Mapping for Methanol in tank
when you get to an advance state of tune like this you will need to drive around with a laptop logging all the time. It will need constant work and development to get it right.
"You think you have a limit, As soon you touch this limit, something happens to you; suddenly you can go a little bit further. With your mind power, your determination, your instinct and your experience as well, you can fly very high."
Laptop is always in there and always connected/logging - much to everyone's hilarity 
The problem is not the wideband sensor as this is telling me lambda, but the narrowband sensor that the ECU uses to auto tune seems to be targeting AFR rather than lambda through the compensation/calibration table within the ECU maps. This map calls for 14.7 AFR based on a certain mA for the sensor output as per image below:
The injection system has loads of room left for the additional flow required.
The problem is not the wideband sensor as this is telling me lambda, but the narrowband sensor that the ECU uses to auto tune seems to be targeting AFR rather than lambda through the compensation/calibration table within the ECU maps. This map calls for 14.7 AFR based on a certain mA for the sensor output as per image below:
The injection system has loads of room left for the additional flow required.
Narrow band sensors are far to inaccurate for tuning even with the ecu. Are you sure the narrow band is used for all ranges. I suspect it will be used for idle and cruise and not for WOT applications.
I think you will struggle to get info or help from the tuning world becasue you want to publish the info for public digestion. I only ever share criticle or info that will give an edge with a few people and some info i will never share.
I can remember people taking the pee out of me for driving about with a laptop in the car all the time but the data can be invaluable.
I think you will struggle to get info or help from the tuning world becasue you want to publish the info for public digestion. I only ever share criticle or info that will give an edge with a few people and some info i will never share.
I can remember people taking the pee out of me for driving about with a laptop in the car all the time but the data can be invaluable.
"You think you have a limit, As soon you touch this limit, something happens to you; suddenly you can go a little bit further. With your mind power, your determination, your instinct and your experience as well, you can fly very high."
The narrowband sensor is used only in closed loop fueling. (i have a separate wideband O2 sensor that I log alongside all datalogs to check actual fuel against fuel mix targeted to enable the maps to be fine tuned)
There are 4 fuel correction levels within the RAM (as opposed to ROM) based on the amount of air flow at each point. These are ever changing and are based upon real time feedback from the O2 sensors. All of these are at low level MAF voltages though. Whatever is learned in these low levels is also applied at the higher MAF voltages as a broad brush approach to correcting fueling/mixture/system differences.
On the normal tesco 99 map that I have developed, these are all within 1.6% of what it wants to add/remove fuel wise. This is the same immediately after a reset as it is after a week or so of running the map.
On the methanol map and with 10% mix it 'learns' to add ~5.5% during this closed loop stage and then applies it to the open loop running. Without this added fuel in the open loop, the lambda is spot on what I am looking for but with it, it runs richer than I would want. This is why I am looking to re-calibrate the std narrowband sensor to cope with the different stoich value of the fuel.
Intricate details wont be published but the general learning process will be as it has been useful to read/apply others findings and come to my own conclusions for what works or doesn't work with my car as we all know that they can react differently to tuning. It will also be a good grounding for switching over to E85 fuel when this is more widely available and may shortcut the experimentation/learning process here.
There are 4 fuel correction levels within the RAM (as opposed to ROM) based on the amount of air flow at each point. These are ever changing and are based upon real time feedback from the O2 sensors. All of these are at low level MAF voltages though. Whatever is learned in these low levels is also applied at the higher MAF voltages as a broad brush approach to correcting fueling/mixture/system differences.
On the normal tesco 99 map that I have developed, these are all within 1.6% of what it wants to add/remove fuel wise. This is the same immediately after a reset as it is after a week or so of running the map.
On the methanol map and with 10% mix it 'learns' to add ~5.5% during this closed loop stage and then applies it to the open loop running. Without this added fuel in the open loop, the lambda is spot on what I am looking for but with it, it runs richer than I would want. This is why I am looking to re-calibrate the std narrowband sensor to cope with the different stoich value of the fuel.
Intricate details wont be published but the general learning process will be as it has been useful to read/apply others findings and come to my own conclusions for what works or doesn't work with my car as we all know that they can react differently to tuning. It will also be a good grounding for switching over to E85 fuel when this is more widely available and may shortcut the experimentation/learning process here.