First off, A big thanks to all of the people who have worked so hard to lower the technical difficulties of getting into Quadcopters. I'm a technical person but this is my first dive into RC aircraft.
I have put mine together, configured it, and got it flying quite nicely. I'm now focusing on being a better pilot, and then plan on moving to telemetry, automation, and aerial photography.
The Bug
I have noticed a bug which I have been unable to troubleshoot. Sometimes when I first power the craft on, it doesn't properly lift off. Instead, two of the motors spin at reduced RPM while the other two spin faster than normal resulting in the craft tilting up on two booms. I have been able to reduce the throttle before the craft flips over in each occurring. I've noticed the following
It only does it on the first lift off of the flying session/battery.
It doesn't do it each flying session, only occasionally.
It doesn't always tilt in the same direction.
It happens more commonly in the X configurations, but has happened in both configurations.
Typically, unplugging the battery and starting over will resolve the bug.
The Fix?
Not sure. Since this only happens on the first liftoff of a session, I have to wonder if there is something in the state machine which has not yet become initialized. eg the state machine thinks that it needs to go hard in a direction to compensate for something.
Any thoughts from the community.
Again, many thanks to all of the people working to bring these kits/code to market.
Replies
I had the same issue. Here is what is going on (at least for me). The yaw axis in stable mode expects your craft to move in response to the motor speed differential (it generates yaw by changing the speed simultaneously on two motors with the same prop direction). When your craft is sitting on the ground it is obviously not generating any yaw movement. Thus, the APM continues to try and generate yaw, which is why only two of your motors are spinning. You would think that when you let off your yaw stick (or it is centered), that no yaw would be applied, but that is incorrect. Search the forum discussions and you can see the explanation given for this which tracks what I've stated above.
You can easily test this by using the configurator. Simply hook up your quad and set the throttle to the mid point and watch the flight data screen. All motors will have roughly the same speed. Now apply a small amount of yaw and watch what happens. If you are in stable mode, two of the motors will drop their speed and stay at that dropped speed until you give it another yaw command, even though you have released the stick. If you were in the air, your craft would turn and the APM would recognize this and then the motor speed would even out again. However, given that the craft can't move, the APM continues to keep yaw applied, hence the difference in motor speed.
There is an easy workaround. Acrobatic mode does not do this. So, for example, before I take off, I simply flip the switch from stable to acrobatic and then back, as this "zeroes" out the commands for some reason.
Sometimes before I take off (in stable mode), I will accidentally apply some yaw, or my yaw is not trimmed properly. Either way, this will make the motor yaw keep increasing until two motors are not spinning at all because the craft can't start moving around the yaw axis until it is airborne.
This is particularly noticeable if your yaw is not trimmed to the expected midpoint of the software, as it will continue to apply more and more yaw, which you won't notice because your motors aren't fired up yet.
In any event, before I take off, I always toggle the acro switch to zero it out. It is very simple to test. As soon as you flip the switch and then flip it back, all of your motors will be spinning at roughly the same speed.
Hope that helps.
the arducopter should sit still on the ground while you see the running light effect on the shield
it's initializing the Gyros while it blinks
Lots of good replies here. Thanks.
I think Im going to put more time in my startup to allow everything to settle down.
Will try this and if the problems continues, I will come back with more information.
Thanks all.
Magellan
Make sure the Arductoper is sitting STILL and mostly level when you connect the main power (and start the IMU).
The only time I've had what you describe is when I tried to connect the battery while hand-holding the Arducopter. Same symptom. I don't know if this is well-known, standard, or a red-herring, but I treat it carefully now.