Innovate MTXL Wide Band Lambda - not accurate
Discussion
I have a Chimaera 500 that is running Ford Gen2 coil packs, AFM removed, Megasquirt 2 ECU, and an Innovate MTXL wide band lambda sensor which is less than 6 months old as the previous Innovate LC1 controller failed. The sensor is located vertically in the main exhaust after the Y piece near to where the bracket attaches the exhaust to the engine block. The previous sensor was in the offside mainfold when the car had its original CAT Y Piece but I took the opportunity to move it so it can read from both banks when it was fully decatted with a Clive Ford Y. The precats were removed a while ago.
The MTXL gauge is tucked up inside the dash so it is not on display most of the time.
The MTXL sensor has been calibrated in free air correctly several times and has not been in the exhaust unpowered while the car is running nor heated up too long without the car running.
I also changed the air intake in the summer and replaced the old system with a silicone smooth bore pipe (similar to ACT kit) that has a large filter and air temp sensor in the nose of the car so that the intake temperate is more accurately reported. The old location of the air temp sensor was right next to the near side manifold and was suffering from heat soak.
I have not used the car much since the new Innovate went in back in Sept 15 and the car was fully decatted. Just before Christmas while on a day out with my Son he asked me to change he engine maps for the 'gunfire' version with all the pop and bangs on overrun. Laptop connected and started to make the change and noticed the car was running lean according to the AFR gauge in Tuner Studio. Checked the MTXL and it said the same thing. I assumed that the decatting and air intake had upset the balance a little.
I also found out that the IAC Stepper motor was shot and needed to be replaced.
So with the stepper replaced and the mapping adjusted to get the AFR gauge back to normal without it appearing to be lean I thought everything was ok. The car did smell a little rich when out and about, but it seemed to be running fine. Given it is the winter the car has not had much use
MOT time came a week or so ago and it failed the emissions tests. I did notice during the test that the AFR gauges both in Tuner Studio and the MTXL I have were showing an AFR value of around 15.3, which is lean, while the gas analyser at the test station was showing a lambda of approximately 0.9 (13.3 AFR) which is quite rich. More adjustments to the mapping and a return visit to the MOT, with a pass this time, but again noticed a similar size variation between my AFR gauges and the MOT analyser.
Other than appearing too lean, the MTXL seems to read as expected tracking up and down as the engines revs and is putter under load etc. The 'too lean' running found on the run out with the my son, in hindsight was probably a false reading as well, but again it will have been playing havoc with the EGO control.
The AFR value in both Tuner Studio and on the MTXL gauge are almost identical so the calibration between these seems to match up ok. They are both set to factory defaults of 0v = 7.35 and 5v = 22.39 AFR.
The invoice I have to the MTXL says it came with a Bosch LSU4.9 sensor, but I am aware that the earlier ones had a 4.2 sensor. I have checked in the LM Programmer that the MTXL is configured to use a 4.9 sensor.
The fact that my MTXL is reading too lean compared to the MOT will be playing havoc with the ECU as it will be trying to richen it up, when it should actually be trying to lean things off.
Any ideas why the AFR reading in the car from the Innovate MTXL sensor is so far our from the MOT reading? Surely they should be close if not identical.
I had originally re-used the live and earth wiring installed by the previous owner for the LC1 he had in for the MTXL I fitted.
I have now re-done the wiring exactly as the manual suggests, adding a relay in to the live feed.
I have the MTXL earthed to the battery terminal connector as this was the most accessible earth point in the foot well.
Both the MTXL gauge and Tuner read almost identical values, but I think they are still too lean.
I have now re-done the wiring exactly as the manual suggests, adding a relay in to the live feed.
I have the MTXL earthed to the battery terminal connector as this was the most accessible earth point in the foot well.
Both the MTXL gauge and Tuner read almost identical values, but I think they are still too lean.
"I have the MTXL earthed to the battery terminal connector as this was the most accessible earth point in the foot well."
I think we've found the problem.
It might be easy to earth the MTXL to the battery ground, but it's very likely to cause the problem you currently have.
You should wire the earth of the MTXL on the same location you have earthed your megasquirt, which should be a bolt/stud on the engine block.
See here for more info, you want to read the bit about wiring up the lc1 or MTXL
https://www.diyautotune.com/support/faq/wideband-o...
Here's a nice video on youtube explaining the importance and installation of good earth connections of your ms2 install.
https://www.youtube.com/watch?v=4sIxbXovD8g
I think we've found the problem.
It might be easy to earth the MTXL to the battery ground, but it's very likely to cause the problem you currently have.
You should wire the earth of the MTXL on the same location you have earthed your megasquirt, which should be a bolt/stud on the engine block.
See here for more info, you want to read the bit about wiring up the lc1 or MTXL
https://www.diyautotune.com/support/faq/wideband-o...
Here's a nice video on youtube explaining the importance and installation of good earth connections of your ms2 install.
https://www.youtube.com/watch?v=4sIxbXovD8g
Yes, the variation was seen at idle during the MOT Test. Before the test I had an opportunity to hold the revs higher to see if that cleared the exhaust and I am fairly sure there was the same variation between the gas analyser and what the MTXL & Tuner Studio were reporting.
The MTXL and Tuner Studio software are both showing the same value already (before and after the rewire). I may have misunderstood this but as they are already showing the same value as each other that would suggest that the wiring is ok. It was previously wired in to the ECU loom, but spliced in to wires in the footwell and without the relay on the +ve power feed. I could understand that sharing the same earth point could eliminate a difference between the MTXL gauge and the TS Software but that is not the problem I have.
