I am a developer and my graduation project (due next week ) is to develop a autonomous quadcopter. I have based it on the APM2.6 and a beaglebone black with mavlink.
To test the stability of the quadcopter platform I tried loiter mode today.
I noticed some troubling behaviour in loiter mode. When I armed the APM2.6 in loiter mode and moved the throttle up slightly the motors started spinning fast. The throttle still was in the “go-down” part of the throttle in loiter mode.
This behaviour does not occur after first putting the throttle into the center (1500pwm) and then all the way down till the rotors stop spinning. When I then put the throttle slightly up the rotors won’t spin until I move the throttle past the center.
I expect that the latter behaviour is the correct one. So the rotors don’t spin until the throttle is above the center.
Is this something I can fix or is it a bug in the arducopter code?
Whilst only fairly new to this myself your description confuses me.
The ‘go down’ and start at high speed in particular, I’d expect your motors to start almost immediately at a low speed when you push the throttle forwards, by 50% the intent is to be at ‘hover’, higher than 50% climbs, lower than 50% which is not enough to sustain lift and it descends.
I take it that you calibrated everything appropriately - transmitter, transmitter / RC to copter, ESC’s if they can be.
Rotors not spinning until past 50% is not normal, I suspect that going back to basics by checking all cables & connections, calibration and tuning my reveal the cause.
I had some issues on my setup with motors doing strange things and it was down to the APM (Pixhawk in my case) updating the ESC’s too often. My ESC’s have a published max refresh of 450Hz but the APM was updating at 490Hz, the symptom was motors stopping when they shouldn’t and going from 0 to max rpm with little to no provocation, it wasn’t predictable how or when they would. I also had issues of the motors not running / starting appropriately because the lowest value from the RC calibration process was 200 units below the start point of my ESC’s - the RC calibrated minimum throttle was 980, the ESC’s needed 1120, my ESC’s can not be calibrated so I had to modify what my transmitter was sending in addition to setting appropriate minimums on the throttle channel, these issues took me two or three days to fix.