PX4 v1.12 Release Coordination

:construction_worker_man: PX4 v1.12 Release Coordination :construction_worker_man:

Community, the PX4 Developer Team is working towards the next stable release (v1.12), we are hoping this release will come out soon, but we don’t yet have a solid timeline. We are trying to get as much testing as possible, and as always, we are really interested in your feedback.

The main focus areas for this release are:

  • Overall system stabilization
  • Performance
  • Enhancing our User Experience

If you have been paying attention to the last developer calls (#weekly-dev-call:px4-dev-call), you will notice that there have been quite many changes coming in, essentially refactors and architecture changes for the better. The team has felt that this work has been long overdue, and we are all very focused on making this release the most stable baseline we can.

Resources:

How can you contribute?

We need all the help we can get with both bench and flight testing on the release, on as much hardware as possible.

Help with release notes

If you have a few minutes to spare, please help us contributing to the release notes. It’s super important we capture all of the new features and changes. Since this is such a big improvement release, we want to make sure everyone knows all the upcoming benefits.

Join the weekly calls

Make sure you join the public dev calls and follow the conversations on the dev-team channel on Slack. We are making our best effort to keep everyone informed on progress.

Bench testing

  • Take-off the props
  • Flash the beta (you can build from source, or use QGC)
  • Try arming, play around through QGC
  • Disarm, and pull the log
  • Upload to Flight Review
  • Share links as a comment below

Flight Testing

  • Put the props back on! :smile_cat:
  • Flash the latest beta (if you didn’t already)
  • Pick a nice location outside, make sure there’s enough space for you to fly (and social distance!)
  • Arm, and Takeoff :rocket:
  • Try the features you would normally try. Maybe your workflow involves missions. Maybe you like flying in stabilized mode and just moving around. Try anything that you would normally do.
  • Make sure to write down any inconsistencies you might catch from past experiences
  • Land, disarm and pull the logs
  • Upload to Flight Review, make sure you write down your comments from the above step.
  • Share the logs below as a comment

Note to Developers

We are going through a transition period. We are not going to accept any new features until we make the final stable release of v1.12, after which we are moving to a much aggressive six weeks release cycle. Don’t worry; your PRs aren’t being ignored. We are just waiting for things to progress on the release, and we will start merging again soon afterward. It’s also important to note that we are thinking of adopting a new branching model that will allow for development not to stall, which we know can be frustrating.

4 Likes

Contributed by @Yuriy_Homyakov

https://review.px4.io/plot_app?log=5942e4fd-d742-4c0d-970e-9bfc89397eb2

https://review.px4.io/plot_app?log=3c889bca-9c95-4d78-9ee4-702d74a79329

Got this error during radio setup on Pixhawk Mini (FMU v.3 bootloader). The rest of setup went fine and it doesn’t seem to have any ill effects Screenshot 2021-03-17 113604

1 Like

I thought the Safety Switch was disabled by default? I still ended up with arming rejected because of a missing safety switch until I manually disabled it in the params. (This is on plane)

Apologies for re-asking, but I’m hoping to not waste too much time.

The kill switch will work under the following conditions:

  1. IOP and FMU are alive - I have verified this is true
  2. FMU is dead but IOP is still alive - I will verify is this claim unless someone tells me otherwise

I would like to avoid testing #2 if no one expects it to work, anyway.

Here is why I have to ask (exposing my ignorance/inability to follow the code):
It looks like the FMU polls (and pulls) “raw” rc status from the IOP and determines whether to issue kill back to the IOP (assuming thrust actuators are connected to MAIN). I cannot see how the thrust actuators will get killed if FMU does not issue the command(s).

Code tracing raw rc channels input on IOP to kill switch functionality on FMU.

Hey Chris, can you by any chance try it on a QGC daily build and let us know if you still get the same message?

