Homebuilt Quad

Let's Make Robots Projects Showcase
1

I just started diving into the world of autonomous robotics. Wow, what a world it is. I started working on my second auto bot and first walker. I purchased the a botboard and Atom Pro from here and dug up some HS-311 servos I got as a gift 2 years back.

Here’s a pic of it so far: flickr.com/photos/15949634@N02/2059774123/

I have intentions of CADing up the entire thing and having the parts waterjet, so the fit and finish is a little better.

I am just learning how to program this thing, so the code leaves something to be desired, but I’m happy it moves. It only works on hard smooth surfaces at the moment because the pointy feet have tend to get stuck in the carpet.

Later,
-Matt-

2

Very nice, I know it can be hard to custom fab you own parts and you did a nice job.

What did you use to cut the parts with?

3

Very cool! Might want to put some rubber onto the bottom of the pointy feet so they don’t scratch the beautiful table :wink:

4

Thanks. I made everything with a hand drill and a bandsaw. The base is delrin and the legs are obviously aluminum.

The wife doesn’t allow bots on the table except for photos. :wink:

I may try cutting the tips off the feet, so the bot will work a little better on other surfaces.

I’m also trying to perfect the code for turning. The bot walks pretty nice in a straight line, but flops around like a fish when turning.

It’s far from being complete, but it’s a nice little platform for learning how to program and perfecting walking gaits.

Thanks,
-Matt-

5

Its been my observation that unless some very complex programming is involved, a fair amount of surface slippage of the feet is required for pods to operate without straining.

6

I have 4 servos left in the bin, so I’m thinking I can turn this into a hex to eliminate the stability problems. I don’t feel the servos are strong enough to make each leg a 3 DOF leg, unless I redesign the mounting method and leg geometry.

Is there a concensus on whether a 3DOF quad has more advantages over a 2DOF hex? I think it would come down to stability vs style. The quad should be able to achieve more intricate movements, but the gait is more advanced to keep it balanced whereas the hex would be inherently stable, but maybe lacks some flash?

Thanks,
-Matt-

7

Hi Matt,

I’ve heard that 3DOF legs are required to do the omni-directional moves (full translation).

I have a fair number of used Futaba 3004 servos that I’m thinking of building a 2DOF hexapod with. Lower torque requirements, I’m told. Might be just right for my nephew.

As 3DOF legs have a greater range of motion, I’m thinking they would facilitate tetrapod (quad) gaits. the greater knee motion might allow shifting the weight from side to side, helping the stability.

Alan KM6VV

8

That’s exactly what I was thinking. If I can drop the corner opposite of the leg being lifted it should help correct the tipping issue. On the other hand adding 2 more legs would definitely correct it. I’m leaning towards a 3 DOF quad, simply because I think the end effect would be more interesting.

-Matt-

9

Hi Matt,

The 3DOF legs are more interesting to watch as well. So we have two diametrically opposed goals, it would seem. Extra mobility, or an inexpensive implementation! I can buy that.

I like it. Not sure how to model it yet! Yes, a hexapod is a good place to start. Seems funny, seems like one would start with a quad, and “move up” from there. You could actually get a normal hex body, add 4 legs, and get something walking. Then add more later. Doesn’t work that way yet, because, as you know, the gait is the tricky part!

Alan KM6VV