did anybody start developing a stabilisation-system for fixed-wing RC-Planes?

StefanL38

Hi,

beneath coding for the propeller-chip I do RC-plane flying.

In all my models I use flight-stabilisation-systems from Horizon-Hobby.
But these stabilisers have their limits. Servo-directions are not adjustable nor gain values for
the stabilisation.

Horizon-Hobby does not offer their new stabilisers as a spare-part. So i started thinking about developing my own.
But maybe somebody else did this already? Of course there is the HoverFlyOpen flightcontroller for quadrocopters
but fixed-wing planes need their own firmware.

I guess some of the readers of my post read my question as "can somebody code a fixed-wing-stabilisationsystem for me for free?"

If you want to read this way do it. I'm free to ask you are free to answer or not. Whenever I came across a question like "does somebody know...?" I answer as best as I can.

best regards

Stefan

evanh

StefanL38 wrote: »

I guess some of the readers of my post read my question as "can somebody code a fixed-wing-stabilisationsystem for me for free?"

Most hobby stuff is done for free first off. The real question quickly becomes one of commercialisation. If one individual starts packaging it for sale without even acknowledging others' contributions, it does tend to garner some negative coverage.


As for a solution, I didn't even know there was any active stabilising for fixed wing craft, I assumed they are normally passive stable the same as your average full scale craft.

Phil Pilgrim (PhiPi)

As someone with minimal eye/hand coordination and who has "tried" flying RC gliders, I can assure you that any help with stability would be greatly appreciated!

-Phil

User Name

I can't speak for Stefan, but there are lots of times I'd rather not incorporate dihedral into a wing or provide pendulum stability. You can still fly such planes (like the Funder & Lightning) but it is a very active sort of piloting, especially in gusty conditions. A little roll-axis stabilization is a great thing in such circumstances. It's also conceptually very simple - just a tiny module inserted between the receiver and the aileron servo(s).

But, yeah, being able to program servo direction and adjust PID coefficients would be great, and is almost de rigueur in the multicopter world.

StefanL38

high-wing airplanes are pretty self-stabilisating and even without stab-electronic relatively easy to fly.

Here is a link to a video showing how much a stabilisation-system with 3-axis-gyro and a 3-axis-acceleration-sensors can do.



You can switch on/off the stabilisation on the transmitter.

best regards

Stefan

Phil Pilgrim (PhiPi)

Heck, even I might be able to fly with a system like that!

-Phil

StefanL38

I got a new idea how the system should work:

The SAFE-System in the Apprentice keeps the plane neutral on each axis when the corresponding transmitter-stick is neutral. This means if you bank the wings to one side with the "aileron"-stick
(plane starts rolling around axle from nose to tail) elevator is added automatically to keep height.

This is a very different behaviour than the plane would show when stabilisation is switched off.
With stabilisation off you - the pilot - has to add elevator yourself. Otherwise the plane would fly a curve downwards. The SAFE-System in beginner-mode limits the roll angle. Without SAFE-angle-limiting the plane would fly a roll faster or slower as long as you give non-neutral-aileron-input.

If you have learned to fly just in what Horizon-Hobby calls "SAFE-beginner-mode" you will be used to pull aileron sticks fully to the left/right and you don't care about the elevator-stick.

Without SAFE activated to fly a curve aileron-input is short to bank the wings as soon as wings are banked aileron-input is neutral and elevator is added to keep height.

Now my idea is a "practising"-mode that works as follows:

Elevator is not added automatically. So the pilot will become used to add elevator. Rolling is not limited to a fixed value of 30 degrees but adjustable. If attitude goes beyond adjustable limits the stabilisation brings back the plane to flat-level flying and as soon as the plane is flat level the stab-system gives back control to the pilot. So the mission of the learning pilot is to fly and keep the attitude within certain limits to avoid activating the auto-leveling. This behavoiur would combine the safety of the Horizon-Hobby-SAFE-System with learning to fly without stabilisation-support.

I have never used gyro or acceleration sensors yet. So I have no idea how complicated sensor-reading is and how much effort it means to code PID-routines for the stabilisation.

What do you think?

best regards

Stefan

StefanL38

I hope it is not too offtopic to insert this video-link. Greatest thing of all would be FPV with a pan-tilt-camera where the camera acts on head-moves like in this video.

I think this would come pretty close to real flying.

Another application for the propeller-chip:
transform head-moves of goggle-headset to corresponding moves of the camera


best regards

Stefan

Duane Degn

As much as I'd like to have an all Propeller solution for flight stabilization, there's a much easier way to accomplish this. Just use one of the many Arm based flight controllers.

I'm very impressed with the Naze32 board. There are a couple different versions of the board and IMO you're better off getting one of the "full featured" boards so the board can detect altitude.

I have this one purchased from Amazon.



There are a couple of different configuration software packages for this board.

Baseflight and Cleanflight are the options I've heard of. I've only used Cleanflight myself.

It's pretty cool to connect to the Naze32 with Cleanflight and see a 3D rendering of a quad in the same orientation as the real one on the bench beside you. The one screen quad moves as you move the real quad.

There's a lot of information about these flight controllers on the internet. Here's a YouTube video series about how to set the Naze32 up for use with a quadcopter.

I'm pretty sure the Naze32 and other control boards can be used with fix wing aircraft. I haven't tried one of these in a fixed wing aircraft yet but I plan to do so in the future.

The Cleanflight software even has WS2812 configuration settings to let you setup how you design patterns with our LEDs.

Hugh

APM Ardupilot does have a fixed-wing version.

I seem to remember a few years ago a simpler stabilisation system using horizon sensors was the Parallax 'Project of the Week' (or similar); it so far eludes my searches...

mmowen

Are your RC models inherently unstable?

Wing dihedral provides roll axis stability. Proper center of gravity location combined with appropriate tail down force, elevator trim and wing center of lift location produces positive dynamic stability on the pitch axis. Yaw stabililty is a function of keel effect and CG location. Of course prop-wash, torque and propeller gyroscopic effects come into play as well.

If the model is designed to be stable, then all thats needed is to zero the power and roll the aircraft upright into a power off glide, no?

Consider the venerable Cessna 172... properly loaded and trimmed you can pull or push the control yoke and let go; it will porpoise a few times and return to level flight attitude. It also tends to roll out of a bank (less than 45 degrees) settling back into a wings level attitude; and if you tromp on one of the rudder pedals and release it the nose returns to a straight path (as when executing a forward slip to landing). I used these actions as simple demonstrations early on for all the people I taught to fly, so they would understand the implications and sources of aerodynamic stability.

Of course if your model has the traits of an F117 computerized stability augmentation is your only option.

This WIKI article (https://en.wikipedia.org/wiki/Relaxed_stability) summarizes it nicely.

JasonDorie

It would be pretty trivial to modify the ESC outputs from the Elev8-FC to make it drive pitch, roll, and yaw servos instead of the ESC controls, but it isn't "built-in".

I've seen it accomplished in the past by using a simple rate gyro on each axis, connected to the appropriate servos, with good results.

J