How do I calibrate X, Y, and Z steps on a 3D printer? (2026)
Quick Answer
To calibrate X, Y and Z steps on a 3D printer, first make sure the printer is mechanically sound, then measure actual axis travel against commanded travel, calculate a corrected steps-per-unit value, save it, and re-check. This only works properly if belts, pulleys, lead screws, wheels and frame alignment are already in good condition, because firmware changes cannot fix mechanical faults.
Overview
Calibrating X, Y and Z steps means checking whether each axis moves the distance the printer thinks it has moved, then correcting the motion setting in firmware if needed. On many hobby printers this is called steps per millimetre, but the exact name depends on the firmware. The process sounds simple, but it is only useful after basic mechanical issues have been ruled out. A loose pulley, slipping belt, bent lead screw or binding Z axis can give inconsistent results, and changing steps in that situation can make accuracy worse rather than better. The safest approach is to start with the manufacturer’s current settings, confirm that motion is smooth, home the printer, command a known movement on one axis at a time, measure the real movement carefully, then apply the standard correction formula using your printer’s present value. Repeat until the measured travel is acceptably close and consistent. For belt-driven X and Y axes, large errors often point to pulley or belt problems rather than a true calibration issue. For Z, errors may come from lead screw pitch settings, coupler issues or binding. If your printer uses fixed factory firmware values that are not intended to be changed, check the manufacturer’s guidance before editing anything.
Who this is for
Owners of FDM/FFF 3D printers who want to improve dimensional accuracy or verify motion settings after replacing motors, pulleys, belts, lead screws, mainboards or firmware.
What you’ll need
- 3D printer with access to movement controls or a terminal
- Printer manual or firmware documentation
- A reliable ruler, steel rule, calipers or a dial indicator
- Masking tape or a fine marker for making reference marks
- Paper or notes app for recording commanded and measured travel
- Access to save settings in firmware, if supported
Before you start
Check that belts are correctly tensioned, pulleys are tight on motor shafts, V-wheels or linear rails move smoothly, the frame is square, and Z lead screws are clean and turning freely. Home all axes first, disable any software settings that could interfere with straightforward movement tests, and confirm the printer is using the expected current motion settings from its manual or firmware readout. If your printer is under warranty or uses locked manufacturer profiles, confirm that changing steps-per-unit is allowed.
Step-by-step
- 1
Inspect the mechanics before changing settings
Power off the printer and manually check each axis for obvious play, slipping, binding or wobble. Inspect belt-driven X and Y axes for loose belts or pulleys, and inspect the Z axis for lead screw misalignment, coupler slip, dirt or uneven resistance. Tighten anything that is visibly loose according to the manufacturer’s instructions.
Why: If motion is inconsistent, any measurement you take will be unreliable. Firmware calibration should correct a consistent scale error, not compensate for mechanical faults.
- 2
Find and record the current axis motion values
Use the printer screen, control software or terminal to read the current X, Y and Z steps-per-unit values. Write them down exactly before making any changes. If your firmware supports storing settings separately from defaults, note both the active values and whether they have already been customised.
Why: You need the current value for the correction calculation, and keeping a record lets you return to the previous state if the result is worse.
- 3
Set up a clear measurement reference
Home the printer. Place a ruler, calipers or a dial indicator so you can measure travel parallel to the axis you are testing. Put a piece of tape on the moving part and make a fine reference mark, or align the tool to a fixed point you can read consistently. Test one axis at a time.
Why: A stable reference reduces reading error. Measuring from a repeatable start point is essential because small mistakes in setup can be larger than the calibration error itself.
- 4
Command a known move and measure the actual travel
Using the printer controls or terminal, command a moderate movement distance on the axis you are testing, making sure there is enough physical room to complete the move safely. Measure how far the axis actually travelled. Repeat the same test at least once more to check whether the result is consistent. Do this separately for X, Y and Z.
