CES eth content sensor installed, flex fuel patch coming soon
Jan 14, 2012, 11:21 AM
#92
Actually I just came from testing flexfuel patch in my IX, I wanted to test it on car with final version before writing here. I did not test it yet with sensor but I simulated ethanol content sensor with potentiometer and everything looked to work like expected. As my car is EDM I made this patch on the top of 8858 v7, of course code is portable to all evo's 7-9 if there is just little space on rom (not needed much). I used naming F=Flexfuel for this mod, ie. 8858F714.
It works as follows:
- ethanol content sensor is needed and ecu needs 0-5V signal from it, I used input pin 64 so map switching does not work anymore from that source (and on the other hand there is not anymore maps to change except boost maps)
- analog voltage is used for table look-up to get interpolation factor (see picture of tuning maps)
- based on value from this table program interpolates linearly between normal and alternate maps: 0%=normal maps only and 100%=alternate maps only, other values between are interpolated so if you tune your car with both std gasoline and E85, all values should be between these two maps and this idea should work pretty well.
- interpolation is made also between primer pulse maps to help starting in cold environment, injector interpolation factor is used for this
Also I made few safety and other features:
- analog input safety (low/high limits, forces interpolator values to safe values in such case if wire is broken). If safety activates, CEL is blinking to tell driver that something is not right.
- force interpolator factor to 0% or 100% for tuning purposes
- bypass safety function (force overrides safety)
How do you guys think? I have planned to port this code atleast to 9653. Also it still needs real testing on car with sensor.
It works as follows:
- ethanol content sensor is needed and ecu needs 0-5V signal from it, I used input pin 64 so map switching does not work anymore from that source (and on the other hand there is not anymore maps to change except boost maps)
- analog voltage is used for table look-up to get interpolation factor (see picture of tuning maps)
- based on value from this table program interpolates linearly between normal and alternate maps: 0%=normal maps only and 100%=alternate maps only, other values between are interpolated so if you tune your car with both std gasoline and E85, all values should be between these two maps and this idea should work pretty well.
- interpolation is made also between primer pulse maps to help starting in cold environment, injector interpolation factor is used for this
Also I made few safety and other features:
- analog input safety (low/high limits, forces interpolator values to safe values in such case if wire is broken). If safety activates, CEL is blinking to tell driver that something is not right.
- force interpolator factor to 0% or 100% for tuning purposes
- bypass safety function (force overrides safety)
How do you guys think? I have planned to port this code atleast to 9653. Also it still needs real testing on car with sensor.
Last edited by donner; Jan 15, 2012 at 05:13 AM.
Jan 14, 2012, 06:37 PM
#94
Donner, this looks great!
It would help a lot if you title it ECA Injector Interpolation instead of EC Sensor or it will get confused with Flex Fuel Sensor, which does not output 0-5V. We will have upset tuners who will think their Flex Fuel sensor is DOE.
It would help a lot if you title it ECA Injector Interpolation instead of EC Sensor or it will get confused with Flex Fuel Sensor, which does not output 0-5V. We will have upset tuners who will think their Flex Fuel sensor is DOE.
Jan 15, 2012, 12:10 AM
#96
We can change table names what ever is the best, this was just my work name for tables. With EC Sensor I meant whole measurement system generally (sensor+some analog output device).
Zeitronix, how fast your ECA is ready to output proper analog signal when you turn ignition on and what it does output when it is not ready? GM sensor readiness seems to be <250ms. In my existing program it will enter safety mode if voltage is outside range (low/high). When proper signal is got, program will automatically go to normal operation mode, look-up correct values for factors and stops blinking CEL as a sign for driver that car can be started. What I mean with this is that measurement should be ready as soon as possible so this wait time would be as short as possible. But if wait time I'm talking about is around 250-1000ms, it is more than ok but this is just good to know how my implementation is done.
Boost tables were not in scope in this first phase but of course possible to do that as well. Before I continue with new features (like boost) I want to make sure our idea of interpolation between maps works properly with sensor and there is sense to continue with this project. I don't have any hardware installed (or even sourced yet, I need to buy injectors and sensor) but I will get those in few months, start real testing and report results. And if it turns out our ideas work in real life I can tell you this patch comes out for sure
Zeitronix, how fast your ECA is ready to output proper analog signal when you turn ignition on and what it does output when it is not ready? GM sensor readiness seems to be <250ms. In my existing program it will enter safety mode if voltage is outside range (low/high). When proper signal is got, program will automatically go to normal operation mode, look-up correct values for factors and stops blinking CEL as a sign for driver that car can be started. What I mean with this is that measurement should be ready as soon as possible so this wait time would be as short as possible. But if wait time I'm talking about is around 250-1000ms, it is more than ok but this is just good to know how my implementation is done.
Boost tables were not in scope in this first phase but of course possible to do that as well. Before I continue with new features (like boost) I want to make sure our idea of interpolation between maps works properly with sensor and there is sense to continue with this project. I don't have any hardware installed (or even sourced yet, I need to buy injectors and sensor) but I will get those in few months, start real testing and report results. And if it turns out our ideas work in real life I can tell you this patch comes out for sure
Last edited by donner; Jan 16, 2012 at 04:24 AM.
