Doing a mapping job for a mine today when my unit started radically climbing and ten dropping. So it would move along on it’s grid, stop and then start to climb at full throttle. Then it would fall back to it’s mission altitude. After a bit, it would continue on it’s mission. I would do this two of three times a flight. I flew to this side, could the high altitude affect the baro in the flight travelling to the job. Weird. Usually I carry a full extra Pixhawk with sensors, but left them at home.
You really need to upgrade from RC2. I can’t recommend using unstable code if you’re going to let it get so old…
I think you have bad vibration issues. You’re getting aliasing. Look at the difference between IMU.AccZ and IMU2.AccZ during those climbs. Your accelerometer calibration doesn’t look great either - look at the difference between IMU.AccZ and IMU2.AccZ while the copter is on the ground.
I thought mission planner told you when there were updates to the firmware… Guess not! I’ll do that right away. Funny about the vibration, it’s always been really low. I’ll test it again. Maybe at full throttle I’m getting a big harmonic. Thanks for taking a look!
Actually, I have to ask what is RC2? I thought I had 3.14 loaded? When I loaded the latest firmware I got “bad gyro health”. Calibrated the first time and if failed. Second time was the charm. I’ll take it up in the next few days and see how it does. Can’t believe I got good maps out of this project! Had to disable the images from the higher altitude when doing the point cloud, but successful considering how insane it was flying. Normally for safety I would have grounded it, but the mine as empty and I flew in with no other window for mapping given the crazy weather : )
You’re running 3.2-RC2 - which could very possibly get false-positive LAND detection that disarm you in mid-air.
You’ll always get vibration on a multicopter - during forward flight in particular. You can’t eliminate it because it is partly an aerodynamic effect of moving a fixed prop sideways through an air stream. Also full throttle might cause more issues even when static of course. The trick is correctly isolating PIXHAWK from it.
If you want perfect vibration isolation, go out and get some 1/2" dubro foam rubber, a little bit of 1/2" x 1/8" balsa or some kind of plywood, cut strips the same length as the pixhawk, affix them to the front and back of the pixhawk so that it has a square base, and then put 1/2" cubes of the dubro foam at each corner of the square. You can CA them on or use an appropriate double-sided tape.
Might be overkill, but it will work properly.