Why: A known commanded movement compared with actual movement shows whether the scale is correct. Repeating the test helps distinguish a real calibration error from a one-off slip or reading mistake.
- 5
Calculate the corrected steps value
Use the standard correction method: new steps-per-unit equals current steps-per-unit multiplied by commanded travel divided by actual measured travel. Apply that calculation only if the error is repeatable. If the error changes from test to test, stop and re-check the mechanics instead of forcing a firmware correction.
Why: This calculation proportionally adjusts the axis scale so the firmware’s idea of distance better matches real movement.
- 6
Enter and save the new value
Input the corrected value for the axis in your printer’s motion settings or through the appropriate firmware command, then save settings if your firmware requires a separate save action. Change only one axis at a time unless you are certain all your measurements are good.
Why: Saving ensures the printer keeps the correction after a restart. Updating one axis at a time makes it easier to isolate mistakes.
- 7
Verify with a second measurement pass
Home the printer again and repeat the same movement test on the adjusted axis. Confirm that the measured travel is now closer to the commanded travel and remains consistent over repeated moves. Then move on to the next axis.
Why: Verification is what proves the adjustment worked. A second pass catches typing errors, unsaved settings and hidden mechanical issues.
- 8
Finish with a practical print check
After all axes are checked, print a simple dimensional test object using normal print settings and measure the finished part. If dimensions are still off, investigate belt tension, backlash, extrusion calibration, shrinkage, slicer compensation or first-layer issues before changing axis steps again.
Why: Axis calibration affects motion accuracy, but printed part dimensions are influenced by more than motion alone. A print check tells you whether the remaining error comes from the printer or the printing process.
Why this works
Stepper-driven axes move in tiny increments determined by motor steps, microstepping, and the mechanics of belts, pulleys or lead screws. The firmware converts requested distance into a number of step pulses. If that conversion value is wrong, the printer will consistently under-travel or over-travel. Correcting the steps-per-unit aligns commanded motion with real motion, provided the hardware transmits movement consistently.
Common mistakes to avoid
- Trying to calibrate around loose belts, slipping pulleys or a binding Z axis
- Measuring from a poor reference point or at an angle to the axis
- Changing several settings at once and losing track of what caused the result
- Using printed cube dimensions alone to set X, Y or Z steps without first checking actual axis travel
- Forgetting to save settings after entering new values
- Adjusting belt-driven axes for large errors that actually indicate wrong pulley tooth count, wrong firmware defaults or mechanical slip
Troubleshooting
The measured travel changes each time you repeat the test
Check for belt slip, loose grub screws on pulleys, binding wheels or rails, inconsistent homing, and Z coupler or lead screw issues. Do not keep changing steps until motion is repeatable.
X or Y is off by a large amount
Verify that the printer is using the correct pulley, belt type and firmware defaults for that machine. Large errors on belt axes often mean the mechanical setup or firmware baseline is wrong, not that fine calibration is needed.
Z accuracy is poor even after changing the value
Inspect the lead screw, nut, coupler and alignment. Make sure the Z axis is not binding and that both sides, if dual Z is used, stay synchronised.
The printer reverts to the old values after restart
Use the firmware’s save settings function if available, or update the configuration in the printer’s firmware or manufacturer tool as instructed by the manual.
Printed parts are still dimensionally off although axis travel now measures correctly
Check extrusion flow, nozzle size assumptions, part cooling, material shrinkage, belt backlash, slicer horizontal expansion and first-layer squish. Motion calibration is only one part of dimensional accuracy.
