Prior to it's last flight, my hexa, which I am thinking of naming after some sort of bug like the queen bee, was finally flying. I had taken it out three times to see how it did. Used a couple simple modes which seemed to work okay (RTL, Loiter). As a novice and not knowing really what to expect from the bird, I was totally stoked that it flew, I had control and it did essentially what I was asking it to do.
The first flight was prior to the Super Bowl here in the US and the second was in the dark after the Super Bowl. I have lights on the arms and it worked out very well. The next day and being inherently impatient, I decided to program all of the modes (6) I could, upload waypoints and alltitudes and see if this would perform. It was a bit windier that night and I got the bird off the ground to trim out the controls at about 4-5 ft alt. There seems to be a little issue with trim as it like to move forward and to the right, but when I bring it back to some center point to continue trimming, it starts moving the opposite way. It isn't out of control, just seems that I can't truely trim it to come to a standstill in the air. The PIDs are all default and I figure there will be some tweeking required after I get a few hours of understanding it better.
I decided to go into auto and see if it would follow the way points. I wrote down the switch settings when I calibrated the mix on AUX1 and Aux2 to get all six modes with reasonable room from the end. I was looking at the stability of the PWM on planner and adjusting the mix on the radio. Stability seemed to be about +/- 5 or so. I gave myself what I could muster up for tolerance and kept the set points >50 away from the edge of any particular mode.
I went to auto and the bird flew up to an altitude and started drifting north. It was programmed and I was expecting it to go east. It drifted further and I figured something is wrong. I went back to Stabilize and brought the bird back. I was about 50 ft up and it pitched front up, which surprised me (I think I was in Loiter). Then it seemed I had little control when I whent back to stabilize. I pushed the throttle up so that it would fall to the ground before going into stabilize, but the bird wasn't really doing what I thought it should. It started to fall. I gave more throttle and in about 3 seconds it cam from fifty feet to right between us, motors whailing and smashed into the ground bouncing about 10 ft from initial impact.
I bent up all of the arms, motors mounts shattered, one blade survived, electronics intact, motors hanging from wires and the show was over.
I looked at the logs when I got home and they do show the flight path as I overlayed it in Google Earth. But the modes I was in were very different than I had thought they would be when I mapped it during calibration. I thought it was in auto and it was really in Alt_Hold. I don't know how to read the logs in depth but I looked at the throttle as it was coming down and it looks like I never gave it throttle. Again, don't know if I interpreted the log right, but if someone could look and tell me what actually happened on that dreadful night, it would be appreciated.
Below are the logs and some before and after pics; (buying a new frame now :( )
The Below is Google Earth Log file showing it's last moments;
but the landing gear definitely proved itself!
seriously though, the flight modes were obviously way off what you expected - i wonder if you noticed that as you set up the modes with the transmitter on it highlights the option youre currently setting? my guess is that you set up modes from 1 to whatever and they dont actually correspond with the switch positions. i only have a 3 way switch on channel 5 and have it configured with stabilize, loiter, auto - but they arent positions 1,2,3 - i think theyre at positions 1,4,6.
i have looked at the log but it seems the values in MP still arent matching up with the log created by arducopter, all i can say is that for the dive into the ground it seems you had got back into stabilize, so i dont know why you couldnt control it :(
very sorry to see the state of your hexa, ive been lucky so far, just three fairly light crashes - ill post here when that changes ;)
I know, didn't that landing gear hold up nicely! :) And when it hit, it basically was roaring past me at about a -30 degree angle which would have had the landing gear hitting first. I have a video, but I have been too busy to do everything with it. Plus it's night and it was tough for my son to keep it in focus.
i did check all of the highlight functions going through the cal and I worte down every setting for the switches. i haven't checked since the crash. I'll do that and see what transpires.
well that has changed, have just suffered what could be a horrific crash - i dont know how bad it is yet because i couldnt get to the hexa to recover it.
i have been struggling with pid tuning and my first flight this afternoon just wasnt stable, i used rtl to get it back but wanted to switch to stabilize for the last metre or so of descent - it didnt switch straight out of rtl and tipped over on a bush.
the next flight i had the tune much better and actually flew about a bit - until a prop came off (i assume it was loosened in the tipped over landing). i guess that it fell about 80m and is at this moment on the mountainside a bit out of reach, we managed to get close enough to see the nav lights and photograph the site but were going to have to go back tomorrow morning and try to get to it from below rather than from above.
will post the details when i have them (bug**r its raining a bit now)
I bet the motors are still working :) test on and keep up your spirit. I too lost mine in auto mission..drifted away and was in a tree for 3 months before i found it.
Wow. 3 months! Yes, I haven't had a chance to look into it, but the motor shafts aren't bent and the dome did a great job of protection of the electronics. The bird ended up bouncing about 10 feet landing on it's top, so I expect this is just the frame. Which is not that expensive compared to everything else. So, lessons learned
1. Take your time
2. no great rush into all of the cool things
3. learn to fly
4. get comfortable
5. try one thing at a time.
6. Gain confidence before the next step.
What type of props are you using they seam very strong too.
I don't want to hurt the sales of the OEM multi rotors frames here.
However I knew that a controller under development may have errors.
They do at times. I built a custom frame that takes hits with less problems.
I have to say I use a Rusty's frame build ;-)