PX4 v1.12.0 Beta - Crazyflie 2.1 Stabilized mode

2 Likes

PX4 v1.12.0 Beta Bench - CUAV V5 nano QAV250 frame

2 Likes

PX4v1.12.0 beta, Pixhawk clone (FMUv3)

STABILIZE / ALTHOLD / POSITION / RTL modes, performs well https://review.px4.io/plot_app?log=e50c1c4f-7940-443a-83b2-ca8f548424f1

MISSION mode https://review.px4.io/plot_app?log=361290b4-8b0b-48c2-a5a8-f280cf43923f
hovering after takeoff, don’t starts mission, need to switch mode from and back to MISSION to start

2 Likes

PX4 v1.12.0 Beta Flight - CUAV V5 nano QAV250 frame

Everything went smooth!

3 Likes

Tested with X500…I had height jumps which was due to no Ranage sensor.

https://review.px4.io/plot_app?log=9f418f89-a797-4fdc-9ffb-20eceeb3fcf9

I updated to the latest daily build, and now the QGC error looks like this (it also takes a long time to download all the params, about 2x what it used to be). Firmware is VTOL latest beta
Screenshot 2021-03-25 162016

PX4 v1.12.0 Beta - Black Cube (FMU V3) F450 frame

there is a problem with the logger. logger starts but does not give an output for some of the tests. the reason is unknown

https://logs.px4.io/plot_app?log=2ee986a9-f9d9-4174-adf4-e79d74247299

Added two other ones before this and before I change the HGT:

https://review.px4.io/plot_app?log=4b18b59c-be3e-4d37-b171-5271620c62be

https://review.px4.io/plot_app?log=ffd7af51-bfc7-492d-bc05-fbffd96c630c

Flew with SITL:
Is it normal?

https://review.px4.io/plot_app?log=462c5f48-8bd5-4fd2-8a84-868931846e06

Another test today. I changed EKF2_HGT_MODE to 0 and enabled EKF2_RNG_AID as @dagar had recommended. Logs seem to report steady height. But the drone was changing height for kinda 5 meters. (The weather was windy a bit as well.)
https://review.px4.io/plot_app?log=f181db72-892b-49d4-b166-168fe9029992
https://review.px4.io/plot_app?log=463cb6fd-fa7a-41d5-8ef3-60022ca69562

My flight with RC was not stable at all at take off (Logs according to it are attached. )
https://review.px4.io/plot_app?log=bf175df7-09ef-48b5-b429-9c752390ea7f
https://review.px4.io/plot_app?log=4719b6c7-ee4c-4064-bcae-07a8ea19f201
https://review.px4.io/plot_app?log=120c61dc-1b5a-4f93-83ab-d92127dc151a
https://review.px4.io/plot_app?log=00703ac0-4004-4c66-b0db-5d36e5c4ddef

Thanks for posting! I’m looking into these. @MaEtUgR it would be great if you could have a look, too.

1 Like

Changed the battery from 5200mAh to 4200mAh and range sensor from TF mini to Lidar lite. Results are just WOW!!!

https://review.px4.io/plot_app?log=ed6a09a1-7d4d-4d09-8a06-3911c5071eec

And here is the MISSION test: I do not understand what these hikes are in altitude. It was kinda furious and fun… :smiley:

https://review.px4.io/plot_app?log=997eec6f-7e0a-44b2-b358-582c7974a444

Hi!

Looks like 1.12 makes IMU_DGYRO_CUTOFF default value to 10, (was 0 in 1.11.3)
I spend 2 hours trying to understand why my quad becomes unflyable whe i switching from 1.11.3 to 1.12.beta. Is this a feature or a bug? My 450mm quad with pixhawk 4 FC can’t fly with default parameters. Here it oscillates and flips on takeoff
With IMU_DGYRO_CUTOFF=0 flies pretty well, with IMU_DGYRO_CUTOFF=30 and IMU_GYRO_CUTOFF=50 also looks good but not tested enough