Servers by jDrones

Dual-motor tailsitters


(Top Gun17) #1124

Turns out that i have a friend who can borrow that.

But what info or where i need to look exactly, how this can help me. Sorry, I just heard 3DR 20 minutes ago :slight_smile:

Thanks guys:)


(letpi) #1125

your log are impossible to analyse for me as they are quite long, about halt an hour for the first one, with continuous switching from manual to q_hover. I hope you dont try to fly manual as this is impossible even you are the best pilot of the world.
I have found that you have disabled arming check. So the pixhawk allow to arm and take off even in the worst case. You should not do that. You definitively need the telemetry link with the 3DR radio, it will give you real time health status of EKF, gyros, compasses…


(Kikislater) #1126

Telemetry radio inform you in real time what is wrong or what is good. You could monitor sensors in flight with GCS to show if something is wrong or not instead of flying then watching logs.
For example, like losawing say, you have disable arming check, if you had telemetry you will know why you could not arm but this option need to be enable.

I’m not a developer, just watching this thread by curiosity, but there is some basics for ardupilot fw. May be it’s not a good idea to start with tailsitters and hope you will fly it in the proper way one day.


(Top Gun17) #1127

I push the manual if my plane goes crazy, then i can control my plane again, when plane is wired up and can’t fell on the ground. That’s why i only use manual mode

Regards


(Mark Whitehorn) #1128

@TopGun17 It looks like your pitch P gain is too high in log 101, since you’re getting large oscillations.


(Bzucak) #1129

Hello I have a question regarding fixed wing logic. I am trying to understand how quadplane.cpp works to be able made some changes. There is a few comments referencing “fw controller” but i was not able to find where is this control for fixed wing modes like FBWA implemented. Can anybody give me some clue? When I looked into quadplane.cpp I can not see any logic for FBWA. Especially I am looking for a place where the servos output for motors is being set. I tough that it is within AP_MotorsTailsitter.cpp but it seems that AP_MotorsTailsitter::output_to_motors() is only for VTOL modes. I will be glad for any clue.


(Mark Whitehorn) #1130

Look at library APM_Control. There are separate controllers for roll, pitch and yaw.


(Top Gun17) #1131

Quess what, my plane can hoover and fly :smiley: At the end i broke my plastic servo but ti worked very well :slight_smile:
Nice, thanks all for your help (Y) Adding video later.

I have question, because it broke down a little bit and i need to have new foamboard, where i can find aileron mixers?
Because now i would like to build almost same, but with rudder and elevator? Where i have to make changes to get those also work?


(letpi) #1132

congratulations. Just to know, what improvement did you implement ?
I don’t understand your question. But when I checked your parameters I saw you have set servo1-function and servo2_function to 77 and 78 which is correct to get elevator and aileron mixing.


(Top Gun17) #1133

Actually my P was too high and i guess the main reason why it didnt hover like i wanted was wrong CG :smiley:

But my guestion is what i do when i add also rudder? What output i have to connect my servo? Where i can find a mixer that my rudder also starts to work like i want :slight_smile:

I try to build tailsitter fuction to a regular plane, at first also two engine to both wing. But i dont know how to put to work elevon and rudder that way that hover also works.

Regards


(letpi) #1134

Rudder function of a dual motor tailsitter is obtained by differential thrust in both VTOL and fixed wing flight. There is a parameter rudd_dt_gain to adjust the amount of differential thrust in fixed wing. But if you want to add a vertical stabilizer with a control surface you simply have to connect a servo to an output and set it to 21.
This is a strange idea to make a tailsitter with a regular plane but the method is always the same, connect a servo and set the output as you wish.


(James Miller) #1135

Here is the hardware for my vectored thrust twin tailsitter. I’m requesting suggestions and/or criticisms.


Please- throw rocks at the design, but not the cats.
This is the third incarnation for this airframe- hovered fairly well as a tailsitter with OAVTOL and a KK2.1.5 board.
The concept has been mentioned here before, of having two lifting motors (bzucak et al) and a more efficient pusher. It seems like the next logical step in development.
The lifters produce about enough thrust with a full battery to lift the airframe- 3 kilos- and with the assist of the main motor, will not be overtaxed, but must run hard enough to offer good control-70% or so during the multicopter phase , perhaps. They will shut down and props will fold in FW flight.
I’ve read the whole thread and taken notes, so I know something about what’s been done so far.
My own observations:

  • if I had it to do over, I’d put the motors a bit farther apart. Too late. But the photo angle makes them look closer than they really are.
  • It will Need the full complement of vectored thrust twin control features, plus some provision to control the pusher motor and shut down the small lifter-stabilizer motors. Can this be done without new code?
    I am asking for constructive criticism, and some advice on initial parameters.
    Thanks in advance

(Mark Whitehorn) #1136

@losawing I just pushed a bugfix to my quad tailsitter PR branch: https://github.com/ArduPilot/ardupilot/pull/8113
I had the top and bottom motors swapped due to a sign error in the VTOL mode pitch control. Sorry if that’s caused as much confusion for you as it has for me :slight_smile:

I also added differential thrust for the top and bottom motors to aid pitch control in forward flight.


(Mark Whitehorn) #1137

Nice looking wing; is that a custom build?
I think code changes would be necessary in AP_MotorsTailsitter to support your booster motor and shut down the other two. Are you able to modify the code yourself? I’m doing similar things in my quad tailsitter branch: https://github.com/ArduPilot/ardupilot/pull/8113


(Kikislater) #1138

It’s an uncrashable Phantom FX61 modded in tail sitter !
@wazoo22 I have the same wing and same blue except th hatch kept in black


(letpi) #1139

My wing is already modified but I need to compile the code again and I am facing an error related to flash memory overflowed. I make some search and found I have to remove modules. Now I have to figure how to do that. This is a little frustrating so if I do not succeed shortly, I will ask your help.


(letpi) #1140

This is the full parameter list of my vectored bellysitter wing. It fly pretty well with those parameters even with some wind
blue wing 21 avril.param (16.3 KB)
probably the most critical is the back transition. Important parameters are Q_m_spin_min, q_m_thst_hover and q_velz_max < 100cm/s. For the beginning, use Q_hover and FBWA modes. Q_stabilize is dangerous and Q_loiter less stable.
You say your wing motors produce enough thrust to lift the wing (just enough ?) . I found that with powerful motors and vector thrust the wing can recover from the worst situation. This is something to consider at least during the tests.


(James Miller) #1141

Yes, it’s a custom build, using parts of the FX-61 phantom, lightly glassed and reinforced.
Unfortunately, I’m not a coder. Slugging my way through “Arduino for dummies”, if that tells you anything.:grimacing:
I’ve followed your ideas, but don’t understand the airframe layout you intend to build once simulation is played out. But I’m interested.


(James Miller) #1142

Cool!
The foam parts of the FX-61 are a good start for a lot of projects. Great design, as is the wing-wing. I like to modify-sometimes radically- the basic airframe.


The drawback here is added complexity, so I solder everything I can, and avoid as many failure points as I can.
The ugly foam goatee is because tailsitters, even with wheels, tend to trip and fall on their faces in real life, and the foam cushions the impact, reducing the shock load on the motor vectoring servos.
Learned the hard way to make servos replaceable.


(James Miller) #1143

Thanks very much for the files. do I just download them, then move them to the .log folder in MP, and review them in the usual way|?
Did I tell you I was a newbie?:man_student:
My thinking on the motors was to keep their weight and power penalties to a minimum, as well as their effect on the aerodynamics, since they will just be dead weight in level flight. Perhaps a mistake. Time will tell.