Amalgam Calibration Guide

This guide walks you through calibrating your Amalgam printer from first power-on to validated prints. The process is designed around Klipper’s capabilities and the Amalgam hardware.

Philosophy: Get it working, then optimize. Don’t chase perfection on step 1 before confirming step 2 works.

Calibration Order

Follow this sequence. Each phase builds on the previous one.

Phase Goal Time
1. Mechanical Setup Frame square, belts tight, Z aligned First build
2. First Layers Consistent adhesion across the bed 30-60 min
3. Dimensional Accuracy Parts come out the right size 30 min
4. Klipper Tuning Input Shaper + Pressure Advance 1-2 hours
5. Filament Profiles Temp, retraction, flow per material Per filament
6. Validation Benchy + victory print Fun

Phase 1: Mechanical Setup

Before any calibration prints, ensure the machine is mechanically sound.

Frame Squaring

Scaffold path (M10 threaded rod): - Use the MDF base as your reference — it’s the squaring jig - Check 90° at each corner with a machinist square - Tighten jam nuts only after square is confirmed

Mill/Lathe path (aluminum extrusion): - Extrusions should be perpendicular to MDF base - Check diagonal measurements — should be equal - Tighten all frame bolts after confirming square

Belt Tension

Belts should be firm but not guitar-string tight.

  • Too loose: Layer shifts, ghosting, dimensional inaccuracy
  • Too tight: Motor strain, premature bearing wear, noise

Test: Press belt at midpoint. Should deflect 2-3mm with moderate finger pressure.

Z-Axis: Triple-Z Leveling

Run Klipper’s Z_TILT_ADJUST to level the bed using the three Z motors.

G28                  ; Home all axes
Z_TILT_ADJUST        ; Auto-level using Triple-Z

Run 2-3 times until adjustments are minimal (< 0.01mm).

Verify Motion

Before printing, verify all axes move correctly:

G28                  ; Home
G1 X110 Y110 Z50 F3000  ; Move to center
G1 X10 F3000         ; Move X
G1 X210 F3000
G1 Y10 F3000         ; Move Y
G1 Y210 F3000
G1 Z10 F1000         ; Move Z
G1 Z100 F1000

Watch for: - Smooth motion (no grinding, skipping) - Correct direction (X+ moves right, Y+ moves back, Z+ moves up) - No binding at travel limits

Phase 2: First Layers

The first layer is the foundation of every print. Get this right before anything else.

Bed Mesh Calibration

Create a mesh map of bed surface variations:

G28                     ; Home
BED_MESH_CALIBRATE      ; Probe the bed
SAVE_CONFIG             ; Save mesh to printer.cfg

Re-run after any bed surface change (new G10 sheet, re-leveling, etc.).

Z-Offset Calibration

The Z-offset determines first layer squish. Too high = no adhesion. Too low = elephant foot.

Using paper method:

G28
PROBE_CALIBRATE        ; Start interactive calibration

Adjust until paper drags slightly under the nozzle. Save with SAVE_CONFIG.

Using baby-stepping during print: Start a first layer test and adjust Z live:

SET_GCODE_OFFSET Z_ADJUST=0.01 MOVE=1   ; Raise nozzle slightly
SET_GCODE_OFFSET Z_ADJUST=-0.01 MOVE=1  ; Lower nozzle slightly

First Layer Test Print

Print the first layer grid — a 5×5 grid of small squares across the bed.

What to look for:

Symptom Cause Fix
Lines not touching Z too high Lower Z-offset
Transparent/rough surface Z too high Lower Z-offset
Ridges between lines Z too low Raise Z-offset
Corners lifting Poor adhesion Clean bed, adjust temp, check level
Varies across bed Bed not flat Re-run BED_MESH_CALIBRATE

Good first layer: - Lines slightly overlap - Smooth top surface - Consistent across entire bed - Adheres without lifting

Phase 3: Dimensional Accuracy

Once first layers are good, verify dimensional accuracy.

Calibration Cube

Print a 20mm calibration cube (or threaded bolt cube).

Measure with calipers: - X dimension (should be 20.00mm) - Y dimension (should be 20.00mm) - Z dimension (should be 20.00mm)

If dimensions are off:

Issue Likely Cause Fix
All dimensions small Under-extrusion Increase flow rate / check extruder
All dimensions large Over-extrusion Decrease flow rate
X or Y off, Z correct Steps/mm wrong Calibrate steps/mm in Klipper
First layer elephant foot Z too low Raise Z-offset slightly

Flow Rate Calibration

Print a single-wall cube (1 perimeter, 0 infill, 0 top layers).

Measure wall thickness with calipers. Should match your line width (typically 0.4mm for 0.4mm nozzle).

Adjust flow rate:

New flow % = (Expected width / Measured width) × Current flow %

Example: If expecting 0.4mm but measuring 0.44mm with flow at 100%:

New flow = (0.4 / 0.44) × 100 = 91%

Threaded Bolt Cube (Optional)

If using the threaded bolt cube, also test: - Thread engagement with a real nut (M6 or M8) - If nut doesn’t thread on: check for over-extrusion or Z calibration

Phase 4: Klipper Tuning

These Klipper-specific calibrations significantly improve print quality.

Input Shaper (Resonance Compensation)

Input Shaper eliminates ringing/ghosting artifacts. This is one of Klipper’s superpowers.

