Open Import Fuel Tuner (VERY long, pls read!)
10-12-2001, 04:33 PM
#1
New User
Thread Starter
Join Date: Sep 2001
Location: Dallas, TX, US
Posts: 5
Likes: 0
Received 0 Likes
on
0 Posts
Open Import Fuel Tuner (VERY long, pls read!)
Well okay, here goes nothing.. I've had this idea in my head for a few weeks which won't go away, and it's time to tell the world 
First of all, let me just say here that I love the JRSC.. There are a lot of turbo, nos, and n/a fanatics around, and that's great.. But I love the JRSC in particular for a few important reasons: 1) It adds a significant amount of ***-kicking power. 2) It's (supposed to be) cheaper than a turbo. 3) It's (supposed to be) easier to install than a turbo. 4) It's (supposed to be) more reliable than a turbo system 5) It preserves the Indy-like torque curve of the Honda motor. Overall it's been a hell of a lot of fun to drive. However despite these advantages, I have discovered in my recent but brief JRSC experiences that the system does have some serious drawbacks.. For the purposes of this discourse, I'll divide the main problems into two categories: Timing Control and Fuel Control. I think that the JR BTC, MSD BTM, and J&S Ultraguard have pretty much taken care of the timing control. But Fuel Control is another story altogether. It seems that the typical scenario goes like this:
1) I think that the JR kit will be complete; ie take care of fuel issues.
2) How wrong I am - to get rid of tip-in detonation, I need the JR MAP Controller - $170.00
3) I turn up the boost a little, and it turns out the stock fuel pump can't handle outrageous fuel pressures, so I need the Walbro - $120.00
4) I decide to use different injectors, to lower the pressure - $300.00
5) But I need an adjustable rise-rate FMU for the injectors - Cartech - $200.00
6) And it won't idle for crap, so I need to lower base fuel pressure with a new FPR - AEM - $145.00
7) Finally I'd like to tune it, so that will require a V-AFC - $400.00
So by the time all this is done, I've invested $1335.00 in fuel management which should have been taken care of from the beginning. This is a bad solution. So of course now everyone will say, "Just get a Hondata" but that thing is insanely overpriced too. By the time you're done paying for the Hondata unit itself, plus the stupid software, plus the OBD-I ECU and $250 wiring harness, plus dyno time, you've spent almost as much $$ as what the supercharger itself cost. That may be the ideal solution from a bottomless-pocket standpoint, but it just doesn't sit well with me.. I think we should be able to get the JRSC Honda system running reliably and well-tuned without dropping another several grand.
This stuff doesn't have to be expensive. A creative mind and an eye for bargains can get you far. When I was in college I designed a digital dash display for a friend with a tricked out 87 Chevy Cavalier. I'm not joking - the thing looked like a space shuttle cockpit. Two flourescent backlit LCD graphic displays, with digital & analog speedo, fuel level, odometer, trip meter, digital voltage meter, temperature display. Plus a Race display which detected "racing mode" (when you hammered it), graphed speed vs time, and displayed calculated 0-60 time and 1/4 mile time and speed. No **** - check it out under hardware @ http://www.fascinationsoftware.com. How much would you expect to pay for this? It cost him $250 for _ALL_ the electronics - embedded computer, displays, all of it. He spent another $700 on a stainless steel plate and fabrication work. But the point is - this stuff doesn't have to be so expensive, and we can put some of ourselves into the work.
Of course now you probably see where I'm going with this train of thought.. I'd like us to make our own special JRSC honda all-purpose fuel controller.. I'm calling it the Open Import Fuel Tuner, for reasons which will become apparent shortly.. Here are my design constraints:
1) Inexpensive
2) Duplicate the functionality of the JR MAP mod (IAT trick)
3) Duplicate the TPS trick
4) Provide fuel tunability vs RPM (ala VAFC)
5) Provide a MAP remapping function (for big injectors or replace JR MAP MOD)
6) All source code and docs should be freely available, and all software tools should be free.
With a system like this a person could could use the stock fuel pump and regulators, and only upgrade the injectors to achieve complete fuel control and tunability while forgetting about the $2k Hondata or other solutions. I'm intending to GPL the code so that anyone who wants will be ablke to tweak it and add more features. I'll also be posting _all_ the schematics and documentation on my website so that anyone may make their own for the cost of the parts. I'm not planning on interfacing a display or keypad for the first version of the device, for two reasons: 1) an small LCD display sucks for tuning purposes - you can't get enough information from it. and 2) You shouldn't be tuning while driving anyway. There will be a high-speed serial port for laptop hookup, and a badass Win32 app for tuning the car.. And of course the source code for that program will also be GPL'd and available on my website.. By foregoing the display/keypad, I think I'll be able to keep the cost of the parts very close to the price of a JR MAP controller alone.
