I'm starting this thread to get input from the srt-4 community.
I am going to develop my own engine tuning controller. This will most likely start as a piggy-back system to allow minor tweaks of the ECU operation. It may develop into a complete stand-alone system as the hardware and software matures. However, for right now it will just have lots of inputs and outputs to "tweak" signals just a bit to fine tune the engine without sacrificing drivability.
This isn't going to be a product. It is mainly for my own use. I will publish all my schematics, board layouts and software, so you're all free to use it at will, but it will require some programming ability and some hardware knowledge.
I realize things like this exist already, namely VEMS and megasquirt, but they just didn't suit what I wanted.
Get Codewarrior v7.0 for the 56800/E. They will also send you an 8k special edition license. That will should be enough for us, given the processor only has 8k words of program FLASH.
Then you will need an updated license file. Let me know when you're ready for it. I'll PM it to you. I had to work with their support team to get the correct 8k license for 7.2.
In later revs I will add the DSP directly to my own design, reducing the processor cost from $65 to $3.
Now, inputs and outputs. What will we need? I'm going to try and design for lots of future uses, but I will only start by using a few of these at a time. So, just because it is on this list doesn't mean you'll have to use it.
PWM Outputs @ 12 volts (6 available)
4 fuel injectors
W/I pump or HSV (see aquamist) or other variable speed motor drive
Ignition coil - 2 outputs
General Purpose 12 volt outputs
Boost solenoid
12 volt PWM inputs
measuring the stock ECU fuel injector drivers
ignition coil drive signal - 2 inputs
General purpose 12 volt inputs
Solenoid drive
Analog inputs, 0-5 V range (For measuring / intercepting analog sensors)
stock sensors like MAP, TIP, TPS
aftermarket sensors like wideband O2, EGT probe, W/I flow meter (self-tuning!)
Analog outputs, 0-5 V range (For outputing "modified" sensor outputs)
MAP, TIP, TPS
5V digital inputs
wheel speed sensors
crank position sensor
A few button inputs for real-time user adjustments.
5V digital ouputs
Driving LEDs, what else?
Serial port
laptop connection
GPS input (I've got some ideas for that one)
OBD scanning for extra sensor information
Other connections
LCD driver for real time displays
Secure Digital or MMC card interface for data logging
Ok. Let me know what you think. I'll update this post to include schematics as I create them.
New! Parts List for Design
DACs - MAX534 (8-bit, 4 outputs) or AD5318 (10-bit) or AD5328 (12-bit) (both 8 outputs)
Digital Design Summary:
On the main board there will be:
14 High Speed Digital I/Os (half inputs, half outputs)
16 Low-speed Digital I/Os (half inputs, half outputs)
11 Analog inputs
8 Analog outputs
2 Serial Ports
On the Remote Data unit there will be:
8 Digital Inputs (buttons, switches, etc)
8 Digital Outputs (LEDs, etc)
LCD
SD / MMC Interface
-Four injectors
-boost selenoid
-water injection(if you can find a fast enough selenoid)
Quote: Originally Posted by phrozen
-6 12 volt PWM inputs for measuring the duty cycle of ECU outputs (measuring the stock ECU fuel injector drivers, for example)
-Four injectors
-one of the two trigger wires on the coil(for rpms)
-boost selenoid
Quote: Originally Posted by phrozen
-8 Analog inputs, 0-5 V range. For measuring / intercepting analog sensors, such as the MAP, TIP, and TPS. Also for connecting things like a wideband O2 and EGT probe.
-You have five good readings sensors above
-water injection nozzle clog
-intake temp
-knock sensor
Quote: Originally Posted by phrozen
-8 analog outputs, 0-5 V range. For outputing "modified" sensor outputs. Think of this as a completely adjustable MAP clamp. You can create any clamping curve you like...
-Nothing else I can think of to use a 5volt out for.
Quote: Originally Posted by phrozen
-Some number of 5V digital inputs for sensors like the wheel speed sensors, crank position sensors (I'm assuming they output 5V logic pulses)
-sounds good, the wheel triggers combined with the rpms can be used to calculate the gear you are in, thus gear specific boost levels.
-crank sensor is a little trickier, we use a messed up tooth count. It can be done though.
Quote: Originally Posted by phrozen
-Some number of 5V digital ouputs
-Serial port for laptop connection or GPS input (I've got some ideas for that one)
- creating maps for boost and injector pulses( or make it self tune giving a specific a/f target.
Quote: Originally Posted by phrozen
-LCD driver for real time displays...
-A kick ass scrolling display of certain inputs across from the odometer display in the gauge cluster!
Quote: Originally Posted by phrozen
-A few button inputs for real-time adjustments.
-switching between stock and predefined boost/fuel maps, 2 stage rev limiter,yada yada yada
Quote: Originally Posted by phrozen
Ok. Let me know what you think. I'll update this post to include schematics as I create them.
First I don't know jack diddly about embedded processors .. but I am willing to help out where I can. I do alot of programming of various control systems utilizing PLC components. C and VB are hobbies. I know they are different worlds.
It seems to me that another alternative you could use to do this is to somehow read the program from the existing ECU and then reverse engineer it. It shouldn't be too difficult to figure out what inputs go where and what lookup tables are used. Then you could just adjust the lookup tables to adjust the exisiting PWM for instance. I imagine it operates on some type of linear curve probably between multiple setpoints to determine the various engine parameters. It may operate on some other formula but I would think it would be easier for DC to develop if they had some fixed points to reference for barometric pressure, boost, temperature etc.
Is this too much or perhaps the wrong path ya think? It would make it more difficult to integrate additional functions such as whatever you have in mind for the GPS (altitude readings?). But if the exisintg CPU is doing all the work already why not let it and just tweak it? This is the reason I have not gone in for a piggy back control of any kind yet. I am waiting for an integrated stand alone system because I think piggy backs introduce alot of potential issues.
ne hows .. my $.02 .. If I can do ne thing to help holler.
__________________
'06 M6 GTO
2006 Times:
13.065 @ 107.78 - 2.091 60' - 100% stock including 17" wheels and tires
12.761 @ 109.07 - 1.979 60' - Toyo T1R and Koni Adjustables Rear Only - Stock Tune
12.704 @ 108.63 Cutouts, K&N Drop in, Koni Rear Shocks, BFG Drag Radials
2007 Times:
??? - Added a 100 Shot ... we will see how it works out
People on here have been trying to reverse engineer this pcm since day one. I think it would be easier to just admit mother mopar got us good on this one and try to find away around it that makes the computer happy.
Ok. Some good ideas. A few things I probably can't do.
Probably can't change the rev limiter without creating my own ignition control. If the stock ECU stops giving me ignition pulses, I'll be at a loss for what to drive the ignition with.
Also, driving the LCD in the stock gauge cluster would require a lot of hacking of wiring and isn't worth it for such a small display. I'm talking about adding some small extra graphical display, perhaps next to the boost gauge.
The crank sensor shouldn't be too hard. As long is it gives some constant number of pulses per rotation... All you have to do is count them once and enter it as a constant.
The GPS input is to make a self-calibrating G-meter. That would take a whole new post to explain that idea. Basically, use GPS and the ABS rear-wheel speed sensors to very accurately determine the tire diameter while driving normally. Then with the calibration, you can use the very accurate wheel rear diameter and position measurement to calculate acceleration. The rest falls out easily. 1/4 mile timer, 0-60, torque (after putting in the car weight).
You just can't use the front wheels for this. Well, for obvious reasons.
People on here have been trying to reverse engineer this pcm since day one. I think it would be easier to just admit mother mopar got us good on this one and try to find away around it that makes the computer happy.
yes its called lawyers on your ass if you release any info if you cracked the code..
If I was an ECU hacker.. I wouldn't even touch the SRT-4 ECU unless it was for personal use only with no sharing with friends
__________________
http://www.hybridsrt.com
In remembrance of Andrew "HybridKid" Chin Shue
In remembrance of Steven "NightShade" Lorenz
DSRt, I thought the same thing too. I'll look into reverse engineering the PCM sometime, but I have a feeling that if DCX has a single good EE, they'll have the PCM locked down really good. The features are built into motorola processors. Just takes a few lines of code and it'll make it damn near impossible for us to get to it.
At least not without some help from the inside. PVO?
You probably can't get the code out of the stock ECU, but you could probably load new code. Of course, I'm not really up to the task of writing a whole engine controller with basically no knowledge of the hardware. Oh well.
So instead, minor tweaks... Change things a little bit at a time and make the ECU do what you want it to do. No?
yes its called lawyers on your ass if you release any info if you cracked the code..
If I was an ECU hacker.. I wouldn't even touch the SRT-4 ECU unless it was for personal use only with no sharing with friends
I'm with you on that one!
Phrozen, as far as the display goes, I'm not saying use the one already in the car, put another one right nest to it in the unused space under the other gauge. I know I was dreaming about the rev limiter, but, oh well.
Hehe. Nah. 4" dot-matrix LCD or so. We're going to have lots of information to display.
Menus, graphs, gauges... ;-)
Now, If i could just find a good color LCD that was easy to interface to... Mainly just to look pimp.
And to FLASH the word "SHIFT IDIOT!" in huge letters right before the rev limit.
Quote: Originally Posted by Sour Kraut
Phrozen, as far as the display goes, I'm not saying use the one already in the car, put another one right nest to it in the unused space under the other gauge.
Just keep an eye on this thread. I'll post code. Maybe even start a sourceforge project so people can contribute.
Quote: Originally Posted by DSrt
First I don't know jack diddly about embedded processors .. but I am willing to help out where I can. I do alot of programming of various control systems utilizing PLC components. C and VB are hobbies. I know they are different worlds.
Of course this whole conversation is simply a fantasy cuz ... Mama DC, you know we would never modify ne thing that came out of you!! Of course even if all of the source code fell off the back of a truck with instructions even I could understand I would NEVER share.
Ok. Some good ideas. A few things I probably can't do.
Probably can't change the rev limiter without creating my own ignition control. If the stock ECU stops giving me ignition pulses, I'll be at a loss for what to drive the ignition with.
Also, driving the LCD in the stock gauge cluster would require a lot of hacking of wiring and isn't worth it for such a small display. I'm talking about adding some small extra graphical display, perhaps next to the boost gauge.
The crank sensor shouldn't be too hard. As long is it gives some constant number of pulses per rotation... All you have to do is count them once and enter it as a constant.
The GPS input is to make a self-calibrating G-meter. That would take a whole new post to explain that idea. Basically, use GPS and the ABS rear-wheel speed sensors to very accurately determine the tire diameter while driving normally. Then with the calibration, you can use the very accurate wheel rear diameter and position measurement to calculate acceleration. The rest falls out easily. 1/4 mile timer, 0-60, torque (after putting in the car weight).
You just can't use the front wheels for this. Well, for obvious reasons.
You could control the REV limiter by poisening the signal right at the end of the stock limit. Make it show just a hair below the limit until you reach the new limit...The fuel and injector adjustment will be nil at these high rpms..also if you in control of fuel you can just adjust for that a bit if need be.
BTW I am down. I am also an engineer and would be interested in how I can help. I have in depth working knowlege of the sensors in this car (specs and such).
PM me with your phone number or I can send you mine. We should talk...I will also help to fund the effort.