Option A: Manual with ringing tower

  1. Print the ringing tower at moderate speed (80-100mm/s)

  2. Examine walls for ringing patterns

  3. Configure shaper in printer.cfg:

    [input_shaper]
shaper_freq_x: 40  ; Adjust based on test
shaper_freq_y: 40
shaper_type: mzv

Option B: Automatic with ADXL345 accelerometer (recommended)

If you have an ADXL345 accelerometer connected:

SHAPER_CALIBRATE      ; Automatic calibration

Klipper will measure resonances and suggest optimal settings.

Verify: Print ringing tower again. Ghosting should be eliminated or greatly reduced.

Pressure Advance

Pressure Advance compensates for filament pressure in the hotend, eliminating corner bulging and improving line consistency.

Calibration process:

  1. Use Klipper’s tuning tower:

    SET_VELOCITY_LIMIT SQUARE_CORNER_VELOCITY=1 ACCEL=500
TUNING_TOWER COMMAND=SET_PRESSURE_ADVANCE PARAMETER=ADVANCE START=0 FACTOR=.005

  2. Print the PA test pattern (square tower)

  3. Measure height where corners look best

  4. Calculate PA value:

    PA = START + (measured_height × FACTOR)

  5. Set in printer.cfg:

    [extruder]
pressure_advance: 0.05  ; Your calculated value

Typical values: | Filament | PA Range | |———-|———-| | PLA | 0.02 - 0.08 | | PETG | 0.05 - 0.10 | | TPU | 0.10 - 0.20 |

Phase 5: Filament Profiles

Each filament type (and brand) needs its own profile.

Temperature Tower

A temperature tower prints sections at different temperatures to find the optimal printing temp.

Method: Use Klipper’s SET_HEATER_TEMPERATURE with layer-based changes, or use your slicer’s temperature tower feature.

What to look for: | Temp too low | Temp too high | |————–|—————| | Poor layer adhesion | Stringing/oozing | | Weak parts | Blobbing | | Rough surface | Discoloration |

Typical ranges: | Filament | Range | Start Point | |———-|——-|————-| | PLA | 190-220°C | 205°C | | PETG | 230-250°C | 240°C | | TPU | 220-240°C | 230°C |

Retraction Tuning

Retraction prevents stringing during travel moves.

Print a retraction tower (multiple columns with travels between them).

Adjust: - Retraction distance: Start at 0.5mm (direct drive) or 4mm (Bowden) - Retraction speed: 25-45mm/s typical

PETG note: PETG is stringy. Use lower retraction than PLA (it’s more viscous) and enable “wipe” in your slicer.

TPU note: Minimal or zero retraction for flexible filaments — they compress and jam.

Per-Filament Settings

Create profiles in your slicer for each filament:

Setting PLA PETG TPU
Hotend temp 200-210°C 235-245°C 225-235°C
Bed temp 55-60°C 75-85°C 50-60°C
Cooling fan 100% 50-70% 50-100%
Retraction 0.5-1.0mm 0.3-0.8mm 0-0.5mm
Print speed 70-120mm/s 50-80mm/s 20-40mm/s
Pressure Advance 0.02-0.08 0.05-0.10 0.10-0.20

Phase 6: Validation Prints

Benchy

The 3D Benchy is the standard “torture test.” Print one to validate your calibration.

What to check: - Hull: Smooth, no layer lines visible - Bow overhang: Clean without drooping - Cabin roof: Flat bridging - Chimney: No stringing, round hole - Text: Readable on stern - Dimensions: Hull should be ~60mm long

Download: 3DBenchy.com

Victory Prints

Once Benchy passes, print your Amalgam victory prints:

  1. Amalgam Maker Coin — Octagonal coin with Logo 10
  2. Amalgam Fidget Bolt — Working threaded nut and bolt

These are your “I built an Amalgam” badges. If the threads work smoothly, your printer is well-calibrated.

Calibration Prints Reference

Print File Purpose
First Layer Grid cad/parts/first_layer_grid.py Bed level verification
Calibration Cube Standard 20mm XYZ cube Dimensional accuracy
Threaded Bolt Cube cad/parts/calibration_bolt_cube.py Dimensions + threads
Ringing Tower Input Shaper test Resonance tuning
Benchy External (3DBenchy.com) Comprehensive validation
Maker Coin cad/parts/amalgam_maker_coin.py Victory print
Fidget Bolt cad/parts/amalgam_fidget_bolt.py Victory print

Troubleshooting Quick Reference

Problem Likely Cause Solution
First layer doesn’t stick Z too high, dirty bed Lower Z-offset, clean with IPA
Elephant foot Z too low, bed too hot Raise Z-offset, lower bed temp
Layer shifts Loose belts, too fast Tighten belts, reduce speed/accel
Ghosting/ringing No Input Shaper Run SHAPER_CALIBRATE
Blobby corners No Pressure Advance Calibrate PA
Stringing Retraction, temp Tune retraction, lower temp
Under-extrusion Flow rate, clog Calibrate flow, check nozzle
Over-extrusion Flow rate too high Reduce flow rate
Warping Bed adhesion, drafts Brim, enclosure, bed temp
Z banding Lead screw issues Check coupling, lubricate

Maintenance Schedule

Interval Task
Every print Check first layer, remove debris
Weekly Clean bed with IPA, check belt tension
Monthly Lubricate Z lead screws, check for loose bolts
Quarterly Re-run BED_MESH_CALIBRATE, check bearing wear
Annually Full mechanical inspection, replace worn parts

References

“Calibration is not a destination, it’s a journey. But at some point, you should actually print something.”