Possible crash due to pixhawk failure? Need help!

ArduPlane
1

Firmware version: 3.5.2
Airframe - X-UAV Talon
Flight controller - 3DR Pixhawk

I was tuning the aircraft in AUTOTUNE mode, and the canopy flew off the aircraft mid-flight. I proceeded to land the aircraft and began the approach. After a short while, I noticed that the plane was no longer in my control. Mission planner showed the plane to be in AUTO mode. By the time I could switch back to FBWA and try to regain control, the plane had overturned and it crashed. I also noticed that the plane didn’t respond to my input in FBWA either.

I checked the logs to see what could’ve gone wrong, and found that there was a brief period in the log where the dataflash log had recorded no sensor data. That period contained only ORGN and ATRP messages. The logging resumed when the plane switched the mode to AUTO. After the switch to AUTO, the dataflash contained only about 2 seconds of logged data.

I checked the telemetry log and it paints a weird picture. The HUD showed (SAFE) which I’ve seen happen when the hardware safety switch is pressed. which clearly didn’t happen mid-flight. Following that, there was no dataflash log. The plane switched to AUTO, rolled right, and went beyond the LIM_ROLL_CD specified angle of 45 degrees. The Tlog also showed the altitude at that point to be 0. But the plane was still in the air! When the plane crashed, the Tlog showed an altitude of - 37m.

I’m thinking that the pixhawk or one of the sensors failed mid-flight. Is an open canopy with oncoming air at 15m/s not good for the pixhawk?

I also noticed that the ORGN message was being updated in the period where there was no logging. The home position was being updated very rapidly from the original position to a new position. What could cause the location to be updated like that?

What cause the (SAFE) message to be displayed? What could have caused the pixhawk to have stopped recording?

This seems like a very weird crash, unlike any I’ve ever seen before. I’d love if some of you would have a look at this and give their thoughts on this.

EDIT: forgot to attach the flight log - https://drive.google.com/file/d/0B-NIddUSff_hSVFDdWxwT3MyNUk/view?usp=sharing

Thanks,
Nihal

2

Found It!

I was reading the EKF documentation. and found this -Extended Kalman Filter Navigation Overview and Tuning — Dev documentation

SH - the ratio of the barometer height inconsistency to the limit set by the EKF_HGT_GATE parameter. This can have the occasional spike to over 1/2, but should never go above 1. If this line goes above 1, then it indicates that the filter stopped using the barometer data for that period in flight. This should never happen with good sensor data. …
Factors that can cause this to be high include airflow past the autopilot affecting the barometer reading and accelerometer errors due to sensor drift or aliasing.

I checked my log file to see the SH parameter. and this is what I saw -

image.png1151×279 17.7 KB

The SH (Barometer Position inconsistency)value has gone over and beyond 1. The GPS position inconsistency did not show such drastic changes.

I still have no answer to why the pixhawk stopped recording the log after this.Or why the (SAFE) message appeared mid-flight. I hope to find answers to this soon.

I’d love to get some of the developers or the community’s thoughts on this.

Best,
Nihal

1 Like