The electrical hook-up for a system like this would be pretty simple. I envision it like this:
IAT ___/----------\_____ ECU IAT
O2 ___|-----------|
TPS ___|---OIFT--|_____ ECU TPS
TDC ___|---BOX---|
IGN ___|-----------|
MAP ___\----------/_____ ECU MAP
So I'm pumped.. It seems to me that if we get together we can create a tuning computer that is: 1) extremely cheap, 2) almost infinitely extensible by anyone who can learn 68HC16 assembly language (piece of cake ), and 3) exactly suited to our engine control needs. I believe that this controller could solve all of our fuel delivery problems and allow us to achieve excellent tunabiliy with only the added cost of the parts for the controller and big injectors. Here's what I'm envisioning for the major software features of the first version:
1) Unlimited MAP sensor re-mapping. By using a piecewise linear translation with a selectable number of points (up to 128), you can just put on big injectors with the stock fuel pressure and re-map the sensor so that a MAP input in the vacuum region will be 'compressed' to a small range of high-vac output for good idling, while the boost range is expanded (but still kept below 3v so as not to trip CEL 03) to dump lots of fuel for boost.. No need for increasing fuel pressure.
2) RPM-based fuel adjustment, like VAFC, for tuning, with a selectable number of RPM points (up to 128 or so).
3) IAT tricking - modifying IAT signal for selectable arbitrary conditions.
4) TPS tricking - modifying TPS signal for selectable arbitrary conditions.
5) Data dump - Ok, hate to say it but - OBD-2 datalogging pretty much sucks. 5Hz is just too slow to fully capture the dynamics of the engine.. But there's no reason it has to be that slow. At 115kbps, it's possible to send 5 or 6 signal values at hundreds of hertz.. And I'm sure the cpu will be fast enough to move that much data - it's the beauty of Assembly Language. I think it would be very useful to send: Intake temp, O2 sensor voltage, RPM, and Timing Advance at 500Hz or so. Coupled with a performance-oriented Win32 logging and tuning program, I think a guy (or a girl) could do some mad tuning to a JRSC honda for low dollars and be very pleased with the results. Trading tuning curves and settings will be a joke - forget about this password-protected BS.
Of course for those of us who don't have the desire or time to build up an electronic device, there is a good possibility that I will find a way to offer pre-built units for sale.
Wouldn't this thing be cool? I probably never would have built this thing alone, but one of my best friends (who is a mad DSM tuner) has really started going gung-ho on this idea, and it's developing some momentum. He's already told several of his DSM comrades, and apparently there has been a strong response. I just got eight different 68hc16 manuals from UPS today - several thousand pages for a grand total of $0.00 (Damn I love Motorola). We'll be making a version for DSMs that interfaces with a Karman sensor instead of the MAP. Since this is going to be a completely open project, anyone with a little EE/comsci experience could port the project to work on a totally different car.
But this will be a very time-consuming and long project if I have to do everything myself. I'm completely open to anyone who would like to donate their time and/or expertise. I need help mainly from these areas:
1. Ideas - what to implement
2. Information & expertise - low-level honda info, the numbers. PSI, volts, etc.
3. Facilities - Dyno time
4. Testing - people with different cars for testing/tuning feedback
It will be a long-term project, since I have a full-time job. Probably 9 months at least before it could be released to the general tuner public. But if you feel like you could donate something to an effort like this, and would like to, please email me..
If anyone has any thoughts on why this would/would not be a good project, speak up!
Richard
SirRichard@fascinationsoftware.com
First of all, let me just say here that I love the JRSC.. There are a lot of turbo, nos, and n/a fanatics around, and that's great.. But I love the JRSC in particular for a few important reasons: 1) It adds a significant amount of ***-kicking power. 2) It's (supposed to be) cheaper than a turbo. 3) It's (supposed to be) easier to install than a turbo. 4) It's (supposed to be) more reliable than a turbo system 5) It preserves the Indy-like torque curve of the Honda motor. Overall it's been a hell of a lot of fun to drive. However despite these advantages, I have discovered in my recent but brief JRSC experiences that the system does have some serious drawbacks.. For the purposes of this discourse, I'll divide the main problems into two categories: Timing Control and Fuel Control. I think that the JR BTC, MSD BTM, and J&S Ultraguard have pretty much taken care of the timing control. But Fuel Control is another story altogether. It seems that the typical scenario goes like this:
1) I think that the JR kit will be complete; ie take care of fuel issues.
