K-series MoT emissions fail
Discussion
My Elise 111S (VVC K-series running a Lotus ECU) has just failed its MoT emissions test on just about everything there is to fail on, by a huge margin. It seems to be running hugely rich, but seemingly only intermittantly or only under certain conditions. The garage said it was popping and banging during the test and the MIL light actually came on for a "multi-cylinder misfire". However it drives fine and fuel economy is roughly where it should be (hard to tell with how little use the car has had recently).
One thing I did notice is that when driving, the ECU (via Torque) is sometimes reporting a fuelling state of "open circuit - under temperature". Is this likely to be the normal code that the ECU reports when the engine is running open circuit due to load (this is Lotus after all), or is it actually a hint as to what's wrong?
Otherwise I can't find anything obviously wrong. There's no code MIL code apart from taht one multi-cylinder misfire which popped up during the test so presumably the lambda sensors are at least in range.
Actual numbers are:
Fast Idle:
Speed: 2445rpm
CO: 11.85%
HC: 1002ppm
Lambda: 0.659
Fast Idle 2:
Speed: 2438
CO: 12.10
HC: 1693
Lambda: 0.628
Natural Idle:
Speed: 7700
CO: 6.863
Any ideas on where to start?
One thing I did notice is that when driving, the ECU (via Torque) is sometimes reporting a fuelling state of "open circuit - under temperature". Is this likely to be the normal code that the ECU reports when the engine is running open circuit due to load (this is Lotus after all), or is it actually a hint as to what's wrong?
Otherwise I can't find anything obviously wrong. There's no code MIL code apart from taht one multi-cylinder misfire which popped up during the test so presumably the lambda sensors are at least in range.
Actual numbers are:
Fast Idle:
Speed: 2445rpm
CO: 11.85%
HC: 1002ppm
Lambda: 0.659
Fast Idle 2:
Speed: 2438
CO: 12.10
HC: 1693
Lambda: 0.628
Natural Idle:
Speed: 7700
CO: 6.863
Any ideas on where to start?
As on one else has commented yet.....
I know nothing about the elise ecu but taking that error at face value, fuel open circuit due to under temperature implies to me that it thinks the engine is cold so is using a fixed cold start map. If this is happening when it is actually up to temperature then it could explain very rich running.
Perhaps monitor the engine temperature with torque and see if it is getting dodgy readings? It could be as simple as a dodgy connection to the temperature sensor?
I could be talking rubbish though
I know nothing about the elise ecu but taking that error at face value, fuel open circuit due to under temperature implies to me that it thinks the engine is cold so is using a fixed cold start map. If this is happening when it is actually up to temperature then it could explain very rich running.
Perhaps monitor the engine temperature with torque and see if it is getting dodgy readings? It could be as simple as a dodgy connection to the temperature sensor?
I could be talking rubbish though
From EX Auto sparks now retired....all those Co readings are way way too High....Co should be 1.3 on hot engine at Idle....change the Temp sensor located as i remember on the Thermastat housing, your saying 2 ??...you might be mixing up with the sensor for cooling Fan...only 1 sensor for ECU...2 PIN Plug with quick release clip on...if faulty will give "Rich running" right through the rev range, make sure pins in plug clean also, best i can help. Pete
Having dug a bit deeper, it does indeed have one temperature sensor for the dashboard and one for the ECU... except Lotus in their infinite wisdom decided to feed the dashboard temperature from the ECU, so the old dashboard sensor is still there but not actually connected up to anything. 
This does, however, mean that I can log the temperature as recorded by the ECU (the dashboard appears to have some sort of deglitch on it so isn't helpful for catching intermittent misreads). However whilst the engine obviously runs rich when it's cold, would it really run rich enough to put the HC values up into the parts per thousand range?
This does, however, mean that I can log the temperature as recorded by the ECU (the dashboard appears to have some sort of deglitch on it so isn't helpful for catching intermittent misreads). However whilst the engine obviously runs rich when it's cold, would it really run rich enough to put the HC values up into the parts per thousand range?
Edited by kambites on Tuesday 12th January 19:19
kambites said:
Having dug a bit deeper, it does indeed have one temperature sensor for the dashboard and one for the ECU... except Lotus in their infinite wisdom decided to feed the dashboard temperature from the ECU, so the old dashboard sensor is still there but not actually connected up to anything.
This does, however, mean that I can log the temperature as recorded by the ECU (the dashboard appears to have some sort of deglitch on it so isn't helpful for catching intermittent misreads). However whilst the engine obviously runs rich when it's cold, would it really run rich enough to put the HC values up into the parts per thousand range?
I suspect the two sensors are a legacy of when dashboards were dumb and so needed a direct sensor drive. I guess lotus thought leaving the sensor in was easier than fitting a blanking plug This does, however, mean that I can log the temperature as recorded by the ECU (the dashboard appears to have some sort of deglitch on it so isn't helpful for catching intermittent misreads). However whilst the engine obviously runs rich when it's cold, would it really run rich enough to put the HC values up into the parts per thousand range?
Edited by kambites on Tuesday 12th January 19:19
I am guessing the ecu records freeze frame data. It might well tell you what (it thought) was going on with the temperature sensor when the faults were recorded. It might need something better than torque to read it though?
Personally I would be checking the temperature sensor carefully. Is it giving a sensible reading? Are the connections (including earthing through engine??) good all the way back to the ECU?
Of course it could be something else, but this sounds a good place to start given the fault codes and the symptoms.
Hmm, so there's some readings from duplicating the fast idle test. You can see where I raise the revs as the intake pressure drops, then the pre-cat lambda sensor starts pulsing roughly as I think you'd expect; the post-cat one looks stable(ish) which I think is also expected? Coolant and intake temperature readings are consistent and sensible (although an intake temp of 26 degrees when teh ambient is about 6 makes you realise how much heat-soak there is when the car is stationary!),
Note that the K-series doesn't appear to have a MAF sensor as such, but rather works out the airflow from a combination of intake temperature and intake manifold pressure.
Anyone see anything unexpected there?
Note that the K-series doesn't appear to have a MAF sensor as such, but rather works out the airflow from a combination of intake temperature and intake manifold pressure.
Anyone see anything unexpected there?
Gassing Station | Home Mechanics | Top of Page | What's New | My Stuff