Aug 5 2005 - Starting to gather information. Downloading Megasquirt manual.
First sub goal is to master the 18F6722. This chip is roughly equivalent to the AVR. I am more experienced with the 16F family. Reviewed "silicon errata" of this chip. Need to better understand the limits of this chip by using an engineering kit issued by CCS. I will also need to etch a practice circuit board using this chip so I know its limits.
I recognize that I'm not going to be able to consult anyone on AVR issues. So be it. At worst I'll be no worse off than now.
Second sub goal is to carefully review Megasquirt manual and AVR source code. Need to better grasp operation of ECMs. Need not use "bleeding edge" version of Megasquirt II.
In particular I need to accomodate pollution controls, especially EGR and Canister. I think that the calculations are doable.
Third Sub goal is to choose appropriate modes of control. I'd be very happy to use an electronic carb on the target. I don't think I'll get my wish, given preliminary findings. I'll accept TBI.
Fourth Sub goal is actual assembly of the unit, followed by its programming, bench testing and then deployment.
Fifth subgoal is make it work and work properly.
Final goal is to enjoy the car and boast to my colleagues about a twenty eight year old car with a one year old brain.
Have been reviewing wide and narrow band Oxygen sensors. Scared off by cost of wide band but attracted by faster response time and wider range of operation. Have to decide when the time is right.
Target is a stock 1977 Caprice classic with a 305 Chevy engine, turbo 350 tranny. Condition is "rough" so I need not worry about keeping it stock. Would like to keep most of the system close to "stock", accepting mediocre mileage and performance as a trade off.
Smog gear is non-negotiable. It all stays. Yes the EGR and fuel vapor canister will complicate calculations - sorta.
Goals:
1. To make it work. My younger brother says I can't do it. I think he's wrong.
2. To make the car work efficiently as gasoline is growing more expensive. These big boats drink gas like an SUV.
3. I'd like the car to operate within clean air standards for its model year or better. I do not intend nor want to remove any pollution controls from the vehicle. Keep it simple, keep it legal.
4. Drivability is important. High speeds and break neck acceleration are unimportant.
5. Bragging rights. I want to be the first kid on my block with a homebrew ECM.
Using the firmware and saying "I did it" sounds interesting, but IMHO it's worth to consider other options. Eg. joining VemsFrontier/ArmUfo / VemsFrontier/ArmEfi or IonSense developments, or the displays or tuningsoftware that make them even more useful. This would mean
- learning more
- earning money instead of spending money
- using more code (all of the firmware is covered by licence, and some part is quite restrictive)
- being useful to the community and vems team
- faster rollout
- less work with the boring stuff (shared between team members. You might notice that getting chips in low quantity is hell, and PNP machines are better in SMD soldering than human)
- exploring functionality that is really new
I'm sure there is more to it, but hopefully even half of it is enough to consider joining the VEMS development team.