2) How wrong I am - to get rid of tip-in detonation, I need the JR MAP Controller - $170.00
3) I turn up the boost a little, and it turns out the stock fuel pump can't handle outrageous fuel pressures, so I need the Walbro - $120.00
4) I decide to use different injectors, to lower the pressure - $300.00
5) But I need an adjustable rise-rate FMU for the injectors - Cartech - $200.00
6) And it won't idle for crap, so I need to lower base fuel pressure with a new FPR - AEM - $145.00
7) Finally I'd like to tune it, so that will require a V-AFC - $400.00
So by the time all this is done, I've invested $1335.00 in fuel management which should have been taken care of from the beginning. This is a bad solution. So of course now everyone will say, "Just get a Hondata" but that thing is insanely overpriced too. By the time you're done paying for the Hondata unit itself, plus the stupid software, plus the OBD-I ECU and $250 wiring harness, plus dyno time, you've spent almost as much $$ as what the supercharger itself cost. That may be the ideal solution from a bottomless-pocket standpoint, but it just doesn't sit well with me.. I think we should be able to get the JRSC Honda system running reliably and well-tuned without dropping another several grand.
This stuff doesn't have to be expensive. A creative mind and an eye for bargains can get you far. When I was in college I designed a digital dash display for a friend with a tricked out 87 Chevy Cavalier. I'm not joking - the thing looked like a space shuttle cockpit. Two flourescent backlit LCD graphic displays, with digital & analog speedo, fuel level, odometer, trip meter, digital voltage meter, temperature display. Plus a Race display which detected "racing mode" (when you hammered it), graphed speed vs time, and displayed calculated 0-60 time and 1/4 mile time and speed. No **** - check it out under hardware @ http://www.fascinationsoftware.com. How much would you expect to pay for this? It cost him $250 for _ALL_ the electronics - embedded computer, displays, all of it. He spent another $700 on a stainless steel plate and fabrication work. But the point is - this stuff doesn't have to be so expensive, and we can put some of ourselves into the work.
Of course now you probably see where I'm going with this train of thought.. I'd like us to make our own special JRSC honda all-purpose fuel controller.. I'm calling it the Open Import Fuel Tuner, for reasons which will become apparent shortly.. Here are my design constraints:
1) Inexpensive
2) Duplicate the functionality of the JR MAP mod (IAT trick)
3) Duplicate the TPS trick
4) Provide fuel tunability vs RPM (ala VAFC)
5) Provide a MAP remapping function (for big injectors or replace JR MAP MOD)
6) All source code and docs should be freely available, and all software tools should be free.
With a system like this a person could could use the stock fuel pump and regulators, and only upgrade the injectors to achieve complete fuel control and tunability while forgetting about the $2k Hondata or other solutions. I'm intending to GPL the code so that anyone who wants will be ablke to tweak it and add more features. I'll also be posting _all_ the schematics and documentation on my website so that anyone may make their own for the cost of the parts. I'm not planning on interfacing a display or keypad for the first version of the device, for two reasons: 1) an small LCD display sucks for tuning purposes - you can't get enough information from it. and 2) You shouldn't be tuning while driving anyway. There will be a high-speed serial port for laptop hookup, and a badass Win32 app for tuning the car.. And of course the source code for that program will also be GPL'd and available on my website.. By foregoing the display/keypad, I think I'll be able to keep the cost of the parts very close to the price of a JR MAP controller alone.
The electrical hook-up for a system like this would be pretty simple. I envision it like this:
IAT ___/----------\_____ ECU IAT
O2 ___|-----------|
TPS ___|---OIFT--|_____ ECU TPS
TDC ___|---BOX---|
IGN ___|-----------|
MAP ___\----------/_____ ECU MAP
So I'm pumped.. It seems to me that if we get together we can create a tuning computer that is: 1) extremely cheap, 2) almost infinitely extensible by anyone who can learn 68HC16 assembly language (piece of cake
), and 3) exactly suited to our engine control needs. I believe that this controller could solve all of our fuel delivery problems and allow us to achieve excellent tunabiliy with only the added cost of the parts for the controller and big injectors. Here's what I'm envisioning for the major software features of the first version:1) Unlimited MAP sensor re-mapping. By using a piecewise linear translation with a selectable number of points (up to 128), you can just put on big injectors with the stock fuel pressure and re-map the sensor so that a MAP input in the vacuum region will be 'compressed' to a small range of high-vac output for good idling, while the boost range is expanded (but still kept below 3v so as not to trip CEL 03) to dump lots of fuel for boost.. No need for increasing fuel pressure.
2) RPM-based fuel adjustment, like VAFC, for tuning, with a selectable number of RPM points (up to 128 or so).
3) IAT tricking - modifying IAT signal for selectable arbitrary conditions.
4) TPS tricking - modifying TPS signal for selectable arbitrary conditions.
5) Data dump - Ok, hate to say it but - OBD-2 datalogging pretty much sucks. 5Hz is just too slow to fully capture the dynamics of the engine.. But there's no reason it has to be that slow. At 115kbps, it's possible to send 5 or 6 signal values at hundreds of hertz.. And I'm sure the cpu will be fast enough to move that much data - it's the beauty of Assembly Language.
