LaserCanvas

Power, speed and passes, and how they trade off

What the material feels is energy per millimetre — roughly power divided by speed. Halving the speed does about what doubling the power does, so many combinations cut the same. Choose between them on edge quality: for cutting, prefer more passes at higher speed over one slow, hot pass.

Updated 2026-09-16

One number matters, and it is not power

Power sets how much energy arrives per second. Speed sets how long the beam spends over any given point. Neither alone determines the result — their ratio does.

PowerSpeedRelative energy per mm
60%600 mm/minbaseline
30%300 mm/minabout the same
60%300 mm/minabout double
100%1000 mm/minabout the same as the baseline, but gentler on the edge

The last row is the useful one. If two settings deliver the same energy, the faster one usually gives a better edge, because heat has less time to spread sideways into the material.

A straight line through the origin on axes of speed and power, with three equivalent settings marked along it Every point on this line cuts the same speed → power → 30% @ 300 60% @ 600 100% @ 1000 above the line: more energy, more char below: not enough to cut through
What the material feels is energy per millimetre, so power and speed trade off against each other. Any point on this line delivers roughly the same energy — but the faster ones char less, because heat has less time to spread sideways.

Which to change, and when

ProblemChangeWhy
Not cutting throughAdd a pass firstEach pass clears debris so the next reaches deeper, without more heat at once
Edges charred and wideRaise speedLess dwell time, less sideways heating
Engrave too faintLower speed slightlyPower tends to widen the mark rather than deepen it
Engrave blotchyLower powerDots are merging into solid burn
Taking far too longRaise power and speed togetherKeeps the ratio, cuts the time
Flare-upsRaise speed, add air assistHeat is accumulating faster than it disperses

Why several fast passes beat one slow one

It is counterintuitive — the same total energy, delivered differently — but the difference in the finished edge is large.

The exception is acrylic, where a single pass gives the polished flame edge and extra passes cloud it. Wood wants multiple passes; acrylic wants one good one.

Minimum and maximum power

Many controllers expose both. They matter when the head slows down — at corners and on small curves, the machine decelerates, and at a fixed power that means more energy per millimetre exactly where you least want it. That is why corners burn darker than straights.

Finding the setting rather than guessing it

  1. Cut a test grid on the actual sheet: power along one axis, speed along the other.
  2. Find the swatch that just cuts through cleanly.
  3. Back off slightly from the extreme — a setting on the edge of working will fail on the next sheet.
  4. Write it down with the material, the thickness and the date.

Common questions

Should I increase power or decrease speed?

They are close to equivalent for how much energy goes in, so choose on edge quality. Running faster at higher power generally gives a cleaner edge than running slow at lower power, because heat has less time to spread sideways. If you are already at full power, the choice is between slowing down and adding a pass — usually add the pass.

Is it better to use multiple passes or one slow pass?

Multiple passes for cutting. Each pass clears debris so the next reaches deeper, heat dissipates between them, and the edge chars far less. Acrylic is the exception: a single pass gives the polished flame edge and extra passes cloud it.

Why are the corners of my cuts darker than the straight sections?

The head decelerates into corners, so at fixed power more energy lands per millimetre there. GRBL’s M4 dynamic power mode scales power with speed and largely removes it. Setting a sensible minimum power helps too, so the beam does nothing while the head is crawling.

Why do settings from someone else’s machine not work on mine?

Because a setting is one point on a curve that depends on optical power, lens, focus, air assist, the exact sheet and how clean the optics are. Even two machines of the same model differ. Use published settings as a starting point for a test grid, never as an answer.

Written by a human with AI assistance and reviewed before publication. Mistakes are still possible — treat settings as starting points and test on scrap.