RomRaider

Been looking at the 07 sti rom more closely and verifying the current defs which were preliminary. Some interesting findings:

1. There was the big hubbub about all the extra tables that these roms had. Like 6 boost and fuel maps, etc. However, some of these extra tables are effectively disabled in the rom. For example, instead of 6 boost maps, only 2 are ever used (same as the other 32bit roms which will be similarily updated in the upcoming ecu def release). Same with some of the other tables as well (fuel 6 to 4, initial wg 6 - 2, etc.)

2. Knock correction was removed from SSM logging for all the 07 models. IAM was added to SSM for the 07 sti (haven't checked the wrx, though). The knock correction logic appears to the same however and therefore knock correction as a parameter can be added to the logger (I will also have an upcoming logger xml update with this and other updates included).

3. Some new tables unique to the 07 sti - Boost Compensation (Intake Temp) and Boost Compensation (Atmosphere). Other 32bit roms also have boost compensation based on atmospheric pressure but they use AP multiplier and offset to determine the multiplier, whereas the 07 sti is simplier with a standard 2d compensation table (the appropriate table will be added to the defs for all 32bit roms).

4. AF learning #1 range on every other 32bit ecu is -15% to 15%, whereas the 07 sti is -25% to 25%.

5. These roms are actually very similar to the A8DH... series instead of the 06 sti. There is very little difference in the logic except for timing and throttle duty. Timing compensation is similar, but base timing\advance determination is strange. Anything involving throttle duty is also quite different. I haven't had a chance to look at these in detail but it might require some volunteer loggers to really see what is going on.

6. It also has the logic introduced on later 05 LGT models where wgd as determined by the initial\max tables is ignored at times and instead wg duty is ramped up from a minimum value until it hits the map target determined by the initial\max tables. Although the limits are different on the 07 sti and it is possible that the wacky wg behavior is disabled. I'll have to look at it closer. If it isn't disabled, it might explain some of the hesitation issues.

Let me know where I can help out.

-Gabe

I have heard over on the legacy board that the 07 heads have 1mm bigger valves, but cannot confirm this anywhere else.

I had asked about this on another board when clark turner was around....
this is what I got.

__________________

I have an 07 STI and don't mind logging if it will help the cause. Just let me know what you want and I'll run it whenever I'm driving.

I have been reaching a limit with the 07.

It appears that there is a limit associated with 4.5Volts on the MAF.

With the factory boost control system it looks like a torque limiting map that cuts the WDC when the MAF limit of 4.5 is reached. I tested this on two differently set up STi's using factory boost control. I thought it was a hard WDC cut limit and put a different solenoid on the car. It exceeded the 4.5 long enough for the ecu to learn the solenoid and quickly fixed the airflow by cutting WDC even more. I was very suprised at how fast it did this. I have not yet tested a manual boost controller to see if it will cut other parameters to help control torque.

I confirmed the existence of a desired torque map with some COBB and EcuPuke tuners (I actually have friends!!!). There is a desired torque map that is a 3d map and a MAF limit setpoint.

I am really very very inexperienced at this .hex .xml stuff.
I am willing to learn and contribute as much as I can in between my two jobs.

I own an 07 STi and have access to 3 others.

Oh and I forgot to add, I used pretty much every map available via RomRaider to see any of them affected the results. The MAF limit and MAX Load didn't do anything for this.

That might xplain a few things and support my conspiracy theory.

My WB is coming in this week. So I can do even more testing, not that testing fuel is that important, but... Helps to have one. There is more and more evidence of subaru just limping this thing.

i have stock roms and I have a rom in .bin form from the ecuT*k guys.

got any advice?

I got a hex editing and bin editing program but I don't know what ANY of the features are.
I don't even understand hex.
I kinda get the numbering system.. but that's it.

Please, let me know what I can do to help.

are the 06 affected also by this "torque control" map?

Wait till Bill gets up and see what he says bout it.

Ok, I'll take a look at it. I'm pretty sure I can find it - no need for Cobb or Ecutek maps as I don't want to be accused of stealing their work.

I caught it this morning. Right when it hit the limit. Post up log in a sec. Its hard to tell though in the log that anything is wrong, but you can see boost lowers a bit, and you can definitely feel it right around 5.6k rpm when MAF hits 250. No timing pulled (at least it looks smooth) but the power just kind of cuts out like it pulled timing.

Edit, actually there was some timing pulled at 4700, I need to look at the KC Map. This happened right after Peak load though. I doubt it was due to knock.

EDIT: It was not pulling timing, map called for that KC. Timing advance C and D call for that drop. Timing advance B does not.

EDIT again: I did feel the drop in power at 4700 (the KC map called for it, WTF), but then again 5600 (where timing looks fine and smooth).

No, the 07 STI ECU (and parts) are very different.