Jan 16, 2012, 10:33 PM
#98
Ok guys, project goes to testing phase faster than I originally planned. I ordered few ethanol sensors last night and most probably will get few sets of FIC injectors too to prepare test cars (VIII and IX) and proove this thing out. When testing is made properly I will report results.
Jan 17, 2012, 08:27 PM
#99
Subscribed! I'm excited by a true flexfuel mod for the 4G63. 
I would be willing to test your mod for 96530706 on my 2005 USDM VIII when you are ready for beta testing. In California we often have to switch between 91 RON and E85 and the difference in tunes is difficult to manage when blending the two fuels. Just let me know what is required when you are ready.
I would be willing to test your mod for 96530706 on my 2005 USDM VIII when you are ready for beta testing. In California we often have to switch between 91 RON and E85 and the difference in tunes is difficult to manage when blending the two fuels. Just let me know what is required when you are ready.
Jan 18, 2012, 10:26 AM
#100
Subscribed! I'm excited by a true flexfuel mod for the 4G63.
I would be willing to test your mod for 96530706 on my 2005 USDM VIII when you are ready for beta testing. In California we often have to switch between 91 RON and E85 and the difference in tunes is difficult to manage when blending the two fuels. Just let me know what is required when you are ready.
I would be willing to test your mod for 96530706 on my 2005 USDM VIII when you are ready for beta testing. In California we often have to switch between 91 RON and E85 and the difference in tunes is difficult to manage when blending the two fuels. Just let me know what is required when you are ready.
If you are 9653 user you are lucky as I will take care of 9653 testing as I just ordered few sets of FIC1100 and will prepare also VIII to test this mod
Jan 29, 2012, 12:37 AM
#104
Newbie
Join Date: Nov 2008
Location: Post Falls, ID
Posts: 3
Likes: 0
Received 0 Likes
on
0 Posts
We are working on almost the same system for the Subaru roms as well.
See here: Subaru RomRaider Posts about Flex Fuel
I am writing code for an open-source ECA.
Mine has the capability to run either the Ford or GM sensors with the flick of a setup switch.
when it's complete and tested, i'll release the source and and parts list, so you can build it.. but to give you an idea here are my costs so far:
Arduino Uno: $35
VFD display: $10
Ford Taurus Flex Fuel Sensor: $8
Misc parts: about $5
Project box from radio shack: $15
Total cost so far: about $75, for a fully functioning ECA that can use EITHER SENSOR!
not to discount Zeitronix here, but I like building my own stuff, they make a great product. if you aren't a tinker-er my solution might not be for you.
See here: Subaru RomRaider Posts about Flex Fuel
I am writing code for an open-source ECA.
Mine has the capability to run either the Ford or GM sensors with the flick of a setup switch.
when it's complete and tested, i'll release the source and and parts list, so you can build it.. but to give you an idea here are my costs so far:
Arduino Uno: $35
VFD display: $10
Ford Taurus Flex Fuel Sensor: $8
Misc parts: about $5
Project box from radio shack: $15
Total cost so far: about $75, for a fully functioning ECA that can use EITHER SENSOR!
not to discount Zeitronix here, but I like building my own stuff, they make a great product. if you aren't a tinker-er my solution might not be for you.
Jan 29, 2012, 01:23 AM
#105
We are working on almost the same system for the Subaru roms as well.
See here: Subaru RomRaider Posts about Flex Fuel
I am writing code for an open-source ECA.
Mine has the capability to run either the Ford or GM sensors with the flick of a setup switch.
when it's complete and tested, i'll release the source and and parts list, so you can build it.. but to give you an idea here are my costs so far:
Arduino Uno: $35
VFD display: $10
Ford Taurus Flex Fuel Sensor: $8
Misc parts: about $5
Project box from radio shack: $15
Total cost so far: about $75, for a fully functioning ECA that can use EITHER SENSOR!
not to discount Zeitronix here, but I like building my own stuff, they make a great product. if you aren't a tinker-er my solution might not be for you.
See here: Subaru RomRaider Posts about Flex Fuel
I am writing code for an open-source ECA.
Mine has the capability to run either the Ford or GM sensors with the flick of a setup switch.
when it's complete and tested, i'll release the source and and parts list, so you can build it.. but to give you an idea here are my costs so far:
Arduino Uno: $35
VFD display: $10
Ford Taurus Flex Fuel Sensor: $8
Misc parts: about $5
Project box from radio shack: $15
Total cost so far: about $75, for a fully functioning ECA that can use EITHER SENSOR!
not to discount Zeitronix here, but I like building my own stuff, they make a great product. if you aren't a tinker-er my solution might not be for you.
Do you have part number and any datasheet for that sensor?Edit: quick google gives me info that only difference between ford and GM sensor is different frequency range, not a big deal for me as I'm building my own freq-to-volt device with microcontroller. Where did you get sensor for that price? Huh, just paid around 300$ for new GM sensor
Last edited by donner; Jan 29, 2012 at 01:28 AM.