DIY: ProjectLambda Stock ECU Reflash Tuning (Master Reference Thread) - Page 16 - Subaru Impreza GC8 & RS Forum & Community: RS25.com
 
Private Support i-Trader Ratings Become a Vendor Rules for VendorsModsUsers Contact
Subaru Impreza GC8 & RS Forum & Community: RS25.com
gc8 RSS GC8 Facebook Subaru GC8 Twitter Subaru YouTube Channel Subaru Sticker
 
Login below or Register Now
Register Mark Forums Read FAQ Advertise
Go Back   Subaru Impreza GC8 & RS Forum & Community: RS25.com > >
DIY: ProjectLambda Stock ECU Reflash Tuning (Master Reference Thread)
Reply
 
Thread Tools Display Modes
Old 09-27-2017, 06:50 AM   #151
Hatecrewdeathroll
1.8L
 
Join Date: Jun 2016
Trader Rating: (0)
Posts: 21
Default

imprezarsc

what is the mod list again? lookin good!
Hatecrewdeathroll is offline   Reply With Quote
Sponsored Links
Advertisement
 
Old 09-28-2017, 08:23 AM   #152
ImprezaRSC
2.0L Turbo
 
Car: 2000 2.5RS BRP
Fav Mod: 8 SEMA awards.
Join Date: Jul 2004
Trader Rating: (2)
Posts: 1,160
Default

Quote:
Originally Posted by Hatecrewdeathroll View Post
imprezarsc

what is the mod list again? lookin good!
Build thread is in the member journal.
ImprezaRSC is offline   Reply With Quote
Old 10-07-2017, 04:02 AM   #153
crjohnson
2.0L
 
Join Date: May 2014
Location: Melbourne
Trader Rating: (0)
Posts: 72
Default

Hey guys I got my hands on some EJ255, blue 550cc injectors to install in my ej251 soon. Looks like installing them will be easy and the load and airflow scaler figures are easy to calculate, but I'm wondering what other affects might happen when I install them.

After doing a WOT test with a zeitronix zt-2 wideband I saw that its pretty lean above 4000rpm, so I have the VE set to 100% (using current 280cc grey injectors) from 4800rpm and 90.7kPa and up.
90.7kPa
3200RPM 82.54
3600 90.00
4000 94.00
4400 97.00
4800 100
5200 100

If I install the 550cc injectors, change the scalers, slope and offset etc, will the ecu still dump a heap of fuel in at these WOT conditions or will I need to reduce the VE table by a decent amount across the whole VE table (and cranking pulse?) ? I don't really want to muck around with the low load and cruise areas of the table too much.

My engine has delta 1500cams, slight head work, EL 4-1 headers and is soon to get a TD04 turbo setup.
crjohnson is offline   Reply With Quote
Old 10-14-2017, 09:23 AM   #154
ImprezaRSC
2.0L Turbo
 
Car: 2000 2.5RS BRP
Fav Mod: 8 SEMA awards.
Join Date: Jul 2004
Trader Rating: (2)
Posts: 1,160
Default

The ECU will do exactly what you tell it, but you have to speak its language to get the right answer. Sounds like you are just up against a restriction, like the injectors . That's a good problem to have.

Get a good fuel pump and a high flow fuel filter while you're at it. It's really never too early to upgrade the fuel systems on these cars. I prefer to let the fuel in instead of forcing it in. These engines are heavily restricted from the factory with both fuel AND air. 280s are also just too small after you do all the typical bolt-ons, even with a better pump. Heck, 440s are too small after atypical bolt-ons, and I'm on a 2.0L! Oversquare engines naturally want more fuel, so it won't have the same injector requirements as an undersquare Honda of the same displacement.

What I'm not sure about tuning with cams is how to get around the overlap issue. The o2 will pick up unburned fuel and then try to lean things out. I'd worry a little more about false readings and unnecessary compensation from the ECU at idle and cruise.

Your 550s will run out after about 10-12psi of boost, FYI.

