Subaru Impreza GC8 & RS Forum & Community banner

2221 - 2240 of 2553 Posts

·
They call me Garrett
98 RBP V4 STi Swapped RS
Joined
·
4,083 Posts
Discussion Starter #2,223
Type R + More Stickers

Tune isn't set right, oil seems too cold after the cooler install, car needs to be raised, and I'm headed to the track in two weeks... better put stickers on the car -_-.

But it looks good at least!















 

·
Registered
Joined
·
162 Posts
Awesome updates! Been a while since I checked in.

That digital display you're working on, I have a few questions for you, I shoot em over in a message. No time to type it all out atm.
 

·
They call me Garrett
98 RBP V4 STi Swapped RS
Joined
·
4,083 Posts
Discussion Starter #2,230
Thanks guys!

With the long weekend I was able to get a few small things done. I finally ghetto installed my AEM wideband, it's so ghetto I don't even want to post pictures of it yet, but the readout was needed. After installing that I noticed my wideband was telling me the same readouts as my stock O2 sensor was, which meant the car was running incredibly lean as compared to the target AFR.

I did some investigating and noticed that the closed loop fuel trims were (metaphorically, but not far from literally) off the chart. My 10-50 g/s range was at like 14% and my 50+ range was 0. I thought that was pretty curious and did some long logs of my drive to and from work realizing that I am almost never in the 50+ g/s fuel trim range as well as in closed loop, so the last column in the map was pretty much never reached. This meant I had no open loop fuel compensation happening as the last column is applied to open loop... boo.

The fuel trim columns are calibratable, and the logic for how the car transitions from closed loop to open loop is pretty complex, have lots of different parameters needing to be met, and not all of them are known. I figured that engine load, throttle position, vehicle speed, RPM, and manifold pressure all had to do with the transition, and how the car's gearing was setup effects how these parameters intertwine. For example, you'll have less engine load, and less throttle needed to get to a higher RPM in lower gears in an STi 6 speed because of how short the gearing is, as opposed to my longer STi 5 speed gearing. This means different MAF values for different throttle positions, etc. Not a problem when dealing with tables that depend on engine load (fueling, timing, etc), because load is load, but definitely when dealing with ranges of the MAF that interact with throttle position and RPM and all that.

At least that's my thought process, I could be wrong, but stay tuned because I feel slightly justified here in a second.

The point I am trying to get at, is that because of my gearing, the ECU pops into open loop before the 50+ g/s fuel trim range can really be used, thus never applying fuel learning/correction to that closed loop region, and subsequently not to the open loop fueling. Like I said, I was hardly ever in the 50+g/2 range and in closed loop and I confirmed this by logging 34,800 data points or so.. about 2 hours of driving. Of those 34,800 data points, 140 or so (I don't have my numbers in front of me) of those were the car in the 50+ g/s range AND running in closed loop fueling.

Yeah... that's less than half a percent of the time. That needed to change.

I changed the values to be 0-5, 5-10, 10-40, and 40+ g/s.

Went for a long drive and was still getting bad fuel ratios, but they were improved. I pulled up learning view after 150 miles of driving and I saw 11% or so correction in the 40+ range. Awesome! I scaled it in such a way it's actually doing something, however not awesome because I shouldn't be adding that much extra fuel!

I had already done a boost leak check, and everything seemed to be fine, so I stole the MAF from one of the Forester's sitting at my house and popped it on. Went for a drive, another 200 miles or there about and fueling has been pretty on point so far. Learning view showed i was only at 4% correction at the max. I can live with that. I guess I'll be buying a new MAF.

Out of curiosity I opened up the FSTi map as it has very similar gearing to my trans in 1-4th (2% different at the most) and what do ya know, the closed loop fuel trim correction MAF values are what I set mine to... which makes me feel justified in my thought process of gearing being the deciding factor in that calibration... but it's also a 2.5L and a 32bit ECU so I could still be wrong.

Anyway:

I've got to tidy up the wideband wiring and get the data into the computer... among other things and then I'll post an official update and add pictures. I do think I need to revise fueling a tad, but at least it's not bonkers crazy now. Car drives SO well now. SO well. Feeling much more confident flogging it at the racetrack in less than two weeks.
 

