You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I used firmware version 1.7.3 for flying an X-UAV Mini Talon in mission mode with hand launch. I tried to analyze some strange roll behavior during takeoff which I did not understand. I analyzed the logfile and found some very strange yaw setpoint behaviour which results in 15° maximum roll setpoint during climbout. After the climbout mode is left at 40m altitude the obviously correct yaw setpoint is computed. I guess the wrong yaw setpoint during climbout is computed with respect to the launch point.
I analyzed the log with flighplot. Image 1 shows the relevant part during climbout.
Image 1: Climbout yaw behaviour
I expect that during takeoff the yaw setpoint is computed from the takeoff waypoint.
I think that there is a problem with the mag initial heading because there is a 35° correction directly after launch. However the problem I describe here seems not to be affected from this because the yaw estimation is corrected afterward. I think the initial left roll is due to the wrong heading estimation but the following 15° maximum roll in the other direction is due to the wrong yaw setpoint.
I did not find a yaw and climbout related issue and I tried to compare the current git master vs. the 1.7.3 code to check if there might already be a fix. The current master has the takeoff code rearranged, so I am not sure. From what I see it seems to be the same.
The text was updated successfully, but these errors were encountered:
I think I found the problem... The takeoff waypoint is only about 60m away from the launch point and the takeoff point is passed at 1:09:50 before the altitude was reached. It seems that during climbout the next waypoint is not selected - so the update happens when climbout is finished. I guess this does not happen when the takeoff waypoint is sufficiently far away such that the climbout is finished before the takeoff waypoint is reached.
I used firmware version 1.7.3 for flying an X-UAV Mini Talon in mission mode with hand launch. I tried to analyze some strange roll behavior during takeoff which I did not understand. I analyzed the logfile and found some very strange yaw setpoint behaviour which results in 15° maximum roll setpoint during climbout. After the climbout mode is left at 40m altitude the obviously correct yaw setpoint is computed. I guess the wrong yaw setpoint during climbout is computed with respect to the launch point.
Flight Review: https://review.px4.io/plot_app?log=b35b3054-ca70-4e56-ad30-206b863e9f83
Youtube takeoff camera 1: https://youtu.be/kpTj9TMslAM
Youtube takeoff camera 2: https://youtu.be/X1NRNYEkSco?t=282
I analyzed the log with flighplot. Image 1 shows the relevant part during climbout.
Image 1: Climbout yaw behaviour
I expect that during takeoff the yaw setpoint is computed from the takeoff waypoint.
I think that there is a problem with the mag initial heading because there is a 35° correction directly after launch. However the problem I describe here seems not to be affected from this because the yaw estimation is corrected afterward. I think the initial left roll is due to the wrong heading estimation but the following 15° maximum roll in the other direction is due to the wrong yaw setpoint.
I did not find a yaw and climbout related issue and I tried to compare the current git master vs. the 1.7.3 code to check if there might already be a fix. The current master has the takeoff code rearranged, so I am not sure. From what I see it seems to be the same.
The text was updated successfully, but these errors were encountered: