Was flying my 800mm quad yesterday, flys well have not had any previous problems. I ran Auto-tune yesterday. It took a good while to run thru the roll portion of the test. Shortly after it began the pitch portion of the test the quad just abruptly stopped and fell from aprox 25 feet. Snapped 2 carbon fibre arms, damaged a motor and cracked the case and circuit board for my fpv cam.
I assumed it was the battery condition, new arms new motor and back in the air today.
This time I used 2 batteries in parallel for a total of 8000mah, a combination I know yields over 20 minutes of flight time with this quad. I made it thru the roll portion of the autotune in around 2 minutes, it then switched to the pitch portion of the tune and I had the exact same results. Cycled thru a few pitches forward and back and then just shut down and dead dropped, this time breaking a brand new fatshark transmitter a motor mount and 2 more carbon fibre arms.
Could someone look thru my log and tell me what exactly I am doing wrong? I put the quad back together this evening and have flown thru 2 batteries, aprox 20 minutes of very aggresive flying and cant seem to get anything to fault or brown out. I am not complaining or in anyway upset I know these things happen when beta testing , or operating flying machine in general, just want to figure out whats up. Thanks Richard
Replies
I am pretty sure it is caused by the bug that Randy and I mentioned.
I would make sure you download rc5 through the mission planner and go from there.
800mm quad! :-0
My V4 is around 1,050mm across the motor centers ;-)
Might this have anything to do with your motors and prop sizes? Certainly seems to be an issue with mine when doing normal PID tuning (seems to be very sensitive, and prefers a less aggressive tune...eg: RATE P and I = 0.090, RATE D = 0.008). Will fly at 1.05 without wobbles, but strange erratic behavior that wasn't vibe related was creeping in. Tuned it down, and flies like an oil tanker, in a solidly and reliably re-assuring way. I've not got as far as autotuning yet, and it sounds like it would get back to the 1.05 tune for me, which was not as good.
I am really sorry about your crash, I suspect that you crashed from the same bug I did.
Did you download your code directly from master on around the 19th. If so you may have got the 12ish hour window where there was a bug that happened as autotune finished.
When you go to master you arn't downloading rc4 any more you are grabbing a work in progress that can have a problem that the dev's missed because they haven't tested it enough yet. We do our best to make sure this doesn't happen but sometime it slips in.
Again, this bug was my fault and I am very sorry to see you crashed.
Richard, I had a look at your logs. I don't really see any obvious problems. The only thing I see is that the logs just end suddenly when you probably were still in the air. There's only a few things that could cause that. If the processor hung, that would happen, but there's no evidence that this is a real possibility. ie: it's not happening to other people. The other cause could be a power failure to the APM.
I wonder if the wiring is marginal, and the jerking of the auto-tune is causing a disconnect.
I noticed that your Vcc is not very stable, and is a little low, with obvious dips down to 4.6V.
How are you powering the APM?
I am a little confused, dev team ask for feedback on autotune, here is a log from a crash during autotune , anybody?
which version of the code were you using? the official release or did you get it directly from github? Also when did you download the code from github? There was a bug in master's autotune code a few days ago but it was bad enough that I don't think you would have gotten as far through the tune.