18 replies to this topic

[MartinS]

As it sais, just had its first MOT after all its tuning work (4-1, 2.4 inlet, bigger injectors, ITG, remaps etc and its failed the MOT. Lambda readings were - fast idle revs 2750-3250 CO .04% HC 65 ppm vol lambda 1.046 (need to be between .970 and 1.030 second fast idle revs 2750-3250 CO .01% HC 2 ppm vol lambda 1.075 Any ideas how to solve this. Did a check with a freind via a diagnostic thing, and it did show that the lambda sensor before the cat was alternating between lean and rich, not static on any particular one. Could that be on its way out? Thanks for any help. Rang Thorney but theyer all racing. Martin S CO2 was

[vocky]

could be either they put the probe too far into the exhaust or the mixture lambda is duff

[MartinS]

Thanks for the notes Vocky. The lambda was taken out yesterday, as I had a 2tubular back box fitted. How do you know youve put it in the right depth? Martin

[vocky]

I've just realised that your new backbox is 'straight through' so the mot emission probe can be pushed in to whatever distance (with the standard or milltek backbox the emission probe doesn't need to go very far in) so it's most likely your mixture lambda is duff

[markv]

Maybe a stupid question, but if you have a 2bular with two exhaust exists, was the second exit blocked off (with a cloth or something) when doing the measurement? If not, you are going to get skewed results.. Mark

[MartinS]

This may sound stupid, but, if I got a new lamda sensor put in (I assume the on before that cat), would that in anyway affect the fuelling of the car, i.e. the current map, could it make it go bad, i.e run richer, weaker than at present? Vocky, your idea on the lambda being poushd too far in, is that possible to fix easily, i.e. is there not only one way to put it in, or can it be backed out a bit and still allw rok ok, but fix my problem. Martin s As ever, thanks all for the help.

[vocky]

don't worry about your lambda sensor, it can only be fitted to a set point defined by the lambda boss I was pointing out the the mot lambda probe can be moved about inside the exhaust for a better reading

[Guy182]

mine showed it was running lean on the MOT. fitted a wideband lambda properly in the exhaust and its sat at around 14.5 on idle so it definately isnt... get them to leave it at idle instead of revving it..

Edited by Guy182, 18 July 2008 - 07:15 PM.

[vocky]

mine was exactly the same, MOT showed lean, innovate a/f gauge showed 13.8

[Guy182]

how did you sort it?? or were your MOT place 'understanding' ? i kinda guessed that because its a 3inch bore system, and at idle there isn't enough flow in the system to give correct readings at the tailpipes? (i.e. too much atmospheric interference?)

Edited by Guy182, 18 July 2008 - 07:20 PM.

[vocky]

how did you sort it?? or were your MOT place 'understanding' ?

i kinda guessed that because its a 3inch bore system, and at idle there isn't enough flow in the system to give correct readings at the tailpipes? (i.e. too much atmospheric interference?)

luckily the mot place are used to the vx220 / lotus exhausts

[MartinS]

I was intrigued by how the lambda sensor could be put in at different depths!! (I am officially a mechanical numptie). Am going to see the mot guy again for a quick chat about how he did the test, but as the test equipment later on, showed the lambda sensor was varying between lean and rich and not stopping on either, that may be on its way out so guess I may have to get a new one. How much are they? Easy to get? Martin s

[VIX]

Probably no use to you but mine failed it's MOT recently for various things including emissions. Turned out to be a loose/rusted clamp on the tail box. A replacement sorted it.

[siztenboots]

Don't worry about the cat sensor, just make sure its in proper tight so no air can get in, check all flanges for same. The mixture sensor on the downpipe must be in good condition , how old is the mixture sensor , they go lazy and slow to react as they get older. Also worth reset ECU learnt values and let it readjust to the freeflow system, the U bend is also a big improvement on the catback exhaust you had on before.

Edited by siztenboots, 19 July 2008 - 08:24 AM.

[MartinS]

Thanks all. Went down again this morning, armed with some of the suggestions. they tried putting the test thing in the lower exhaust instead of the upper, that lowered it a bit, then plugged one exhaust then both and it passed this time. The mixture lambda is original so may get that changed at some time and interested in the resettingg ecu idea Sizeten, does that not mess up remaps etc? Any way, all ok now. Thanks for all advice, as ever a great bunch of people on here, and I'll see a lot of you in Wales next week or the chase tomorrow. Martin s

[siztenboots]

Resetting it just puts it back to how it left TMS, ie. before it adjusting itself to your driving style and fuel in the tank. Reset it with either a scantool or the fuse method, then go for a steady run using all gears and some steady throttle inputs, ie. 20% in 4th gear let it pull through the rev range, then repeat at 40% so it can learn. This will get the car in to closed loop mode, so it is only using the lamda sensor to fine adjust the fueling , and adjust itself to the throttle body flow characteristics.

[Arno]

CO .04%
HC 65 ppm vol
lambda 1.046 (need to be between .970 and 1.030

CO .01%
HC 2 ppm vol
lambda 1.075

CO is good, HC is good. Primary lambda should be fine.

Calculated lambda value being too high would mean it's running lean, but in this case means you have an exhaust leak after the primary lambda sensor that's allowing oxygen from the ouside into the exhaust system and it picks up on the sniffer equipment.

Sniffer lambda value is just calulated from the distribution of the % of gasses (CO, CO2, O2, HCx) in the exhaust gasses and NOT a reading from the lambda sensor itself.

If a leak were present in front of the sensor you'd get high CO values as the ECU would think it's running lean and dumping in fuel. This would give you an 'impossible' result with a high lambda reading (lean) and high CO (rich)..

Bye, Arno.

[MartinS]

Thanks Arno. Steve, dont have that kit so will live with it as it is at the mo. All the help and commenst much apprecaited. Martin s

[EdButler]

CO is good, HC is good. Primary lambda should be fine.

Calculated lambda value being too high would mean it's running lean, but in this case means you have an exhaust leak after the primary lambda sensor that's allowing oxygen from the ouside into the exhaust system and it picks up on the sniffer equipment.

Sniffer lambda value is just calulated from the distribution of the % of gasses (CO, CO2, O2, HCx) in the exhaust gasses and NOT a reading from the lambda sensor itself.

If a leak were present in front of the sensor you'd get high CO values as the ECU would think it's running lean and dumping in fuel. This would give you an 'impossible' result with a high lambda reading (lean) and high CO (rich)..

Bye, Arno.

Great info Arno; thanks for sharing. I either have an Exhaust manifold leak or very duff lambda and this little nugget will help the diagnosis