openecu.org

Using the posted 32 bit map addresses, i've been making an ecuflash xml for the 05 sti. I'm curious/concerned about the boost maps, as I think that I am missing something.
AJ930 8791 US STI 2005
Desired Boost A 000788DC 8 12 000c0e18 000c0e38 000c0e68 8 3f800000 00000000
Desired Boost B 000788F8 8 12 000c0f28 000c0f48 000c0f78 8 3f800000 00000000
Desired Boost C 00078914 8 12 000c1038 000c1058 000c1088 8 3f800000 00000000
Desired Boost D 00078930 8 12 000c1148 000c1168 000c1198 8 3f800000 00000000

Using this, I get a max boost target of 1.510 bar in the 80(load?) column, from 4000-6000 rpm. Problem is, the sti i thought maxes at 14.5 psi, and 1.510 bar is ~21 psi. That is either 7 psi too much or too little, depending on if it is relative or absolute. Any thoughts?

Divide by 93.7 from ECU to your MAP, reverse for map back to ECU, 16bit values, it seems to solve for psiG. Checks out on the 710J AJ243 and 500J AJ241 (2004 USDM STI's). Stock maps peak at 16.12 psiG.

ok, that looks good, but where does the 93.7 come from? also what do we think the x axis is? ranges from 0 to 80
edit: does the 93.7 come from the boost amto comp rpm multiplier value?
edit #2, to clarify the above comment.
There is are 2 boost atmos comp multipliers, that carry 4 values each vs an rpm value. These tables are duplicates. They carry a multiplier (from the struct) of 14208. That number, divided by 10000, (ie 1.4208), multiplied by the values in the stable, gives a value of 93.7728, hence my question.

gregsachs wrote:also what do we think the x axis is?

Most likely TPS.

I dunno where 93.7 came from, it is the standard built-in conversion in EcuEdit. But it seems to be right.

X axis is TPS. The axis goes from zero to 80%

I have it fully defined in this XML for Ecuedit:
http://freon.shackspace.com/car/ecu%20a ... ZJ500J.xml
There are a few 500J rom files here you can look at:
http://freon.shackspace.com/car/ecu%20a ... ng/openecu

Make sure to use the latest ecuedit, 1.2.0.66b, since it just added 32bit float/IEEE-754 support.

All the axes are in 32bit float. I believe only Ecuedit supports this so far, and it only supports it if you manually go hack the XML--it isn't in the GUI F9 editor yet. Either way it works.

Freon wrote:I dunno where 93.7 came from, it is the standard built-in conversion in EcuEdit. But it seems to be right.

I don't know about that... the boost maps in the 2.0L WRX ROMs are in absolute pressure (2.0 in ROM = 1.0bar = 14.5psi)

Dividing by 750 seems to be much more like what I would expect to see.

If you *really* want to see psi, try using ([value]/750-1)*14.5

edit: based on other stuff in the ROM, maybe 760 would be better?

Yeah for some reason I want to agree, I think it should be in absolute, but it works out perfectly with that conversion. The map definition only has a multiplier of 1 (3F800000) and no addition factor (0000000). I'm not sure what that unit represents exactly.