The Gimbalotl: A 3-Axis Joystick Gimbal

A friction-free, cam-based, and durable joystick gimbal based off the mechanisms in those expensive ones I can't afford.
63
315
1
2611
updated August 22, 2023

Description

PDF

Credit to my son for the name; it's fun to say.

EDITS: See bottom for details

A long time ago I got excited and found that gimbal design by ulokelo and asked my neighbor to print it. He did, then I learned it had no Z axis and I got sad, bought my own printer, learned that Fusion 360 was hard and so was designing things, and gave up. 2 years later and now I know how to use make CAD bend to my will (somewhat), then my ADHD got me all hyperfocused on this gimbal and I've ridden that wave for 3 months. I can feel the inevitable end to that wave and so I better post this while its nearly done, lest I never do.

This design has been printed and assembled and lightly refined, but as of this publishing date (8/15/23) I still have some desired refinements I would like to work out. So far my first iteration has been tested brutally by multiple 7-year-olds pretending it was a walking stick for a few hours, and then by me really yanking it around for at least 4 cumulative hours. No signs of extra deflection or damage yet, which I gotta say is surprising. (4 perimeters, 20% infill, PETG). It is about 4.5 inches tall, not including the aluminum rod.

The aluminum tube serves as the shaft to which the grip will later be attached. It allows for wires to pass through safely out the bottom into the (soon to be designed? Maybe?) gimbal shroud and the arduino/leonardo board ideally kept within. For now, this published design only contains the gimbal mechanism itself.

There are some things to keep in mind when printing:

  1. Only one part requires supports - The Z Axis Top Cam Mount, it prints upside down and the three flanges need support. The top of the flanges (which will be inverted for the print) will thus be the support interface area and doesn't have to be perfect. 
  2. The rest of the model is designed to be supportless! Yes, even the Lever Body halves, although this really depends on how clean your printer prints. No unsupported angle over 45 degrees, except for the arches of the Lever Body halves, and they should be fine.
  3. I recommend outer brims. Warping is not helpful here and this project really has made me fond of brims. Just use a razer blade to trim them off, no brimmed surface in this design needs to be perfect, I made sure of it.
  4. I also recommend to print outer perimeters first. Dimensional accuracy is pretty important here.
  5. I have mounting slots for each of the 3 magnets, but nothing specifically for the hall sensors yet. I want to do some testing, and honestly the sensors themselves can just be glued anywhere. Perhaps the next iteration once I get the correct ones tomorrow. I intend the X & Y axes to have the sensor placed directly below the bearing mount for each, and similar for the Z axis, but on the Top Cam Mount. 
  6. Designed tolerances are around 0.2mm (+/- 0.2mm depending on print location\orientation). 
  7. I'm including the STEP files for the design, feel free to modify it to adjust tolerances.
  8. See the third photo below and the notes above it about printing the cams. I also hit the curved face of the cams where the roller bearings glide with 600 grit sandpaper. This really makes the action glide like butter, it's crazy.
  9. Feel free to redesign and repost, please provide attributions though!

This print requires other hardware: 

  • 7x 6702 Bearings
  • 18x 685 Bearings
  • 15mm x 10mm Aluminum tube, at least 20 cm long, (ext. diameter is most important)
  • M4 flat wafer head screws - 10x 12mm, 10x 18mm, 10x 30mm (not exact quantities yet)
  • M4 sleeve nuts - 10x 8mm, 10x 14mm, 10x 18mm (not exact quantities yet)
  • 3x 30mm long, 6mm wide tension springs (8mm wide might be better), 0.8mm thick
  • 3x 3mm x 6mm disk magnets 
  • 3x Linear Analog Hall sensors (NOT digital ones as I've just learned haha)

I bought all the above from aliexpress for ~ $40 (…and a bunch more stuff I didn't use, but don't tell my wife. I'll use it eventually!)

Example wishlist HERE - THESE ARE EXAMPLES, the list items don't save my chosen variations like the dimensions and stuff. 

See the Screencaps below for print orientation reference:

For the cams, you'll need to print 4x of the XY cams and 2x of the Z Cams. They are marked for easy reference after the print. Notice how my print layout has the curved surfaces that will contact the roller bearings facing outwards. This is to prevent the nozzle from crossing those specific perimeters. You want them as smooth as possible, and so I have also painted on seams using prusaslicer along the sections that surround the mounting bearings:

 

Some Assembly Photos:

Here you can see the hole drilled into the rod, this allows for the Z Axis Roller Arm to be affixed to the rod. I used a cheap drill for the this, thought I'd need a drill press but luckily not.

 

 

 

 

A little push is required to get the Top Cam Mount to click and slide over the Z Axis Roller Arm, and sandwiched betwixt the two is a 6702 bearing.

 

 

 

 

 

I don't have pictures of the Lever Body and Z Axis Bottom Mounting Plate being assembled, but see the next gif for an example of how it works. You can see one bearing here, another is farther inside. Then the Z Axis Bottom Mounting Plate slides over the top of two Lever Body halves and holds them together. Getting this fit snug is pretty important, a little wiggle is ok but the tighter the better. Then you slide the tube into the lever body half and attach the top and bottom Z Axis Mounting plates together, sandwiching the Z Axis Cams (seen below)

 

Yeah I hope this makes sense!

 

 

 

 

 

 

 

 

 

 

 

Ready for the springs!

 

 

 

 

 

 

 

Springs like so!

 

 

 

 

 

 

Now we start the Gimbal Core assembly…

 

 

 

 

 

 

 

Prepping the cams…

 

 

 

 

 

 

 

Mounted! Do this before assembling the core, and same for the other axis onto the larger frame side.

 

 

 

 

 

Ready to be mounted into the frame! That's pretty much more of the same so there you go. Last pic for now.

 

 

 

To DO (we'll see):

  • A damper system - I have some good ideas, one may be perfect and require no modification…
  • homogenizing the fasteners. I got it down to a short list already, but I'd like to do better
  • Test hall effect sensor locations, etc.
  • Housing/ a case
  • Test more cam profiles. 
  • Test more spring strengths
  • other stuff

(8/15/23) Gimbalotl Version 1.0 - The complete & functional gimbal (mechanism only) posted.

(8/22/23) Discovered I need bigger magnets, Increasing the slots to ~ 5mmx10mm. Hall sensors 49E work pretty great with that configuration. Files to be updated tonight.

 

Tags



Model origin

The author marked this model as their own original creation.

License