I don’t know if this is new to 1.12, but the preflight arm timeout (COM_DISARM_PRFLT) is WAY too short (10 seconds). I’ve now had two crashes because it timed out before I could throw it and the plane’s motor died just as it was leaving my hand.
COM_DISARM_PRFLT should default to 60 seconds IMO
Slight logic bug with Rover: When mission planning, GQC disables the waypoint button until you add a Takeoff waypoint, but there’s no such thing as a Takeoff with Rover. So it’s impossible to plan a mission 
Temporary workaround until QGC allows mission planning with Rover without a Takeoff: fortunately, Mission Planner works well enough with PX4 that you can plan the mission there and then edit it in QGC.
VTOL manual flight with transition, maiden on this airframe. Flew okay and transitioned well, but it’s still not well tuned in MC mode and tends to “falling leaf” when descending. Modifed EasyStar. Probably a bit underpowered with a 3S battery; will switch to a 4S
https://logs.px4.io/plot_app?log=01ceddca-5699-4769-a24d-3ddaf11c4e20
Thanks for sharing! I can see that the system tried to yaw very hard after the back transition and then you ran into the actuator limits - do you have a picture?
https://logs.px4.io/plot_app?log=01ceddca-5699-4769-a24d-3ddaf11c4e20#Nav-Yaw-Angular-Rate
And here is then where it is running into full limits:
https://logs.px4.io/plot_app?log=01ceddca-5699-4769-a24d-3ddaf11c4e20#Nav-Actuator-Outputs-Main
We did push too hard for performance in the default tuning and we will need to go back to robust defaults. Thanks for sharing this!
Update. I switched to 4s batteries on that modified EasyStar and it worked SO MUCH better. No “falling leaf” on descent and otherwise very nice with the default tune. The wing loading is a bit high in fixed-wing mode on this model, so it tends to start off a bit fast, but VTOL aircraft aren’t meant to be gliders.
Tested in MC and FW modes with transitions in manual flights, and RTL in auto
So big thumbs up!
https://logs.px4.io/plot_app?log=2d2a7ccc-b5bc-4a1a-b714-4088f80fbe6a
PX4 Rover auto mission. Worked okay but needs some tuning. Also need to shrink the waypoint hit radius because it’s cutting off the corners
https://logs.px4.io/plot_app?log=c9903cdb-ef38-422b-9725-324b3d1213f2
Community, the developer team has a new beta out (v1.12-beta4) and as always we would love your help with testing!
For some reason Crazyflie 2.1 stoped working after beta 4. Might be related to the PWM changes.
After arming motors don’t spin and when I increase throttle it just makes a beep noise.
I changed the PWM_MAIN max and min rates and I was able to make the motors spin but the annoying beep wouldn’t stop.
Anything I need to fix on PWM_MAIN_RATE?
I tested on my X500, although basic commands and not missions or safety since I was concerned about clippings. This made me to test at low altitude.
My overall experience from all flights:
And one more thing, I did not notice any clippings again. I had made some hardware changes this time to make Pixhawk isolated from vibrations.
https://review.px4.io/plot_app?log=0b7217d4-92e0-4479-bfb3-6b1b4422f249
https://review.px4.io/plot_app?log=5d360f50-f447-444a-93f4-a01bbf9763e5
https://review.px4.io/plot_app?log=9318a556-8487-4a8c-81e2-72a77bf4952b
https://review.px4.io/plot_app?log=633868e6-c281-4776-9e95-e846e60a4d2b
https://review.px4.io/plot_app?log=028f6c5f-058a-4fee-ae5f-0a0f8df24774
Interestingly, starting at 1.11.x we started having to simply disable it all together. Since we taxi the planes out, if we stop for a bit or are crossing paths and need to wait then the whole re-arm etc needs to happen again but I’m definitely with you on the 60 seconds. That makes pretty logical sense.
Did two small tests with QAV250 and Beta5. Twitches are experienced while in position mode. One log has a link to the video as well.
https://review.px4.io/plot_app?log=6560d81a-daed-4783-bcb1-7a80bda80e79
https://review.px4.io/plot_app?log=84a12a54-6641-415c-aa4b-24cb1e3632e3
