Skip to content
All projects

Recycled Plastic Pelletizer

One motor, two belts, one sensor — a pelletizer that turns waste 3D-printing filament into 3 mm pellets.

Summary

What it is
A single-motor pelletizer that turns waste 3D-printing filament into 3 mm pellets
Role
Mechanical design and drive ratio, CAD, ESP32 firmware and testing
Team
5 (ES 1050 Design, Western)
Dates
Winter 2026
Tools
ESP32, C++, AccelStepper, A4988 stepper driver, IR beam-break sensor, GT2 timing belts (×2), Onshape
Results
Gear-drive prototype: fed a 1 m PLA strand autonomously at 3 mm/s, under 50 dB, 12 V input, 0.5 A idle draw; first pellets cut cleanly, then the cutter jammed. Final GT2 design: built, no measurements taken.

Problem

Turn waste 3D-printing filament into 3 mm pellets, with one motor doing both the feeding and the cutting.

Gear-drive prototype CAD model, exploded view

Early gear-drive prototype

What I built

  • We designed a single-motor pelletizer: two GT2 belts drive both the feed wheel and a Forstner-bit cutter at a 7:1 ratio, so feed and cut stay mechanically synchronized.
  • ESP32 firmware (C++, AccelStepper): a debounced start/stop switch, potentiometer speed control from 6 to 120 RPM, and an IR beam-break sensor that runs the motor only while filament is present.
  • We picked the design from 7 concepts with a weighted decision matrix, and produced the Onshape CAD assembly, exploded view, BOM and drilling template.
  • I worked across mechanical design and drive ratio, CAD, ESP32 firmware and testing.
PotentiometerwiperGPIO 34Start/stopswitchother leg to groundGPIO 18IR beam-breaksensoroutputGPIO 19ESP32GPIO 26STEPGPIO 27DIRGPIO 14ENABLEactive LOWA4988 stepper driverStepper motor
Final build wiring (from firmware pin map). Signal connections only.

The 7:1 drive

Final GT2 design · built, not yet tested

Illustration of the design ratio, not a measurement.

Feed wheel ×1Forstner-bit cutter ×7Filament
Cutter : feed wheel
7 : 1

On-screen speeds are scaled. Only the 7:1 between the cutter and the feed wheel comes from the design; each belt's ratio to the stepper isn't shown, so neither shaft speed is given.

How we chose the design

Pugh matrix weighted totals against the baseline concept
ConceptWeighted score vs baseline
Drill-Powered Attachment+0.55
Razor Cutter+0.10
Baseline0.00
Guillotine Cutter−0.05
Cam-Driven Scissor−0.10
Dual-Motor Rotary−0.10
Solenoid Guillotine−0.15
Pugh matrix: Drill-Powered Attachment scored highest. We kept its drill-bit cutter and drove it from a single stepper instead of a drill.
Tri-fold project board and the gear-drive prototype on its bucket at the design showcase
Design showcase, gear-drive prototype

Results, by build

All measurements are from the gear-drive prototype. The final belt-drive design was built but not measured.

Build 1 of 2

Gear-drive prototype

Tested. Every measured result on this page comes from this build.

Gear-drive prototype, side view on the bucket lid: filament spool, electronics, stepper, printed spur-gear train and cutter housing
Early gear-drive prototype (all test results)

Measured results (gear-drive prototype only)

3 mm/s
Fed a 1 m PLA strand autonomously
<50 dB
Noise, by phone sound-meter app
12 V
Input
0.5 A
Idle draw

Gear-drive prototype

Drive
3D-printed spur gears
Firmware
Earlier version, no longer available
Gear-drive prototype CAD model
CAD model. Early gear-drive prototype

Build 2 of 2

Final GT2 design

Built, and the firmware in the repo ran on it. No measurements were taken on it.

Final design exploded-view CAD with 14-item BOM; the two GT2 belts are items 6 and 12
Final design (built, not yet tested)

Measured results

None. Built, no measurements taken.

Final GT2 design

Drive
Two GT2 belts from one stepper
Ratio
7:1
Driver
A4988
Speed
6–120 RPM, set by potentiometer
Firmware
ESP32, C++, AccelStepper (in the repo)

What went wrong

  • On the gear-drive prototype the first pellets cut cleanly, then the cutter jammed. We traced it to poor chip clearance at the bit.
  • All measurements were taken on the gear-drive prototype. None were taken on the final GT2 design, so nothing on this page claims how the final design performs.
Gear-drive prototype, front view on the lid
Early gear-drive prototype

Next steps

Film the final build running. It would be the first evidence from that build.