I think it would be very useful to send: Intake temp, O2 sensor voltage, RPM, and Timing Advance at 500Hz or so. Coupled with a performance-oriented Win32 logging and tuning program, I think a guy (or a girl) could do some mad tuning to a JRSC honda for low dollars and be very pleased with the results. Trading tuning curves and settings will be a joke - forget about this password-protected BS.Of course for those of us who don't have the desire or time to build up an electronic device, there is a good possibility that I will find a way to offer pre-built units for sale.
Wouldn't this thing be cool? I probably never would have built this thing alone, but one of my best friends (who is a mad DSM tuner) has really started going gung-ho on this idea, and it's developing some momentum. He's already told several of his DSM comrades, and apparently there has been a strong response. I just got eight different 68hc16 manuals from UPS today - several thousand pages for a grand total of $0.00 (Damn I love Motorola). We'll be making a version for DSMs that interfaces with a Karman sensor instead of the MAP. Since this is going to be a completely open project, anyone with a little EE/comsci experience could port the project to work on a totally different car.
But this will be a very time-consuming and long project if I have to do everything myself. I'm completely open to anyone who would like to donate their time and/or expertise. I need help mainly from these areas:
1. Ideas - what to implement
2. Information & expertise - low-level honda info, the numbers. PSI, volts, etc.
3. Facilities - Dyno time
4. Testing - people with different cars for testing/tuning feedback
It will be a long-term project, since I have a full-time job. Probably 9 months at least before it could be released to the general tuner public. But if you feel like you could donate something to an effort like this, and would like to, please email me..
If anyone has any thoughts on why this would/would not be a good project, speak up!
Richard
SirRichard@fascinationsoftware.com
10-12-2001, 05:48 PM
#2
Honda-Tech Member
Join Date: Apr 2000
Location: Toronto, ON, Canada
Posts: 996
Likes: 0
Received 0 Likes
on
0 Posts
Re: Open Import Fuel Tuner (NeedToGoFASTER)
This sounds like what ZIP has already done.
http://www.clubsi.com/whats_hot.shtml
ZIP22 EBL (Electronic Boost Limiter):
a) TPS mod: to help eliminate tip-in detonation
b) MAP mod: to richen/lean out fuel delivery (i.e. similar to the Apexi VAFC)
c) VTEC controller: adjustable VTEC engagement from 2500-5300rpms
d) IAT mod with intelligence: allows enrichment of fuel in the low and mid-range rpms but will cut-off at a user programmable rpm before the Honda ECU fuel dump to allow a perfect mixture across the motor’s rev range.
e) AF warning alert: audibly and visually warns the driver if the air and fuel mixture is getting to lean at WOT
f) VTEC light: this mountable LED informs the driver that the VTEC solenoid has begun to spool
g) Scramble boost: this feature allows maximum boost at press of a button
h) Valet feature: keep other drivers from tampering with your settings and allows no boost to be produced by the supercharger
http://www.clubsi.com/whats_hot.shtml
ZIP22 EBL (Electronic Boost Limiter):
a) TPS mod: to help eliminate tip-in detonation
b) MAP mod: to richen/lean out fuel delivery (i.e. similar to the Apexi VAFC)
c) VTEC controller: adjustable VTEC engagement from 2500-5300rpms
d) IAT mod with intelligence: allows enrichment of fuel in the low and mid-range rpms but will cut-off at a user programmable rpm before the Honda ECU fuel dump to allow a perfect mixture across the motor’s rev range.
e) AF warning alert: audibly and visually warns the driver if the air and fuel mixture is getting to lean at WOT
f) VTEC light: this mountable LED informs the driver that the VTEC solenoid has begun to spool
g) Scramble boost: this feature allows maximum boost at press of a button
h) Valet feature: keep other drivers from tampering with your settings and allows no boost to be produced by the supercharger
10-12-2001, 07:13 PM
#3
New User
Thread Starter
Join Date: Sep 2001
Location: Dallas, TX, US
Posts: 5
Likes: 0
Received 0 Likes
on
0 Posts
Re: Open Import Fuel Tuner (DirtyLude)
Right - it looks like Zip22 has a really sweet device.. But there is one very cool thing about this project which no-one has done yet: it will be free to be modified/tweaked/improved by anyone.. For instance, I'm not sure whether the EBL can translate the MAP sensor signal in the way that I envision. But I know that I'll be able to code this thing to do whatever I want.. Plus it will be really inexpensive, and I'll be able to write really cool tuning software.. Think about it: if someone got the bright idea that they wanted to jack an EGT sensor in and wire it up the box, they code do it and write code to support it.
Richard
Richard
Thread
Thread Starter
Forum
Replies
Last Post