Heph Posted March 15, 2006 Posted March 15, 2006 Hi everybody, I just bought a 1998 Lotus Elise with a known lambda problem. It failed MOT with the following results: CO Idle: 0.2 CO 2500 rpm: 6.7 Lambda: 0.81 It passed the same test in september 2004 (3000km / 2000 miles since then) with the following results: CO Idle: 0.0 CO 2500 rpm: 0.0 Lambda: 1.00 So it was pretty much perfect !!! From that I doubt the CAT is to blame right ? I took a look at the Lambda sensor and I get 0.6V (flat) at 2500 rpm. A sensor seller told me voltage was low considering the lambda I got from MOT. According to him the sensor is faulty ... but he sells the sensor. I read somewhere else that 0.5V was already a "Rich" signal from the sensor so who is right ? I'm concerned about the Idle/2500rpm différence, what does it tells? I'll check for exhaust leaks tonight but I was also wondering if a bad airfilter could do that, gonna check that too. I've heard a bad air intake tempature sensor could do that. How to chek it and where to buy one (cheap)? Thanks
dobbo Posted March 15, 2006 Posted March 15, 2006 To be honest, those 2004 results look fake ! When running correctly, the sensor should flick between 0 and 0.8-1 v about twice a second at idle revs. New sensor is around £55 from motorfactors, but is a bitch to get at on a standard exhaust. There's a thread from early January about my fun and games with the MoT.
Heph Posted March 15, 2006 Author Posted March 15, 2006 Hi Dobbo31 I don't know about the 2004 results. They are official results done at the same place with the previous owner there... If he had faked the 2004 results, he could have managed to fake this one... Anyway thanks for the answer, my lambda sould be dead then... Just wonder, what would happen if the engine run with the lambda sensor unplugged ? I wonder if the ECU would comme back to a basic setting wich would be beetter than the one with the faulty sensor. If so I could have CO and Lambda mesured without the sensor and see if it induced the error... Might be a stupid question there but I'm beginning ....
dobbo Posted March 15, 2006 Posted March 15, 2006 I don't know about the 2004 results. They are official results done at the same place with the previous owner there... If he had faked the 2004 results, he could have managed to fake this one... <{POST_SNAPBACK}> Not if the MoT station has now been computerised - there's too much at stake now for (most of) them to be interested in faking them. Anyway thanks for the answer, my lambda sould be dead then... Just wonder, what would happen if the engine run with the lambda sensor unplugged ? I wonder if the ECU would comme back to a basic setting wich would be beetter than the one with the faulty sensor. If so I could have CO and Lambda mesured without the sensor and see if it induced the error... <{POST_SNAPBACK}> I think that'll fail miserably. If there's no output from the sensor, it won't matter whether it's dead or unplugged - still no output and the ECU won't know the difference. The only other connections are the 12v supply for the preheater, but all that does is get it up to temperature quicker. Have a search for the thread I mentioned earlier - it went on a bit and there were a few things there to think about and the contact details for where I got my new sensor as well.
Heph Posted March 15, 2006 Author Posted March 15, 2006 Good point: 0V is 0V no matter if it comes from an unplugged sensor or a dead one... I've read the post you're talking about (and all the post Lambda sensor related on the site). Now I've got to replace it... Good, I didn't know what to do next week's evenings Thanks
Tango190 Posted March 15, 2006 Posted March 15, 2006 What test centre were the 2004 results obtained from ??
