Automated Mirnov Coil Winder

Developed for General Fusion through my SFU mechatronics capstone, this automated coil winder turns a delicate manual process into coordinated mechanical motion. The system winds fine wire onto bobbins for magnetic-field sensors used in fusion research. As the CAD and electronics lead, I brought together iterative SolidWorks design, fabrication, electrical assembly and operator-interface integration to help deliver a working, tested prototype. 

Project type: Academic - SFU Industry-sponsored Capstone

Sponsor: General Fusion

Timeframe: January - August 2026

Status: Complete

My role: CAD and electronics lead; operator-interface design and integration; shared prototype assembly

Main tools: SolidWorks, PETG 3D printing, Teensy 4.1, TMC2209 drivers, NEMA 17 stepper motors, soldered protoboard electronics

Developed through SFU’s General Fusion-sponsored capstone, this automated benchtop winder produces Mirnov coils used in fusion-plasma diagnostics. The project addresses the difficulty of manually coordinating bobbin rotation, wire placement, tension, and layer changes while handling delicate 36 AWG Kapton-insulated wire.

The system combines a stepper-driven spindle and lead-screw traverse with a Teensy 4.1 controller, TMC2209 motor drivers, and low-friction wire guides. It automates homing, synchronized winding, layer reversal, and cycle completion, with an OLED interface and emergency stop.

As the CAD and electronics lead, I:

• Led iterative SolidWorks design of mounts, supports, and assembly interfaces.

• Integrated PETG-printed components with machined hardware.

• Performed most soldering and electrical assembly.

• Designed and integrated the complete operator interface.

• Shared prototype assembly and electromechanical integration.

Testing and troubleshooting addressed alignment, electrical reliability, and wire bunching at layer boundaries. Calibrating the traverse endpoints improved winding quality. Three consecutive four-layer trials were completed without wire breakage, demonstrating predominantly side-by-side placement. The prototype met the under-15-minute cycle requirement and remained within the $2,000 CAD budget.

The final system uses open-loop control, with turn counts calculated from commanded motion rather than independently measured. Future improvements include encoder feedback, active tension control, wire-break detection, and enhanced guarding.

This project strengthened my skills in mechanical design, electronics, fabrication, and system integration through practical iteration and testing.