Compare your options
Direct axis travel measurement
Best for: Setting or checking true motion calibration
Pros: Measures the axis itself, avoids mixing in extrusion and material effects, more suitable for steps-per-unit adjustment
Cons: Needs careful setup and a reliable measuring method
Printed calibration object
Best for: Checking overall dimensional accuracy after motion is verified
Pros: Shows real-world print results, easy for most users to do
Cons: Dimensions are affected by many factors besides axis steps, so it can mislead if used as the only calibration method
Dial indicator method
Best for: Users who want higher confidence in fine motion checks
Pros: More repeatable than reading a ruler by eye, useful for short precise moves
Cons: Requires extra equipment and careful mounting
| Option | Best for | Pros | Cons |
|---|---|---|---|
| Direct axis travel measurement | Setting or checking true motion calibration | Measures the axis itself, avoids mixing in extrusion and material effects, more suitable for steps-per-unit adjustment | Needs careful setup and a reliable measuring method |
| Printed calibration object | Checking overall dimensional accuracy after motion is verified | Shows real-world print results, easy for most users to do | Dimensions are affected by many factors besides axis steps, so it can mislead if used as the only calibration method |
| Dial indicator method | Users who want higher confidence in fine motion checks | More repeatable than reading a ruler by eye, useful for short precise moves | Requires extra equipment and careful mounting |
Alternatives
- Use the manufacturer’s official calibration routine or service menu if your printer provides one
- Restore factory motion settings after replacing parts, then verify travel instead of calculating from scratch
- If your printer is stock and measures consistently, leave the factory X and Y values alone and focus on mechanical tuning and slicer compensation
Pro tips
- Use a longer commanded move when space allows, because proportionally it reduces reading error
- Move the axis in the same direction when comparing readings to reduce backlash effects
- Record every original and revised value before saving changes
- If only one axis is inaccurate, compare it with a known-good matching axis before changing firmware
- After hardware upgrades, confirm the correct firmware profile first
Safety notes
- Keep fingers, tools and loose clothing clear of moving axes during powered tests
- Make sure the commanded move will not drive the carriage or bed into the frame limits
- Power off before tightening pulleys or inspecting wiring near motors
- Be cautious around a heated bed or hot end if the printer has recently been used
What this guide does not cover: This guide covers axis motion calibration for typical FDM/FFF printers in general terms. It does not provide model-specific menu paths, firmware commands, or exact default values, and it does not cover delta kinematics, CoreXY belt path diagnosis in depth, resin printers, or advanced geometric compensation.
Cost considerations
This job usually costs little if you already have a measuring tool, but repeated failed calibration attempts can waste filament and time. If the real issue is worn bearings, damaged belts, bent lead screws or incorrect replacement parts, fixing the hardware is usually better value than chasing settings.
Frequently asked questions
Should I calibrate X, Y and Z steps on every printer?+
Not necessarily. Many printers are already close when assembled correctly. It is most useful after replacing motion components, changing firmware, or if direct travel measurements show a consistent error.
Can I use a calibration cube to set axis steps?+
It is better to use direct motion measurement first. A printed cube includes effects from extrusion, material shrinkage, cooling and slicer settings, so it is not a clean measure of axis travel.
Why are my X and Y dimensions wrong even though the steps look right?+
Common causes include belt backlash, loose pulleys, too much or too little extrusion, slicer compensation settings, and material behaviour during printing.
Is it normal for Z to need more attention than X or Y?+
Often yes. Z commonly relies on lead screws and can be affected by binding, misalignment, coupler slip or dual-Z synchronisation issues.
Do I need to change firmware files, or can I do it from the printer screen?+
That depends on the printer and firmware. Some let you edit and save motion settings from the screen or a terminal, while others require a manufacturer tool or firmware update process.
Sources & references
Guidance on this page is traced to documented sources. Last checked 24 September 2026.
- Marlin Firmware Documentation · official
Supports the concept of steps-per-unit motion settings, terminal-based configuration, and saving settings in Marlin-based printers.
- Prusa Knowledge Base · manufacturer
Supports manufacturer-led calibration practice, the importance of mechanical condition, and the distinction between printer calibration and print quality issues.
- Duet3D Documentation · manufacturer
Supports axis movement configuration concepts on another widely used firmware platform and reinforces that motion values depend on hardware configuration.
The core method is stable, but exact menus, commands and whether settings are user-editable vary by printer model and firmware version.