Heph Posted March 15, 2006 Author Posted March 15, 2006 Well ... I forgot to mention, I'm french so it's from a french test center... Both tests where computerised .... Anyway, doesn't matter, I love this car already Correct me if I'm wrong, lambda sensor don't oscillate on it's own. It the ECU correcting the Lambda constently that make it oscillate right? That means, if the Lambda sensor isn't able to send a high enough voltage so the ECU takes it as a Rich mixture, the ECU won't correct the setting to a leaner mixture and the ECU won't oscillate...
chorton_1 Posted March 16, 2006 Posted March 16, 2006 On these points above:- It is indeed the ECU oscillating the amount of fuel it is adding, which in turn makes the lambda sensor output oscillate. Depends on the ECU as to what happens under different conditions and I'm afraid my stock ECU knowledge is not that good however in general it will do one of two things: *) If the sensor is in range but faulty, the ECU may wind in as much fuel as it is allowed (there is a tight limit set within the ECU, so this is not dangerous to the engine) and hold it there. If the ECU does this for an extended period there may be a function in the code to realise something is not right and ignore the sensor. *) If the sensor is out of range then the ECU will ignore the sensor and go off the basic map value. It is true to say that you are likely to get better fuelling without the dodgy lambda sensor, however it is unlikely to be good enough to get through the MOT.
Tango190 Posted March 16, 2006 Posted March 16, 2006 Feck a French MOT !! Anyway, sorry Heph joke for UK folks Can you confirm standard ECU is fitted not Emerald ECU ?? Regards Bob
Heph Posted March 17, 2006 Author Posted March 17, 2006 Yes I can confirm: My car is its original configuration so standard ECU fitted. I did some more tests Yesterday: First I took a look at the sensor reading without changing anything: +0.65V I then took the air filter out as it looked dirty: The voltage changed sign: -0.65V This make sens as the volt-meter was plugged backward: Black Lead connected to black lambda wire (signal), Red lead connected to grey lambda wire (earth). But How can that be possible... Multimeter also gave me some oscillations... At last, I unplugged the Sensor and the reading felt a bit: -0.45V to -0.55V. Redaing where more stable too. Anyway, the new sensor is on it's way and I started to take out the manifold ... And yes, it's a nightmare Manifold is totally unscrewed but I still have to take the flex-pipe out so I can get the manifold Out. I first thought I could get it out by the top... no-way So the correct order seems to be: take the Heat-Shield out, Unscrew the manifold-flexpipe flange, Unscrew the Cat-Flexpipe flange, take out the flexpipe then you can unscreww the manifold from the motor and take it out... This might help someone in the future One last question, the manifold nut where so corroded, it came out with the screw... Why not replace it with a simple Stainless bolt ?
chorton_1 Posted March 17, 2006 Posted March 17, 2006 As far as I know you can replace with a stainless bolt - I would imagine it was purely cost that prevents this being standard fitment (at least as a stainless stud).
Heph Posted March 17, 2006 Author Posted March 17, 2006 I've got the answer from a guy at my work. Stainless is not very compatible with Aluminum, it will cause electric corrosion. The best solution is to use mild steel with a special coating (no galvano, same problem as Stainless Steel) and / or a grease that can take such High temp. I took some of this grease at my job, It can take up to 1400°C !!! That should do
chorton_1 Posted March 17, 2006 Posted March 17, 2006 Yep - the electrolytic reaction can indeed be a problem and it does depend on what you use to coat the threads and how long you expect the assembly to be together for.
Heph Posted March 19, 2006 Author Posted March 19, 2006 I might use two things: the grease I was talking about or some duralac if it can stand such high temperatures. I might use both if I go for mild steel or only Duralac if I use Stainless. As I haven't planned changing the exhaust anytime soon, I hope this will work if it stays together for a long time... I finally took of the flexpipe and the Cat but I need to take the oil filter out so I can take the manifold out... I might go all the way and take the silencer out to have a good inspection... Beside the Lambda sensor, I will change the air filter as the one in place is dirty. I'll also take a look on the spark plugs but I wonder if it can have great influance on lambda readings ???
chorton_1 Posted March 19, 2006 Posted March 19, 2006 Unless the engine is misfiring, then I would doubt the spark plugs are the problem (or have any influence). They are low cost though, and if the air filter looks old then maybe the plugs are too.
Recommended Posts
Archived
This topic is now archived and is closed to further replies.