RomRaider

05 sti
Pacific North West(rainy and 50 degrees)
92 octane pump gas
dom 2.0xt
process west tmic
perrin big maf
tgv delete
AEM 320 fuel pump
FIC top feed fuel rail conversion
FIC top feed injectors (i believe they are 1000cc)
Tial EWG 38mm
Grimmspeed 3 port
Catless downpipe
3in exhaust
Supposedly upgraded pistons and rods

Hey guys and gals. I mostly tune evos but occasionally help out my friends that own subarus and need any tuning done. One of my close friends recently acquired an 05 sti with the mods listed above. The car was previously tuned by a local shop near by but needed to be retuned because it was running 26psi on straight pump and was stumbling bad due to knock. Supposedly this car has upgraded pistons and rods but I have not confirmed this myself. I started to tune from scratch because the rom tuned by the shop needed their definitions to be able to view it. As I got the car running and tried to do some pulls I found out that ignition advance multiplier drops to zero pretty much immediately after .61 g/rev. I have tuned several subarus over the past few years and never had the iam drop to zero. I inspected the knock sensor to make sure its tourqed down properly which it is, nothing is loose around it. Has anyone experienced anything like this? On the evos you can desensitize the knock sensor, is there a way to do that on the sti? I know for a fact that there is no knock at that low of an engine load with the timing its running so it shouldn't drop like that. The engine itself runs good, no abnormal noises or anything like that. I know that built internals might trip the knock sensor into picking up false knock, how do you guys go about that without disabling the knock sensor completely? Any input/help would be greatly appreciated guys! attached is a log and map of the current tune which isnt finished but I cannot go any further without addressing the current issue.

I had a similar issue but with a 16bit ECU, it was an electric issue. There are two green connectors identical to the test connectors but at the passenger side, under the plastic cover in the bottom by the door. Check if they are disconnected.

is there a chance you've uploaded the wrong log here? There's really no useful data in there. Also grab the LTV info too if you can it might have something useful in there.


The knock sensor tables have been defined on some but not all roms so it's in there if you were busting to play with it - but i've yet to come across a build so noisy that it's absolutely needed. What is possible is that the knock sensor is cracked around it's housing - once water gets in then it's had it and will report all kinds of drama.

Your feedback correction minimum load is enable above 0.61 - so that's likely why your seeing something as soon as you cross that point, you could increase this to 4.0 if you had a death wish (honestly, it shouldn't be needed - the standard sensor if working should be fine) and I definitely wouldn't do it without a set of knock ears on the motor as a backup.

No I uploaded that log to show where the IAM drops. Yeah that is exactly what I did today, set feedback correction minimum load to 3.5 because I incrementally bumped it but each time that load was reached it went to zero again. Definitely not something I wanted to do. I also ran it in knock correction advance alternate mode because the 3 port wont work. And it still didn't work so I disabled it and after i set feedback correction minimum load to 3.5 I got it to work and while I was doing a third gear pull it just randomly cut out at like 5k rpm going back to wastegate pressure. That log is attached below. What is LTV?

LTV is the long term values, ECU saved corrections over time since the last reset. In romraider there is a "show LTV" button on the left hand side of the screen which will retrieve them for you.

You've not logged wastegate duty or target boost so hard to see what's happened at 6000ish there when the boost drops off - doesn't look to be knock related but you can definitely see the air flow taking a nose dive. The AFR is quite rich on spool up as well, do you need some MAF scaling or is the fuel target rich there as well?



Have you got a set of cans on it to listen to what's actually happening in there, i'd be more inclined to go back to basics - get the knock control working as it should then sort fuelling/timing at wastegate pressure and build up from there.

Oh i got ya, yeah the thing is this laptop sucks, when I select even a single additional perameter the laptop freezes lol. I still need to finish scaling maf but could only go so far due to low boost, I spent a little time yesterday. Waiting for hallman to arrive then will mess with it again. Thanks for the input!

Can you reflash with this rom and see if it still happens? There was a bunch to clean up. Log the same 2nd gear pull and log a idleing car then start to drive it to 30mph. tip in a little like your climbing a small hill on the highway and present us back those two logs.

If you hate it, just flash it back

That would be a symptom of logging too many parameters in slow poll mode.

Really? how do you change poll mode? Logging protocol? Weird because my other laptop doesn't do it but I don't use it because I deleted all RR stuff and just keep evoscan and ecuflash.

Ill try it, but other than changing the fuel/timing map and leaving one of the kca tables zeroed how is it supposed to not drop IAM? I kept the fuel table flat like that to help scale maf, I do it manually instead of using the spread sheet, once scaled then I make it look presentable lol. Regardless Ill try it out and will report back, wont happen till next week tho because my next day off is Tuesday. I greatly appreciate everyone's input here guys!

Is your laptop from the early 80s? It should have no problem reading in a data stream of 4800 baud.
Use fast poll mode in the Settings menu.

No sir but it’s not that good of a laptop, lenovo thinkpad t410si.

That computer has more performance than this one I use everyday.

Any update? We needed you on something that made sense before we start suggesting whats wrong. You should try that map and get back to us please.

I will be off this upcoming Thursday and will try the map out, as soon as I do I’ll update with results.