Arduino based fuel injection

[Bullet_theory]

Yes i know there is megasquirt, a diy freeware based efi system, and other systems like it. And i know you can program an arduino (or similar) microprossesor boards as a piggy back system on a oem ecm module. I have a Arduino mega, that i used to make a freeware psp, i was bored. But i want to run an arduino mega board as a stand alone system (like megasquirt minus the 120-700 dollar cost), i know the crystal on the arduino board would need to be replaced with a 24mhz crystal to run the megasquirt freewarez, and id probably need to connect a mini lcd hud, and a few potentiometers, and some other daughterboards to connect to the wideband and such (in all the arduino board plus extras should run about 60 bucks)... im just wondering aside from the obvious (injectors fuel rail and such) does anyone have any thoughts on what id need to get the board to run the megasquirt warez to run as a diy efi management system?

[brodster]

It sounds like you are going to use all your own hardware right? Are you wanting to basically port the MegaSquirt code to your Arduino?

 

You would be using C++ for Arduino I take it. And I think the new MegaSquirt controller utilizes the same for changing the embedded code.

 

Just off the top of my head you will probably run into some issues with the port assignments and timers etc. So you would have to go through all the headers and main code. But again, I haven't researched either chip. It sounds like you know how to program just fine.

 

Also I would take a look at what hardware is on the new MegaSquirt controller and compare it to your Arduino Mega. Again, I am not sure what either is using for buffers, ADCs and DACs.

 

But all in all, if you can match the input/output functionality of the controller (assuming it can keep up with the demand), it should be a fun project. I have considered doing it myself. I couldn't tell you without researching the actual code, but something like this may be easier to completely redo using Megasquirt code as a rough model.

 

A couple of things: I am not sure your Arduino mega can support higher than 16 MHz operation. I googled it, and the board looks like it comes with a ATMEGA1280 processor. But that may not be a problem. I think the old chip was at 8 MHz. But it depends on how far you want to take.

 

[Bullet_theory]

It sounds like you are going to use all your own hardware right? Are you wanting to basically port the MegaSquirt code to your Arduino?

 

You would be using C++ for Arduino I take it. And I think the new MegaSquirt controller utilizes the same for changing the embedded code.

 

Just off the top of my head you will probably run into some issues with the port assignments and timers etc. Also I would take a look at what hardware is on the new MegaSquirt controller and compare it to your Arduino Mega. Again, I am not sure what either is using for buffers, ADCs and DACs.

 

But all in all, if you can match the input/output functionality of the controller (assuming it can keep up with the demand), it should be a fun project. I have considered doing it myself. I couldn't tell you without researching the actual code, but something like this may be easier to completely redo using Megasquirt code as a rough model.

 

A couple of things: I am not sure your Arduino mega can support higher than 16 MHz operation. I googled it, and the board looks like it comes with a ATMEGA1280 processor. But that may not be a problem. I think the old chip was at 8 MHz. But it depends on how far you want to take.

 

yeah i have my own hardware.. i was looking for a short cut in programming by using the megasquirt freeware

yes c++

i hadnt even had a single thought on the buffers (crap)

yeah i wanted to steer away from making a new code... but im aware that might happen

there is a website that specializes in changing the clock speed for the arduino boards, and they promise up to 64mhz speed (im unconvinced)

yes it dos have the atmega1280

 

and ive just stumbled across a site that has an article about a guy using a similar board with his bmw 325i (inplace of his standard ecu) but there seemed to be an issue...the board was maxed out and couldnt track throttle position, like the origional megasquirt that could only sence either full open or full close and just guessed the rest of the way thru the range.... but unlike the mega squirt there was hesitation, were the motor sounded like it was about to lean out

 

and id deffinately love to see what you do (if you ever actually do it)

[brodster]

I will do a bit of research on it and let you know what I find. Again, I am not sure how the Megasquirt directly interfaces to the sensors, etc. But I think it wouldn't be that complicated.

 

To be upfront I have NEVER installed or designed any kind of EFI system. I have researched Megasquirt a bit, but nothing specific to the code. However, I have started researching EFI systems in general, and building a basic system like Megasquirt in theory isn't that complicated: A bunch of input sensors, some computations and table lookups, and some injector output (and maybe ignition control). However the years of smoothing out the code and getting it to perform correctly on an engine is a whole other story...

 

Regardless, it could be a really fun project to learn all about it. I don't really have time to dive into something like that at the present, but it is something I have wanted to do. To be honest I just want to get my Datsun fuel injected first, by any means!:D

[Bullet_theory]

yah, it was kindof a spur of the moment idea this morning, so i havent done too much research... my wife asked me what the thingy (carburetor) was on the side of the engine and what does it do... and then she asked where the computer was... (shes blonde)... and finnaly flipped when i wast looking at the fast efi system (mostly due to the price)....and asked why i couldnt build one... *light bulb*

 

meh, ill play around with it for a while untill she caves in and i can but a megasquirt with out her killing me in my sleep

[brodster]

Haven't looked into the Arduino yet, but I started this thread the other day: http://community.ratsun.net/topic/29631-programmable-datsun-dizzy/

 

Hacked521 got me thinking about ecus. Then I started thinking about reprogramming ecus... Basically, I am kicking around the idea of using an old school Datsun/Nissan ECU and using it as a base. My idea was originally to make a programmable dizzy for L-Series engines. The KA24e dizzy looks very promising. It is easy to tap into an utilize. However, someone already built all the circuitry and it just needs to be tapped into.

 

In your case, the ecu should already have all the transistors to fire the injectors, etc. You could probably figure out how to mount your Arduino to it too. This will be a difficult project, but it MAY be a cheap way. Unfortunately someone is going to have to reverse engineer the circuit to figure out how it works. Maybe I will grab an old 280zx ECU or 200sx ECU and see how complicated it is inside.

 

Just some thoughts.