Brad Burnett

As adam said, hard to tell from screen shot. I would start by adjusting the calibration. I never try to rely on the closed loop control under wot. The control is just not fast enough. You need almost an additional 15% fueling to achieve the desired lambda for the given situation. If you really wanted to rely on the CLL to do this, you would need to crank up the gain value for the algorithm. Side note: if this is a car you recently did, it most likely might have a mechanical issue arising. Maybe something like a fuel pump failing, pressure regulator lost its reference line, etc.

on the exhaust cam sensors, i would tie the sensor ground and associated shield into the same ground that is utilized for the factory crank/cam sensors. but if that is not viable, the standard sensor ground should be sufficient.

I did a fury on a C30A not terribly long ago. Once i get home ill glance at my notes.

the ecu needs to see a couple/few rotations of the crank shaft to see it sync, then it will try to fire spark/fuel.

you will need to add a relay. This relay is part of the failsafe strategy if there is a fault with the ethrottle. If there is a fault with the ethrottle, the ecu will cut power to pin B5 via that relay thus removing power from the throttle actuator.

straight from the help file: " In some cases it may not be practical to wire the ECU in this manner or there may not be sufficient spare digital inputs. In these cases, Stepper Reset must be set to Key-ON. In this mode, when the key is turned on, the stepper motor is moved to its fully open position, then returned to its correct position for startup. During this period (up to four seconds) the idle valve can be open far enough to produce a very high engine speed (eg up to about 4000 RPM on a V5/6 WRX/STI). This produces an undesirable over-rev on startup. The Key-On Fuel Lockout allows the engines starting to be delayed by certain time period to prevent this. If the stepper reaches its starting position before the time period is up, the engine will be allowed to start immediately. There are two options for the driver: 1.Crank the engine until it starts. The engine will be allowed to start when the stepper is ready or the Key-On Fuel Lockout time is up. 2.Wait a few seconds for the stepper to reset, then crank. The engine will start immediately in this case."

If you know the calibration, you can set it up in one of the cal tables 1-3 or 7-10. You could also measure resistance in a bath and record the values to fill out the cal tables.

Im not ruling this out, but with the Link ecu, the car will still run fine with some moderate vacuum leaks. If it was on a stock Subaru ecu, pending tuning, it may run shit with vacuum leaks.

Do you see the DI change status in the runtime values?

If you want you can email me a copy of the cal, a log, and some pics of the dyno graphs and engine setup. [email protected]

They should be fine on the same bus as they communicate at the same 1 MBit/s and their generic IDs as per the help file are definitely different. How is the wiring configured? It may be that you need to either turn on or off the termination resistor in the dash pending lengths of the wiring. Aside from that, it may be necessary for you to post a copy of your cal file to see if there is any conflicting config data.

shoot me an email at [email protected]

where are you located? I have a couple spare. Im in south florida.

Can wire that wire into a digital input on the ecu. Then could configure the ecu to change boost, timing, and fuel values when that is active.

Will need an external egt amplifier for the EGT sensors. I would look into a canbus based unit.

That is what i did. just removed the resistor and replaced with a piece of wire.

When i modified my stock tach in my ae86, i didnt even use a resistor.

you need to sync the timing between the motor and the ecu to start. Open the PCLink software. Click on help, -> contents, -> tuning functions, -> triggers, -> Calibrate. This will explain how to sync the timing. I would recommend verifying that the distributor is locked down securely and not move it afterwards. In some small cases, it is possible to have the rotor phase angle set poorly. In this situation the spark can jump from the rotor button to the incorrect cap tower. To rectify, it is best to set the motor to roughly 15-20 degrees BTDC and adjust distributor so that the rotor button is centered on the #1 cap tower(mark the dizzy body of #1 tower location before pulling cap to set) and lock dizzy down and resync via PCLink software.

regardless of what base map you start with, you will need to configure all I/O to match your application. You will still need to calibrate all fuel and ignition settings for your specific engine. Did you set the base timing? Most base calibrations are solely intended to be used to get the vehicle started and onto the dyno.

I would set your wastegate tables to have target boost on the y axis as opposed to MGP.

sensor pin 1 is signal pin 2 is sensor ground

the 2.0 diagram that you posted the pic of seems to be a usdm non avcs diagram?

Ive seen all the variations of subaru avcs. a)The early 2.0L variant that had crank sensor, cam sensor, and the 2 avcs cam angle sensors b)Then the later 2.0 and 2.5L single avcs had crank sensor and the 2 avcs cam angle sensors c)lastly the quad avcs motors. in variant a) you would wire crank to trig 1, cam to trig 2, and avcs sensors to di 1 and 2 variant b) crank to trig 1, lh avcs to trig 2, rh avcs to a di channel 1-4 variant c) you need not worry about in your case.

If you are going with the r8 coils, you would really be better off getting a Storm at minimum. You can make the r8 coils work with the atom but it will be configured as wasted spark and semi sequential fueling. Also with an atom on a turbo car, you are going to consume every bit of I/O and wish you had just ponied up for a larger ECU from the start.

DBW will be dependent on board revision. AVCS, oil and fuel pressure shouldnt be an issue once you shuffle the I/O around. Turbo speed will work if you get a frequency divider as it will get higher than the digital input will allow.