Adamw

If the boost doesnt increase with the boost hose disconnected then there is a mechanical issue.

Is this a standard single throttle body or ITB conversion or something? I notice in your most recent log the E-throttle motor DC very rarely goes negative, it is mostly working at quite high positive DC. It is like what I would expect to see it there was some stiff springs fitted or high friction linkage?

Yours is not related to the OP which has an internal hardware issue. If you want help solving yours it would be best to start your own thread and attach a log and tune exhibiting the problem occurring.

I think the main problem is your firing order table, it should be set like this: I would also change the trigger mode to the "Chev LS2/LS7 (90 offset)" mode, as the existing one wont reliably sync during cranking. This wouldnt make it run incorrectly, but sometimes you would probably find it will just crank but not start. The following arent related to your problem, but a few suggested changes from our LS3 to give you a better starting point: Set your accel fuel up like this: Idle control:

For the VVT, set your PWM frequency to 250Hz on all camshafts, then with the engine warmed up and at fast idle about 2000RPM, set the VVT cam angle test to "Calibrate", this will automatically populate the tooth count and tooth offset tables and it should turn itself back to off after a few seconds if successful. Your VVT should then work. Quite often with subarus you will find the cam control quite different on each side of the engine, if you dont get good control you will have to change the VVT "PID setup" setting to custom and optimise each bank independently. Our WRX11 base map vvt settings should be a pretty good starting point for an EJ207.

The Ign Sw function will cut all fuel and spark when the signal on DI2 goes off, so you must have something else connected to the ignition switch circuit that is keeping it high. Possibly the alternator lamp? Provided those fan relays are connected to ignition drives I dont see any way they could cause it.

Ok, odd, they are not there by design as far as I can see from our manufacturing files. My test ecu which is the only one I have handy doesnt have it, but it is a pre-production one. I suspect possibly the manufacturer had a concession to add the thread to facilitate work holding while machining for the first batch before proper tooling was made or something along those lines. Head office is on holiday now for a couple of weeks so I cant confirm the full story.

That suggests the wiring is ok. How is the sensor attached to the rail or fuel line? Is it possible the port or flow path to the sensor is being blocked somehow when you thread the sensor all the way in?

Load source should be set to off for MAF use as you are measuring the mass directly, so the ecu does not need to know/calculate density. Having said that, I suggest you stick to speed density for most applications, MAF measurement isnt particulary good during transients due to the inherent lag, most OEM ecu's will have a speed density type calculation in the background that overrides the MAF measurment during transients, Link and most other aftermarket ECU's dont have this functionality. I would only use MAF for logging or in applications where operation is relatively steady state such as aircraft or cruise boats.

I dont see any obvious reason in your log. How often do you get this error? Can it happen at anytime or only after sustained WOT, or long periods of foot work?

The link to the log works. What is this log meant to be showing? The fuel pressure sensor is not working correctly in that log but the engine is running, I thought the claim was the engine wouldnt run when the fuel press sensor was assigned? I guess that could be the case if they mean when they change the fuel system type to FP sensor which would upset things since the sensor is reading no pressure. You have a constant 0.59V showing on AN Volt 2. To get that voltage it would likely be the sensor pinned wrong. Is it definitely an AEM sensor? If you unplug the sensor does AN Volt 2 then show 0.00-0.02V in PC Link? With the sensor unplugged, if you short the 5V and signal pins together using a paperclip or similar like below, does AN Volt 2 then show 5.00V?

To make it easy to get started on the forum each user gets a small storage allowance to allow them to attach log files, tune files and pictures to their post. When you have used up that allowance and try to attach a file you will get an error message stating your file is too large to upload. The way forward is to upload the file to an online drive service and use a share link to the file. Google Drive and Onedrive are probably the most common free options in western countries so I have done a quick guide on those two here. There are many other options such as Dropbox, iCloud, WeTransfer, etc that usually follow a similar process. How to share a file using Google Drive and Onedrive.mp4

Can you email me a photo of the underside of the board where the wires have been added. [email protected]

That doesnt just look like the typical drop you see with a restrictive turbine, this looks more like the wastgate is being bled open by the controller, or a plumbing/mechanical issue. Since we have no idea what the standalone controller is doing this is pure speculation. You will either need to connect it to the ecu or verify some other way such as disconnecting the wastegate hose.

Typically there will be some method to force the analog output to a specific voltage. For example with an AEM wideband when you unplug the sensor the controller will output a constant 2.5V, you can then look at the voltage reported on the ECU AN Volt input to confirm it matches. You can then plug the sensor back in but leave it hanging in free air so the controller then outputs the full lean signal of 5.0V to give 2 calibration points. Then you check the same 2 voltage points with heavy chassis electrical load working (ie head lights on full beam, all fans, ac and demisters on). The problem with many lost cost analog devices is the output references the power supply ground so no matter how good your wiring is, the calibrated value will change as electrical load through the chassis changes.

