We completed 5–6 test flights for Basic & tuning and PID optimization. After tuning, the drone performance was stable, so we proceeded with an endurance hover test with full charge battery 50.1v
Battery failsafe settings
42 V → Low battery (no action)
40 V → RTL
Reserve capacity: 7000 mAh
Critical: 4000 mAh
Incident
During hover at ~38 minutes flight time (battery ≈ 40.2 V), Motor-3 stopped suddenly in mid-air, similar to an ESC desync.
There was no error message or warning in the GCS.
This occurred after several successful flights with the same configuration.
I have attached the flight log.
Could you please help identify the cause of the sudden Motor-3 stop?
Please define precisely what you mean. Did you do initial PID tuning and stopped there? Did you tuned the notch filter? Did you do quick tune? Did you do autotune? Did you do mag fit? In which order did you do them?
Initial PID tuning was performed manually to achieve stable hover, and attitude response.
We did not perform the following advanced tuning steps yet:
Notch filter tuning
AutoTune
QuickTune
MagFit
After the above basic tuning and several successful flights, the drone performance appeared stable, so we proceeded with the endurance hover test during which the Motor-3 stop occurred.
@amilcarlucas PID tuning flights were completed in the morning session, and we were satisfied with the hover stability and performance during the autonomous waypoint mission. All tests were completed in the morning session. In the afternoon, we conducted an endurance test, during which Motor 3 stop unexpected.
@Adam_Borowski After the crash, we inspected all the electronic components. There was no burnt smell, and we powered on the drone. We tested all the motors, and they were running properly. After reassembling the same setup with propellers, we conducted a static thrust test. Each motor was tested up to 3.5 kg thrust upto 38 V with continuous operation. During this testing, the motor did not stop. (Static test means the drone arm and landing gear were mounted on a thrust rig.)
You may want to check with T-Motor directly and ask if Flame ESC’s are the best choice for those motors. Not for the current rating but the firmware they run.
I think interesting is also what @Adam_Borowski figured out, the extremly raising current just before motor 3 maxed out. So first current raising than FC reaction. Unfortunatley there is no feedback from the ESC showing ESC temperature or RPM.
Where are your ESCs located. Are the cooled by enough airflow, just hoovering for long time is relative hard for the ESC
During last year I tested flame 80A v2 esc with various firmwares from T-motor and had 1 clear case of desync. Since then I’ve moved away from them and pwm escs.
They have indeed a bad track record. With some searching there are a lot of users who experienced problems with the flame ESC and big motors. But also with the more expensive alphas.
Why you do think this was a desync? @amilcarlucas At the time of the crash, I saw that Motor 3 stopped. It looked like a desync, not a disarm stop. After the drone fell to the ground, Motor 3 did not rotate.
@Juergen-Fahlbusch From the image below, you can understand the ESC placement. The ESC is attached to the fuselage CF side plate with proper heat dissipation. The ESC-to-motor wire length is 65 cm, the ESC-to-PDB wire length is 10 cm, and the signal wire is the default length no extension.
@Dr.Gadget In my case, this is the third occurrence. Previously, I used another motor with similar gains to the Antigravity motor, but not a T-Motor. During a mid-air turn after hovering 3–4 minutes of flight, Motor 4 stopped in a similar way. I want to clarify that in both crashes the motors were different, but the ESC, TBS PDB and battery were the same brand. If this is an issue with the Flame 80A V2.0 ESC, is any T-Motor ESC developer available on this forum for discussion?
Desync is the most likely reason for this. An over temperature shutdown shouldn’t produce a current spike.
If commutation doesn’t fit motor orientation, thats similar to a short.
65cm ESC DC wires is quite long. What’s the wire size? Did you add additional lowESR Caps to the ESCs?
From the power supply side its better to have short DC wires and longer AC wires as they are less critical - yes, this contradicts heat dissipation aspects.
air cooling is not the issue, because I run for 40 minutes at a constant 3.5 kg thrust per motor in direct sunlight. After the flight, I measured all ESC temperatures, and they were only 41–48 °C, not higher. I designed the ESC heatsink facing outward, so I hope this is not the cause of the motor stopping.
I don’t think the cooling is a real problem in this build. Airflow isn’t perfect but that shouldn’t matter too much. Its roughly 15W per ESC thermal loss. (7A x 40V x 0,05)
@KowsiganCould you please tell us about your wiring?
Edit: Got your numbers wrong. All fine in this case.
Edit2: Might be worth to disassemble the problematic ESC and check heat transfer pads/paste applied correctly (if any). If heat isn’t transfered to the aluminium case sufficiently, the case will stay cool but the PCB/Mosfets get too hot.
