Lost RC signal leads to crash

Flight Testing & Log Analysis
1

Hi,

I designed and built a 7-inch enclosed-frame drone and took it outside for testing, but I noticed some strange behavior with the radio link.

At first, the drone flew really well and was very responsive to my commands. I landed it and took off again, but this time I experienced a brief loss of throttle for just a fraction of a second. I kept flying, assuming it was a minor glitch.

Later that day, I flew it again, and right after takeoff, I experienced another throttle drop, this time I completely lost control of the drone. It drifted away and crashed into my pool wall (lol). After the crash, it even tried to take off by itself and ended up mowing a bit of my lawn.

For some context:

  • I had a SiK 915 MHz telemetry radio mounted relatively close to the RC receiver (FrSky R-XSR).
  • The GPS was mounted far from the receiver.
  • The RC antennas were outside the enclosed body of the drone.
  • The receiver’s signal wires were twisted together and routed alongside the motor wires.
  • My flight controller is a Matek H745 Slim, and the ESC is a Skystars KO45.
  • My RC loss failsafe was set to land

Here’s the log file.
Thanks in advance for any help! :slight_smile:

2

Sorry to hear about the crash.

So, had you flown it in altitude control or position control yet, or only stabilized? I wonder if altitude control was not tuned yet and hence becoming unstable once it went into RTL/land mode when it lost RC.

Looking at the log it seems like it was stable for the period of the loiter (waiting for RC) and only got into trouble once trying to land, right?

3

Thanks for the quick response.

I already had successful and stable Position Mode flights earlier that day. Before the problematic flight, I changed the EKF2_HGT_REF to the TF Mini to test terrain following. During that flight, Position Mode drifted significantly, so I took over and landed the quad manually.

After landing, I changed EKF2_HGT_REF to 1 GNSS (the recommended option) , armed the quad, and proceeded with the critical flight.

The goal of that flight was to test Position Mode and RTL with the height reference set to GNSS, since it normally defaults to the barometer. Because of this, I can’t confirm whether Position Mode was working correctly before the crash.

The main question that arises from this event is: Why did I get the “Manual control loss” error message?

After a quick log inspection, I realized there was no RC_LOSS failsafe triggered; instead, it was a manual control loss failsafe. I don’t fully understand the difference between the two, or how it’s possible to lose manual control while still maintaining an active RC connection. I also checked the failsafe parameter on my transmitter, and it’s set to No PWM.

Thanks a lot for the help,
Gabriel

4

Ok, so this could be a regression with terrain following in the random alpha software version that you’re using. I suggest to use stable v1.15.4 or the latest v1.16.0 release candidate, and not some random version in-between. Otherwise, surprises like this are more likely.