I designed a DeviceNet interface for the controller that I am using for industrial applications. It was based on a canbus chipset and the hardware layer is the same as canbus, just need to figure out the canbus protocol . . . if it were easy, girls could do it
I am using a
motion controller board that has a Motorola DSP56303 processor and it has a PC104 bus on it as well so that an Intel based single board computer can host the system. I have all the code roughed out and tested on the engine control side and have loaded FreeDOS on the intel processor and am running a small graphical user interface from it. The OS and application files all fit on a 32Mb compact flash.
I know it doesn't sound very practical, but I have dozens of these systems that we have developed for the semiconductor industry to control robotic workcells for Seagate hard drive assembly lines. I know the architecture and it has great digital and analog I/O capabilities as well as deterministic scan times (I can set up a routine to manage spark / injector timeing and set/reset of digital I/O with a bandwidth of up to 20kHz scan rate.) The distributor is now a 1024 line incremental encoder with a marker pulse for absolute reference after the first rev. Three coils will fire 6 cylinders with 2 plugs firing at a time (one on the compression stroke and the other on the exhaust stroke).
I'll look at the references you put up and map out a strategy. On the one hand, I would probably rather use a 5 spd overdrive, on the other hand, I really want something my wife can drive. Either way, the electronic hacking I am doing is a lot of fun, just hope I don't screw it up and create the worlds "second" most expensive boat anchor.
