Got everything tuned a flying well without the spinning lidar, and got it to fly well. When I hooked up the lidar, it induced a crazy wobble on takeoff and several moments after takeoff. I unplugged the lidar to prevent its spin, and it flew great again. I verified with another flight with it plugged in and the wobble was back. Looking at the logs, of course the yaw, pitch, and roll are crazy, as this is what it did. The desired yaw is steady, while the actual yaw is going back and forth.
Any ideas on how to remedy this? It is just the rotation or could it possibly be magnetic field related due to the motor for the lidar? It is mounted in the center on the top of the copter. The base is 2 inches above the copter arms and the top of the lidar is 6 inches up. Next to the lidar are two gps antennas on polls. One has a compass built in. The flight controller (PX4 cube) is mounted on another shelf just above the arms, below the lidar. The top of the cube is about 1/2 inch below the lidar.
Must be the gyroscopic effect of the spinning mass. Ever tried moving around a powered-on hard disk drive ?
Back when the Hero2 was brand new, I used a hard drive motor spinning 3 platans to stabilize it while it was hanging underneath the quad on some silicone tubing. Those were the days… !
Anyway, back to the task at hand. The spinning doesn’t affect your yaw. It’s the pitch and roll that bear the brunt. And if you dampened the unit in any way, remove the dampening, because it makes wobbles worse.
Your best bet would be a 2-axis gimbal. A fast servo-based one will do as long as it’s trimmed well, 'cause the spinning lidar “wants” to stay level. A brushless is more difficult because you end up with more hardware obstructing your scan surface, usually.
Or, you can try autotune with the lidar spinning. But if I had this problem, I’d first try it with something cheaper installed ontop, like the aforementioned HDD motor, It’s a simple BLDC, I was driving mine with a cheap 10A Turnigy ESC. And if you can find older, thicker drives, with more platans, you can even tune the weight.
Slowing down the scan rate seems to have helped. (It defaults to 4.5 Hz) Also, the top plate was coming loose, obviously that would add to the problem. I’ll have to ramp up the speed again to see if it stays good.
I’m not entirely certain it is the lidar. It sometimes takes off ok with the lidar spinning. Also, it takes off great in position hold and does this twitching in alt-hold on takeoff. Looking at the logs, vibration is within limits, roll and pitch vs desired looks like roll and pitch are more than desired. I’m wondering if it needs an autotune as it is overcorrecting for problems, which are more likely on takeoff. Anything else I should look at? Could a motor be failing/sticking on takeoff? All props don’t start at the same time, but it is the side that is lower that starts up, which I would expect.
