I think I read a post somewhere stating that some hose or wire, or something like that is on the bottom of the intake manifold - supposedly next to one or two EGR valves. I wonder if anyone can verify this.

Well, there appears to be two communication lines - the CAN and the "data link".. you think the "data link" might be CAN as well?

 

If there is an EGR system, its not controlled by the ECU in any way.

Ah, I forgot about the CAN data lines. At any rate, that stuff is not what I'm interested in.

 

There's a bunch more ADC lookups than for the Evo 8/9. There are two throttle position sensors, two accelerator pedal position sensors, two intake air temperature sensors, two O2 sensor offsets, and more.

 

Out of curiosity, have all the addresses that correspond to physical inputs been found (including what sensor it's for)?

 

Do you mean RAM addresses or something else, maybe pins on the processor itself?

 

Forgive my poor terminology. I'm not sure whether they would be ram addresses, or processor registers.

Example: if I want to check the state, or particular value of an input from an external sensor like a TPS.

 

ok, you mean the RAM addresses. I'm working on it now. I've deduced some, but still several more to deduce. I'll post up a full list of the sensor ADC RAM addresses when I figure it all out.

 

Awesome!

 

I've working on verifying existing tables. For the DBW throttle maps, the axis labeled as "Load" is not load. Its something completely different, but I haven't yet figured out what it is.

 

Maybe pedal position?

 

That would make sense, but nope. Its some value from another 2D or 3D table. The 3D lookup is based on load and rpm. Its bizarre. I'm wondering if these DBW throttle maps are are actually something else.

 

I noticed that these maps were a little different between the stock factory flash and the stock factory reflash to correct idle problems (JDM 52360016 & 52360018). So far, it seems like these are the only maps that were touched between the 2 revisions. Maybe it is something else?? I don't have the USDM revisions to checkout but I wonder if they adjusted the same maps with the USDM revisions.

 

Sheesh! It almost seems easier to just find out all the input/output addresses, and rewrite the ECU program. Bah.. I'm just yapping.

From the little that I do know, much of the system seems bizarre.

 

Brick wall, meet head. Head, meet brick wall. There is a big bloody spot on the brick wall next to my desk

Disassembly is can be super challenging, so it's impressive to see guys like y'all pick it up and run with it on your own.

 

I moved onto the boost control subroutines for now. The boost control algorithms are identical to the Evo 9, just coded in the m32r language. I've written ECUFlash definitions for all tables and passed them on to Tactrix.

As a side comment, since the Evo X already has a 3-bar MAP sensor, it should be a piece-of-cake to convert to psi-based boost control.