A detailed STLBEAST repair guide to calibrate tower geometry, endstops, and bed plane systematically. Learn how to recognize the symptom, rank the likely causes, apply safe fixes in order, verify the result, and prevent the failure from returning.
Start hereStart with check joints and towers, then equalize belt tension. Confirm the result with a short representative test before changing additional settings.
First safe actionRun the first safe check and change one variable at a time.
Proof targetA small representative test no longer shows the original symptom.
Match the failed area to the visual before changing a setting.
Repair progress0 of 8 complete
Step 1 of 8
Document the failure and confirm that it matches this guide: Nozzle height or dimensions vary by position on a delta printer.
Keep the problem and target visual open while you make this change.
Why this matters
Each carriage starts from a different reference.
Setting or checkAcceleration
Return to a conservative baseline before diagnosing mechanical motion.
Expected result
The effector remains level and nozzle height is consistent across the build area.
Step 2 of 8
Return extreme overrides to a known printer, nozzle, material, and slicer profile so the diagnosis starts from a stable baseline.
Keep the problem and target visual open while you make this change.
Why this matters
Firmware geometry does not match hardware.
Setting or checkTravel speed
High speed can expose play, binding, or resonance.
Expected result
The same test succeeds at least twice without a new artifact appearing.
Step 3 of 8
Check tower/endstop offsets incorrect. Check joints and towers.
Keep the problem and target visual open while you make this change.
Why this matters
Carriages move unevenly.
Setting or checkInput shaping
Apply only after belts, frame, wheels, rails, and pulleys are sound.
Expected result
No safety warning, unusual noise, heater error, binding, or material damage is introduced by the change.
Step 4 of 8
Check delta radius or rod geometry mismatch. Equalize belt tension.
Keep the problem and target visual open while you make this change.
Why this matters
The nozzle position changes under load.
Setting or checkMotor current
Use only manufacturer-supported values; excess current creates heat.
Expected result
The successful values are recorded with printer, nozzle, material, slicer, and date.
Step 5 of 8
Inspect belt tension differs by tower. Run supported delta calibration.
Keep the problem and target visual open while you make this change.
Why this matters
Each carriage starts from a different reference.
Setting or checkAcceleration
Return to a conservative baseline before diagnosing mechanical motion.
Expected result
The effector remains level and nozzle height is consistent across the build area.
Step 6 of 8
Rule out effector or joints have play. Verify center and edge tests.
Keep the problem and target visual open while you make this change.
Why this matters
Firmware geometry does not match hardware.
Setting or checkTravel speed
High speed can expose play, binding, or resonance.
Expected result
The same test succeeds at least twice without a new artifact appearing.
Step 7 of 8
Change only the single setting or hardware condition supported by the evidence, then run a small test that reproduces the original failure.
Keep the problem and target visual open while you make this change.
Why this matters
Carriages move unevenly.
Setting or checkInput shaping
Apply only after belts, frame, wheels, rails, and pulleys are sound.
Expected result
No safety warning, unusual noise, heater error, binding, or material damage is introduced by the change.
Step 8 of 8
Compare the test against the target condition, record the successful value, and save it in a printer/material profile before repeating the full print.
Keep the problem and target visual open while you make this change.
Why this matters
The nozzle position changes under load.
Setting or checkMotor current
Use only manufacturer-supported values; excess current creates heat.
Expected result
The successful values are recorded with printer, nozzle, material, slicer, and date.
Controlled test log
Change one variable, then record what happened
This keeps temperature, retraction, speed, mechanical, and material tests from blending together. The history stays private in this browser.
Motion & MechanicsLoose Belt Symptoms and Adjustment
Motion & MechanicsV-Wheel Adjustment: Too Tight or Too Loose
Motion & MechanicsGantry Not Level or Sagging
