Adamw

Preliminary support will be in firmware 6.19 which is due for release any day now. So far only widely tested in a JDM 2007 STI (same as 2008 in most other countries). I believe it should work with most cars to 2011. Lots of torque based messages as well as fuel consumption etc so the engine needs to be mapped in modeled mode and the torque management function needs to be set up so that the ecu estimated torque is realistic. I can give you a base map in a few days when the included features are publicly released. We are working on a plug-in ECU for these cars but it is still at prototype stage so quite a few months off still.

I dont have much else to suggest. So, to confirm, your white wire that is shown in the pic above in ECU B connector, is connected to the white wire that is in position 8 in the AIM 14pin plug? And the green wire connects to the Blue wire in pos 9 at the AIM end? terminals are pushed fully home in the housings etc?

Plug-in ECU's with E-throttle dont have Aux 4 available as it is used as part of the E-throttle control circuit. You can use any other aux from 1-8 for VVT solenoids.

Ok, the next step I would suggest is remove the ecu from the car and power it up on the bench with a 12V battery and a couple of test leads, then try connecting. All it needs to power up is +12V on pin 5 and ground on pin 34. There are two possibilities; It is not powered up properly - i.e a bad ground or a back feed etc. So enough current leaking through to power the LED but not enough to get it to come alive. Some of the very early metal case Atoms had a relatively sensative USB circuit and it could be damaged easily. So it could potentially be damaged. If you dont get comms when powered up on the bench then that would be my suspicion and it will need to come back to NZ for an upgrade.

It is hard to tell without seeing the map, but it looks like the 0% AP row in your E-throttle target table must have some variable numbers across it. It should generally be flat - at least near the idle area.

Did you attach the correct file? It has all the same mistakes in it and none of the changes I suggested.

Attach your map and a log of a start up.

That is not an error, just notification that you are trying to command a PW that is less than the min eff PW that you have set so it is being clamped to the minimum.

I answered last night in your similar other post. I have merged them together.

Your auxes are fine set up the way they are, they can work either way around - the only important thing is the correct cam position input is assigned to the correct cam solenoid output (which yours is). I will get the manual or basemap changed to remove that confusion for the next user. I suggest the first step is to perform the VVT calibration. Go to VVT setup, hold the engine above 1200RPM and change the Cam Angle Test to "Calibrate", it will take a few seconds then switch itself off when finished. See if that solves the "missing pulses" error. If still no joy, please attach a log.

Unplug both the ecu and dash, and check continuity from end to end. ECU pin 27 should have continuity to AIM pin 8 (14pin connector), and ECU 28 should beep through to AIM 9.

Please attach your dash config and ecu map again.

Both are configured correctly now. The transmit and acknowledge errors suggest there is a problem on the bus. Can you give us some photos of the wiring so a second set of eyes can confirm.

Select the correct ecu then give me the configs again. The ECU is showing acknowledge errors on CAN 2 so we need to see if they go away when the dash is set up correctly. Oh, in the ECU, turn the OBD off in the CAN setup screen too.

Set up instructions for the ECU are in the Haltech ICC software help file. You dont need a separate CAN bus for each device, assuming they all run at the same bit rate or have a configurable bitrate you can connect all devices to the same CAN bus. So GPS can likely connect to either bus. Let me know what GPS you have and I will confirm bit rate.

Change bit rate to 1MBit. Set OBD to off. Change channel 2 mode to off.

I think the problem is the dash config. In your dash config, change the "ECU" to "LINK - CAN_BUS_BASE_LCC". I think the CAN BUS FULL stream you are trying to use is something very old from G3 or G4 days.

@Davidv New PC Link & firmware version on Link website today should have this fixed.

Does that mean you solved your problem? What was the problem?

Try this, I have added boost switch and fixed the lambda scale. Dew MXP Strada new.zconfig AfterDyno10-29-20CANAiM new CAN.pclr

Is the USB cable plugged into the correct USB socket (on the underside of the top board) and not one of the CAN sockets on the bottom board? I have seen users plug them in to the wrong socket before.

Yes, I normally have fuel system type set to none if Im using modelled mode with NA ITB. I think as Mapper suggested the FP sensor method would probably work fine with some MAP filtering but just I have never tried to confidently say go for it. I think the average MAP in each runner would usually be fairly similar so the fact you are only measuring MAP in one runner doesnt concern me much. Yeah maybe overthinking it. You usually have pretty odd looking VE numbers at low loads with ITB's anyhow and I have never had much drama getting them to work well with consistent fueling.

Yeah, attach your ecu map and the dash config, and maybe some photos of you can wiring may help us spot something you missed.

Not really, the only important thing is you dont want the dizzy tooth to pass its sensor at the same time the missing tooth on the crank is passing the crank sensor. So anywhere else is fine.

No, changing load source to BAP/MAP X-over or MAP with TP on the fuel table axis look up will not solve your problem - your VE will still be the same when manifold pressure is 100KPa. Most of your fuel settings look right, so large VE numbers usually means the engine is not getting as much fuel delivered as the ECU thinks it is delivering. Fuel pressure looks ok so I would be suspicious of injector flow. Those 2000cc CNG injectors are very prone to partially seizing and or blocking so I would start with a flow test on those first.