Speed Density ROM - A8DH201X 2006 WRX

Merp · **Posted:** Fri May 11, 2012 9:38 pm

The 04 FXT is in the queue, and the 08/09 USDM STi will be released very soon

mrf582 · **Posted:** Wed Aug 01, 2012 4:13 pm

td-d has defined some more things for a USDM 2006 WRX that aren't in the standard ECUFlash definitions. I'll copy/paste them here if you want to incorporate them into your Github.

Code:

   <scaling name="overrun" units="units" toexpr="x*0.1953125" frexpr="x/0.1953125" format="%.0f" min="0" max="255" inc="1" storagetype="int16" endian="big"/>

   <table name="Over run fueling " address="c97f0" type="2D" level="1" scaling="overrun">
      <table name="Engine Coolant Temperature" address="c8974" type="X Axis" elements="16" scaling="CoolantTemp(DegreesF)"/>
   </table>

   <table name="Overrun_Enrich_RPM_Delta_Activation" address="C8744" type="1D" level="1" scaling="rawecuvalue"/>

   <table name="Overrun_initial_enrichment" address="C8748" type="1D" level="1" scaling="ms"/>

   <table name="Overrun_Fueling_Cut_Counter_RPM_Threshold" address="C8790" type="2D" level="1" scaling="RPM">
      <table name="RPM" type="Static X Axis" elements="3">
         <data>Low</data>
         <data>Med</data>
         <data>High</data>
      </table>
   </table>
<table name="A/F Correction #3 Limits" address="2FF5C" type="2D" level="1" scaling="A/FLearning#1Limits(%)">
      <table name="X" type="Static X Axis" elements="2">
         <data>High</data>
         <data>Low</data>
      </table>
   </table>

 <scaling name="AF3 Lambda" units="Lambda Additive" toexpr="x" frexpr="x" format="%.4f" min="0" max="255" inc="1" storagetype="float" endian="big"/>

         <table name="CL Fueling Target Compensation - Rear O2 " address="2E020" type="2D" level="1" scaling="AF3 Lambda">
      <table name="X" type="Static X Axis" elements="2">
         <data>High</data>
         <data>Low</data>
      </table>
   </table>

<table name="A/F Learning Max threshhold" address="C8584" type="1D" level="1" scaling="MassAirflow(g/s)1"/>
<table name="A/F Learning Min threshhold" address="C8574" type="1D" level="1" scaling="MassAirflow(g/s)1"/>

   <table name="Fuel Pump Duty" category="NSFW Beta" address="3d4e4" type="2D" level="1" scaling="rawecuvalue">
      <description>The ECU will set fuel pump duty cycle to 0%, 100%, or one of these two values.</description>
      <table name="Fuel Pump Duty" type="Static Y Axis" elements="2">
         <data>Medium</data>
         <data>Low</data>
      </table>
   </table>


   <scaling name="IDC" units="Injector Duty Cycle" toexpr="x*0.7371358727094787" frexpr="x/0.7371358727094787" format="%.2f" min="0" max="255" inc="1" storagetype="float" 
endian="big"/>
   <scaling name="Pressure (bar)" units="units" toexpr="x/750" frexpr="x*750" format="%.2f" min="0" max="255" inc="1" storagetype="float" endian="big"/>

   <table name="Fuel Pump Duty High Injector Duty Cycle" address="CCBA8" type="3D" level="1" scaling="IDC">
      <table name="Battery Voltage" address="CCB98" type="X Axis" elements="2" scaling="rawecuvalue"/>
      <table name="Manifold Relative Pressure" address="CCBA0" type="Y Axis" elements="2" scaling="Pressure (bar)"/>
   </table>

<table name="Fuel Pump Duty Medium Injector Duty Cycle" address="CCBC8" type="3D" level="1" scaling="IDC">
      <table name="Battery Voltage" address="CCBB8" type="X Axis" elements="2" scaling="rawecuvalue"/>
      <table name="Manifold Relative Pressure" address="CCBC0" type="Y Axis" elements="2" scaling="Pressure (bar)"/>
   </table>

td-d · **Posted:** Wed Aug 01, 2012 4:34 pm

Speaking of fuel pump duty, check out the tables I just posted - viewtopic.php?f=34&t=8655 - these determine when low, medium and high fuel pump duty is switched to.

Merp · **Posted:** Thu Aug 02, 2012 1:33 am

Great work, I'm going to add the alpha branch to the Git and will start incorporating some of these

mrf582 · **Posted:** Thu Aug 02, 2012 10:31 am

Edited my post above to include the new fuel pump duty determination tables td-d mentioned.

mrf582 · **Posted:** Mon Aug 06, 2012 12:28 am

What is going on here? After an ECU reset, the car runs really lean and the ECU has to add a ton of fuel. Then after driving for a few minutes it suddenly changes the IPW and the ECU ends up pulling fuel. I captured the following log during cruise. You'll notice RPM, TPS, MAP, IAT, SD VE airflow are all pretty much constant. But suddenly the ECU decides to alter its Final IPW calculations by 35%. The Engine Load also shifts by roughly the same percentage for no apparent reason.

This doesn't happen after a hot restart. Once the ECU goes into 'mode B' where its calcs result in a higher IPW, it stays that way until the next time the ECU is reset.

I don't think I've noticed this with a stock MAF ROM. I was wondering if it has something to do with the 'Hot Restart' tables you have in the experimental section of your Github but it doesn't do it after a hot restart. Also, all those tables say 'lamba additive' and have positive values of 0.2 , 0.3 in there. Does that mean that after a hot restart, it targets 1.2 to 1.3 lambda instead of 1.0? That's quite a big difference!

