I have been flying my C8 3axis with 5D2 and 16-35mm lens with 8000 4s. Gross weight of 14.7 lbs I have checked the motors for vertical alignment and have the aircraft and gimbal balanced (I believe) I have been noticing that my on some flights my even motors have a much higher current draw than the odd motors. I have attached the GPX for a flight where I can't figure out why the even motors are so out of whack with the odds. Of interest is that this was flight 9 or 10 and while some of the other flights display a similar pattern others don't ... anyone have any idea what might be causing this disparity? Thanks
Scott if you have the wireless XBee system you could use MKTool to check the the copter is balanced in flight.
I must confess Gary that I don't truly understand what exactly the XBee system is. I thought it was a telemetry device and since I was using the Graupner radios I probably wouldn't be needing the XBee... outside of the MKTools and the balance issue what else does this do? I did google it but havn't really found much that I can seem to make sense of .... well at least at this hour (east coast) Thanks Gary
Scott it allows you to do changes to the FC system wirelessly. So you can change settings and see things like the balance window real time. Essential if you ever plan to do waypoint flying in the field.
Scott I looked at your data and performed some statistical system analysis on it using a regression model and some data massaging techniques. What I did was regress motors 1 and 5 against pitch , motors 3 and 7 against roll and the rest of the motors against the resultant combined pitch and roll angle using simple trig theorem Pathagorian. This allows me to pool current changes due to pitch and roll into error. I can now come up with a normalized current for each motor by removing any response from attitude changes This in theory models the copter on a continuous flat plain. Now motor response will only be due to outside factors such as how the the motors effect that change. I took the average of the normalized currents for the even motors. Then with that result I regressed it against each resultant motor current to find trends in the motor response. With this method we should expect to see that the individual even motors match the response of the average with almost perfect statistical significance. We should expect that odd motors have a negative response with good correlation. What I found was an oddball (no pun intended) to be the culprit motor. Number 3 an odd motor, tested to a 100% statistical significance to the response of the total average of the even motors. Motor number 7 had almost 0 correlation value on the even motors. Number 1 and 5 tested to a statistical value of 45%. They did negative leverage on the response which is what we would expect. Although they are not efficient. 7 Is of concern it may be providing some yaw inertia in the same direction as the evens cancelling out its own torque force and therefore not provoking a response one way or another. The remaining odd motors 1 and 5 are working but not as efficient as they should. Below are the graphs for each adjusted individual motor current against the average of all adjusted even motor currents. You can see that numbert 3 matches perfect. Meaning what ever 3 is doing it is providing a direct perfect linear change in all of the even motors.