Home Creators Posts Import Register Favorites Logout
haven't archived this post yet. have a subscription? use the importer!

Content

Here's part 2 of openDog, early for Patrons!

Obviously there are a few things to improve with the design and a few decisions to make, but that's the reason for building a prototype before cutting metal. I have a few more videos and tests like this to do before we get on with the actual build.

Files

Previews only

openDog Dog Robot #2, Leg & Joint Prototype | XRobots

Part 2 of my openDog open source hardware & software Dog Robot. This time I'm prototyping one leg using Odrive and a ball screw to check joint angles and control using the remote and code from part 1. This robot dog project is open source! - get part 2's CAD and code from: https://github.com/XRobots/openDog https://odriverobotics.com/ https://github.com/madcowswe/ODrive https://github.com/madcowswe/ODriveArduino XRobots is Supported through Patreon with LulzBot 3D Printers from Aleph Objects, Inc. Help me make more AWESOME projects, and get exclusive REWARDS with Patreon!: http://www.patreon.com/XRobots Get my T-Shirts, US shop: https://shop.spreadshirt.com/XRobots Get my T-Shirts, Europe: https://shop.spreadshirt.net/XRobots Instagram: https://www.instagram.com/xrobotsuk/ Facebook Page: http://www.facebook.com/XRobots.co.uk Twitter: http://twitter.com/#!/xrobotsuk Web: http://xrobots.co.uk G+: https://plus.google.com/104659063664991081890/posts Facebook Group: https://www.facebook.com/groups/287089964833488/

Comments

Anonymous

You may also want to create a form of harness to support the dog while you work on the code to make it walk. This way it will support the weight and keep some form of balance so you can see the legs work (or not) while you tweek the code.

Anonymous

Awesome progress! Glad to see it's coming along so nicely. Couple questions/thoughts. 1. why not put the encoder on the actual pivot joint for the knee so you're measuring the angle directly on the linkages. 2. some of the CUI AMT encoders allow you to save the zero position in non-volitile memory so it's remembered on a power cycle, perhaps this would help with some of the calibration issues. 3. have you considered modeling the robot leg with D-H (denavit hartenberg) parameters? At that point the forward kinematics become simple matrix multiplications, and inverse kinematics can be worked out with a "pseudo jacobian / gradient descent" approach. If you're interested, I'm happy to provide a pointer to some reference code for that. Keep up the great work!

XRobots

The Odrive will work best with the encoder on the motor, I considered extra absolute position encoders on the joints as well, but we'll keep it simple for now. I really want to do the kinematics the long way so I can explain it easily which makes the entry point to understanding for the viewer a bit lower.