Last edited by ImprezaRSC; 10-14-2017 at 09:32 AM..
ImprezaRSC is offline   Reply With Quote
Old 10-22-2017, 07:10 AM   #155
ImprezaRSC
2.0L Turbo
 
Car: 2000 2.5RS BRP
Fav Mod: 8 SEMA awards.
Join Date: Jul 2004
Trader Rating: (2)
Posts: 1,160
Default

Added fuel enrichment strategy explanation to post #3.
ImprezaRSC is offline   Reply With Quote
Old 10-23-2017, 01:52 PM   #156
crjohnson
2.0L
 
Join Date: May 2014
Location: Melbourne
Trader Rating: (0)
Posts: 72
Default

G'day guys,

Just thought I post some before and after results of mine lately with some wideband readings if anyone was interested in the results or comparing to yours. The logs were done in 3rd gear on flat road, had to back off before 6000RPM due to getting a little too quick for the conditions..

Current setup: ej251 sohc, match ported in/exh manifolds, 10:1comp, Delta 1500 cams, long runner/big plenum intake manifold, 4-1 EL headers (approx. 39"), stock 2.25"ish exhaust, 3" SS intake pipe+K&N Pod, manual gearbox, 98Octane fuel, standard grey 280CC injectors, lightweight flywheel (think it was around 14Lb).

The only changes between the 2 datalogs is changes to the VE table figures as highlighted in blue. Intake air temps were also within 10C between the runs and on cool <20deg C ambient temp. days.

Some interesting stuff I've noticed mainly surrounds the timing figures. I've never been able to make sense of these figures when taking into account any knocking, fine learn and IAM etc. The IAM is always 0, so the Ign. Advance should just be Base spark - Knock Retard right?
Would it be possible to set the entire Learned Spark table to 0deg, so hopefully only the Base minus Knock Retard is acting? I'm not sure what to then do with the Spark Fine Learn - Engine Speed Index or Load index tables?

Anyone have any idea what else might be happening here as we would like to start playing with the timing.

Next is the knock retard, has anyone ever turned down the sensitivity?






crjohnson is offline   Reply With Quote
Old 10-24-2017, 10:59 AM   #157
Hatecrewdeathroll
1.8L
 
Join Date: Jun 2016
Trader Rating: (0)
Posts: 21
Default

hey cr,

I set IAM to start at 1, so that the ecu will use the full learned spark table immediately after flashing and I can quickly see if spark is too advanced and causing knock.

I use a base spark table that is very conservative, and utilize the learned spark table on top, being much more aggressive. adding 8 to 12 degrees in various parts of the map (especially WOT).

If you set your IAM to 0, that only makes IAM = 0 for when you first start after flashing (or maybe initital startup, not sure). Either way, if the ecu doesn't detect knock, the IAM will start climbing up to 1. That is why sometimes your total ignition timing is greater than base. Some times its less than base (knocking!)

You should probably log spark coarse learn and iam.

Having IAM set to 1, allows me to see if timing is too advanced immediately when I start tuning, instead of starting at .5 and working my way up to 1 (full learned spark advance).

You can set the learn spark table to all zeros (tried it) but I don't think it 100% meshes with the ecus logic, and might make it kinda spazz when you get knock. May also cause issues with spark fine learn and other stuff. To me, it makes sense to just use the tables as designed, or at least try to.

"Spark Fine learn - load index" is (i believe) for saving long term ignition advance corrections with kpa as a threshold or whatever. Prob only want to change this if boosted.


I tinkered with the knock sensitivity, but ultimately I do not believe it is worthwhile. The knock sensors seem fairly accurate.


I hope i answered you questions with out being too confusing. I also hope I'm not sharing any kind of misinformation. Good luck!


p.s. I typically log this stuff:

Time (s) Manifold Absolute Pressure (kPa) Engine Speed (rpm) Throttle Position (%) Vehicle Speed (km/h) Air/Fuel Sensor #1 (λ) Power Mode - Fuel Ratio Target (λ) Intake Air Temperature (C) Fuel Trim - Short Term (%) Fuel Trim - Long Term (%) Fuel - Acceleration Enrich Knock Retard () Ignition Advance () Knock Cumulative Noise - Bank 1 Knock Cumulative Noise - Bank 2 Ignition Advance Multiplier Injector Pulse Width (ms) Ignition Advance - Base (BTDC) Ignition Advance - Fine Learn () Ignition Advance - Coarse Learn () Mass Air Flow (g/s)

Last edited by Hatecrewdeathroll; 10-24-2017 at 11:08 AM.. Reason: readability
Hatecrewdeathroll is offline   Reply With Quote
Old 10-24-2017, 11:20 PM   #158
Impreza boy
1.8L
 
Car: 2002 Subaru Impreza
Join Date: Sep 2016
Trader Rating: (0)
Posts: 16
Default

I'm still a newb tuner so take what I say with a grain (or two) of salt, nevertheless I though I'd offer up my opinion.

I'd agree with setting the IAM close to, if not at, 1 while tuning. Like Hatecrew said, it helps with making sure the tune is safe.

As far as datalogging goes, try not to log too many things at once. The time intervals between your logs are almost 0.5sec which makes the logs harder to read and get a good idea of whats going on. My logs (which aren't that great) have intervals of 0.05sec which gives about 10 times more data. This gives a more fluid idea of what's going on.

From what I can tell, between the first two rows of your log, it looks like there is a large increase of change in your MAP and IAT, which gives me the impression of a lot more airflow. However, there is not a big change in your wideband reading between those first two rows. This gives me the idea that the ecu pulled timing (-3.5 degrees) because of the quick increase in airflow. Whether it knocked or the ecu was expecting knock and trying to prevent it, I do not know. This is just my assumption. I could be totally wrong.

Here are a couple good reads helping to explain Subaru's knock control strategy:
RomRaider View topic - Subaru's knock control strategy explained
A Subie Newbie Tuning Guide - there is a section which goes into depth on the various forms of knock control and how they work.

Edit: The articles go into depth on the newer models' ecus, so I don't think that everything stated is explicitly applicable to our cars; however, I think it helps to get a better understanding in general of what is going on.

I hope this is helpful in any way.

Last edited by Impreza boy; 10-25-2017 at 07:50 PM..
Impreza boy is offline   Reply With Quote
Old 10-25-2017, 01:48 PM   #159
crjohnson
2.0L
 
Join Date: May 2014
Location: Melbourne
Trader Rating: (0)
Posts: 72
Default

Cheers for the replies and suggestions!
Apologies but I forgot to mention an important bit of info in my previous post. That picture is only showing every 6th row of the datalog. Otherwise the time steps are usually around the 0.07sec mark and I usually log every bit of engine related data possible. It looks as though you guys have a fair few more features than my ecu does. Ie. Ign.Advance course and fine learn?

I did a test tonight and set the IAM to 1 which gave me about 4 to 6deg more advance and much more knock retard! Up to -4.5 to -5deg knock retard in 2nd and 3rd gear around 5700-6000rpm.

I might reset the IAM to 0.5 for now to reduce the knocking a little.
Does anyone elses Ign.Advance Fine Learn ever log as 0 like mine? Its never logged as anything but zero, unless it is actually referring to another fine learn table that I don't actually have?

I'm also confused as to why the injector pulse width reduced by 3-4% between the 2 runs when ONLY the IAM was changed. Intake air temp was 27degC vs the previous 35degC.
In the mean time I'll see what I can learn from the romraider thread (Thanks Imprezaboy)



Last edited by crjohnson; 10-25-2017 at 03:12 PM..
crjohnson is offline   Reply With Quote
Old 10-25-2017, 05:42 PM   #160
ImprezaRSC
2.0L Turbo
 
Car: 2000 2.5RS BRP
Fav Mod: 8 SEMA awards.
Join Date: Jul 2004
Trader Rating: (2)
Posts: 1,160
Default

It looks like you're using a little too much base timing. My own combination never runs more than 32deg total on the top end. I see yours around 37. I don't run that much timing even on E85. That's more for light cruise and trailing throttle.

Last edited by ImprezaRSC; 10-25-2017 at 05:51 PM..
ImprezaRSC is offline   Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 11:41 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
vBulletin Security provided by vBSecurity v2.2.2 (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright © 2019 VerticalScope Inc.
Designed & Powered by Domain Architect