HOG Drive System aka Hemispherical Omnidirectional Gimballed Drive System

Let's Make Robots Robots
1

All my projects & progress are available at marginallyclever.com.

about

They were first researched in 1938. Curtis Boirum built one that got on Hackaday. Now I bring you the plans for a model that you can print and build yourself. Use one to make an awesome RC car. Use two to make an RC car that can move in any direction. This system has about 15 degrees of movement on either axis.

Instructions

Long way: Get the designs from thingiverse and 3D print a copy. Drill out the missing holes Add - two hobby servos - one Tamiya 70171 kit (cut to size) - a pair of single-arm servo horns (or amputate a double arm) - 3 10mm M3s - a pair of zip ties - a 5v0.2a DC motor

Short way: Get the parts in a kit from marginallyclever.com. Assemble as in the pictures.

Expect to file a bit because it's a very tight fit. I haven't found a good way to attach the hemisphere, please comment with your thoughts.

Final thoughts

You'll need two free-spinning wheels to make a triangular base and allow the HOG drive to move about.

I'm using one to make a line-following robot. The body of the vehicle will be made from folded cardboard with a pair of free-spinning wheels on the back. I'm using QRD1114 IR sensors for the line detection and an Arduino to be the controller.

Moves a vehicle forward, back, left, and right; teaches about gimbals

  • Actuators / output devices: 5v DC motor
  • Control method: PWM control to the servos; 5v to the DC motor.
  • Power source: 5v1a power supply.
  • Target environment: smooth flat surfaces
This is a companion discussion topic for the original entry at https://community.robotshop.com/robots/show/hog-drive-system-aka-hemispherical-omnidirectional-gimballed-drive-system
2

Floor Buffer

Janitors have been “steering” large floor buffers using this method for years. Its amazing how effortlessly one can control a very large, heavy, powerful machine simply by changing the angle of attack and thus, which “edge” is grabbing the floor. 

I never thought of using this method for robot control, but I have to tell you I love it and for the best reason --Fewest moving parts. In my mind, there is no better way to determine the quality of an idea or project other than by simply counting the number of moving parts.

Incredible, dude. I know what is on my list of “next things I want to build”. I think I need to get to the supermarket so I can get some of those rubber super-balls from the gumball machines. 

Awesome. Just awesome.

3

I’m going to try and print a

I’m going to try and print a sphere with a hole already in the right spot for the shaft. I doubt I can get a set screw in there.  Anyways, once I print it I thought I’d use dip & grip to put a rubbery coating on it for traction.  Please let me know if the super balls work.

4

I will indeed.

I gotta tell you, this system has really bitten me. I have been rolling the idea in my head for most of the afternoon. So far, this is what I got:

  • Dia of ball and thus, gear ratio
  • Superballs from gumball machines
  • Spray-on bedliner from a rattle can
  • Those red rubber kick balls (pnuematic problem)
  • Racket balls --again, not solid
  • That red stuff that comes in the tall can that you dip the handles of pliers in
  • Is there a max size for the ball?
  • If really low gearing is introduced, could 3’ dia balls (yard size) be used?
  • Advantages/ disadvantages of total number of units --Basically, is anything better than a triangle?

This is my personal list to start playing with. I will keep you in the loop if I hit anything new and/or exciting.

 

5

Whoa, everyone slow down a wee bit…

Umm, wow. Ok, somehow we went from, “this is a great project” to “stealing proprietary stuff” and selling kits. Let’s nip this guy in the bud…

I have no connection to 6677, and he is not speaking for me or my business. I think he was just making a friendly suggestion. I have no intention of scooping, stealing, borrowing etc. I have no intention of developing this into a comercial project. And I sure as hell would not make the smallest of moves in that direction without asking permission first. 

I get sad when it is suggested that I would even consider such a thing. 

Of course, I am open to partnerships and I am always looking for new stuff. In this instance, however, I don’t know how we jumped from “Hey I am going to play with this” to “business relationships and selling kits”. Everyone just take a breath and slow down a bit here…

Alright, done with that. Christ.

******************************************************

The racket balls are working great. I have one cut in 1/2 with a semi-half-round aluminum hub stuck inside. I drilled for the axle of a little geared motor and added a set screw. The motor itself is geared quite low (<100 RPM), but with the diameter of the ball, the speed is not bad.  --I.e. robot speed, not quite your gym floor speed. The grip is amazing. Note, all tests were done by simply holding the motor/ball assembly by hand and running it along the desk.

I have decided to print the frist set of gimbal brackets instead of doing metal or plexi. Even with my dislike of sketchup, I think I can draw them up faster than I could machine them. I am doing a single unit to start and moving on to a triangle with one guy in each corner. --I am wondering if I am going to have to do some kind of counter-rotation using 3 balls and no fixed wheels of any kind.

 

6

This really helped me! I am

This really helped me! I am going to atempt to use this in the butt of my robot I am building to save me linkages, power, and two extra servos i don’t need! Im using this design (loosely) as a ball joint. I think. But It’s really neat tho!