PDA

View Full Version : 2.4, CHR6DM and Big Quad



Honk
04-12-2011, 05:35 PM
Well, I finally got some courage and fired up the fat murdering machine once again. This is with CHR6DM and APM boards only, mainly flying Rate mode, but at 4:06 I switch to Attitude mode, and it seems to level very good for hands free flight for about 5-6 seconds, then the oscillations entirely take over. Why?

PID's are:
Rate P: 65
Rate D: -50

Attitude P: 3.5
Attitude I: 0.60

...?


http://www.youtube.com/watch?v=f1VmiFMurwE&feature=youtube_gdata

aadamson
04-12-2011, 06:32 PM
I think you need to go back to square 1...

first it's has the oscillation in rate mode, but you know that... so turn off the D and dial in the P... is this using Honk modified code? Any chance you still have some of that funky delta time stuff that you were using and I removed all of?

Specifically make sure you haven't changed anything in Pid.h.

The oscillation starts out slowly and then continues to gain in magnitude... Can only be PID's at this point, so you'll just have to play

Honk
04-12-2011, 07:42 PM
Nah, it's the 2.4 vanilla. Only difference is I rescaled the CHR6DM to radians instead of degrees and fixed the int/float bug Ala found, so mainly that unit slid in quite nice with the changes you and John has made recently.

With these PID's:
Attitude P = 2.80
Attitude I = 0.60
Rate P = 50.00
Rate D = -60.00

I get this (attitude mode starts at 3:00, getting a little hands-off at 6:30 and forth):

http://www.youtube.com/watch?v=d9hBBpdvIg4&feature=youtube_gdata
Next thing is as you say to remove negative D totally and get I down a bit... But now it's definitely able to handle itself again.

ziojos
04-12-2011, 09:23 PM
Nice Honk The Acro seem very smoot, I think stable need a bit more tune,
BTW your very nice snow video seem better in difficolt environment (Acro or Stable ?)

http://aeroquad.com/showthread.php?1482-Who-says-you-can-t-fly-in-snow-!&p=14517&viewfull=1#post14517

Honk
04-12-2011, 10:38 PM
Yeah, I feel that the I term is making it drift quite a bit sideways before compensating when I let the sticks go... So of course I'll have to tune it more. This was the second test flight/tuning with 2.4 and CHR6DM.


BTW your snow video seem better in difficolt environment (Acro or Stable ?)
Don't quite remember actually... I think it was with 2.1 code and Oilpan, but I think I modified quite a few things in that code to suit me somehow... But now very many things have changed. I mean I could surely get that back by using the old code and stuff but I wanna get forward.

ziojos
04-13-2011, 04:54 AM
I mean I could surely get that back by using the old code and stuff but I wanna get forward.

Absolutely agree, I know Your style and Your perseverance :) ...... me too !

lokling
04-15-2011, 06:25 PM
Hmm, Honk - recon I should go for the CHR6DM or the Wii sensors when first trying out the 2.4 codebase?

Honk
04-16-2011, 12:49 PM
Lokling... you've missed so many flame discussions! :)

Anyways, I had to fight a lot with the CHR6DM but you know what? Ala42 of course found a real bug we had with it! We used gyro/accel data as partly int's all along! We just didn't notice it so much when using degrees, but now we're using radians and radians/second, and then I actually felt the truncation when hand testing the quad! :)

So I changed scalings to rad and rad/s, added "zero_rate_gyros" function as a serial command and it flew! Code is a zip (yeah, have not yet despite all help gotten really gitting) in the big 2.4 release thread.

Don't actually know status of our Wii code... But CHR works at least! The code change for CHR6DM I did isn't yet in the official code I think but Kenny should do that he said...


Strange thing is that you saw the video before this (post #3) where I got handsfree time.... Then I changed PID's to:
Attitude P = 2.80
Attitude I = 0.25
Rate P = 50.00
Rate D = -20.00

And got a much worse Attitude Mode result...? Why?

http://www.youtube.com/watch?v=n3wjW-JehlE

Honk
04-16-2011, 03:27 PM
Well, it apparently handles gusts of wind quite bad... I used about the same PID's as I had when I first got it to handsfree, but... Nope...

Attitude P = 2.80
Attitude I = 0.55
Rate P = 50.00
Rate D = -60.00


http://www.youtube.com/watch?v=w1lpf7P7Xy4&feature=youtube_gdata

Honk
04-16-2011, 06:59 PM
http://www.youtube.com/watch?v=3Utoszm0Ynk&feature=youtube_gdata

I wish the video resolution was better:
http://www0.mediafire.com/imgbnc.php/0c23d816bb2acd81302203b57884071b4f36de3c974a0254cd 49d23cd6891c586g.jpg (http://www.mediafire.com/imageview.php?quickkey=9yawzrjvczjwea1&thumb=4)

Tomorrow I'll have the cam on the quad instead and take some summer pic/vids.

krat
04-16-2011, 07:00 PM
what did you change from the previous flight?

lokling
04-16-2011, 07:11 PM
Nice! Looks like I picked a good time to come back for some flying :D

Honk
04-16-2011, 07:20 PM
what did you change from the previous flight?
I seriously don't know :) Mixture of wind, stiffer sensor mounting and some PID's... Really strange thing but before this nothing worked on any of my two differently sized quads. It's still quite sensitive to wind gusts. I guess it comes from the high negative D on the gyro signals.

Yes, Lokling! You did. Now it's really much more streamlined code and well, nicer allover. So get flying!

ziojos
04-16-2011, 08:39 PM
Per aspera ad Astra

Nice Honk, I'm happy with You :)

Ciao

Honk
04-16-2011, 10:01 PM
Through blood, sweat and cuts I would say :P

Thanks, I'm happy with you too! :)