Scratchbuilt gimbal?

[Axbri]

I want to have a gimbal to stabilize the GoPro camera on my next multicoper project. I understand that for best results I should have a brushless gimbal, but most of them are pretty expensive. Therefore I am thinking of building my own. HobbyKing has motors and a “gimbal controller”.

Motors: http://hobbyking.com/hobbyking/store/__42247__Quanum_2208_Precision_Brushless_Gimbal_Motor_GoPRO_size_100_200g_.html

controller: http://hobbyking.com/hobbyking/store/__41386__2_Axis_Brushless_Camera_Gimbal_Stabilization_Control_Board_w_IMU.html

Has anyone tried those motors or this controller? Do they work well?

I have made a simple Google SketchUp design:


It is built out of high quality plywood and some aluminum parts.

What are the important things to think about while building a gimbal? I understand that the camera needs to bee “in balance” so it can be rotated with as little torque as possible.

Share your ideas and thoughts!

 

[Mustang7302]

I dont have any experience in such a project, but it looks like a fun project to try!

Later versions of the KK2 software will also control a camera gimbal from the M7 and M8 ports. Fancier controllers like the Naza and ArduPilot also support the a gimbal as well, I believe. It could save yourself from using a dedicated controller and all of the additional wiring/clutter.

 

[Craftydan]

Thoughts about the design (which may be totally wrong) :

Nice design from a construction and service standpoint -- assembly/disassembly looks to be easy, especialy if your joints are pinned and the angle brackets are tapped. mapping out cable runs would be a good next step.

If those are brushless gymbals, holding position against force requires current through the motor. As currently designed, gravity is torqueing that roll motor when straight and level, and not resisted by the berrings. If the roll arm was designed vertical down instead of horizontal, you'll probably burn virtually no power when holding the camera static -- you'll only burn powr when compensating for roll motion. You'll have to have a complicated bracket to install the pitch axis w/o a yaw, but the added weight from the structure has got to be less power drain than holding an arm up against gravity.

 

[DDSFlyer]

I think that's where the balancing of the gimbal pieces and the camera is key...in balancing the roll from always needing to be compensated, that's why most gimbals are made in this configuration, but that's also just an uneducated guess.

Look up RC Manchild or dhdsracer on youtube for his custom 3dprinter designed gopro3 gimbal. I'm sure he'll be open to questions about the process.

I'd like to see if someone could make an inexpensive one for the GoPro2 (since that's the one I have) that would be controlled by the KK2 board (also which I have)

 

[Craftydan]

True, balancing can accommodate for it, but if it's built right, balance is built into the symmetry of the frame instead of it requiring extensive counter balance, or continuous draw.

Also, for the frame mount on pitch axis, it might be good to place it on a vertical slotted mount, so it can slide up and down to adjust for different camera heights. Ideal is always to put the lens at the center of rotation.

BTW, It's my understanding is that for best performance the control board (the sensors, really) need to be mounted on the platform with the camera. Otherwise, it's just dead-reckoning any compensation -- it knows it's just been rolled 5 degrees, so it commands the gimbal what it thinks might be 5 degrees, instead of knowing it's off by 5 degrees and rolling the right direction until the sensor reads the right value.

The control board linked has external sensors, so it's a matter of making sure the cable is routed to it and sticking it on the camera, but a KK2, Naza, multiwii, etc., would need space on the mount the whole board. And preferably mount right at the center of rotation . . . right where the camera lens should be.

 

[Axbri]

I have considered using servos and driving them from the kk2.0 board or the Ardupilot2.5, as I will use one of them for my next multirotor project. But I am afraid that servos may have to bad resolution. I also believe that a servo-powered gimbal is going to be bigger and heavier than a brushless one.

I would like to see future versions of those flight controllers with built in drives for brushless gimbals.

One key aspect of my design is that I want all parts to be easy to make or repair, not requiring 3D printed or laser cut parts. I want cheap and easily available materials, therefore wood and aluminum angle brackets. I also want to keep the camera in the casing to protect it in crashes.

I have adjusted some of the dimensions to make it more balanced. Here are some images:

 

[DDSFlyer]

The only thing about keeping the case is it would make it extremely heavy. You could probably make a case/mount to surround the gopro that would be lighter. I'm very interested in your results, keep up the good work!

 

[Tomtom]

Hello there!

That's funny, I wanted the exact same thing for my Knuclke H-Quad (KK2.0 board powered) but be aware that the KK2.0 gimbal feature works only with servos, not gimbal motors! I read that on the rcgroup forums when I was doing some research (see last post of the thread).

As I started to design my own gimbal mount plate for the H-Quad (using sketchup as well) I had to rethink how to do it. I can share it once it's finished if you are interested