Loiter feature added to position hold mode

[peabody124]

So I recently added a loiter feature to position hold mode - when you can use the roll and pitch sticks to move your setpoint during the hold.  It works fairly well and you can see some video and details from testing on my blog if you are interested:

http://buildandcrash.blogspot.com/2014/01/loiter-mode-and-more-rth-testing.html

If anyone else would like to test it and has something set up to try, drop me a line.

[Carsten Böhme]

Is the landing rate static for the whole descent or is there a way to start with a higher speed and reduce it gradually? Otherwise landing might take too long in an emergency.

[peabody124]

You can adjust it in VtolPathFollowerSettings.LandingRate

[Carsten Böhme]

I know, but I was thinking that a landing rate of say 2 m/s which may be well suited for going from 50 to 5 m altitude might be the very wrong choice for going from 5 to 0 m... I'll test 1 m/s next.

[Carsten Böhme]

Tried PH on a new copter today and caught two things on log that might be interesting.

1) PH would work for about 30 s very well, then the UAV would start spiralling out from the holdpoint quickly, see attachment between flight time 800 and 840. This happened every time, without obvious external influence.

2) sometimes PH would behave like (1), sometimes it would start spiralling right away when engaged. Holding the position manually before PH is engaged may help, but not in every case. I can look up the flight times if necessary.

[peabody124]

I'll try and analyze the log better later, but my first question is how far away is your mag from your wires?  This sounds like toiletbowling when the heading estimate gets rotated.

[Carsten Böhme]

That is a possiblity, the Quanton sits about 3 cm above some motor wires. I'll see if I can try an external mag for comparison. I believe mounting the whole flight controller significantly above the center of gravity is not a good idea?

I was puzzled because PH worked so well for a while, so I guessed some parameter might become saturated over time.

Btw, do you recommend tuning the pathfollower to the individual airframe, or is it even necessary to do so? I did not see any change from small modifications.

[pug398]

Updated all my quads (CC, CC3D, Sparky, Quanton) to newest release and must say they are all flying beautifully. Decided to give autonomous a try again after experiencing the same thing Carsten did and after reading quite a bit decided my mag was located to close to power distro board (~25cm). Got a LEA6 with active ant and ext mag coming as well as box and stick to put it on so hopefully it will hold position better this go around. They fly so well in atti and horizon mode I really don't care but still like to keep up :)

[José Eduardo S. C. Xavier]

25cm is too close?

--
You received this message because you are subscribed to the Google Groups "phoenixpilot" group.
To unsubscribe from this group and stop receiving emails from it, send an email to phoenixpilot...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[peabody124]

There is no absolute distance. It's largely going to be a function of the amount of current your motors take and the shape of the wires. Keep the positive and negative close to possible, or twisted if you can. Higher current motors will create a stronger field that weaker ones.

A simple test would be to watch on the PFD and run the motors up to a near hover, see if your mag starts drifting.

[pug398]

Oops sorry metro_Ro that should have read 25mm. Just a little bit of difference there :D

 

Should be ~15cm setoff with the post so hope that is enough. If anyone has had experience otherwise please speak up. In everything I have read since trying it with the first GPS module it sounds like definite mag interference. Bought a second Neo6 GPS and had drift, then a third with ext mag and big active ant. and that one was DOA. Ticked me off and have been flying atti or horizon ever since. They are all flying very well so time to give autonomous a try again.