[SUCCESS] MT7615E "Driver own failed" fixed and TxPower optimized to 19dBm on Kernel 6.1.x**

Hi everyone,

I wanted to share my progress and a working workaround for the MediaTek MT7615E (DBDC) on the Banana Pi R2. After struggling with the dreaded driver own failed (timeout=1000) and low signal strength, I managed to get it fully stable and optimized.

My Environment:

  • Hardware: Banana Pi R2
  • OS: Ubuntu / Debian
  • Kernel: 6.1.159-bpi-r2-main
  • Driver: mt7615e

1. The “Driver Own” Timeout Fix

Surprisingly, the timeout issues disappeared after ensuring a clean firmware load and specifically patching the EEPROM. If you are facing initialization errors, check your dmesg. For me, the card now initializes perfectly: mt7615e 0000:01:00.0: Load eeprom OK, count 1024 byte

2. TxPower Optimization (The 19dBm Mod)

The default EEPROM often limits the card to 13-16dBm. I manually patched the Target Power registers in the mt7615e_rf.bin firmware file to reach the hardware’s stable maximum of 19dBm (0x26) across all bands, while staying within ETSI (HU) limits.

The Workaround: I used dd to patch the hex values at specific offsets in the firmware file:

`cd /lib/firmware/mediatek/

Patching Target Power registers for Band 0-3 (Offsets 329-339)

Value 0x26 corresponds to 19dBm

for offset in 329 331 333 335 337 339; do
  printf '\x26' | dd of=mt7615e_rf.bin bs=1 seek=$offset conv=notrunc
done

Reload driver

modprobe -r mt7615e && modprobe mt7615e

iw reg set HU

3. Results

  • DBDC Mode: Working perfectly. Running 2.4GHz and 5GHz concurrently via hostapd.
  • Signal Strength: Solid 19.0 dBm on almost all channels.
  • Stability: No more firmware crashes or interface hangups.

A note on 2.4GHz: If you see Fallback to 20 MHz in hostapd logs, it’s likely due to OBSS (overlapping networks). Moving to a cleaner channel (like 11 or 13 in EU) usually helps.

If anyone needs the patched .bin or specific hostapd.conf examples for DBDC, let me know!

Special thanks to the community for the pointers.