Back to roadmap
Print quality and tuningIntermediate16 min

Speed, acceleration and input shaping

Find the melt-rate ceiling, see why short moves never reach top speed, then watch ringing appear and disappear as you change acceleration and the input shaper.

Speed has a ceiling: the melt rate

A hotend can only melt so much plastic per second. Multiply speed, line width and layer height to see what you are asking of it.

Hotend
Melt rate demanded12.6 / 12 mm³/s
Over the limit

The hotend cannot keep up. The extruder may click or skip, and lines under-extrude. Lower the speed or raise the temperature a little.

Highest speed at this setup

143 mm/s

Melt rate (mm³/s) = line width × layer height × speed.

Typical limits for PLA: roughly 10 to 15 mm³/s for a standard hotend and 20 or more for a high-flow one. Find your own with a max flow test.

Max volumetric speed test: measure where the print starts to thin out, then convert that speed to mm³/s.

Why short moves never reach top speed

The head has to accelerate before it can cruise and slow down before it stops. On small features that is all it ever does.

Time (ms)mm/s5000
  • Head speed
  • Top speed setting

Time

167 ms

Peak speed

200 mm/s

Cruising

33%

This move is long enough to reach top speed and cruise. Both top speed and acceleration matter here.

Ideal straight-line model. Real printers also smooth corners and have other limits.

Ringing, and how input shaping removes it

Every sharp corner makes the head change direction, which shakes the frame. The ripples that follow are called ringing. Change speed, acceleration and the frame's stiffness, then switch input shaping on.

Ripple depth

75 µm

Ripple spacing

3.3 mm

Visible ringing

Obvious ripples. Lower the acceleration, stiffen the frame, or tune input shaping to the printer's real frequency.

Teaching model of one dominant resonance with a ZV shaper. Depth is estimated from acceleration and frequency; real printers have several resonances.

Ringing on a real test print: ripples start at the corner and fade. The wavier the lines, the stronger the vibration.
An input shaping test: each band is printed with a different shaper setting. Find the band where the ripples nearly vanish.

Measuring your printer's frequency

Input shaping works only when its frequency matches your printer. There are three ways to find it.

1

Accelerometer

Klipper can read a small accelerometer on the head and bed, shake the printer, and pick a shaper for you. It is the most accurate method.

2

Print a test

Orca and Marlin can print a test piece whose bands use different settings. You measure the ripples with calipers or by eye and pick the best band.

3

Let the printer do it

Some printers, such as Bambu Lab models, run a vibration compensation check at start-up and set it for you.

Which shaper?

ShaperTrade-off
ZVSmallest smoothing of corners, but the most sensitive to a wrong frequency.
MZVA balanced default: good reduction with little smoothing and forgiving of small errors.
EIRemoves more vibration and tolerates larger frequency errors, at the cost of rounder corners.

A different problem: fine vertical banding at certain speeds (VFA). It comes from motor and belt vibrations, and a speed test shows which speeds to avoid.

Stronger shaping rounds corners slightly, so high-detail prints may prefer a lighter shaper and lower acceleration instead of the strongest one.

The VFA test: a star with tall walls whose print speed changes with height. Fine vertical banding at some speeds points to resonance.

Check yourself

Pick an answer for each question.

1. You raise the speed to 400 mm/s but the plastic thins out. What is the likely limit?

2. Why does doubling the top speed barely shorten a print of tiny details?

3. What does input shaping do?

4. After enabling input shaping, ringing is still visible. What is the first thing to check?

Sources (5)