0

I'm working on an ECU reprogramming and coding app (iOS, Android, etc.) for a client who is part of the automotive tuning scene. I have been trying out dozens of OBD2 adapters, but most of them fall short in one or the other way. I wonder whether an OBD2 adapter manufacturer is reading here or whether anyone of you have contacts to someone in that area. We would be interested to bundle the tuning software with a white-labelled OBD2 adapter.

If we could just plug an USB adapter, we would not have a problem in the first place, but we absolutely need mobile connectivity.

As long as you only transmit OBD2 commands, even the most shabby ELM327 clone will do. Once you start sending UDS commands, it gets rough. Many adapters don't handle intermediate (NRC78) responses and most will fail sending large ISOTP buffers (4095 bytes in "one go") which are required for boot loader, application, and calibration data transfer.

The requirements are:

  • Support all CAN-based protocols
  • Mobile connectivity (WiFi and/or BLE5) with a simple serial protocol on top
  • Powered through the OBD2 port
  • Hardware ID filtering
  • Enough memory for buffering a few thousands of CAN bus frames during intermittent loss of connectivity.
  • Automatic ISOTP handling up to the full payload size (4095)
  • Voltage reading

The STN22xx-based OBDLINK MX+ ticks a large number of these requirements and would almost be perfect, if it was available as a white-label solution.

I'm not sure whether mechanics.stackexchange.com is the proper forum to post such a question, but I figured it might reach the appropriate audience, so bear with me. Do we really need to build our own adapter?

8
  • Rather than building an OBD2 adapter it may be easier to build a box that will wirelessly forward your traffic to a USB based devices. You could use a Raspberry PI clone with a battery to do that, and it's probably easier than replicating the OBD and CANBUS functionality.
    – GdD
    Commented May 3, 2021 at 8:25
  • @GdD - Wireless and tuning are not a good combination. If for some reason, while writing the tune to the ECU it lost connectivity for even a millisecond, you stand the risk of bricking the ECU. It's the same reason why they say to ensure you have full battery power in the car and in your computer (usually a laptop) before you start writing. Wireless solutions are required to accept interference (at least the FCC says they do), which means it could drop the signal at anytime. Not a good thing when trying to tune. Commented May 3, 2021 at 9:45
  • Wireless is not a problem, if the adapter is well designed. If it has enough memory, it could easily hold the data transfer content before sending it over to the ECU. Commented May 3, 2021 at 10:23
  • I'm not going to argue with you on that one @Pᴀᴜʟsᴛᴇʀ2, it seems a more sensible approach than trying to build the adapter itself though. Another approach would be to use a micro-computer to run the software and then remote desktop into it wirelessly.
    – GdD
    Commented May 3, 2021 at 18:48
  • 1
    Sounds like you want to invent a new device, because to my knowledge, nothing like that exists right now. Whatever you devise will have to be a lot smarter than what is available right now. I wish you luck. Commented May 4, 2021 at 14:33

1 Answer 1

1

I actually work in this field, have been the sole developer and done this. A custom PCB is the best way, also allows for security to be built in and add on data buffers for the transfer data. I have not heard of anyone white labeling a wifi or Bluetooth dongle, only USB. If you want contacts for the USB send me a DM and I can send you some people. On the other hand if you want a full time job instead of developing whatever company you are working for also let me know as I’m looking to grow our team fast.

1
  • I’m curious. Contact me.
    – zipzit
    Commented May 5, 2021 at 0:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .