2008 USDM STI: Per-gear ignition timing [BETA]

g00fy · **Posted:** Wed Sep 14, 2011 7:27 pm

Question, are these per-gear tables also hidden on older ECU's, say an 05 FXT? I'm having issues with the 4eat and the different gears causing the engine to do wonky stuff like randomly knock.

dschultz · **Posted:** Wed Sep 14, 2011 8:59 pm

Not hidden as much as not defined for your ROM in the definition. I believe in most non-STi ROMs these tables are inactive and set to all 0's.

Fiend · **Joined:** Tue Apr 24, 2007 10:49 pm **Posts:** 243

It's been a while since I looked at the code but all 08+ cars seem to have similar per-gear logic but, as dschultz said, it is skipped or disabled for some. It wouldn't take much to enable it and modify the tables (or add tables if the existing ones are too small.)

Not sure if the code varies for older turbo cars like a FXT.

Airboy · **Posted:** Wed Nov 02, 2011 11:18 pm

Hope I did this right...

I separated the conversion multiplier from the compensation so that the tables have the "raw" compensation values instead of degrees. The multiplier address (CF85C) is given in the first post, toward the end.

Code:

   <scaling name="PerGearIgnitionCompensation (raw)" units="units" toexpr="x" frexpr="x" format="%.0f" min="0" max="31" inc="1" storagetype="uint8" endian="big"/>
    <scaling name="PerGearIgnitionCompensation multiplier" units="units" toexpr="x" frexpr="x" format="%.3f" min="-1" max="1" inc="0.01" storagetype="float" endian="big"/>
   <table name="Timing Compensation Per Gear Conversion Multiplier" category="Ignition Timing - Compensation" address="cf85c" scaling= "PerGearIgnitionCompensation multiplier" type="1D" level="1"/>
   <table name="Timing Compensation Per Gear (1st)" address="d23e4" scaling="PerGearIgnitionCompensation (raw)" >
      <table name="Engine Load" address="d23c8"/>
      <table name="Engine Speed" address="d23d8"/>
   </table>

   <table name="Timing Compensation Per Gear (2nd)" address="d240c" scaling="PerGearIgnitionCompensation (raw)" >
      <table name="Engine Load" address="d23f0"/>
      <table name="Engine Speed" address="d2400"/>
   </table>

   <table name="Timing Compensation Per Gear (3rd)" address="d2434" scaling="PerGearIgnitionCompensation (raw)" >
      <table name="Engine Load" address="d2418"/>
      <table name="Engine Speed" address="d2428"/>
   </table>

   <table name="Timing Compensation Per Gear (4th)" address="d245c" scaling="PerGearIgnitionCompensation (raw)">
      <table name="Engine Load" address="d2440"/>
      <table name="Engine Speed" address="d2450"/>
   </table>

   <table name="Timing Compensation Per Gear (5th)" address="d2484" scaling="PerGearIgnitionCompensation (raw)" >
      <table name="Engine Load" address="d2468"/>
      <table name="Engine Speed" address="d2478"/>
   </table>

   <table name="Timing Compensation Per Gear Activation (RPM)" address="cf84c">
   </table>

   <table name="Timing Compensation Per Gear Activation (Load)" address="cf854">
   </table>

Disclamer: Have not been tested

td-d · **Posted:** Thu Nov 03, 2011 12:56 pm

Airboy - I just did this too on an 08 EDM Sti, and as suggested above also broke out the multiplier setting it as 1 (and of course changing the scaling definition within Ecuflash as well). I subsequently changed the multiplier to 0.12, and revised the tables and scaling accordingly, as I felt it gave more resolution than going up an whole degree everytime.

All I can say is wow... it really brings the lower gears to life. You have to pay close attention to the base timing tables, as in most people's cases, if you've been road tuned, most of it would have been done in 3rd or 4th gear. I've bumped up timing by 3.5* in 1st; 2.5* in 2nd, and 1* in 3rd, and I suspect I could push it more.

What would be nice is if it were possible to define the scaling by referring to the actual multiplier (like a variable) as opposed to hardcoding it. As Dale said, one slip in terms of the multiplier could lead to bad doo-doo.

justune1 · **Joined:** Sat Mar 26, 2011 2:52 pm **Posts:** 34

Nice logic for the DD

thanks

Easty · **Posted:** Thu Jul 19, 2012 12:36 am

Hi,

Are these tables active in ADM 09 WRX’s if not would it be hard to activate them?

My Cal ID is AZ1G400W.

Regards

Dave.

Easty · **Posted:** Thu Aug 09, 2012 4:36 am

Does anyone know what the parameter is that causes it to use the 3rd table regardless of what gear you are in?

Easty · **Posted:** Sat Oct 13, 2012 12:45 am

If anyone is interested?