I still have a copy of the original modified firmware so have attached it below if you want to give it a try in the meantime. As far as I know it worked ok for the other rover. This will be a 6.23 based firmware so you will lose some of the more recent firmware additions. G4xApplication-6.0.1958 Rover Triggers.img

They are not threaded holes, they are/were plan drilled/milled through holes with a smooth internal diameter of 5.6mm, designed for screws to go through. So it sounds like you have made your own threads by winding a larger bolt into the hole. As long as you only gave them 3 uga duga's each you should be good

The service notes say it was reverse battery voltage damage. The large barrel-shaped TVS diode and the main FET that is just on the back side of that damaged track you show was fried and the main 14V track which goes to pin 35 & 27 was burnt. That track was replaced with a wire. That was a NZD150 repair which is our minimum service charge that most common repairs will fall under. I guess it is possible that burnt track you now have was actually mostly damaged in the original incident but remained hidden under the solder mask and still had a weak connection that passed our test but has since heated up with higher loads in the car. Its hard to imagine that not being obvious though.

The 3 through holes are 5.6mm dia, so yes, intended for an M5, 3/16", USA#10 or #12 screw. The 4 blind tapped mounting holes on the sides are M5 x 0.8.

Is there anything actually wired to aux 10? I suspect that function might be a legacy control for a separate stepper motor driver box.

Try these. Evo 9 V1.1.zconfig Evo idle issue tune 3 - idling V1.1.pclr

Have you just purchased this ecu? That ecu serial number was originally unlocked in 2018, it has come back to us for repair in 2021 with reverse polarity damage, several components fried and a track burnt through. It was repaired and tested as all functional back then. If there is 12V on pin 34 but pin 6 is not grounded then it sounds like the power circuit has been damaged again, it will need to come back to Link for inspection and repair. Contact [email protected].

In that case I wouldnt hesitate going for the NB1 plug-in Vs a Link wire-in ecu, this particular plug-in is a fairly recent design so has nearly all available IO exposed. After a quick glimpse, I think Ign 5-8 are the only outputs missing, but you do have all 8 inj drives, and 6 more aux outputs compared to an xtreme wire-in so it will have plenty for even a very serious race car. I have no recent experience with the ME442 so perhaps my comments arent relevant, but about 4 years ago I tuned a ME221 and a couple of "Nodiz" ign ecu's from the same company and based on those I wouldnt be rushing to try them again. The nodiz just stopped working within about an hour of running, it took over a month for them to fix it, but it eventually came back and I got the tune finished. About 3 months later the car came back running like crap and when I connected to it I found the ign map and several other settings had all scrambled themselves, I loaded the old tune back in, it started and ran good until it warmed up and then decided to fry itself again. The ME221 was slightly better but was very basic and the software was buggy as hell considering it had almost no features. Would crash often so you lose anything that hadn't been saved or stored, wouldnt reconnect if ign was switched off and back on etc. The EMU black I have tuned a few, again a very basic ecu designed with low cost as the primary intent, it has very basic low-performance hardware but at least it seems relatively stable on the ones I have tuned. They have actually done quite well to get as much functionality from the tiny processor that they have, but you do really notice the limitations once you start working with it compared to what im used to. Many tables have to share the same axes, interpolation between table cells is very low resolution, analog inputs are low resolution and slow sample rate, timing control is poor under transients, only 2 slow PWM outputs, 2 slow frequency inputs, no internal logging etc... I have never looked closely at how well it takes care of important scheduling of things like cuts and recovery from cuts etc as they have only been used on very basic road cars but I suspect from the tiny 64pin 8bit xmega running it there will be significant compromises there too.

Your throttle looks to be controlling quite well in that log, I doubt you would be able to feel the +/-0.1% movement in most conditions. I would think the lambda target of 1.1 in that cruise area would more likely be the cause of any surging or torque variation you can feel, you will need more like 45deg ign advance and nice big spark plug gaps to run with stable combustion at 1.1 lambda. You can possibly slightly improve throttle control with the latest firmware, released a couple of days ago. In FW6.25 a user-configurable deadband has been added to the E-throttle control which will quieten E-throttle movement down when it is close to target. 0.05 - 0.10% has worked well on the couple of engines I have tested. Also in 6.25 the max limit on E-throttle derivative gain has been increased so you could try also bumping the derivative up to say 80 which has resulted in improved stability on some problem throttles for me.

And did you check the other pin has continuity back to the ecu knock wire? Polarity doesnt matter for most Bosch doughnut sensors including the KS4-P, but there are a few odd exceptions so it would pay to check just in case you have got something different. The ones you need to be careful of have an internal connection between the centre metal mounting boss and one of the pins, so these are polarity sensitive. Use a multimeter to check there is no continuity between the centre boss (or engine if still mounted) and either of the pins.