P2 based Dog Bot walking (much better)

Rayman

This robot has been on the shelf for a couple years, but finally had some time to work on it.
I posted before on this, but it wasn't walking very well.

This time used SolidWorks to extract angles from joints.
Idea this time is to keep body at the same height and level the whole time.
Creating a table of angles and then interpolating from table while moving forward.

The battery needs some work, can't go much faster without stalling. Just four AA batteries now. At least need some big caps, maybe need to switch to LiPo pack...

Anyway, the walking motion looks much better now.
https://www.youtube.com/shorts/eRXr38CXFz4

Rayman

Here's the view from SolidWorks.
I'm not the best at it but got the angles of leg to body and foot to leg to show up as "sensors".
I'm using that measurement that says 9.13 as the X in the table of angles.
So, the body height is fixed and the bottom of the feet are fixed to holes in the green baseplate, except when advancing.

One pain is that SW sensors always report the angles from 0 to 90 degrees, so sometimes have to manually fix angle, A, to be 180-A degrees.

Rayman

Put the angles into this Excel table. Then, adjusted to be 180-this angle where needed.
Then, multiplied x*100 and angles *10 to make good integers for Spin2 to handle in a table.

Tubular

Wow, Ray!

I also like that its 2-tone

pik33

First steps Good puppy...

Give him shoes Some rubber (TPU? - it seems to be 3D print...) on these legs to avoid slipping.

Rayman

@pik33 There are grooves on the feet for o-rings to prevent slippage on slick floor. Have to try that later...

BTW: Here's how the angles are defined and extracted from SolidWorks for this.
Turns out the angles for left side are just 180-angles for right side.

Christof Eb.

@Rayman said:
This robot has been on the shelf for a couple years, but finally had some time to work on it.
I posted before on this, but it wasn't walking very well.

This time used SolidWorks to extract angles from joints.
Idea this time is to keep body at the same height and level the whole time.
Creating a table of angles and then interpolating from table while moving forward.

The battery needs some work, can't go much faster without stalling. Just four AA batteries now. At least need some big caps, maybe need to switch to LiPo pack...

Anyway, the walking motion looks much better now.
https://www.youtube.com/shorts/eRXr38CXFz4

Thanks for sharing! Nice scenic view too!

From the picture the upper limb and the lower limb have length (center-center) proportion 8,5:8 ?
What length would be needed to climb stair steps, if their height is up to 200mm?

Where we live, there are some stairs. It would be nice to have a "robot" to carry loads up and down those stairs. It should be able to carry 20kg to be useful, I think.
Do you think, that this principle of movement would make sense for such application?
I am trying to do some rough estimations for feasibility.....

Good to have the geometry in a CAD system. I sometimes use Excel to plot results of equations or measurements. I can then let Excel calculate trend lines in the plot and have their formula displayed. This I can then use with microcontrollers for approximation.
Christof

Rayman

@"Christof Eb."
Think already pushing limits of what hobby servos can do here...
Do have a dream of it walking up stairs, but don't know how realistic that is yet.
Going to have to adjust servo rotation to even make it touch top of stair...

Rayman

Tweaked front arm rotation limit as much as wire would allow.
Looks like it might be possible to climb..