Adamw

Ok so it looks much happier without the cam. That means it either has more than one tooth on the cam or it is sitting right on the razor edge of the wrong spot. Have you actually pulled the cam cover off to confirm there is only one tooth on the cam or are you assuming?

The are no cuts being applied in your log (except RPM limit once at 6900). If it feels like a cut and it isnt the ECU applying a cut then it is likely a misfire. Maybe need a bit more dwell or smaller plug gap? The only thing I see happening around that RPM is the cam turning off at 5400RPM.

The V1-6 trigger mode is the correct mode for the common EJ 6 tooth crank/7 tooth cam. What coils does it have? Does it have an ignitor? If you pull the plugs out and crank it do you have a spark?

It looks like some sort of trigger error, I know we have looked at your triggerscope before and said it looked ok but can you do another triggerscope at about 2000RPM.

The Aux outputs are capable of 2A current. The most commonly used valve is the MAC 3 port, something like 35a-aca-ddba-oba. They are widely counterfeited though so make sure you are getting a real one.

Ok, so your new injectors are about 45% bigger, so we need to reduce the pulsewidth by about 45% to end up with the same fuel volume injected. The quickest way to do this is left mouse click on the top left corner of the fuel table (this selscts the whole table), then on your keyboard type in "-45%" and hit enter. This likely wont get it perfect due to different deadtimes etc but it will get it in the ballpark.

Are you using traditional fuel equation mode or modelled?

In your "semi successful start" log the trigger error counter is continuously counting up and it is bouncing off the rev limiter the whole time. So I suspect something is not what you believe it is on the trigger. Most likely either it is not 36-1 or there is more than 1 tooth on the cam. Can you set trig 2 sync mode to none and do another log. If we still get trigger errors that will confirm it is not happy about the crank - if we dont get errors and it doesnt bounce off the limiter at idle that will probably confirm it is an issue with the cam. Note it may be hard to start with the cam sync off. if it doesnt fire let it stop before trying again, it will only sync on the correct stroke approx 50% of start attempts with no cam sync.

I dont have a base map to share, but the EvoX trigger pattern is definitely in there. Note, if you didnt already know, at the top left corner of any selection list there is a little A-Z button, clicking this will sort it into alphabetical order and make it easier to find things.

Our manual shows 36-2-2-2 but there may be some different models. Once you get the triggerscope to show something then you will be able to confirm. You have the pin numbers correct in your notes above, so it should work but the fact the triggerscope is showing nothing means you have something wrong.

You have the CAN lambda configured on both CAN 1 & CAN 2, im not certain but that could potentially upset things so I would start by turning off the one that is incorrect. If your Lambda is connected to the DI9/10 pins then that is CAN 2 so I would turn off CAN 1. Once you have done that then reboot ECU, go to the runtimes screen (f12) and have a look at the CAN tab, see if there are any read errors showing on CAN 2.

The triggerscope shows no signal at all, just a bit of noise. Just in case you done it wrong - can you confirm that you clicked the capture button while the engine was cranking? The K3VET should use the Subaru EJ20 quad AVCS trigger mode. There is no user configurable RS232 in our ECU's (or any other ecu that I know of), only a couple of preconfigured streams designed to be used for older motorsport dash displays.

We dont have OBD2 available as an input at this time. I dont think there are any plans to at this stage either. Not a bad idea though, it is not something I had considered before. I would expect your car to have CAN bus though. You might have to buy a cheap CAN sniffer and learn how to find it. Have a look at the original ecu and see if it has any wires that are twisted pair connected to it, if so that will likely be the CAN bus. I have a liitle bit of CAN info for a couple of late Toyota's and they both have wheels speeds on the same CAN ID's.

Accuracy and repeatability are two different things. The absolute accuracy may not be too great as they probably dont go through any type of calibration check against a known standard at the factory but I would expect being a solid state device the repeatability to be good. Your dif fuel pressure seems to sit consistently around 295-300Kpa for the vast majority of your log, the only times it drops is when MAP increases so I think the sensor is reading fine and it is showing you have a control problem. It could be a regulator issue or the pump not keeping up. Note dif fuel press will never be dead flat, all regulators have a "gradient" and never control fuel press at exactly 1:1 with boost, however some are much better than others. You are seeing about a 30kpa drop in differential for less than 100Kpa increase in MAP so that is worse than I would typically see.

