rusefi.com

Hello everyone!

I'm new to the rusEFI platform and recently got a Hellen 154 ECU for my Hyundai Genesis Coupe. The seller recommended that I use an AEM X-Series Wideband (model 30-0300) for easy CAN Bus integration, which would allow the ECU to receive data directly from the sensor.

Based on their advice, I connected the AEM's CAN High (AEM+) and CAN Low (AEM-) lines to my harness via the OBD2 connector, as shown in the diagram they provided: The issue:
After completing the wiring exactly as shown, the ECU is not reading the wideband values reliably. The lambda/AFR readings in TunerStudio are erratic and fluctuate wildly. I’ve double-checked the wiring and confirmed the polarity of CAN H and CAN L. I also attempted to configure the CAN settings in the rusEFI ECU - Enable CAN Wideband set to True - but nothing seems to stabilize the readings.
This is a quick video of the readings seeing in TunerStudio.

https://youtube.com/shorts/GAg-5wjlOQo

Things I’ve already tried:
Verified wiring and pinout of the AEM 30-0300 against what the seller provided.
Experimented with different CAN configurations in the rusEFI.

My setup:

ECU: rusEFI Hellen 154 for Hyundai Genesis Coupe
Wideband: AEM X-Series 30-0300 (CAN output version)
TunerStudio + rusEFI console
Wiring via OBD2 connector

I’d really appreciate any help or suggestions from those who are running the AEM X-Series via CAN. Maybe I'm missing a configuration step in the rusEFI firmware, or perhaps there's something unique about the AEM 30-0300’s CAN protocol that needs tweaking?

Thanks in advance!

Appreciate nicely articulated question!

hellen154hyundai has on-board wideband controller which is fighting with your external CANbus. https://github.com/rusefi/hellen154hyundai/issues/153 looks like this was now yet addressed on software side?

seems like so far most people do some evil hardware mods to kill on-board WBO? On the bright side I expect a firmware-only fix in the next 7 days

GNS4 wrote: ↑

Thu May 15, 2025 8:15 pm

I’d really appreciate any help or suggestions from those who are running the AEM X-Series via CAN. Maybe I'm missing a configuration step in the rusEFI firmware, or perhaps there's something unique about the AEM 30-0300’s CAN protocol that needs tweaking?

https://github.com/rusefi/rusefi/pull/7917 now you should be able to manually select which AFR controller to use.

Hello guys,

Thanks a lot for the updates!

After going through the link shared by Andrey (https://github.com/rusefi/rusefi/issues/7631), I was able to partially resolve the erratic readings from the AEM X-Series wideband.

What worked for me was setting the WBO (Wideband Oxygen) index to 1. Once I made that change, the AEM readings started working correctly — just as described in the reproduction steps within the GitHub issue. Interestingly, the readings continued to work even after disconnecting the car battery and rebooting the ECU via TunerStudio.

Now, After trying to update the ECU firmware, I got a Java error and looks like the board is no longer recognized by my windows as a COM port.

Attached the image of the Java error and the device manager no longer showing the RusEFI COM port. Any idea how to solve this?

Thanks to @ggurov in discord for the help with flashing the ECU in DFU mode!

It looks like the erratic readings from the AEM X-Series Wideband via CAN are now resolved. The values are stable and accurately match what the physical gauge displays and remains after rebooting the ECU and disconnecting the battery from the car which is good. That might help with https://github.com/rusefi/rusefi/issues/7048 However, I’m running into a new issue. When I try to start a fresh tune from scratch and updating TS, I get a configuration error in TunerStudio (see attached image). The error specifically mentions a problem with the "WastegatePosition":

"open 0.00 / closed 0.00 cal values are too close together" Alongside that, TunerStudio also throws multiple other errors — TPS, Pedal, CLT, and IAT are all flagged, which is odd. TPS calibration doesn't seem to be capturing the "grab up and grab down" value, doesn't appear that the pedal is doing anything. That didn't happened starting a fresh tune in the previous firmware version as far I noticed it. What's more interesting is that if I load my previous working tune, none of these errors appear, and the car starts up without issues.

Any idea what might be causing this difference between a fresh tune and the working tune? Could there be default calibration values that conflict with my hardware? I can attach both files if needed.