Hi, I need help diagnosing a repeated propulsion failure on a heavy-lift ArduCopter.
Setup:
- Quad, ~12 kg AUW
- Cube Orange
- 4× T-MOTOR U8 Lite KV190
- G29×9.5 props
- T-MOTOR ALPHA 60A 6S FOC ESCs
- 6S LiHV battery
- T-Motor DataLink telemetry
Flight 117: same heavier setup, flew and landed normally.
Flight 118: front-right Motor 1 suddenly almost stopped, then recovered violently. DataLink showed:
- FC motor command was still present
- ESC input throttle was still present
- but ESC output throttle collapsed and RPM dropped
Other three ESCs did not show this behavior.
I replaced the front-right ESC with a brand-new ALPHA 60A, checked the propulsion system, and did a short takeoff/landing test. Everything looked normal.
Flight 119: on the next real flight the aircraft again lost thrust on the front-right side and crashed. ArduPilot logged:
Potential Thrust Loss (1)
Motor 1 command went close to maximum while the aircraft continued to roll/pitch toward the front-right side. EKF/GPS errors only appeared after the aircraft was already out of control.
Unfortunately the aircraft then burned completely, so I cannot inspect the failed hardware or read the final DataLink.
My main question is:
What could cause the same Motor 1 / front-right propulsion channel to lose thrust twice even after replacing the ESC?
Possible common points I am considering:
- Cube Motor 1 output / PWM pin
- signal or ground wiring
- front-right power distribution / PDB connection
- motor phase wiring
- FOC commutation/desync or ESC protection
- prop/hub slip
Could RCOU.C1 still look normal in the log if the actual PWM signal physically reaching the ESC is interrupted?
Also, how much confidence should I place in Potential Thrust Loss (1) identifying the actual failed motor?
The forum does not allow me to upload the log files directly, so I uploaded all relevant logs to Google Drive. The Drive folder contains ArduPilot logs 117, 118, 119 and the T-Motor DataLink logs from Flight 118.
Google Drive link: log - Google Drive
Any help analyzing these logs would be greatly appreciated.