Help please, because the car is impossible to tune like this.

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

Load is ordinarily just MAF divided by RPM, so I'm really puzzled by the way load value increased from below 0.5 to over 0.7, even though MAF stayed right around 20-21 and RPM stayed right around 2700. It seems like there's a load compensation table at work here, but I don't think that any of the known compensations would do this. MAP and IAT are flat, so it's not those.

Does your fuel table (or your CL compensation (Load) table) go richer around 0.7 g/rev? If it does, that would explain FFB going richer. If it doesn't, then I'm stumped.

mrf582 · **Posted:** Wed Aug 08, 2012 10:15 am

CL Target Load compensation changes just a tad around that engine load but it's very minor. Not too worried about it right now.

What is worrying though, is the sudden change in the engine load calculation. I did disable A/F # 3 correction limits and A/F #3 CL Target tables. i wonder if that has something to do with it? I doubt it.

Another thing different about this car is that the MAF sensor is unplugged completely and the wires cut... Wonder if that is causing any issues.

I'm attaching the full log of the snippet I quoted above as well as the ROM and definitions. Maybe another set of eyes looking at it will catch the issue, because I'm stumped.

Merp said I might want to try using a resistor across the MAF signal wire and see if that helps. Maybe leaving it floating is causing issues?

Anyway, thanks for your guys' input.

mrf582 · **Posted:** Mon Aug 20, 2012 12:36 pm

To update this problem, we switched to Cobb AP2 in SD mode. The car seems to be consistent. Have a rough tune on it to 15psi so far and everything seems to be okay. I really wonder what the problem was with the ROM.

But now the only problem is, Cobb's logging is so damn slow! Is it possible to log Cobb AP2 with an OpenPort cable and RomRaider?

Merp · **Posted:** Mon Aug 20, 2012 1:01 pm

Did you attempt anything with the MAF wiring? Was there some other issue other than post-reset?

Does this car also have the rear-o2 removed? Iirc, fuel learnings are applied 'after' engine load, so I wouldn't suspect that to be the issue.

Engine load is changing while SD VE airflow remains relatively flat, so whatever is going on is happening after MAF/SD calcs, but before fueling calcs. One thing that comes to mind is the MRP compensation map, although MAP is fairly constant in the log.

mrf582 · **Posted:** Mon Aug 20, 2012 3:43 pm

We didn't change anything with the hardware. Just married the Cobb AP2 and started tuning.

I think rear O2 is still there but I've disabled all the limits/corrections tables. Did this for both OpenSource and Cobb SD.

I've also flattened the MRP compensation map in both since SD does that anyway.

Merp wrote:

Engine load is changing while SD VE airflow remains relatively flat, so whatever is going on is happening after MAF/SD calcs, but before fueling calcs.

Is it possible that the ECU is reading the airflow value from the wrong address for a while?

This is definitely a software issue. Either in the ROM itself or in whatever tables I've modified. I just don't know how to find the root cause.

EDIT - 11/13/2012
Just wanted to update this thread. We kept seeing a fluctuating battery, heard random misfires from the exhaust, felt them in-car, noticed slight AFR instability at low load etc. and decided to go after a possible grounding issue. After diagnosis, I found this:

That's Pin 6 of ECU plug B137 melted! Apparently Pin 26 and Pin 6 of B137 are ignition coil grounds. Is it possible the ignition voltage was backfed through the wires and overloaded that pin? Anyway, I've soldered in a replacement wire and will see if that fixes the symptoms. But who knows what else is fried in the ECU. We'll still keep it on the Cobb SD stuff for now. If problems are solved and I can get the tune perfect on the Cobb stuff, I'll try to convince the owner to let me have a go at it with your SD ROM.

EDIT - 11/14/2012
Battery voltage still fluctuates. While digging around the engine wiring harness for hacks by the previous owner, I came across some exposed wires where it looks like someone had installed something by splicing into an existing harness and then just snipped the wires and left them there to hang all exposed. But, the real kicker was whoever painted the intake manifold did not protect the ground points! And this feels like some pretty thick paint too! This is the main ground on the RH side of the manifold. I believe its for all the engine sensors and ECU etc.

Subaman · **Joined:** Tue Feb 05, 2013 2:18 pm **Posts:** 3

Does this patch reference the maf sensor? I notice it changes my fueling when i have the sensor plugged in. When the MAF sensor is not plugged in i am always pulling 25% fuel at idle, no matter how low my values in the VE table are, but when it is plugged in things change drastically.

Merp · **Posted:** Fri Feb 22, 2013 12:10 am

Subaman: the patch does not use the MAF voltage to calculate engine load, however... removing the MAF triggers some alternate control logic that appears to affect fuel trims and probably more. Best bet is to leave it hooked up, or replace with an appropriate resistor.

mrf582 · **Posted:** Mon Mar 04, 2013 1:41 pm

Subaman wrote:

Does this patch reference the maf sensor? I notice it changes my fueling when i have the sensor plugged in. When the MAF sensor is not plugged in i am always pulling 25% fuel at idle, no matter how low my values in the VE table are, but when it is plugged in things change drastically.

Do you have a separate air temp sensor wired in? If not, when you disconnect the MAF plug, you are removing the built-in air temp sensor as well. So the ECU reads full cold then adds a crap ton of fuel and then AF correction tries its best to remove the extra fuel.

Subaman · **Joined:** Tue Feb 05, 2013 2:18 pm **Posts:** 3

Yeah, I have an aem temp sensor right before my throttle body

RomRaider

Forum rules

Speed Density ROM - A8DH201X 2006 WRX

Who is online