Tom-Guy

If you see it as viable, go for it. There is a reason it was never done. The dataflow of a poorly documented processor with lack of support by way of IDEs, compilers, tools, and a bank of readily tweakable code snippets is not something you can go have fun reverse engineering on your own - it's a goddamn PITA and it's only been done once, correctly, despite all the pretenders (some multi-million dollar) to the throne.

0b00st0

DING DING DING! We have a winner.



Why no one has simply replaced the MCU is beyond me. Any Motorola, Microchip, or Atmel product would fit the bill nicely. Sigh, I wish I had some free time.

Tom-Guy

If you think the OKI MCU is the only, or even main, problem then you need to learn about undocumented ASICs./
/
If you want an easy reverse engineer, clone an AEM EMS. If you want it to be worth a ---- with VR/mag sensor cars (Honda, Toyota, others) then clone the H1C1 board on OBD1 Honda and incorporate it into the adapter header. Incorporate an FTDI chip if you want USB like the v2.0. In quantity you should be able to get the boards, populated, for sub-$100. That makes much more sense as now you have a universal unit with the ability to configure different pickup configurations.

0b00st0

Oh ----, don't get me started on that POS ASIC in there.


JD, give me few years. If I still care then, we'll see.

Tom-Guy

Have fun with that, the components on them are starting to decompose from age. I suggest you start from scratch and utilize known good sensor/actuator circuits (refer to p28.pdf), and incorporate your own wideband if you want. Don't expect it to sell any better for having an onboard wideband, I have yet to see a suffix -U AEM box as only tuner geeks care about that ---- not the sort of people who buy parts for their cars.