RTL failure, resulting in crash

Hi all,

Just had an interesting crash. Flew my large X8 (28in props, 6S) and went to test RTL. It worked well, but when it went into the landing phase, the motors dropped to idle at 30m and pancaked into the ground. The craft is totaled, and the SD card is cracked. I have a Tlog and the param file, but I couldn’t get the SD card to work. From the Tlog, it all looks fine until it begins to descend home. Can anyone help? Is this a known bug? This craft is overpowered, it’s meant to fly a larger payload. My plan was to do a short test flight with no payload, then add my payload for a more comprehensive test.

2016-11-01 10-06-52.param (9.3 KB)
2016-11-01 10-06-52.tlog (553.1 KB)

I’ve had a look at the tlog, and I’m afraid I don’t have a clear answer on what happened. The RTL was going fine until it suddenly stopped sending telemetry packets for 2 seconds. The radios were still communicating, but no data was coming through from the pixhawk for a couple of seconds at the crucial time. That is unlikely to be coincidence.
I can’t even tell for certain if the vehicle disarmed as it doesn’t send any more HEARTBEAT packets. It does send a few other packets (such as one GLOBAL_POSITION_INT which shows the altitude). It also sent one VFR_HUD packet showing throttle at zero, which probably means it did disarm, but that is not certain.
Note that disarming should not stop it sending telemetry.
Without the DF log I can’t tell for sure if this was a hardware or software failure. I know that doesn’t help you.
I can tell you that it didn’t reboot as the timestamps didn’t reset.
Sorry I can’t be more helpful!

the last VFR_HUD indicates a throttle demand of 0% , during the climb, there is some sensor clipping due to vibrations, but I doubt that was the reason
Motor PWM is also lower during the last part of the climb, than hover… unusual.

I am pretty sure there’s an error that explains it in the dataflash log (not all messages make it over the radio)
there is enough time between the “good flight” and the “throttle demand = 0” lines to have space for some serious error or something that trigged crash detection.

try to get the dataflash log. (not sure you had trouble with the SD after the crash, or before, and have no logs)

Thanks for the reply. The vehicle flipped over and pancaked from about 25m. At 13kg AUW, the entire top of the drone was destroyed. Thankfully, the batteries didn’t rupture, but they were also badly damaged in the crash. The SD card that was inserted into the pixhawk was cracked in half. I might look into data recovery on the SD card, but the Tlog is currently all of the data I have.