24 calculators for calibrating a printer and planning filament: e-steps, rotation distance, flow rate, volumetric speed, shrinkage and more. Each one updates as you type and shows its formula, a worked example and its source.
Heat the nozzle, mark the filament 120 mm above the extruder, ask for 100 mm at a slow speed, then measure how far the mark moved.
new e-steps = current e-steps × length you asked for ÷ length that moved
Worked example. The printer is set to 93 steps/mm. You ask for 100 mm and 97 mm goes in: 93 × 100 ÷ 97 = 95.88 steps/mm.
Read the current value by sending M92 with no values (older firmware: M503).
Set it with M92 E and the new number, then save it with M500. Saving needs firmware built with settings storage (EEPROM); without it, the value is lost at power-off.
Measure the move, not the mark. If you marked at 120 mm and 23 mm is left, 97 mm moved.
Klipper uses rotation distance in place of steps per mm: the distance the axis or filament moves in one full turn of the motor (or of the last gear, when a gear ratio is set).
extruder: new rotation distance = current rotation distance × length that moved ÷ length you asked for
from steps per mm: rotation distance = full steps per rotation × microsteps ÷ steps per mm
belt: rotation distance = belt pitch × pulley teeth
leadscrew: rotation distance = screw pitch × number of separate threads
Worked example. Extruder: set to 7.5 mm, you ask for 50 mm and 48 mm moves: 7.5 × 48 ÷ 50 = 7.200. From steps: 200 × 16 ÷ 80 = 40. Belt: 2 mm pitch × 20 teeth = 40. Leadscrew: 2 mm pitch × 4 threads = 8.
The extruder test: mark the filament about 70 mm above the extruder, send G91 then G1 E50 F60, and measure. The slow speed matters: a faster one builds pressure and skews the result.
Round the new value to three decimal places. If the next test is still more than about 2 mm out, repeat.
It runs the other way to e-steps. Too little filament means a smaller rotation distance.
Geared drives: set the gear ratio in the config as well. This page does not work it out.
For firmware that counts steps: how many motor steps move the axis 1 mm.
belt: steps per mm = motor steps per turn × microsteps ÷ (belt pitch × pulley teeth)
leadscrew: steps per mm = motor steps per turn × microsteps ÷ screw lead
Worked example. A 1.8° motor (200 steps) at 16 microsteps on a 2 mm belt with a 20-tooth pulley: 200 × 16 ÷ (2 × 20) = 80 steps/mm. The same motor on a screw with an 8 mm lead: 200 × 16 ÷ 8 = 400 steps/mm.
These are the theoretical values, and for belts and screws they are normally right. If a printed part is the wrong size, look at flow and shrinkage before changing axis steps.
Print a hollow cube with a single wall (vase mode), then measure the wall.
new flow % = flow used for the test × wall width set in the slicer ÷ measured wall
Worked example. The slicer’s extrusion width is 0.45 mm, the cube was printed at 100 % and the walls average 0.48 mm: 100 × 0.45 ÷ 0.48 = 93.8 %, or a multiplier of 0.938.
Use the width from the slicer, not the nozzle size. The two are often different.
Measure the middle of each wall at least three times and average them. A micrometer is better than low-cost calipers for a wall this thin.
Calibrate per filament. Flow changes with the material and sometimes the colour.
maximum speed = maximum flow ÷ (layer height × line width)
flow from a test = filament speed × π × (filament diameter ÷ 2)²
Worked example. 0.2 mm layers, a 0.45 mm line and 100 mm/s: 0.2 × 0.45 × 100 = 9 mm³/s. A hotend that manages 12 mm³/s tops out at 12 ÷ (0.2 × 0.45) = 133 mm/s.
We don’t list a figure for your hotend. It depends on the hotend, nozzle, material and temperature, so it is an input here.
To measure it: mark 100 mm of filament, extrude 100 mm into the air at rising speeds, and note the fastest speed at which the full 100 mm still goes in. Put that speed in the third box, then set your slicer a little below the result.
The rectangle is a small overestimate. Slicers model a line with rounded ends, so the real flow is slightly lower than layer height × width × speed.
Signs you are over the limit: a clicking extruder, thin lines at speed and weak layers.
Densities vary between makers and blends, and filled filaments differ most. Use the figure on your spool’s data sheet when you have it. Empty spools have no standard weight: weigh one of the same make, or weigh a new spool and subtract the filament weight on its label.
07Shrinkage and scale compensation
Print a simple test piece, let it cool fully, and measure it. The scale that corrects it is the designed size over the measured size.
scale % = designed size ÷ measured size × 100, for each axis
hole compensation per side = (designed diameter − measured diameter) ÷ 2
Worked example. A 100 mm part measures 99.6 mm: 100 ÷ 99.6 × 100 = 100.40 %. A 5 mm hole measures 4.8 mm: (5 − 4.8) ÷ 2 = 0.10 mm per side.
