MECHANICAL DESIGN & MACHINE DESIGN

This week was a group project week. All 10 students at Fablab Kochi came together, brainstormed machine ideas, then split into two groups of five. My group built the Pixel Art Machine, a CoreXY bead-placement CNC that physically recreates pixel art by dropping black and white beads into a perforated acrylic sheet, one bead per pixel.

My roles in the team were Electronics and Documentation.

Ideation

We started with a full group brainstorm. Everyone put forward machine ideas on the whiteboard, anything from fun to functional. The ideas ranged widely:

• Kurian's Lights — floating light display
• Automatic Scene Changer
• Origami Paper Creasing machine
• Cornhole Conveyor Belt
• Ola Leaf — ola pantu machine
• Foldable machine with Iron
• Paint Mixture machine (R + G + B → colours)
• Metal 3D Printer
• Auto Soldering machine
• Manual Drip

Alongside the ideas, we mapped out the generic structure of any machine — end effector, mechanism, electronics, firmware, sensors, communication, interface — to make sure whatever we built could be broken down into clear roles.

After discussing the ideas, the 10 of us divided into two groups of five. My group chose the Pixel Art Machine.

Our Group

The project was split across five roles:

• Nadec Biju — Machine Gantry (CoreXY design in SolidWorks)
• Kevin J Jijo — End Effector (bead dropper mechanism)
• Merin — Electronics & Documentation
• Architha B K — Firmware & Presentation
• Kurian — Interface & Fabrication

Concept Sketching

After settling on the Pixel Art Machine, we sketched out the machine on the whiteboard to figure out how all the parts would work together. The main things we needed to resolve at this stage were:

• How the drop mechanism would work, we sketched several funnel shapes and gate ideas for the end effector
• How the bead would travel from the container down to the hole in the acrylic sheet
• What the perforated acrylic grid would look like and how the machine would index across it
• Which team member was responsible for each subsystem

The sketches helped us align on the physical layout before anyone started designing in CAD. The funnel-shaped containers on the left explore different ways to hold and feed beads into the drop channel. The grid on the right represents the acrylic substrate and how the machine would traverse it row by row.

My Assignments

Within the group, my two roles were Electronics and Documentation. Here is what I was personally responsible for:

• Writing and publishing the group documentation site
• Project management — creating and maintaining the Gantt chart
• Drawing the electronics block diagram
• Designing the PCB in KiCad
• Testing the limit switches

Documentation

I was responsible for documenting the entire group project, not just my own subsystem. That meant writing up the gantry design, end effector, firmware, and the web interface as well as the electronics, pulling together notes and photos from the rest of the team. The full site is on GitHub.

Project Management

To keep track of who was doing what and when, I put together a Gantt chart mapping out the full project timeline across all five roles. The main challenge was that some things were blocked by others, the firmware couldn't be tested until the electronics were ready, and the end effector couldn't be assembled until the gantry was built. The chart helped make those dependencies visible early.

PCB Design

Before starting the schematic, I drew a block diagram to map out all the electronic components and how they connect to the MCU.

The diagram has three output branches from the MCU:

• Motor Driver 1 → Stepper Motor 1 — one of the two CoreXY drive motors. The limit switch feeds back to the MCU for homing.
• Stepper Motor 2 (via Motor Driver 2) → Stepper Motor 2 — the second CoreXY drive motor, also with a limit switch for homing on the other axis.
• Servo Motor — directly driven by the MCU, used to actuate the bead drop gate in the end effector.

The dotted lines represent sensor feedback (limit switches back to the MCU). The solid lines are control outputs.

I designed the control board in KiCad. The board brings together the RP2040, two DRV8825/TMC2209 stepper drivers, limit switch inputs, and feeder control outputs onto a single board.

The main things I had to work through during layout:

• Keeping motor driver traces short and away from the RP2040 signal lines to reduce noise
• Adding a logic-level voltage divider
• Breaking out all the limit switch and feeder pins with enough clearance for connectors

The RP2040 was chosen specifically because it has built-in Wi-Fi — this means the machine can receive G-code wirelessly from Kurian's browser-based interface, with no USB tether needed during normal operation.

Bill of Materials — Electronics

The schematic and PCB layout were done in KiCad. The board routes stepper driver signals from the RP2040 GPIO pins, provides 3.3V logic level compatibility, and breaks out the limit switch inputs and feeder control outputs.

How the Machine Works

The software pipeline runs from image to physical bead placement:

1. Image Input — A monochrome image is loaded into Kurian's browser-based Pixel Bead Placer tool
2. Thresholding — Each pixel is converted to black or white based on its brightness
3. G-code Generation — The tool maps pixel coordinates to hole positions and generates a G-code file
4. Wireless Streaming — G-code is sent over Wi-Fi to the RP2040 via the NeoPI_Wireless approach
5. Bead Placement — The CoreXY gantry moves to each hole, activates the correct feeder (M3 for black, M4 for white), and drops a bead

Machine Specifications

Group Documentation

The full group documentation site — covering all subsystems including gantry design, end effector, firmware, and the web interface — was also put together by me as part of the Documentation role.