Hi guys!

Perhaps someone has already faced the task of replacing individual blocks in a GM Ultium battery?

They operate at a frequency of 2.4 GHz like Bluetooth, but the MAC addresses are too large.

What you’re seeing with the Lyriq/Ultium pack is that GM’s using a dedicated 2.4 GHz wireless BMS mesh, not “normal” Bluetooth, so MACs don’t behave like you’d expect from consumer gear. Each section has its own local module that talks wirelessly to the K291 pack communication module, and K291 then forwards everything over the wired side to the rest of the car.

Instead of exposing big, friendly MAC addresses, the system really thinks in terms of pack position + module serial + VIN. When you replace a section, that new module has to be “learned” into K291’s table and then programmed so it matches the vehicle VIN and calibration, usually as: program the new section first, then run a vehicle‑wide programming step so all modules stay in sync. If that doesn’t happen cleanly, you get the typical Ultium drama: DTCs and the pack going offline, sometimes even no‑start.

So yeah, the MAC‑style angle is basically hidden under GM’s abstraction.

The interesting questions are: has anyone sniffed the Ultium RF traffic or used ADI eval hardware to map the addressing scheme, and in real‑world section swaps, does K291 always re‑adopt the new module cleanly or sometimes refuse it and brick the pack until engineering intervenes?

Question - how to program the new section first.... ? Another - no problem)