RC Loss Triggering Often w/ Joystick

[jcharette]

Hi all, I’ve seen similar issues brought up in many posts, but there are usually no solutions.

https://github.com/PX4/Firmware/issues/13702
https://github.com/PX4/Firmware/issues/5807

Manual control lost issue

But curiously, how could the sitl report the same issue?

Manual control lost

Each time the UAV will be armed and the joystick is activated at that time, any call to the configuration file for reading parameters or recording - results in interruptions, and the UAV issues a manual error. That is, when the UAV is armed, and at this time press any configuration button in the program (it accesses the configuration file to display the last settings) - the UAV loses connection with the joysticks and pixhawk does not receive the signal and switches to filesafe mode (at this tim…

Basically, when using a joystick, QGC reports RC_LOSS often, and triggers failsafe procedures. I have been working around this by disabling the failsafe, but really would like to understand what is happening.

REPRODUCE: I’ve noticed that when I open the mavlink inspector it will cause the issue every time.

I have also noticed that with the failsafe disabled, I maintain full control of the vehicle, so it doesn’t seem to be an actual loss.

I have read that spikes in CPU usage can cause a delay in RC which can be seen as an RC loss. I have disabled mavlink streams in the firmware to drop the cpu usage(PixRacer) to about %60 and logs do not show a spike in CPU. I have also tested with a Holybro Kakute f7, which idles at ~%35 and also has the same RC loss issue, so the CPU usage doesn’t seem to be the culprit.

I also read the commander can get bogged down, and then read old RC data with an old timestamp and report it as a loss, but there no mention of we might work around that.

If anyone has any insights or is also having this issue please let me know! I will update as I learn more, and hopefully find a solution.

Thank you!
-Jake