PDA

View Full Version : 30cm AeroQuad, 1st testflights



ziisco
10-22-2010, 06:34 PM
Hi,

We just finished our 30cm prototype quad (prototype because cables, ESCs and receiver still look really ugly :D).

specs:
- 32cm Motor-Motor
- 4 x 8x4.7 GAUI props
- 4 x KDA20-22L
- 4 x 20A ESCs
- Arduino Due / Shield v1.8
Weight including 4000mah 20C lipo: about 1kg

stable mode not yet working but we'll get there soon I hope :)

http://www.youtube.com/watch?v=bQ031XFMrkI

Hamburg
10-22-2010, 06:41 PM
Who says "small quad are not stable" is wrong! :D Maybe we can help each other getting stable mode properly working. Still setting pid up :(

ziisco
10-22-2010, 06:53 PM
yes they can ;) I had my doubts while we were building it, but it's really good in acro now. PIDs are something like : 1.5 / -8 roll and 1.6 / -9 pitch. pitch needs a little more because lipo is oriented front to back, balanced.
it has way to big motors for a quad in this size. motor weight is about 280g including props. frame about 150g and lipo 200-340. if we would mount smaller motors we can use smaller lipos. but i like this power... we'll do some test flights on a big field tomorrow (if weather allows it). I guess it'll be quite fast in forward flight :D

UndCon
10-22-2010, 08:46 PM
Cool little machine :)

Hamburg
10-23-2010, 08:42 AM
yes they can ;) I had my doubts while we were building it, but it's really good in acro now. PIDs are something like : 1.5 / -8 roll and 1.6 / -9 pitch. pitch needs a little more because lipo is oriented front to back, balanced.
it has way to big motors for a quad in this size. motor weight is about 280g including props. frame about 150g and lipo 200-340. if we would mount smaller motors we can use smaller lipos. but i like this power... we'll do some test flights on a big field tomorrow (if weather allows it). I guess it'll be quite fast in forward flight :D

Nearly like mine! I fly 1.13 and -9

Honk
10-24-2010, 02:58 AM
Seems very easy to handle in the air! Good.

Have you experimented much with negative D values in acro mode? I still can't really figure out how this parameter is optimized and used fully in dynamics with the P parameter...

Hamburg
10-24-2010, 09:13 AM
Speaking for my setup, which is very compareable to his. P is very strong on small quads - setting P to 1.00 and D 0.00 itīs nearly uncontrolable and hard oscilating. My goal was to gain P as much as I could but let D as "high" as it doesnīt "pressed plained" the reaction curve. So it, became that values. I f I raise P to e.g 2.00, D has to be -20.00 - but then the quad is going "wobbly". Thatīs the aco part which works great - but I have the feeling that stable mode reacts too slow for such small quads. Working on PID, itīs oscilating hard (i term) either or hardly floating in all directions (i term too low) - Couldnīt find the golden middle, yet.

Hamburg
10-24-2010, 12:18 PM
Found it! A good point to start tune stable is D:2.50 I:1.50 with P and D settings from Acro mode! Have to finetune a bit but these are the first good results

Honk
10-24-2010, 01:10 PM
Hmm... Yes I've understood that negative D should be more negative when raising P, but what happens with too much negative D? I understand it numerically (it is the delta between set-value and is-value) but not in flight characteristics.

Also I is tricky (cumulative adding of the error each new loop), putting it high makes for a big number fast, but also goes to 0 faster. Putting it low makes the values go up very slowly if an angle error is present but since no big numbers are reached, it can come back to zero as quick.

But I really can't decide if I need much more or much less I in my stable mode. It feels like it takes a minute of hands-off to get it to actually stop itself and level out. Before it wanna level out it seems it want to go very slowly from angle to angle you know...

Please tell me what you think is happening here, I think its pretty clear at times that it wanna like go around in a circle (which I'm most of the time am trying to manually compensate for), then settle for 0 angle after a long time:

http://www.youtube.com/watch?v=zYMVnxtrfHA

Hamburg
10-24-2010, 02:32 PM
Honk - P value seems it gains the amplitidude - the -D value flatens the gain curve on same amplitude level - That itīs how it "feels" - like a normalizer in an audio unit. It pulles silent noises up to the loudest noise so you get a leveled noise

Honk
10-24-2010, 02:41 PM
I know what you mean if we're talking audio.. But please translate into quad characteristics! :) Can you suggest anything I should change to the current characteristics you saw?

kevinlei324
01-31-2011, 04:56 AM
@ziisco
I actually just built a quad that's about the same size as yours, but a lot heavier (stupid pvc). I'm still waiting for the props before I can tune it. Have you flown in stable mode yet? Can you post the PID values you used? I would really appreciate it.
Thanks!

ziisco
02-20-2011, 01:17 PM
Hi,

No, we didn't get stable mode workingon the 30cm quad yet. I guess the weight/size radio is making stable flight impossible. It's way easier to get 60cm quad stable than a small one. It's not a good idea to use same motors, lipo, electronics, etc. on a small quad, size of the arms does not make it really that lighter. But that's the mistake we made, maybe you found very light motors and electronics. Good thing is the quad is very fast with the big motors though... :)