Ok , this what happened, I was doing a test flight after I put the quad back together. I flew it most of last summer, maybe 15 flights with no real issues. Occasionally it would lose GPS lock, but I could see that in the logs. the quad is a Walkera QR X350 Pro with a RTFHawk installed. I was doing some low level flight, <10 feet. I was hovering in loiter, looked at the laptop screen, and then it just took off full loiter speed forward, I tried full back on the pitch, but no reaction, by that time it hit the ground, tumbled forward breaking the front props, landed upside down. I didn’t get a chance to switch into stabilize.
When I looked over the logs, I didn’t see anything out of the ordinary, but I am not a pro at log analysis. Can someone take a second/third look? Maybe i missed something
at log line ~57600 you do get some speed, I do not know why, but it does seem like an navigation thing. (8sats, 2.22HDOP , not a very good fix)
Then you move RC2 to max PWM 2006 , and keep it there, while DesPitch drops and actial Pitch follows nicely, it seems to me like to applied full pitch forward yourself.
you also seem to be using simple /super simple mode (not really sure I got the bitmask right, but it seems so) and such mode is very dependent on accurate position, especially when near you, an position error can result in changed controls, as if you and the RPAS were in a different orientation.
Anyway - switching to stabilize would clearly save the day, as it performed (controlled attitude) as expected until it hit something at altitude (or broke apart for some reason)?
if you are new, learn to fly stabilize first, a.k.a. “learn to fly” make it a habit to switch to stab whenever in doubt.
[quote=“Andre-K”]at log line ~57600 you do get some speed, I do not know why, but it does seem like an navigation thing. (8sats, 2.22HDOP , not a very good fix)
Then you move RC2 to max PWM 2006 , and keep it there, while DesPitch drops and actial Pitch follows nicely, it seems to me like to applied full pitch forward yourself.
you also seem to be using simple /super simple mode (not really sure I got the bitmask right, but it seems so) and such mode is very dependent on accurate position, especially when near you, an position error can result in changed controls, as if you and the RPAS were in a different orientation.
Anyway - switching to stabilize would clearly save the day, as it performed (controlled attitude) as expected until it hit something at altitude (or broke apart for some reason)?
if you are new, learn to fly stabilize first, a.k.a. “learn to fly” make it a habit to switch to stab whenever in doubt.[/quote]
Thanks for looking… Something does not sound right.
Not really ‘new’ to this. I have been flying this quad since last spring, and multirotors about 2 years or so.
I do not believe I have simple/supersimple set up on this one, but I’ll look. Once it veered off I wanted to switch to Stabilize, but was only ~4feet off the ground, first reaction was full back on the pitch. BUT you are saying the logs said full forward??? Then I think I pulled the throttle back and let it run into the ground, then it tumbled. I’ll look over the logs again…
your RC2 (pitch) is at 2006 shortly before crash, and desired pitch is leading higher than actual pitch, so it seems like high PWM means “forward”.
your SIMPLE paramter is 2 so is SUPER_SIMPLE Unless I remember wrong, third bit (2) means third mode (your loiter) is super-simple .
That can influence “which way is what” on the radio depending on GPS relative to your (takeoff) position.
[quote=“Andre-K”]your RC2 (pitch) is at 2006 shortly before crash, and desired pitch is leading higher than actual pitch, so it seems like high PWM means “forward”.
your SIMPLE paramter is 2 so is SUPER_SIMPLE Unless I remember wrong, third bit (2) means third mode (your loiter) is super-simple .
That can influence “which way is what” on the radio depending on GPS relative to your (takeoff) position.
I hope that helped somehow…[/quote]
Maybe, I’ll review the logs again when I get home later.
OK I had some time to review the logs a little more.
I did have both simple/supersimple on one of the flight modes, controlled by RC8, however logs show it was not engaged, RC8 never changed from 982. Also you mentioned RC2 (pitch) went to 2006 like you said, however that is full pitch back on my setup.
I also created a kmz file, and the last 2 flight segments were about 150’ south of the actual location.
I also ran an auto analysis and this stuck out at me
Test: Compass = FAIL - WARN: Large compass offset params (X:178.00, Y:237.00, Z:91.00)
WARN: Large compass offset in MAG data (X:178.00, Y:237.00, Z:91.00)
Large change in mag_field (83.13%)
Max mag field length (598.86) > recommended (550.00)
OK I probably need to calibrate the compass better, BUT the “Large Change in mag field” got me wondering. Any ideas???
One thing I didn’t think of earlier is that just before the crash it was flying about 3’ above the ground pretty much parallel and above either a 4KV or 13KV feeder that feeds the transformers in the condo complex. Maybe that messed up the compass???