Need some help on Tricopter (yaw)

[vpb]

Hi everyone!
This is my very first multicopter, as I only fly fixed-wing. I've spent so much time on reading and building this tricopter, almost time is balancing propeller & reducing overall vibration. The 2nd flight test is fine this evening, windy field, start with acro mode for trimming & pid tuning. The PID seems to be ok. I can sometime be hand-free.

My current PID is 8.0 0.025 30 for pitch & roll, 8.5 0.045 0 (default setting), 2.1 version. Every time I try to rudder-turn (yaw), my tri looses height like I throttle down. I dont know whether it's tri behavior (as I must increase throttle each time I turn), or I must edit the config somewhere. Please help

http://www.youtube.com/watch?v=PrqajwU0FRU

On board camera test (keychain camera, but it's fake one, not good frame speed)
http://www.youtube.com/watch?v=hfeZcT5Qq7I

[wilco1967]

Congrats on your first flight.
Seems allright....

You could try increasing roll and pitch 'I' setting. My Tri seems about the same size as yours, and it is a lot more solid on I around 0.05 to 0.07 (on both)....
With very low I, it seems to sag a lot.

[vpb]

Thank you, wilco1967! I will try next test day. Can you explain why I need to increase roll&pitch I?
My Tri uses alu arm, 50cm length, weight is about 1000kg including 3S2200 25C bat. DT750 kv motor with emax 10x4.7 prop.

[wilco1967]

The I on a PID works as an automatic trim..... it will keep increasing / decreasing the output until the desired effect (zero roll/pitch/yaw) is reached.

I had a lot of trouble with my tri sagging the tail when rapidly throttling up, or when coming to a stop from fast forward flight. If it drops a lot when yawing, it is probably the same effect. Could not really tell from your video (I don't know when / how you move the sticks from your video).
Increasing the I solved it for me....
Not saying this will 100% solve your issue, but hey, nothing to loose, so you might as well give it a try...

>1000kg

Hmmmmm..... that is one heavy tri.... about the same weight as a small family car..... Or did you mean 1000 grams

[vpb]

Hehe, I mean 1000g, sorry for the typo. Can't wait to try your solution, maybe next morning . Thanks again, I always think I as the sampling average speed, but automatic trim, it sounds more useful to tune.

[vpb]

Hi wilco, I've just test increasing I term recently, it really solved about 80% the problem. It still sagged a bit but I stopped at 0.064 on both yaw, roll & pitch, it's dark and I couldnot fly anymore . Now my tri turns slowly. Maybe I'll do hard test & tuning tomorrow.

[wilco1967]

Keep increasing 'I'.... I think I went up to 0.1 before (no, did not forget a zero).....
It is very difficult to get any kind of vibrations from 'I' alone

You might find it will become slow in level mode though..... that's what was stopping me from turning it up even further.

You don't have to be too gentle on adjusting PID values.... anything under a factor of 1.5x previous value is hardly noticable. Get it roughly in the ballpark value using large steps, and then tweek it to your liking.... and remember, there is no single 'correct' value.... As long as you're happy with how it flies....

Does your tri turn slowly without stick input, or do you mean it will only turn slowly ?
For the latter, increase pitch/rol rate or yaw rate in GUI. (default = 0... I use about 0.5, and it becomes a lot faster)
If it yaws without stick input, check your stick centre is at 1500 +/- 5 (much more, and it becomes noticable).
If you moved you tri while powering up, the gyro calibration could be off.... just calibrate again using the sticks before you take off (throttle down, yaw left + pitch backwards/down/towards you).

[vpb]

Hi Wilco, I mean the tri turns slow with stick input. Without it, tri is almost tail-locked. 0.1 I is a big high value. Btw, do you think that higher I will decrease the effect of P?

For PID tuning, I do the big step, then I stopped at 8.5 for P, when it flies, it oscillates a bit, so I decreased a step of 0.2, and I stopped at 8 for pitch & roll P. You're right, do a big step first.

For gyro calibration, I never move my tri when powering it up, I always put in on a flat ground and re-calibrate gyro. I mount a small button spirit level to check that the tri frame is on a flat ground. But I'm still wondering that, if we dont have a flat surface, like a rock ground, grass or even on our hand, will it do well-calibrated? The tri need static position when it calibrating gyro, or a flat surface?

And Wilco, what is your idea about D term? Now my tri's little slow to response stick input. I saw the funny thing that when I ordered it move forward, it slowly pushed the tail up and slowly went ahead, and the speed was fast later. I found that it's difficult to stop the tri, I must "flatten" the kit, and do a opposite direction stick to cancel the current moving momentum. Maybe it's the multicopter behavior, and in level mode, it will be better.

Do you think we can brake the tri with only PID tuning? Or it must have the gps module. I make this tri mainly for FPV, so I need stable and gentle movement

[carter]

vpb

a basic question, what is your tail servo min/max throw values? in my tri, I have limited them a bit so that at extreme stick movements, the prop don't hit the ground. doing that, it would also 'sort of' lessen the YAW, but copter should not drop in altitude when YAWing.

[vpb]

Hi Carter, min 1250, max 1870, middle 1560, I never using the max input even with that settings.

[vpb]

Today I re-celebrate 3 ESCs (tur plush 30) with 2.1 code (ESC_CALIB_CANNOT_FLY in sketch), reduce MINCOMMAND from 1000 downto 900, increase max-throttle to 2000. The Tri now have more power, it seems to be lighter, I just can use about 40% of my throttle for taking off instead of 60% before. I reset "I" term to 0.035, it flies great, just sags a bit when yawing.

Then I kept increasing P term to 15, tried to rock it by roll stick but no problem (as my RC rate is 50%), just a little wobble with the wind. So I decreased down a couple of times, until 13.3, no wobble at all. I think if I keep increasing P further, it'll still fly ok. So I dont know that it's good P tuning? Keep increasing P until it really has problem, or just a sign of wobble then back off?

For I, when I increased to 0.07, 0.1, 0.121 I couldn't really notice any difference. So I should increase I or decrease I? And what is the sign for me to stop changing? Any different sign than the wiki or common direction .

Thanks!

[vpb]

Esc(s) are changed to HK F-30A reflashed with simonk fw, what a difference! After PID tuning, my tri is super stable now, even it's just in acro mode

Some boring clips
New esc(s) test http://www.youtube.com/watch?v=HmLxQGptJhY
Hover http://www.youtube.com/watch?v=YwYAJhonDIg
High hover http://www.youtube.com/watch?v=TsfgON2fZAw
Stable mode (trimming & test) http://www.youtube.com/watch?v=W0bawafsm-g

Love multiwii!

[kjetil]

HELLO
i am new to this forum and i need some help from you guys

i have a tricopter and a afrofligthcontroller with tricopter firmware on it
my radio is a graupner mx 20 with a gr16 hott reciver

my problem is1:to get the radio setup rigth,so it wil work with the ppm on the afro!
2: what program on the pc do i have to download to get this thing to work?
3: i have some castle esc on my tricopter,and i wounder if i anly shoud have the red cabel from one esc going too the afro,or can i have all 3 red cabel from eatch esc into the afro

i hope someone can explane to me how to connet this thing so i can go out flying

thanks for answering
kjetil

Posts: 2
Joined: Wed Sep 05, 2012 7:37 pm

[vpb]

Never use AfroFlight so I cant help, why you dont ask this on AfroFlight forum?
For the BEC on ESC, you should remove all red wire and use an UBEC to power the board, or you just keep only one red wire from 1 ESC.