Check which number your slicer wants. Some shrinkage settings ask for the printed size as a percentage (99.60 % here), not the scale. Both are shown.
Hole settings usually move the outline, so the hole’s diameter changes by twice the value. Positive values make holes bigger. Print the test again to confirm.
Fix flow first. Over-extrusion also makes holes small and outsides large.
A plucked belt behaves roughly like a guitar string: its note depends on the free length, the tension and how heavy the belt is.
frequency = (1 ÷ (2 × span)) × √(tension ÷ mass per metre)
tension = 4 × mass per metre × span² × frequency²
Span in metres, tension in newtons, mass per metre in kg/m, frequency in hertz. The boxes take the span in mm and convert it.
Worked example. The numbers here only show the arithmetic. A belt of 0.010 kg/m with a 150 mm (0.15 m) free span at 20 N: (1 ÷ 0.3) × √(20 ÷ 0.010) = 149.1 Hz.
Belt mass is an input. We found no belt maker’s data sheet giving mass per metre for a 6 mm wide, 2 mm pitch belt, and figures quoted online differ by more than double. Weigh a known length of your own belt and use the third box.
No target is given here. Use the frequency or tension your printer’s maker publishes.
It is an estimate. A real belt is stiffer than an ideal string, so treat the result as a guide and compare both belts of a pair.
A Z motor holds its position best on full steps. Layer heights that are whole multiples of one full step avoid relying on microstep positions.
one full step = leadscrew lead ÷ motor steps per turn
full-step layer heights = 1, 2, 3 … × one full step
Worked example. An 8 mm lead screw on a 1.8° motor (200 steps): 8 ÷ 200 = 0.04 mm. So 0.08, 0.12, 0.16, 0.20, 0.24 and 0.28 mm land on full steps, and 0.10 or 0.30 mm do not.
This is a refinement, not a rule. Microstepping does hold the positions in between, and many printers show no visible difference. It matters most with inch-size screws and unusual layer heights. Belt-driven Z axes need the steps-per-mm figure instead.
Rules of thumb, not limits your printer enforces. They are a starting range for a new nozzle size.
layer height: about 25 % to 80 % of the nozzle diameter
line width: about 100 % to 150 % of the nozzle diameter
Worked example. A 0.4 mm nozzle: layers from 0.25 × 0.4 = 0.10 mm to 0.8 × 0.4 = 0.32 mm, and lines from 0.40 mm to 1.5 × 0.4 = 0.60 mm.
Layer height: one filament maker’s guide gives about 25 to 75 % of the nozzle as the reliable range; one slicer maker says to stay below 80 %, and that a layer can never be taller than the nozzle.
Line width: one slicer’s guide gives 100 % of the nozzle as a good start, warns that narrower lines can bond poorly, and that above 150 % can over-extrude.
Print time comes from your slicer; it depends on acceleration and travel moves, which no simple formula captures. For the money side, and for drying settings, use these.
Extrude slowly for extruder tests. Fast extrusion builds pressure in the hotend and shortens the measured length.
Measure more than once. Take at least three readings and use the average.
Let parts cool to room temperature before measuring size. Plastic shrinks as it cools.
Use the right tool. Calipers are fine for a 100 mm part; a 0.45 mm wall needs a micrometer or very careful caliper work.
Change one thing at a time, then print the test again to confirm it.
Questions, answered
Are these calculators free?
Yes. They are free, need no account and run in your browser.
E-steps or rotation distance: which do I need?
It depends on your printer’s firmware. Marlin and firmware based on it use steps per mm, set with M92. Klipper uses rotation distance in its config file. They describe the same thing, and the page converts between them.
Why does the Klipper formula look upside down?
Steps per mm goes up to move more filament; rotation distance goes down. So e-steps multiplies by asked ÷ moved, and rotation distance multiplies by moved ÷ asked.
What is a good maximum volumetric speed?
There is no single figure. It depends on your hotend, nozzle, filament and temperature, so the calculator asks for it. The volumetric speed section explains a quick way to measure your own.
What frequency should my belts be?
Use the figure your printer’s maker publishes. The calculator only converts between frequency and tension for the span and belt mass you enter; it does not recommend a tension.
Do I have to use “magic number” layer heights?
No. They keep each layer on a full motor step, which can help on some printers, but microstepping holds the positions in between and many printers show no difference.
Is anything saved or uploaded?
No. Nothing is sent anywhere. The values live in the page, and in the link if you copy one.
Sources
Print advice is a starting point drawn from these references, checked on 2026-11-11. Community posts are one maker's experience, not a specification.