200 of parts and a few hours of my time and a good result. If I were a betting man I’d say the lambda was failing but I’m a bit short of evidence!
1 Like
This sounds similar to what I’ve just been through although your car setup sounds different to mine.
It sounds like you’ve got a cat in the manifold so (assuming it’s the same as my '06 car) you’ll find one lambda above the car and one below it.
If your specialist suggests looking at the lambda and you’re happy to have a go yourself then I’d suggest a OBD reader (£15 Eurocarparts) and the “Car Scanner” app for about £3 and you will be able to trend the behaviour of your lambda sensors.
If you’re anywhere near Bristol I might be able to help in a socially distant way!
Glad to hear you got it sorted. With such a large drop in emissions, I’d be inclined to think O2 sensor as well.
Very late reply, Apologies.
Thanks for the advice, my car is a 2009 NC2 but I think they are fundamentally very similar.
I will heed your advice and see if I can work out if its the sensor and get it sorted.