·
I ❤ BOOBIES
99 Coupe
Joined
·
16,997 Posts
32 bit stuff its totally different than 16 bit, and thats odd yours goes into open loop that qucik. I havent touched my OL/CL transition values and it holds CL for quite a while into load/boost before switching to OL. Factory 02 values, your on the JDM ecu right? Its probably different, theres a thread by Merchgod or NSFW on Romraider about the OL/CL process. Its not as complex as you think it is, the ECU just looks for different values of different tables depending on what the current load/rpm/boost etc is at, then theres a timer for how long it holds it as well as a timer for how long the engine has been running so if you flash it and drive, it takes like 10 mins before the corrections are where they need to be ill do some funky stuff.
 

·
They call me Garrett
98 RBP V4 STi Swapped RS
Joined
·
4,083 Posts
Discussion Starter #2,234
32 bit stuff its totally different than 16 bit, and thats odd yours goes into open loop that qucik. I havent touched my OL/CL transition values and it holds CL for quite a while into load/boost before switching to OL. Factory 02 values, your on the JDM ecu right? Its probably different, theres a thread by Merchgod or NSFW on Romraider about the OL/CL process. Its not as complex as you think it is, the ECU just looks for different values of different tables depending on what the current load/rpm/boost etc is at, then theres a timer for how long it holds it as well as a timer for how long the engine has been running so if you flash it and drive, it takes like 10 mins before the corrections are where they need to be ill do some funky stuff.
I meant complex in the sense that there are many different parameters that determine when the transition occurs mores so than complex that no one understands it. It's more than just 2D table, blah blah blah. I've left all of that stock in the ROM and don't really plan on touching it unless I REALLY need to.

After about 400 miles driven now, I am only seeing 6% correction added in the 40+ range and my AFR matches my wideband, close enough at least.

I don't really like that much correction... but it's better than 14%

0 knock of any kind indicated now as well, and I think I can add more timing in.

GD one no work, right? I have a open box new one...

2002-2007 Denso MAF Sensor (197-6040) [22680AA310]
They are all the same, I bought a new OEM one already
 

·
I ❤ BOOBIES
99 Coupe
Joined
·
16,997 Posts
Honestly i used to be paranoid about AF corrections, if your under +/- 10% your fine. The whole point of the ECUs corrections is to compenstate for atmospheric conditions and you cannot control any of that. +/- 5% like everyone says is fine but dont sweat it if you cant get much closer. It took me well over a week straight to get the ID1000's to sit within 6% and i cant get any better.

Im going to start adding in some timing as well, particularly down low to maybe get some better off boost response and quicker spool time, i got a big housing on this lol
 

·
Experimenter
Tubaru Pickup
Joined
·
3,910 Posts
You could in theory get it close to perfect. There's IAT, ECT and MAP compensators. Mine on e70+ SD it's usually under 1.8%. On pump gas it was pretty high but I never really tuned that table on low E. Never ran low E till my laptop broke and I put it away for winter. Now all my intercooler pipes and exhaust is different so I'm sure I'll need to tweak a bit
 

·
They call me Garrett
98 RBP V4 STi Swapped RS
Joined
·
4,083 Posts
Discussion Starter #2,237
Yeah, ethanol gasoline really makes that correction go one way or the other. I'm not worried about it as long as my fueling looks fine in open loop!

Put my brand new MAF on the car last night, my factory and wideband are reading pretty much identically minus some lag on the wideband part being behind the cat and all.

I haven't seen a lick of knock since I swapped the MAF out, so i really want to add a bit more back in and get some more beans out of this thing. Maybe after the first day of Gridlife when I'm inevitably not happy with my laptimes... we will go full send then :)
 

·
They call me Garrett
98 RBP V4 STi Swapped RS
Joined
·
4,083 Posts
Discussion Starter #2,238
Small update:

Swapped in some new exhaust gaskets, as well as a new t stat gasket. The T stat gasket I forgot to get when I did timing, and it was dripping slowly. The exhaust gaskets had clearly been replaced before with some garbo aftermarket ones and were leaky.. good lord..








Car is giving some me some more closed loop fuel learning.. so... i'm sure it's happy to not be leaking preturbo.

No more coolant leaks either.


All this talk of "JDM engines with 50-80k KMs" .... yeah.... sure...



Race time on Thursday.
 

·
They call me Garrett
98 RBP V4 STi Swapped RS
Joined
·
4,083 Posts
Discussion Starter #2,240
Back from 4 days at the race track.

Had a panic radiator moment before I had to leave.. but we got there.

Engine and transmission were flawless, and stayed super cool.

Brakes were great for sake of a bit of judder.

Proper update and pictures soon!
 
2221 - 2240 of 2553 Posts
Top