"FMIC Rich Dip" partially caused by the ECU

td-d · **Posted:** Fri Jun 22, 2012 7:54 am

Pm me your current Ecuflash def, and I'll plug it in for you.

td-d · **Posted:** Mon Jun 25, 2012 5:24 pm

Ok, I've PM'd you with the definitions added.

bnguyen · **Joined:** Tue Sep 15, 2009 2:01 pm **Posts:** 5

I tried the new tables with NSFW's edits(Throttle, Innovate, FFB and OL Map):

NSFW and td-d, if y'all were local, I'd shake your hand. Thanks a ton! Power comes on smooth. I'm in the process of retuning/messing with a new mbc and getting rid of the rich dip is going to be great.

td-d · **Posted:** Sat Sep 01, 2012 4:19 pm

These are the addresses for A8DK100P, as requested by Patrick.

Code:

Table_Input_A_Rising_Load_Activation_ECT_Load 16 x 3

x axis                 .data.l h'C7A24
y axis                 .data.l flt_C9B50
Table data                 .data.l word_C9B5C      ; = 1.211e1


Table_Input_A_Falling_Load_Activation_ECT 16

x axis                  .data.l h'C7A24
Table data                  .data.l word_C81B2      ; = 6.906

Input_B_Activation_Load 16

x axis                 .data.l flt_C82A8
Table data                  .data.l word_C82D4      ; = 2.0 

td-d · **Posted:** Sat Sep 01, 2012 4:58 pm

AZ1G101N - note, for the '08 up, there are 3 input a rising, and 3 falling tables.

Code:

<scaling name="Input A Falling Load2" toexpr="x*0.00048828125" frexpr="x/0.00048828125" format="%.2f" min="-2.14748e+09" max="2.14748e+09" inc="5" storagetype="uint16" endian="big"/>

   <table name="Input A Falling Load Activation (ECT) A" address="C8D16" type="2D" level="1" scaling="Input A Falling Load2">
      <table name="X" address="C7C54" type="Y Axis" elements="16" scaling="rawecuvalue"/>
   </table>

   <table name="Input A Falling Load Activation (ECT) B" address="C8D36" type="2D" level="1" scaling="Input A Falling Load2">
      <table name="X" address="C7C54" type="Y Axis" elements="16" scaling="rawecuvalue"/>
   </table>
   
      <table name="Input A Falling Load Activation (ECT) C" address="C8D56" type="2D" level="1" scaling="Input A Falling Load2">
      <table name="X" address="C7C54" type="Y Axis" elements="16" scaling="rawecuvalue"/>
   </table>

   <table name="Input B Activation (Load)" address="C91C8" type="2D" level="1" scaling="Input A Falling Load2">
      <table name="X" address="C919C" type="Y Axis" elements="11" scaling="rawecuvalue"/>
   </table>

   <table name="Input A Rising Load Activation (ECT, Load)" address="CC388" type="3D" level="1" scaling="Input A Falling Load2">
      <table name="x" address="C7C54" type="X Axis" elements="16" scaling="rawecuvalue"/>
      <table name="y" address="CC37C" type="Y Axis" elements="3" scaling="rawecuvalue"/>
   </table>

   <table name="Input A Rising Load Activation (ECT, Load) B" address="CC3F4" type="3D" level="1" scaling="Input A Falling Load2">
      <table name="x" address="C7C54" type="X Axis" elements="16" scaling="rawecuvalue"/>
      <table name="y" address="CC3E8" type="Y Axis" elements="3" scaling="rawecuvalue"/>
   </table>

   <table name="Input A Rising Load Activation (ECT, Load) C" address="CC460" type="3D" level="1" scaling="Input A Falling Load2">
      <table name="x" address="C7C54" type="X Axis" elements="16" scaling="rawecuvalue"/>
      <table name="y" address="CC454" type="Y Axis" elements="3" scaling="rawecuvalue"/>
   </table> 

MFB · **Posted:** Sat Sep 01, 2012 8:46 pm

Thanks, looking forward to adddressing the rich afr on increased load. :-)

Patrick P931 · **Posted:** Sun Sep 02, 2012 12:55 am

Thank you!

So much new stuff in one weekend!

td-d · **Posted:** Sun Sep 02, 2012 3:48 am

Patrick P931 wrote:

Thank you!

So much new stuff in one weekend!

Enjoy

Let me know if you're looking for other 'alpha' defs for your rom - the only thing that's stopping me from back populating all the roms is the sheer volume of disassembly involved...

td-d · **Posted:** Sun Sep 02, 2012 2:03 pm

Interesting - the reason why the scale is different in the 08's up for the Tau tables is this routine:

byte_C72A6 is set to zero in all the roms which use the alternative scaling.

EDIT: the scaling is actually independent of Engine Load - it is MAP (in mmHg) * 0.01

Merp · **Posted:** Sun Sep 02, 2012 2:11 pm

Interesting.

This is next on my list of alpha definitions to be added to the git repo. I'll have the ones posted here knocked out this weekend along with an xml template.

WolfPlayer · **Joined:** Mon May 01, 2006 2:15 pm **Posts:** 124

In case there are any AccessTUNER Race Cobb-o-philes reading this, I've formally requested these tables to be added to ATR. That said, my personal guess is that we won't have access to them for quite some time. The effort to add those tables will need to be prioritized and there's higher priority things in the pipeline at the moment. Once it's in the pipeline, the normal project management milestones will have to be done: research on the tables across ROMS, understanding their impact, testing, etc.

I'm really looking forward to having access to these tables again. They are really helpful for those that have a FMIC.

t

td-d · **Posted:** Sun Sep 02, 2012 4:40 pm

Silly me, I missed the float value in there - so it looks like the scale is actually independant of engine load altogether.

A value of 0.01 is loaded and then multiplied by the MAP (in mmHg).

So in other words, for the AZ1G101N scaling of 5, 7 and 10 it's the equivalent of 500mmHg, 700mmHg and 1000mmHg.

Merp · **Posted:** Sun Sep 02, 2012 9:06 pm

Just pushed a bunch of these to the git repo alpha branch here: https://github.com/Merp/SubaruDefs/tree/Alpha

The template is set up for the three tables, with variants for the pre/post 2008 roms including the switch to the MAP Y axis.

I pulled everything from this thread, and did sanity checks on the data using ECUFlash for the following roms:
AZ1G101N
A8DK100P
A2WF200N
A8DH200X
A2ZJE11J
A2ZJB10J *note* all tables do not seem to port directly to the A2ZJB11J
A2WC522N

td-d · **Posted:** Mon Sep 03, 2012 1:31 am

Thanks Merp.

Btw, Input B activation (load) also uses the same pressure based scaling in the newer roms (where the axis extends from 1 to 16).

NSFW · **Joined:** Wed Nov 22, 2006 10:23 pm **Posts:** 2273

Nice job on the definitions at github, that's really cool!

RomRaider

Forum rules

"FMIC Rich Dip" partially caused by the ECU

Who is online