This video provides a detailed update on the development of the "Ball Butler" robot, focusing on critical mechanical design and control system challenges. The speaker addresses issues with the pitch axis string mechanism, proposes a dowel pin solution, and evaluates motor positioning precision, concluding that a gear reduction is essential for the yaw axis. He details the iterative CAD design process for the base and belt drive, including overcoming a design flaw and seeking community input on joining GT2 belts. A significant portion covers power management concerns, particularly the high current draw of the hand motor, which could overload the shared power supply with the Jetson controller, leading to the exploration of hydraulic dampers for deceleration and a spring-based throwing mechanism. The video also shares insights into optimizing an OnShape CAD library for parts management and organizing an electronics component binder, while inviting feedback on several unresolved technical dilemmas, including safer power routing and off-axis encoder implementation.
Exploring the "Ball Butler" Robot: Design Challenges and Solutions
0:40 Pitch Axis Mechanism Refinements: The speaker addresses a user's feedback about the string's perpendicularity on the pitch axis. He implements a design modification involving a slot and an M8 dowel pin to allow the string to move laterally, aiming for smoother operation.
13:24 Motor Positioning Precision Test: Despite enabling anti-cogging on the ODrive Pro, the motor's positioning accuracy was still around half a degree to one degree off, confirming the need for a gear reduction stage for the robot's yaw (base) axis.
21:28 Base Design and Belt Drive Evolution: The video showcases the CAD design for the robot's base, which includes a large bearing for rotation. The initial belt drive design with idler pulleys was flawed because the idlers were fixed to the moving part.
30:34 Revised Belt Drive Design: The idlers were removed from the belt drive. Tensioning will now rely on precise belt length and potentially reprinting the main red component with adjusted dimensions. A 3D-printed pulley (using a 0.2mm nozzle) demonstrates excellent print quality and belt fit.
33:10 GT2 Belt Joining Challenge: The pre-looped 200mm belts were too short for the current design. The speaker seeks advice from the community on reliable methods to stitch GT2 belts into a continuous loop, expressing skepticism about simple super glue techniques.
40:48 Cable Management & Off-Axis Encoder Investigation: Routing cables through the rotating base without tangling is a major concern. An off-axis encoder was considered to create a clear central path, but commercial options were deemed too expensive ($480), and DIY magnetic off-axis encoders are complex to implement accurately.
49:02 On-Axis Encoder and Mechanical Rotation Stop: The current plan is to use an on-axis encoder for absolute position feedback on the base, but this necessitates a mechanical stop to limit the rotation to just under 360 degrees to prevent cable entanglement.
52:40 Hand Deceleration & Power Spikes: The speaker discusses the challenge of decelerating the hand motor after throwing a ball. Using the motor for regenerative braking would send power back to the 12V 5A power supply, which is shared with the sensitive Jetson controller, risking overload.
54:53 Hydraulic Dampers as Deceleration Solution: He proposes using off-the-shelf hydraulic dampers (similar to RC car shocks) to absorb the hand's kinetic energy, thus avoiding regenerative power spikes back into the system. He plans to test these.
59:33 Spring-Based Throwing Mechanism Explored: An alternative to the motor-driven hand for throwing is a spring-based mechanism. This would simplify power requirements but introduces mechanical complexity for varying throw heights and a quick-release system.
1:06:42 Power Supply Routing Risks: A significant discovery was a previously detached ground wire on the Jetson's power input, raising concerns about the reliability of the current power routing. The speaker worries that high current draws from the motors could damage the Jetson due to shared power lines.
1:13:18 OnShape Library Best Practices: The speaker shares a refined method for creating an OnShape CAD library for standard parts using "derived parts" and "configured properties" to ensure accurate naming (e.g., "M5 washer") when imported into assemblies.
1:23:47 Electronics Binder Organization: An updated electronics component binder is presented, featuring individual pockets with flaps for resistors, capacitors, and LEDs. A minor issue noted is the "top-heavy" nature of the binder due to component distribution, and plans for labeling and integrated circuit storage are discussed.
1:26:17 Spring Throw Calculation & Simulation: He presents detailed calculations and a Python-generated plot that model the relationship between spring stiffness, stretch distance, and ball throw height, providing a valuable tool for designing a spring-powered throwing system.
1:32:30 Project Scope Management: The speaker emphasizes the ongoing challenge of managing scope creep to ensure the "Ball Butler" is completed effectively, balancing perfection with progress, and prioritizing key design elements.