April 21, 2021
Join us
Agenda
- Community Q&A
- Roadmap
- High priority queue
- Release
- In-Depth discussions
Release
MAVSDK v0.39 release 8 days ago. Contribution with a transponder plugin to talk to an ADSB device.
Feedback from testing the PX4 0.12 release:
-
HITL seems broken. Framerate and simulation rate goes down and breaks normal functionality.
@JulianOes working on it: mavlink receiver fixes by julianoes · Pull Request #17438 · PX4/PX4-Autopilot · GitHub- Need a fix to successfully arm because of battery checks (check battery is 4 cell like SITL):
https://github.com/PX4/PX4-Autopilot/pull/17438/commits/86f6d30570746743005d31afdbb92de526d6bb33 - Problem with QGC connection
Make sure to run HITL on the autopilot first and have it connected to the simulation before you start QGC.
Mentioned in docs? Simulation | PX4 User Guide
- Need a fix to successfully arm because of battery checks (check battery is 4 cell like SITL):
-
Follow up on feedback about hand launched fixed-wing disarming before throttle was increased:
commander_params: lower COM_DISARM_PRFLT but disable for fixed-wing by MaEtUgR · Pull Request #17421 · PX4/PX4-Autopilot · GitHub -
Found bug where you can ramp the velocity setpoint in position mode while disarmed:
Fix velocity rampup based on stick while disarmed by MaEtUgR · Pull Request #17437 · PX4/PX4-Autopilot · GitHub -
Reported PWM output rate issues. Possibly because of reconfiguration after using DSHOT.
Fix the repeated configuration of PWM rate after initialization by CUAVcaijie · Pull Request #17427 · PX4/PX4-Autopilot · GitHub
Blockers need to be labeled with v1.12-Release-Notes: Issues · PX4/PX4-Autopilot · GitHub
Community Q&A
-
Kalyan Sriram: Question about progress on Durandal magnetometer calibration issues
CUAV Nora mag calibration crash · Issue #16813 · PX4/PX4-Autopilot · GitHub
Suggestion to test with [DO NOT MERGE] commander: temporarily increase worker thread stack substantially by dagar · Pull Request #17439 · PX4/PX4-Autopilot · GitHub, see the comment. -
Jacob Christ: Pixhawk 4 DS-009 document has a 6 pin JST-GH power connector and a 6 pin telemetry connector. Isn’t there a risk of plugging it in the wrong way around and powering the telemetry port?
Yes that’s good feedback and it’s in theory possible to cause damage. In practice the pins are usually tolerant to a 5v power source. Ideally a product could use a different connector for power e.g. Micro-Lock Plus (specific part number). It’s not consumer proof and there are other ways to wire it up wrong. There’s more work to be done if we want to get all connectors bulletproof. We are not aware of a widespread power connector issue with the current documented standard. Issues are reported in Issues · pixhawk/Pixhawk-Standards · GitHub with prefix DS-009. Issue to mention Microlock Plus connector in standard.
-
Jacob Crabill: Reason for F4/F7 builds did run UAVCAN at high rate was found (minutes ago). Canard cdev was polling with 10ms interval. David mentioning to keep priority inversion issues with the current driver in mind which is only an issue with high bus usage.
Courtesy of Peter van der Perk PX4-Autopilot/CanardNuttXCDev.cpp at 789d91900df6ac826da450abc0c6becad09e8799 · PX4/PX4-Autopilot · GitHub -
David is working on getting PX4 working on current upstream NuttX. There were quite some changes introduced which result in effort. Major advantage on the horizon is dynamically allocated file descriptors (no fixed limit) and improved flash and RAM size.
Roadmap
Discussion based on board: Roadmap · GitHub
High priority queue
Discussion based on board: High-Priority Queue · GitHub
In-Depth discussions
For smaller groups expanding technical discussions, stay until the end and follow up.
If you have any feedback or corrections please comment on this post.