ajayel

Just learnt that there is a new open source way to change Subaru BIU settings similar to dealership software. I was able to change a few things already but it can probably do way more. It's based on the FastECU project which has ECU logging & mapping capability. Kudos to the people who created these projects; https://www.subaruoutback.org/threads/communicating-with-the-biu.552361/ https://github.com/miikasyvanen/fastecu-oem

hi @logan25 a few months on how is that IAM and fuel trims ?

same, amayama now have an auckland import warehouse. So no import duty to pay and NZ shipping is only $5

Turns out it is possible to DIY switch ECU's, at least on 2005-2009 leggies. Found an ECU with the same part # on e-bay then was able to code it to the vehicle using the instructions at https://www.subaruoutback.org/threads/immobilizer-reverse-engineering-2005-obxt-probably-other-year-models.501753/

OK to close this out, seeing as it's lock-down. The problem was that someone doing previous work left the plastic panel above the drivers feet hanging by the OBD-II socket & wires. So that got damaged causing an intermittent connection to the ECU. Fixed by installing a new OBD-II port from iWire https://www.iwireservices.com/product-page/obd2-port no troubles since.

Good to read this @flokie, any chance you can share the modified ROM file ?

Did you have any luck getting this car running any better @BMCracer ?

Yeah just 4 bolts but you'll lose a little coolant disconnecting those lines so may need to bleed air out again and it's recommended to use a new TB gasket. You can also hold the throttle open for cleaning with engine off/ignition on & a brick on the accelerator.

if you post up examples of mismatches someone here maybe able to chime in with knowledge or experience

try https://jdmfsm.info/Auto/Japan/Subaru/Legacy_Outback/ On that site the PDF manuals tend to be US market , the web versions match JDM better. Check the identification section under general description against your chassis number, my imported 07 is an 08 JDM model year .

IAM 0.0 is the ECU needing more data to make up it's mind but somethings definitely wrong here, as you know! FLKC protects the engine up to 7 or 8 degrees which did get touched on albeit spiked very briefly. Could be so many things but airflow is interesting. After you let off the pedal the throttle closes appropriately but it's taking three or four seconds for the mass airflow reading to reduce. - is that intake chamber crack fixed ? - the pipe between the the air filter and intake chamber has a tab that must line up on the chamber kinda underneath and towards the driver - the pipe on the air filter box & where the chamber hooks onto the throttle body has rubber ends which can get kinked or tear causing leaks ? - is the MAF sensor dirty, loose or wiring damaged ? - is the MAP sensor (front of intake manifold) loose and have it's small rubber hoses to the filter connected (should be able to blow air gently thru filter & hoses) - are the 2 knock sensors (under the intake manifold, blue connectors) loose or corroded, cabling ok ? - any loose debris rattling around under the intake manifold ? - any gunk around the fuel injector seals ? - the orange gaskets under each side of the manifold and on the throttle body have tabs do they look perished ? - is the throttle body clean ? (added with edit)

Just one more ; fbkc (feedback knock correction) I've thought of others that may help, not all available on all vehicles; gear (could be calculated gear) fuel tank pressure (if available) cpc (canister purge) egr steps (exhaust gas re circulation) You don't need these; Engine Load (Relative) (%) Mass Airflow Sensor Voltage (V) Main Accelerator Sensor (V) Sub Throttle Sensor (V) Sub Accelerator Sensor (V) Main Throttle Sensor (V)

Might be worth a Romraider test run but prolly don't collect a heap of data until the intake is sealed as that's a vac leak. Please inc. both air fuel ratio sensors, both oxygen sensors and the others mentioned above.

Feedback Knock Correction = FBKC (aka FKC) changes Fine Learning Knock Correction = FLKC changes Rough Correction = IAM changes If FBKC happens often it is stored to FLKC. To calc total timing base timing, (KCA*IAM), FBKC and FLKC are combined. The knock control strategy doc posted previously explains this properly but each of the three correction modes are applied within preset load/rpm ranges. A log file after the intake crack is fixed from normal driving, cruise and harder acceleration (joining the highway) might show us more around the conditions at accel/cruise/decel (load, rpm, pedal,throttle sensors) and airflow (VVT,MAF,AF,O2 sensors). The logs show how the params are increasing or decreasing rather than the freeze frame btssm is triggered to collect.

The 3.6 has a lower compression ratio (10.5) so can use regular fuel and would prolly ping less but not had any hands on time with the EZ36. If you can get btssm to log another CSV that would be interesting. So similar to your previous logs posted maybe drop the OCV, OSV stuff and add in Feedback Knock correction and Fine Knock correction. The computer only applies knock correction above certain thresholds of load & rpm because some is just noise but these parameters tell us when its detected and learnt.