Hello, after reading a lot about this project, and in the need of choosing a new FC system, I decided to go Open Source. My choice is APM. 


After assembling my mk hexa xl frame with all the rest from Martin (buildyourdrone.co.uk) (jdrones 30 amps esc, 2836 850 motors). 


I managed to put it to fly. But it's far from stable and after moving a little the stick it oscilates.


The problem I think is coming from the PID tuning. For now my pid are:

Which values should I correct to have decent stabilization?


Thank you and sorry for the newbe question.


Views: 1630


Reply to This

Replies to This Discussion

Does it do big/slow oscillations, or small/fast oscillations?


Watch this video then read this. They use different terms. Attitude (video) == stabilize (manual), Rate == rate.

If it is large/big/slow oscillations then try reducing stabilize p for both pitch and roll (keep them the same) 

If it is fast/small oscillations, then try reducing the rate p for both pitch and roll.

You do not have to reduce by big numbers... just because it shows two significant digits... i.e., you can go to 3.45, it is not just 3.4 or 3.3!


Look at the video, see how he tunes his pid by focusing on one axis only? and mounting between chairs? Good idea! You can see easier what is happening, get better tuning... 


Please come back and post on what you did and what you found so others can also learn, if this works. And if it does not, then maybe you can make some videos?

Thank you for the answer, I will try the chair mode tomorrow trying to find the issue. 


I will try to do a video and of course once I found the settings I will post them here.




I like to tun with CH6 using my proportional control on my radio. I start with Rate_Roll_P/Rate_Pitch_P, which for you seems too high.



So I can assign ch6 to Rate Roll P / Pitch? 

I didn't know that. Good to know, I will try if I find how.




Hello, well I have managed to make a video using the same method as the video Mike point me. 


As you can see it doesn't stabilized to much. 

I'll have been playing with PID and arrived at something not to bad I think. So here are my values:

If I put less rate_roll/pitch P I don't have almost any movement from when moving the sticks.


I will make a fly test, outside and try to film.


Jason, how does it work the tune feature? I test it but I can't find any info.



I hope the video is only an example, and you also did more. In the video, it is too much stable w/o throttle, and too little throttle to see what behavior you have, I think. You will want more throttle, let it lift up. Also, maybe add more distance between the tied off points, maybe or some other way to allow the copter to spin easier so the ropes do not "control" the copter as much.


Also, maybe remove PIDs for the axis you are not testing. Try to create the problems he has in the video, because it is a balance, and when you have wind or what to fly with speed, you do not want totally stable, so you will have a better tuning if you first make it very wrong, and repeat the "problems" in the video.

Thanks for the advise. 


I went ouside and video it, but it's crappy video with very little images of the hexa (I'm retrying with the gopro).


It handles a little bettter than yesterday, on take off is quite stable, but then when I give him an input, he moves and then starts oscilating and circling.


You think maybe I'm too high in P parameters? Should I go lower? I'm not touching I for now. 


Another problem is my yaw, even though it's stable (maybe too stable) i can't make it turn, it's very very stable in I values. Should I lower P? 


Sorry for the questions but it's tough to find some PID settings for start. (I know some other use similar configuration, but little share the pid, or I don't know how to search).




I'm only guessing, because I am making assumptions.

I think maybe a tiny increase in stabilize I, and maybe even increasing stabilize P also. I think you want to increase stabilize P until, in a hover, it started to oscilate, then back it off a bit. 

And if it is oscillating when you move it, I think (?) your stabilize I is a tiny amount too low?

I might have that backwards, but in any case, make tiny movements in I, thousandths of values.

And if Jason recommends something different, listen to him first ;)

Ok, here a little idea of what it does with the gopro.


Jason told me to do the tune with ch6 (but I don't know how to make that work).


I will change some params and see what happens.




And I think you might want to try increasing rate_p until you get an oscillation, then back it off. I was thinking about this after watching your new video, and I think i would do that before messing with any I values. But PID tuning is still magic to me, so be sure to write down your values and any notes so you can correct my mistakes :)

thanks for that, I will try to compile but I can't access the Git, maybe windows problem. 

I'm getting better with stability, but it always try to go right (maybe I'm not really level but just did it). Anyway a new video

with these settings: 


Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service