legs-still trying


I have had to make the feet larger and larger so the legs can stand on one foot. Tilling the legs to the right, so I can pick up the left foot caused the left foot to land on the right foot on the way down. Breaking things mostly knees. I added metal guards to keep feet apart on the way down. The hip action requires a lot of force. My first attemps were the rack and pinion idea, but the ABS hip parts would break/separate. The legs are still controlled by a arduino mega and am using roboclaw motor controllers for the DC motors with encoders.  I have been trying to get the legs to take several steps with out breaking something.  Thought I would show the current status. I am not sure this will ever work, so I have not posted any part updates.

hairygael's picture

Oh Wow!!Nice work John!Can

Oh Wow!!

Nice work John!

Can the legs go faster or because it is tests and you are actionning them slowly?

What a thrill it must be when doing the first foot step!

harland's picture

They can go faster. The hip

They can go faster. The hip motors can just keep rotating and very fast. The hip motors don't have to go forward then reverse as now. to move the legs back and forth.  If the leg is not up and I try to move the hips something breaks (legs are so long, big lever). It turns out the legs are not very strong when tilting them to the side, to put weight on one leg. That is why the feet can step on each other.  So slow for now to slow down the breakage.

hairygael's picture

Ah good! I understand, safety

Ah good!

I understand, safety first!

When I saw the legs design posted on Thingiverse, I suspected the structure might be a bit weak for to support the upper body.

But anyway, I think it's fantastic to see your progress!

Ray.Edgley's picture


Do you have any Inertial Measurement Units (IMU) installed to assist with the ballance?

Walking is an exercise in arresting the falling action.

I noticed that top of the leg assembly was tilting back quite a lot, with an IMU this should remain reasonable level.

The use of PID loops using both the IMU, and the current motor leg position and combined with a desired leg position may help here.

Don't be disheartened, you have made very good progress. And don't feel you shouldn't post the parts you think are failing, just add a note with the parts, someone else may be able to provide you with a small change that might make all the difference.  As a community working together, we will have walking robots soon.


Keep up the great work.