In firmware 5.6.6 we added 3D CLL trim tables. You get one table for adding fuel and a seperate table for removing fuel. So, you could turn these on, then put your switch status on one axis. In the example below, I have my CLL switch on DI4. I have put DI4 on the Y axis of the trim table (0 means switch is off, 1 means switch is on). So for instance when my switch is on and at 4000RPM the CLL is allowed to add up to 20% fuel. When the switch is off it is not allowed to add any. Obviously the other axis doesnt necessarily need to be RPM, TP or MAP might be more suitable but it is up to you.

It will need to be tuned by a professional. Ecu settings>Auxilary outputs>Engine fan. Most commonly 300Kpa with engine off or vacuum hose off. But really it can be whatever it needs to be for the fuel system to work properly. Again this is best left to the tuner. I would just set boost control mode to off for now. It needs to have a safe tune in it before you go adding boost.

The value you use to check/set base timing is irrelevant, all we are trying to do is set up the ECU so when it is commanding say 10deg, that the engine is actually receiving 10deg. The procedure is explained in the help file page: G4+ ECU Tuning Functions > Triggers >Calibrate Start by adjusting the master fuel up or down a little and see if it gets it any better. Otherwise you are going to need to adjust specific regions of the fuel table which im reluctant to advise yet.

Something like below would be the most basic setup. All numbers are just examples, you will have to see what works best for you. In this example a virtual aux is used to enable launch control when wheel speed is <10kmh. The launch control limits RPM to 5000RPM and retards timing by 20deg. As soon as the clutch is let go and the wheel speed goes above 10Kmh launch control will turn off.

The ECU end is easy, Rackpak not so much... If you have the Universal EFI VNet module or the IQ3S dash then both those are actually user configurable. If you have one of the other VNet modules they are hardcoded and the end user cant change anything. However, even the ones that are configurable are majorly painful to configure and Racepak's non-existent CAN documentation makes a lot of it guesswork. I have done a custom stream on one before but it took hours to get right and I had the dash in front of me for trial and error. I dont think I would be able to do it without a dash to test on.

Measure resistance between pin 1 (APE) and Pin 5 (MES), the normal range is 30-300ohms. I would also suggest you check this at the ECU end of the loom (ecu unplugged) as this will verify wiring is correct and all connections/crimps are sound.

There is a large TVS diode inside that should protect the ecu, assuming you have an appropriately sized fuse on the main supply as the wiring diagram shows. If there is no fuse then it will possibly fry the diode.

My understanding is something like this: For a rotary there are three separate combustion chambers. On ported engines especially there is also significant intake port/chamber overlap so that fuel/air charge from one chamber can bleed through to the next one following. If you only cut fuel but there is still a spark then there can potentially be enough air fuel in the chamber (that had no fuel injected) to allow (lean) combustion. If only spark is cut then the first cycle after the cut is removed potentially has a partially fueled chamber that could cause a lean combustion cycle. The safest way to limit them is to first cut fuel, then two rotations later cut spark, then to remove the limit you need to first re-initiate fuel, then two cycles later re-initiate spark. For a piston engine the fuel is added to the same cycle that the spark is in so it doesnt have this complexity. I believe the reason to not have spark and fuel cut at the same time in a piston engine is due to possibility of injector and spark timing crossing over or overlapping causing a similar effect where you might have a cycle that ends up getting a partial fuel dose.

You shouldnt be messing around in there. As per my original picture above this pull-up resistor would be done in the wiring external to the ECU. Example below.

The unfiltered file I cant really tell. The filtered file sounds pretty odd however it definately has all the spatter and pop characteristics that I would normally associate with knock noises. I would suggest deliberately making it knock at some relatively low load area to see if you get a similar sound. You can do that by putting some really large timing numbers into a small group of cells in the ignition table and use your throttle to move the engine operating point into one of those cells.

I remembered to ask Simon about this today. His only thoughts were possibly some early firmwares had this feature missing. Can you pull the cover off and give us a photo of the sticker on the micro (hopefully there is one).