Replies to This Discussion

3D RoboticsPermalink Reply by Alan Sanchez on February 21, 2012 at 5:41pm

Im having the same problems, for now I just edited from arduino but its annoying having to do the setup everytime. Those params are in config.h btw.

Theres a thread going on about this on the dev list.

• ▶ Reply

3D RoboticsPermalink Reply by Alan Sanchez on February 21, 2012 at 7:27pm

Thomas, out of curiosity what gains (cam_r_g / cam_p_g) do you use?

• ▶ Reply

Permalink Reply by Thomas Kramm on February 22, 2012 at 12:13am

On 2.3 I had CAM_P_G,2 and CAM_R_G,2.2 for my setup.

• ▶ Reply

Permalink Reply by Tomas Soedergren on February 21, 2012 at 9:24pm

Thomas, DZ probably stands for Dead Zone.

• ▶ Reply

Permalink Reply by Bill Jowitt on February 22, 2012 at 3:43am

I asked that question on another discussion, but got no reply. What does Dead Zone mean and how might we use it?

• ▶ Reply

Permalink Reply by Tomas Soedergren on February 22, 2012 at 5:37am

I assume Dead Zone is synonymous with dead band. Say you enter 30 in that DZ parameter,then there would be a dead zone +-30us around the neutral of that channel where nothing happens. Until command exceeds neutral +-30us, then the response would come. Mostly used to surpress inadvertant stick movement at neutral and to surpress noise/gitter.

• ▶ Reply

Permalink Reply by afernan on February 22, 2012 at 1:12am

2.4 flight test: LOITER

Yesterday I was flying 2.4 concetrated in LOITER and got very good results (with baseline code) after a lot of tunning (small 650g quad, APM1). In summary:

RATE-P has a MAJOR influence in loiter and it´s very critical. I´ve put it in Ch6 and there is a very narrow value zone where loiter is good:
            RATE-P = 0.09           bad, have local oscillations diverging
            RATE-P = 0.1             perfect LOITER
            RATE-P = 0.11          circles (small)
            RATE-P > 0.11          circles, increase the radius

LOITER_P aparently have no effect (or very small). I´ve used 1.5.
NAV_P = 2, but I guess we can up a bit in compromise with RATE-P. Rest of pid´s are the defaults

Finally I´ve got a very good "compromised" tunning (see attached file)

Hope this help to tune-up others

Angel

Attachments:

• 2.4 Quad Small V1.param, 3 KB

• ▶ Reply

Permalink Reply by Bill Jowitt on February 22, 2012 at 3:23am

Hi, what weather conditions were you testing in? When you write about loiter being "good", are you talking about precision, or smoothness, or both? Personally I am interested in a smooth and stable platform so would prefer to have a smooth 3m box rather than a twitchy 1m box. Any hints on going for smoothness?

Would love to test as well, but the wind is very strong here at the moment - should be 10mph tomorrow morning I hope. Regards, Bill

• ▶ Reply

Permalink Reply by afernan on February 22, 2012 at 4:07am

@Bill

weather was calm wind with some light breeze. The parameters tuned are for a compromise between smooth platform and small box (I had about 1-2m box). 

Please consider this is a small quad (650g including 2.2A batt) wich is very critical in loiter. Bigger quads should be easy tunning.

I can suggest you to start with deffaults and use Ch6 for tunning in flight RATE_P. You will see how sensitive LOITER is to it. So for a "stable" setting, low down as much as possible until oscillations occur. Then you need to check if STAB is still good for you with so low RATE_P.

Angel

• ▶ Reply

Permalink Reply by Bill Jowitt on February 22, 2012 at 4:51am

Thanks - if the weather does let up I will give it a go. Bill

• ▶ Reply

Permalink Reply by Dave C on February 22, 2012 at 5:43am

It's a shame we share the same weather Bill. I'll give all this ago too (on my very touchy small quad frame) as soon as things clear up. Thanks for the test results Angel :)

• ▶ Reply

Permalink Reply by Johann Van Niekerk on February 22, 2012 at 9:09am

Hi Angel

 

I just came back from fying field , i tested Marco's PIDs and yours and a combination of the two , with your PIDs my quad wants to flip over the whole time , but recovers and swings from side to side untill it hits the ground,  did you start tuning with standard PIDs?

 

thanx JOhann

• ▶ Reply

• ‹ Previous
• 1
• …
• 10
• 11
• 12
• 13
• 14
• …
• 90
• Next ›
• Page