# sv06-belt-recovery-stress-tests Belt tension recovery and motion stability test models for bed-slinger 3D printers (e.g., SV06-class machines). --- ![](img/DSC_5098.webp) ## Purpose These models are designed to help recover from situations where printer motion accuracy has been degraded, typically due to: * Adjusting belt tension without a reference baseline * Experiencing layer shifts during prints * Replacing belts or modifying mechanical components This project provides: * Parametric OpenSCAD models * A structured test procedure * Guidance for interpreting failure modes --- ## When To Use This Use these tests if you observe: * Sudden layer shifts (e.g., 3-5 mm offsets) * Inconsistent dimensional accuracy * Audible belt skipping or motor stutter * After adjusting belt tensioners Here's an example of a case for a hand held radio where the top suddendly shift at the 20mm height. ![](img/20260318_215156_Wed.png) --- ## Models Included ### 1. Cube (Baseline Test) * Dimensions: 40 × 40 × 40 mm * Purpose: * Verify basic printer stability * Detect gross mechanical issues --- ### 2. Y-Bar (Axis Stress Test) * Dimensions: 20 × 150 × 50 mm * Purpose: * Stress the Y-axis (bed movement) * Detect belt slip under momentum and direction reversal **Important:** * Orient the 150 mm dimension front-to-back on the bed --- ### 3. Tower (Long-Duration Test) * Dimensions: 18 × 18 × 120 mm * Purpose: * Reveal issues that appear only over time * Detect cumulative drift, resonance, or late-stage layer shifts --- ## Print Procedure (Recommended Order) Run each model as a **separate print job**: 1. **Cube** * Confirms system is not fundamentally broken 2. **Y-Bar** * Directly tests axis tension and inertia behavior 3. **Tower** * Tests stability over time and height --- ## Suggested Slicer Settings (Orca Slicer) ### Cube * Layer height: 0.20 mm * Perimeters: 3 * Infill: 20-25% (grid or gyroid) ### Y-Bar * Layer height: 0.20 mm * Perimeters: 3 * Infill: 15-25% (grid recommended) ### Tower * Layer height: 0.20 mm * Perimeters: 3 * Infill: 0-10% * Brim: optional (only if adhesion is an issue) See the settings folder for settings used in Orca Slicer. --- ## Belt Tension Guidance This project assumes GT2 belts (typical for most hobby printers). ### Frequency Method (Recommended) Using a phone-based spectrum analyzer: * Target range: **90-110 Hz** * Reference: ~110 Hz ? ~2 lb tension (Voron baseline) Procedure: 1. Move bed so belt span ~ 150 mm 2. Pluck belt 3. Measure frequency 4. Adjust tensioner 5. Repeat until stable ### Notes * Too loose: * Low frequency (<70 Hz) * Layer shifts likely * Too tight: * High frequency (>120-130 Hz) * Motor strain, possible missed steps --- ## Interpreting Results ### Good Result * Clean vertical alignment * No sudden offsets * Consistent layer stacking --- ### Failure Modes #### Sudden Layer Shift * Cause: * Belt too loose * Pulley slipping * Action: * Increase tension slightly * Check motor pulley set screws --- #### Repeated Shift at Same Height * Cause: * Mechanical obstruction * Cable snag * Action: * Inspect motion path * Check wiring harness clearance --- #### Gradual Lean / Drift * Cause: * Frame or axis alignment issue * Action: * Inspect rails, wheels, or rods --- #### Ringing / Ghosting (No Shift) * Cause: * High acceleration * Action: * Reduce acceleration (not a belt issue) --- ## Notes on Methodology These tests intentionally: * Emphasize repeated motion in one axis * Increase inertial load with height * Create conditions where marginal tension fails visibly This approach isolates motion system issues rather than general print quality. --- ## OpenSCAD Usage Generate models: ```bash openscad -D 'model="cube"' -o cube_$(date +%Y%m%d_%H%M).stl stress_test_models.scad openscad -D 'model="ybar"' -o ybar_$(date +%Y%m%d_%H%M).stl stress_test_models.scad openscad -D 'model="tower"' -o tower_$(date +%Y%m%d_%H%M).stl stress_test_models.scad ``` --- ## Provenance Generated with: * OpenSCAD * Orca Slicer (recommended) * Frequency-based belt tuning derived from: * Voron Design tuning methodology This project was developed with ChatGPT. --- ## License Recommend: MIT or CC-BY 4.0 for broad reuse. --- ## Contributing Improvements welcome: * Additional test geometries * Axis-specific stress models * Automated calibration workflows ---