Hi everybody, today I’ll do an article about a method to debug a Navio2 drone, and some basic recalibration and configuration you can do.
Creating a connection between the drone and a ground control
First of all, you’ll have to setup some connections between the drone and the ground control. To do so, I will use the WiFi capabilities of the raspberry we connected to our drone. I will setup an automatic Wi-Fi connection to an access point I will setup on my smartphone (tethering) and configure the wifi using wpa_supplicant on the raspberry.
Doing that, the drone will automatically connect to a specific access point once powered on.
After that, in /etc/default/arducopter, you’ll be able to configure the telemetry (ground control) IP, I will setup my smartphone IP (192.168.x.1), after that all I’ll have to do is power on the drone and launch a ground control software on Android, such as QGroundControl.
The calibration process is pretty straightforward, you’ll first have to calibrate the accelerometer by setting the drone on various position (on its sides, nose down, nose up and on its back), the you’ll have to calibrate the compass by moving the drone in various position.
We had some issue with our drone the last day even though we managed to make it fly the previous days. Even though the raspberry pi got powered on, ardupilot didn’t seem to launch. To debug that issue, I took a SSH client on my phone, such as JuiceSSH and connected to the raspberry pi.
I had two ways to see what was wrong, either in the /var/log/syslog, which showed that Ardupilot couldn’t run, or by launching the configuration of ardupilot (emlidtool ardupilot, if i remember correctly). It said that some module couldn’t be launched. Due to that error message I think there are two possibilities, either the navio is not properly connected to the raspberry (low probability, since the navio is supposed to power on the raspberry), or some issue with the power module (we had one cable disconnected from the DF13 module, maybe that couldn’t power some modules so the startup of ardupilot failed.)
Overall I’m still happy that the drone we built was able to fly and even gets stabilized in AltHold mode, it means that the preparation we did earlier this year were correct and that the only thing that was wrong was that we put the props the wrong way (and that we didn’t setup correctly the motors/ESCs).