Adamw

As far as I know there is nothing stopping you from using the MXG along side the PDM32. Did the dealer give you any reason why it wouldnt work?

If ecu is off and has disconnected then anything displayed on the screen is just what it was at when comms was lost. Is the fuel pump still controlled by Aux 6? Does the forester have the same 3 speed PWM control module like the V10 WRX? The WRX controller actually needs a high signal to run the pump so it sounds like either your car has something different or there is a wiring problem.

We also have data from the GT86 one. I can set one of these up as a receive channel if you attach your map.

No, it the chipset is being seen correctly by windows but PC Link cant get coms then it suggests a hardware issue.

I would suspect then that the USB chipset has possibly failed. Contact [email protected] to arrange inspection and repair.

The Link CAN lambda works fine without the termination. At least it does if you are using our CAN interconnection cables. Your OBD tool likely wont though.

Do you have anything plugged in to the CAN 1 port?

Yes 1/4W is fine.

Dont think so, you last log still shows a significant trigger issue, low battery voltage and low cranking speed.

Ok, that is good then, the Link one can be reprogramed to operate at 250 or 500k. The CAN bus is generally constructed as a main "trunk" with short branches off it. The trunk should be twisted, the short branches dont need to be. The trunk needs a termination resistor at both ends - the ECU already has one built-in so you only need one at the far end of the bus. So to confirm, your orange/green and Orange/yellow wires will need to be twisted by hand from the ecu all the way to the far end. There are many ways you could layout the bus and still met the correct architecture. I would typically do it "daisy chain" style like drawn below where I run the main trunk directly to each device and fit the resistor across the connection to the last device. You should be able to crimp two wires into each pin on the OBD2 socket to save doing any splices etc. Heres how I normally do the termination resistor in a road car: Seal removed from connector, fill back of connector with RTV instead of the seal:

The settings below will likely be close. Only the "Sync position" is unknown. You will just have to try it on the 6 possible cylinders to see which one gets the spark at the right place. Yes this trigger is capable of sequential.

The first question is what lambda device do you have? OBD2 needs to run at 250 or 500kbit so it wont be compatible with some Lambda devices.

Yes you can measure voltage to ground. When there is no data both CAN H & L will sit at about 2.5V. When there is data you will see higher voltage on CAN H and lower voltage on CAN L. It varies with how much data is being sent but typically you will see about 3.5V on CANH and 1.5V on CAN L if there is data.

It shouldnt have. The AC relay does get is power from the same circuit that supplies 12V to pin 45 - but so does the idle valve and boost solenoid. Do those click if you run a test on them? Does the AC request show active and the AC clutch aux show active in the software when you turn AC on? Note the AC relay trigger wire also has the pressure switch in line so you could check that too.

Each ecu has a unique unlock code. You need to contact the dealer you purchased the ecu from to get the unlock code.

Yeah I have similar thoughts. Mostly looks like a trigger issue. Do a cold trigger scope when you can and that should give us some clues where the problem is. The battery voltage is very low so it will help if you fix that. Your trigger setup is odd - does it have a VVT crank wheel and a non VVT cam trigger?

To see the cam and crank signals you need to use the trigger scope. >ecu controls>triggerscope. Click the capture button while the engine is cranking. If you do a PC log everything is logged. If you havent already I suggest you load one of the "1920 x 1080" or "1366 x 768" layouts as these have a much wider range of parameters shown on each page. Unlikely to damage anything. You can fry coils etc but unlikely to damage the ecu.

I think that would be a difficult car to put an aftermarket standalone ecu in. There will be a lot of CAN bus and it will likely be quite complex. Does your can have a manual gearbox?

It sounds like you've got a ground offset coming from somewhere, like sensor ground is coupled to chassis ground somewhere or similar. It may be ok to just set the deadband high enough to ignore it but it is hard to know whether that is going to make it too insensitive to actual throttle movement. You will have to try it and see.

Here you go: CAN Lambda coldstart.mp4

If accel enrichment is set to TPS mode it is triggered by TPS delta (i.e how fast the throttle is moving). Your TPS Delta at idle is upto around 19 so set your deadband somewhere above this. Note this is unusually high (more typically less that 1% delta at idle), it may suggest there is no cable free play, or throttle is not sitting on the stop, or there is a bad ground or similar. But increasing the deadband to more than 19% will fix the immediate problem.

Is the com port number in device manager match what you have assigned in connection settings? What message do you get when you try to connect? Can you also right click on the com port and show us the hardware ID's like my pic below.

Is the Blue LED on the ecu board glowing? Does the ecu show in windows device manager under the Ports - COM and LPT section?

The oscillation is from the wideband device. Suspect it is a bad design with the heater sharing the same ground as the analog output reference as that looks like heater noise you can see (frequency is constant but duty cycle varies). If you unplug the sensor you will probably see it flatline. Pic below from the log with the engine off. I have set up the 3 analog inputs to display the same 0.25V range, you can barely see the dither on the MAP and TP inputs as expected (they are stable within 0.002V if you set scaling to auto) , but ANV4 has 0.3V oscillation - so more than 100 times the others. It should be dead stable with the engine off. You cant really. The best indication would be the parameter "Engine Cycle Count" which will increment by 1 each compete cycle.

Sorry missed this one. The V7-9 idle valve works like a "two wire" valve. It should have +12V, Gnd and a single aux output position signal ( it can use any aux, ign or inj drive since all flyback etc is handled by the internal valve driver). Only the earlier WRX's (I think V1-4) had the old style 3 wire valve that needed 2 aux outputs.