Project Logs

Collapse

• Quick orientation overview

  Kyle Isom • 03/30/2022 at 02:57 • 0 comments

  I drew up a quick overview of the qu4druped's orientation. This numbers each leg, as well as the segments and servos in each leg.

• Basic progress

  Kyle Isom • 03/29/2022 at 07:36 • 0 comments

  I have an OpenSCAD model done up for the base plate; I'm new at 3D CAD so this is mostly a matter of trial and error. I targeted a basic plate with M3 mounting screws (matching the Qu4druped model base plate) and a pair of mounting holes for an 18650 battery pack. The first print shows my screw holes are a little small.
  

  I have one leg (a mirrored leg) printed and assembled correctly. I've had a lot of trouble printing parts for this, which I'll talk about in a future log.
  

  The picture also shows the control system that I'm using to experiment: a Sparkfun Blackboard and a Sensor Shield 5.0 that I got in some kit at some point. It works, but I've got a Teensy 4.0-baesd control board that just left for the fab tonight. The first experiments were... a little twitchy.
  

  
  
  My plan is to use this to develop a serial-based gait test platform. Basically, it will receive lines over the serial console in the form
  

  3 15 15 145

  That's the leg number (number three is the back right leg), the servo angle (0-180º, with 90º as the "neutral" position) for the first (innermost) to third (outermost) servo. I'm building out a chart of general servo settings to help develop the gaits, and I'll probably use a Python program on my laptop to test out gaits - that will help me figure out gaits for walking, turning, obstacle navigation, etc.
  

  Eventually, I might go to a Raspberry Pi/ROS-based system with maybe a TFLite model for motion planning. This will go hand-in-hand with a general pan/tilt sensor system that I'm designing. I have some drawings, but my CAD skills are not up to par yet.
  

  I still need to print another mirrored leg, and I have segments 1-3 of the non-mirrored legs. I'm debating printing out section 4 (the black leg tip) in PETG because I've had a lot of problems with cracked parts, even right out of the printer. I think PETG might be helpful because it's more flexible but I need to install the high-temperature nozzle in order to do that.
  

View all 2 project logs