Today is the first day of testing the autopilot using APM 2.0 board. Although I had the APM 2.5, it suddenly cease to work. Perhaps something is wrong with the diode. For today’s test, I would like to have the quadcopter lock the GPS, take-off up to 6 metres and return to launch (RTL). The settings of the waypoint definitions are as in figure below:
I have made some mistakes prior to the launch of the autopilot. Firstly, since we are approaching monsoon season, the situation here is very windy. The first and second manual flight was OK, however the third manual flight, the quadcopter crashed on the grass.
Now, after the third flight, I tested the autopilot. First I gave 5% of throttle (autopilot won’t work without initial manual throttle) and I flipped the autopilot switch on my Devo 7 transmitter. The quadcopter instantly flies to 6 metres high and being pushed by wind (or maybe the gyro calibration went berserk due to the initial crash?) to the other side of the field. The quad tried to compensate that, howewer I made a mistake of turning off the autopilot to regain control. Lack of throttle made the quad turn up side down and fell on its back.
Further inspection showed that the frame was OK, electronics are all good, however two of the props were bent due to the crash (see figure below). Observing the Mission Planner, it shows that the gyro settings were very badly off.
Moral of the story are:
- When testing the autopilot for the first time, make sure to operate in the most appropriate whether. Avoid flying in windy days.
- When regaining manual control, approximate the amount of throttle that needs to be given in order to stabilize the quadcopter.
- Having a telemetry is a good idea.
- Having a FPV is also a good idea.