Thinking this through, if disconnecting the yellow analogue output wire from the MTXL to the ECU effectively breaking the link, there should be a shift in the reading on the MTXL gauge if the earth is causing the fault. Obviously the ECU will not have a lambda value, but I have temporarily turned off the EGO control. Just tried this and there is no change in the MTXL gauge
I have noticed today that there appears to be a small purple switch inside the plug on the sensor with the words open and closed at each side of the switch. Could that be of any relevance. Reluctant to just change it as there is no mention of this in the innovate manual.
The MTXL and Tuner Studio software are both showing the same value already (before and after the rewire). I may have misunderstood this but as they are already showing the same value as each other that would suggest that the wiring is ok. It was previously wired in to the ECU loom, but spliced in to wires in the footwell and without the relay on the +ve power feed. I could understand that sharing the same earth point could eliminate a difference between the MTXL gauge and the TS Software but that is not the problem I have.
Thinking this through, if disconnecting the yellow analogue output wire from the MTXL to the ECU effectively breaking the link, there should be a shift in the reading on the MTXL gauge if the earth is causing the fault. Obviously the ECU will not have a lambda value, but I have temporarily turned off the EGO control. Just tried this and there is no change in the MTXL gauge
I have noticed today that there appears to be a small purple switch inside the plug on the sensor with the words open and closed at each side of the switch. Could that be of any relevance. Reluctant to just change it as there is no mention of this in the innovate manual.
Edited by Smokey Boyer on Saturday 27th February 15:07
Edited by Smokey Boyer on Saturday 27th February 16:13
Over the last 48 hours I have been reading a lot of information, trying to gain a better understanding what could be wrong and why the Lambda sensor could be reading lean while the car still smells like it is very rich.
This was perhaps the best website http://www.crypton.co.za/Tto%20know/Emissions/exha...
From this I concluded that I had a spark issue that was causing loads of unburnt fuel and oxygen to be dumped in to the exhaust. This could in theory cause the Lambda to be reading lean, while the car smells rich. I had previously ruled spark issues out, but decided to check again.
Today I found that there was no spark to cylinder 1, with the fault being the coil pack. Interestingly the plug in 1, seemed ok when I took it out last weekend to put a fresh set of plugs in.
I had always assumed a coil pack caused at least a pair of plugs to go off if it was faulty until I read this http://easyautodiagnostics.com/ford/1.9L-2.0L/how-...
It is staggering how little difference one cylinder being down makes to the engine in terms of sound and feeling. I would have expected it to sound and feel a little lumpy. I know the power will have been down and it has not been running efficiently, but I have not really gone mad over the winter and it was still quick enough to be lethal on wet and slippy roads.
Working back through my log of drives and work I have done to the car, I think the coil pack started to pack up around October/November time, and the car has not been used much since then. It is really the MOT emissions test that has highlighted an issue. I assume that the ECU has been working its socks off trying to keep the car running as best it can, and I have made adjustments to the fuel maps and ignition thinking that some of the other changes I had made to the exhaust and air intake had upset things.
This was perhaps the best website http://www.crypton.co.za/Tto%20know/Emissions/exha...
From this I concluded that I had a spark issue that was causing loads of unburnt fuel and oxygen to be dumped in to the exhaust. This could in theory cause the Lambda to be reading lean, while the car smells rich. I had previously ruled spark issues out, but decided to check again.
Today I found that there was no spark to cylinder 1, with the fault being the coil pack. Interestingly the plug in 1, seemed ok when I took it out last weekend to put a fresh set of plugs in.
I had always assumed a coil pack caused at least a pair of plugs to go off if it was faulty until I read this http://easyautodiagnostics.com/ford/1.9L-2.0L/how-...
It is staggering how little difference one cylinder being down makes to the engine in terms of sound and feeling. I would have expected it to sound and feel a little lumpy. I know the power will have been down and it has not been running efficiently, but I have not really gone mad over the winter and it was still quick enough to be lethal on wet and slippy roads.
Working back through my log of drives and work I have done to the car, I think the coil pack started to pack up around October/November time, and the car has not been used much since then. It is really the MOT emissions test that has highlighted an issue. I assume that the ECU has been working its socks off trying to keep the car running as best it can, and I have made adjustments to the fuel maps and ignition thinking that some of the other changes I had made to the exhaust and air intake had upset things.
Edited by Smokey Boyer on Sunday 28th February 20:47
Glad you found it, thanks for letting us know what the issue was.
Sorry for sending you down the wrong path, TBH your first post was a little long so I just skimmed over it before replying.
It looks like the cat was burning the unburned petrol and responsible for the change in lambda.
I think a failing injector(leaking or not closing) could cause the same symptoms.
I'm gathering parts for an MS2 conversion for my 500.
After numerous issues with the original ECU I decided to make the swap.
Do you have a build thread somewhere? I would be interested in reading it.
Fred
Sorry for sending you down the wrong path, TBH your first post was a little long so I just skimmed over it before replying.
It looks like the cat was burning the unburned petrol and responsible for the change in lambda.
I think a failing injector(leaking or not closing) could cause the same symptoms.
I'm gathering parts for an MS2 conversion for my 500.
After numerous issues with the original ECU I decided to make the swap.
Do you have a build thread somewhere? I would be interested in reading it.
Fred
Gassing Station | Chimaera | Top of Page | What's New | My Stuff