Discussions

Hi folks, I'm relatively new to RC aircraft with 3 months experience. My goal is to have a low cost, reliable platform for aerial photography of forests which I can do on a weekly basis for my research. I've gotten a lot of valuable advice from this forum on how to operate my Arducopter and I'd like to share my issues now so hopefully everyone can benefit from a solution. My current setup is:

- Next Level 650mm Hex-Copter Frame Kit
- 1 x Flightmax Zippy 5000 mAh 4 cell 25C battery
- APM 1 board, magnetometer and mediatek GPS
- jDrones PDB, 20A ESCs, and AC2830HE-508 motors

After a hard crash where I broke many frame parts and bent several motors, I decided to try the sturdier Next Level frame, which while heavier (about 1 kg versus about 0.5 kg for the Arducopter frame), has stood up better to hard landings and crashes. I've also moved to the high efficiency motors which are supposed to be less likely to be damaged in a crash due to their design.

After replacing these parts, balancing props, and redoing all the setup steps in the Arducopter wiki, the aircraft is still having difficulty navigating in stabilize mode. I think this may be due to accelerometer damage from the crash. The APM board was only protected by the dome pieces that come with the jDrones kit and sustained visible damage including a bent CLI <-> Flight Mode switch and some bent pins, as shown in the picture.

Here are a few issues I'm noticing:

1. The accelerometer doesn't appear to be holding calibration very well. Right after calibration, the arducopter takes off straight up in the air. However after a minute or two of flying, it becomes less stable. If I land and take off again it will pitch to one direction. It will tend toward this direction until I recalibrate. Recalibrating temporarily fixes the issue, however after a few minutes it will choose another random direction, and start pitching that way.

2. If before take off I give the arducopter a little throttle, and let it sit there with props spinning and not moving the throttle, the props will start to spin faster despite lack of any change in throttle. They will sometimes then start to spin slower. I'm guessing this is because the accelerometer is already losing calibration and the Arducopter is trying maintain what it thinks is a stable position.

I wonder if I should try to replace only the accelerometer or get a whole new APM board. Does anyone have any thoughts about whether I'm diagnosing this correctly?

I've also been wondering if I would need to change the PID settings because of the heavier frame.

Any advice is greatly appreciated. Thanks!