ADM 09 WRX AZ1G400W

</table>
<table name="Timing Compensation Per Gear (1st)" storageaddress="0xD0DDC" sizey="3">
<table type="X Axis" storageaddress="0xD0DBC" />
<table type="Y Axis" storageaddress="0xD0DD0" />
</table>
<table name="Timing Compensation Per Gear (2nd)" storageaddress="0xD0E0C" sizey="3">
<table type="X Axis" storageaddress="0xD0DEC" />
<table type="Y Axis" storageaddress="0xD0E00" />
</table>
<table name="Timing Compensation Per Gear (3rd)" storageaddress="0xD0E3C" sizey="3">
<table type="X Axis" storageaddress="0xD0E1C" />
<table type="Y Axis" storageaddress="0xD0E30" />
</table>
<table name="Timing Compensation Per Gear (4th)" storageaddress="0xD0E6C" sizey="3">
<table type="X Axis" storageaddress="0xD0E4C" />
<table type="Y Axis" storageaddress="0xD0E60" />
</table>
<table name="Timing Compensation Per Gear (5th)" storageaddress="0xD0E9C" sizey="3">
<table type="X Axis" storageaddress="0xD0E7C" />
<table type="Y Axis" storageaddress="0xD0E90" />
</table>
<table name="Timing Compensation Per Gear Activation (RPM)" storageaddress="0xCE4D8" />
<table name="Timing Compensation Per Gear Activation (Load)" storageaddress="0xCE4E0" />
<table name="Timing Compensation Per Gear Multiplier" storageaddress="0xCE4E8" />

Log Param's

FFFF7B60 Pergear Ignition Timing Compensation
FFFF7B68 Pergear Ignition Timing Compensation status

rtbrjason · **Posted:** Tue Oct 16, 2012 2:16 am

Does anyone know if this is available in the 08 spec b rom? (A2TB001L)

td-d · **Posted:** Fri Mar 15, 2013 8:30 pm

For endrswrd

Insert the following into your AE5IB00V.xml

Code:

<scaling name="Timing Comp Multiplier" units="multiplier" toexpr="x" frexpr="x" format="%.2f" min="0" max="1" inc=".1" storagetype="float" endian="big" />
<table name="PerGearTimingCompensation ConversionMultiplier" category="Ignition Timing - Compensation" address="ce2d0" type="1D" level="1" scaling="Timing Comp Multiplier"/>

also remember that if you want the numbers in the table in Ecuflash to show up as the actual degrees that will be added, you must find the following scaling in the 32bitbase.xml

<scaling name="PerGearIgnitionCompensation (degreesofcorrection)" units="Timing Compensation (degrees)" toexpr="x*-0.12" frexpr="x/-0.12" format="%.2f" min="-30.48" max="0" inc="0.12" storagetype="uint8" endian="big" />

And change the parts I've highlighted to whatever multiplier you choose - so if it's 1, replace -0.12 with 1. This is VERY important - otherwise you may forget that you've changed the multiplier, and see negative numbers in the table which are actually being applied as additive! To give you a worse case scenario, in some roms the tables can deduct close to 2 degrees - which reflects as 16 in the raw table value. If you change the multiplier to 1, and forget about it, your table (because of the scaling) will still show ~-2 degrees - but you would in fact be adding 16 degrees. Boom.

Merp · **Joined:** Thu Jul 23, 2009 5:46 pm **Posts:** 863

IMO, scalings should never reference another user exposed variable like that. Ought to just be the gradient/offset used in the lookup table and any universal unit conversions (in this case, just x), with a description that explains the logic.

td-d · **Posted:** Fri Mar 15, 2013 8:54 pm

Merp wrote:

IMO, scalings should never reference another user exposed variable like that. Ought to just be the gradient/offset used in the lookup table and any universal unit conversions (in this case, just x), with a description that explains the logic.

Yeah - I've been thinking we should change to the method that COBB uses - which is basically the raw value. That way you are forced to think in terms of the actual multiplier.

endrswrd · **Joined:** Thu Oct 25, 2012 6:09 pm **Posts:** 49

For my car this gives me enough to work with. But I agree with you guys to expose the multiplier instead of hard coding similar to the Cobb implementation. I gained awareness of this stuff through tuning a customer in ATR and once I realized it could be made positive vs the current implementation in my Definition I realized its usefulness!

Ziggyrama · **Posted:** Tue Apr 02, 2013 3:13 pm

In danger of spamming everybody, since I already opened a thread to request the defs, I'd like to request USDM 2011 defs for these. We do have per-gear timing tables already in defs but the multiplier is not exposed yet. It's likely because the formulas for the timing tables have the -0.12 multiplier hardcoded so having the multipler independently changed without changing the comp table view can be very dangerous.

Also, within the context, the boost tables would also be immensly useful to stage2 tuning. Greatly appreciate the support.

RomRaider

2008 USDM STI: Per-gear ignition timing [BETA]

Who is online