This is a discussion re the ArduCopter-3.1 AutoTune feature. A place for people to report successes and failures, ask questions and get answers.
The AutoTune wiki page is here.
You need to be a member of diydrones to add comments!
This reply was deleted.
Replies
Bernie,
It looks like a brownout although it could be anything that suddenly causes the board to completely stop functioning. Basically all we can see is that the board suddenly stops logging. There doesn't appear to be anything unusual in the board voltage, battery voltage or current although we normally don't see anything in these messages because brownouts happen very quickly.
Randy,
I do agree that the data stopped suddenly. The hex spun about the left/right axis and when it hit the ground, the battery disconnected. When you suggest a brownout, are you referring to the receiver or the Pixhawk? I believe that I clearly heard a couple of the motors go to full speed. Probably the front 2. I had thought that up to the point that the battery disconnected, that the esc's would have voltage but the pulse width would probably go to not zero but shut off in case of a brownout. Wouldn't that prevent any esc's from going full tilt? I thought that all of the motors would come to a stop.
Barney Liftoff
Bernie
Bernie,
I think the Pixhawk browned out. During a brownout I wouldn't expect the ESCs to go to full throttle but it's possible they would freeze at their last known value. Ideally the motors should be shut-down if the board has enough power to do that. The motor levels change so quickly on a multicopter that it's possible that some motors were working hard to correct a momentary rotation, the brownout happened, and then they were left in that state which made it continue with that correction.
Let me check around a bit with others in the team on how the Pixhawk handles brownouts.
Randy,
My Pixhawk is powered by number 1 esc which happens to be a Castle Talon 35. The power gets to the pixhawk by the output channel 1. I read that I can power the pixhawk through the power port. Would that be the preferred method?
Bernie
Bernie,
I'm not an expert on the Pixhawk hardware but if it's the same power-supply I wouldn't expect it to matter whether it's fed through the power port or where the ESCs connect at the back of the Pixhawk.
I think the more reliable method is to use a power module (assuming you're using a 3S or 4S battery and current is under 90amps) and an ESC/BEC plugged into the back.
I've had this condition happen twice recently, the logs stopped and the vehicle just plummeted, but it was with Plane and we're pretty sure related to the condition that is fixed by the addition of the Zener diode. Not sure if that applies to copter?
Bernie, any chance you have some digital servos plugged in to the servo rail, or only the ESCs?
Hi all,
I am experimenting with different vibration damping solutions for long endurance hexas (low kv, 15-17''). The problem is that although I have relatively low vibrations the copters are sometimes (if vibrations of AccX and AccY are greater +-1 and +-1.5 for AccZ (1/10G)) twitchy in loiter. Hence, I have to further soften the damping.
Since auto tune works with these copters, I am wondering if auto tune can be used to detect if the damping is too soft?
Thanks a lot in advance!
Thorsten
PS: Some related discussion about this problem can be found at drones-discuss.
Hi all,
Someone did some work on autotune to let the user choose what tune they wanted to do using a bitmask and also implemented yaw and some speed improvements.
Unfortunately I can't find the emails where I discussed this.
If you know who this person is, where this work is, or if this person is you. Please get back in contact with me.
Thanks,
Leonard
Legend mate, THANKS!!!!!!
Hi all,
Just to give you a little update. I have been working on the problems auto-tune seems to have with very small copters, very powerful copters, flexible copters and copters with a big flexible gimbal hanging off them.
To this end I am building a QAV 250 and a Tarot 960. I have a number of theories, filtering settings may be preventing auto-tune from working effectively, large D terms may be causing overshoot on their own and we may be able to structure the tests and adjustments differently to handle a wider range of copters.
Thanks for everybody that have taken the time to understand auto-tune and help others on this forum. I have been reading your posts and suggestions for inspiration.
Thanks!
-
23
-
24
-
25
-
26
-
27
of 39 Next