Let's Make Robots!

Almost a robot page worth

1069d1_putting_together_640.jpg

My Theo Jansen Walker is getting closer and closer to a robot-project-page-worthy machine/creature. The port side is now complete but not finished.

Complete in the sense that I have all components designed, purchaged, built and even assembled. Not finished in the sense that it will not move! Hence the lack of video. Photographs will have to do.

1069d1_belts_too_tight_640.jpg

There is friction in every moving part of this finger trap. Too much friction for the pulley on the motor shaft to handle. It is grinding through the teeth on the timing belt. I will have to figure out which part is the big trouble maker. And then sand down some of the plywood parts as well. And then lubricate all the pivots with bees wax. And find a better way to tension the belts. Or maybe start shopping for much biger ones.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I think you want to check this out:

http://code.google.com/p/sketchyphysics/downloads/list

will do
what about a chain system ?
But I invested heavily in these pulleys. So I am going to milk for everything they're worth.

Just seen the video, nice final touch  :p

Your work is amazing! Can't wait to see what will be the next move :)

I got the first leg running (again). I disconnected it from the other legs. This is the one that had the most filing and sanding done on it. Therefore, the pivots are very loose. And I lubricated them all with bee wax, as Chris advised. (He is a carpenter, you know!)

tj-pt10_1s0.jpg

I attached an LED on the tip of the foot and placed the camera in front of it. Long exposures show the path of the foot. In the above picture, you can vaguely see the LED leads running left to right. Here's the close ups:

tj-pt10_four_traces_500.jpg

The numbers are exposure times. Notice the different path in the bottom photo. It is running in the opposite direction and displays a few bumps in the upper part of the trace. And the bump in the right hand part (which is the rear of the robot) disappears. Also notice how flat the bottom part of the trace is. That is Theo Jansen's work. His genetic algorithm was set to find the flattest step motion. This means the creature is never lifted by the leg, thus saving energy.

Where the trace is brightest, the LED is moving the slowest. Notice the small gap in the 1.0 s exposure (top). I guess this motor can push one lubricated leg around its path once per second. The reduction ratio in the pulley system is 1:1 in this experiment.

Video to follow. As if that needs to be mentioned.

Not much to explain. Lubrication works. Changing the pulley ratio is not a problem. For just the one leg....
Amazing! I may loose a bet ;)
If the wheather holds ;-)
But it goes way over my head. Im always amazed at how neat and precise your work is.