Heavy Quad arm (MotorsX8 coaxial multirotor) - Crashed

Good Day!

MotorsX8 coaxial multirotor Drone crashed during its 1st flight test. What could be the possible problem that I need to check before I perform another flight test for this drone?

Params and Logs Files : RP-U515B _ MT50 Delivery Drone - Google Drive

Battery Used is (1s2p) Tattu 14s 30000mAh ( so 2 pcs of 14s in parallel connection)

FC : Pixhawk 6c + Holybro PM08-CAN Power Module 14S, 200A

RC : Siyi Mk15

Telemetry : Holybro SiK Telemetry Radio - Long Range 1W 915MHz

GPS : M10 GPS/GNSS

Drone Frame Specifications

I followed the PX4 guidelines in the Qgroundcontrol.
I was able to configure and tune drones with total weight of 10kg ( quadrotor Drone) without problem. but this heavy drone that can carry of upto 50kg fails on its 1st flight. We have done calibrations too before flight test.

What could be the problem and what should I be looking for?

Im thinking this might be the problem..
Ill check it .

This is the exact frame we used.

How to measure the motor position for KM?

Also the PY/Z axis value?

When posting logs, use Flight Reivew, like this: https://review.px4.io/plot_app?log=09cbd0ca-0d4a-4015-8d7d-7988e8f564df (this is your log)

It looks like your Mag was not happy

Looking at thrust (current to motors) vs mag field is very correlated, this is bad.
This will casue the EKF2 internal estimator to get wrong readings and ignore the magnetometer.
Most likely, the motor cables are too close to the MAG. I would suggest using an external (e.g in gps) or standalone.
The orange should be mostly flast, only 5-15% changes.

Lots of saturation on the motors aswell

I haven’t done the math but with such high potential thrust you might be running into the problem of not being able to give low enough thrust on the motors. The motors have to spin a certain speed to not stall or produce uneven lift. There is also a minimum thrust MPC_THR_MIN and MPC_MANTHR_MIN

One of the best tips I can give you is to build a good testing rigg where the drone can be attached and spin withuot damanging anything.
It can be as easy as a steel bar where the drone can only rotate in 1 way.
This will let you test props direction, thrust, mag, some initial flight stability etc before doing real world tests.

This is VERY large, costly and dangerous drone and the according tests have to be done

1 Like

A few other notes regarding large platforms:

  • Large drones create alot of noise, both on gyro and acc. There needs to be sufficient filtering, often both hardware and software. Motors are only 70 KV meaning they produce very low frequency noise, especially on low throttle. Those blue dampener might be too stiff.
  • Gyro filters too high to filter noise atm, try IMU_DGYRO_CUTOFF = 12Hz, IMU_GYRO_CUTOFF = 20Hz, prob need to calculate for best performance but it is a start. Lower accel filter aswell IMU_ACCEL_CUTOFF = 10`
  • Notch filters (RPM based) could help alot (requires DSHOT telem or CAN telem). They allow higher LP filters for less delay while still filtering well.
  • With large propellers and linear throttle ESC (usually all are) the throttle to thrust relationship is not linear. This creates a huge effect at low and high throttle. I would suggest setting THR_MDL_FAC = [0.7,0.85]. Here is some good info: Tuning Process Instructions — Plane documentation
2 Likes

Hello thank you for the reply.

Ill try to change the dampener on the flight controller
also will try to make some changes on the layout of the components to reduce the noise

also the frame we used is JIS M50 which is custom made to 8 motors