Laser Rotary Attachment Sizing: Diameter Range, Step Calibration and Tumbler Fit
Roller rotaries suit straight-walled items between roughly 20 mm and 120 mm; chuck rotaries handle tapered tumblers, stemware and anything under 20 mm that a roller would simply spin past. Calibrate in two steps — a full 360-degree rotation test to fix steps-per-rotation, then an arc-length test to fix the entered diameter — because a single test cannot tell those two errors apart.
The wrong rotary isn't a slightly worse rotary — it's a rotary that physically cannot hold your workpiece square to the beam. A roller unit will happily accept a 30 oz tapered tumbler and then engrave a logo that climbs uphill, because the item is sitting on two parallel shafts and the taper tilts it.
Geometry decides the purchase. Diameter range decides whether it fits at all, taper decides whether it sits level, and the drive ratio decides how much calibration work you're in for. Price barely enters it.
Pick by geometry, not by price
Rotary type Workpiece diameter range Taper tolerated Typical steps per workpiece rotation Holds well Fails at Two-shaft roller, fixed spacing 25-100 mm Under 1 degree without shimming Depends on roller diameter — commonly 8,000-14,000 Straight-walled tumblers, can coolers, cylindrical bottles Anything tapered, anything with a handle, small diameters that fall between the shafts Roller with adjustable shaft spacing 20-130 mm 1-2 degrees with a shim under one end 8,000-16,000 Wine bottles, larger tumblers, wide mugs Slippery powder-coated surfaces at speed; sub-20 mm items Three-jaw self-centring chuck 6-90 mm gripped, larger with jaw extensions Any — the taper hangs free 6,400-20,000, gear-dependent Tapered tumblers, stemware, pens, small bottles Very long items without a tailstock; thin glass if you overtighten Four-jaw independent chuck 5-100 mm Any 6,400-20,000 Square and irregular sections, off-centre work Speed — every job needs manual centring Collet or pen holder 3-16 mm None 3,200-12,800 Pens, vape bodies, ferrules, drill blanks Anything you can't push into a collet Roller plus tailstock cone 25-120 mm 2-3 degrees 8,000-16,000 Long tapered tumblers and flasks Items with no open mouth for the cone to grip Match the row to the widest and narrowest items you actually sell. If both a 12 mm pen and a 95 mm tumbler are on the list, that's two attachments or one chuck with jaw extensions — no single roller covers it. The honest recommendation for most people: buy the chuck. It costs more, it's slower to load, and it handles roughly everything. Roller units are quicker for a production run of identical straight-sided items, which is a real use case but a narrower one than the marketing suggests.
Measure at the band, not at the widest point
Software asks for a diameter. It means the diameter at the height where the engraving will sit, and on a tapered vessel that's rarely where you instinctively measure.
A 20 oz skinny tumbler might be 73 mm at the lip and 63 mm at the base. Enter 73 and engrave a design centred 90 mm down and the artwork comes out compressed along the circumference by around 8%, which on a 100 mm wide design is 8 mm of visible squash. Calipers at the band, every time.
For anything with more than about a degree of taper, level the item so the engraving band is horizontal under the lens — a shim under the narrow end on a roller, or simply letting the chuck hold the base while the taper hangs in free air.
Two calibration tests, in this order
Nearly everyone runs one test, gets a wrong answer, adjusts the wrong parameter, and spends an evening chasing their tail. There are two independent errors — steps-per-rotation and entered diameter — and one measurement cannot separate them.
- Test one, full rotation. Put a strip of masking tape across the workpiece and draw a fine pen line along the axis. Command exactly one full rotation at low speed with the laser off. If the line comes back to the same place, your steps-per-rotation is right, whatever diameter you entered. If it stops short or overshoots, correct steps-per-rotation by the fraction of a turn you missed and repeat.
- Test two, arc length. Only once test one passes. Engrave a straight 100 mm line running around the circumference at low power. Measure the burnt line with a flexible rule or a strip of tape you then lay flat.
- If the measured line is longer than 100 mm, the diameter you entered is smaller than the true diameter at that band. Multiply your entered diameter by measured ÷ 100 and re-run.
- If your controller doesn't let you enter a diameter and only exposes steps-per-rotation, use the arc test to correct steps instead — the table below does that arithmetic.
- Write the final numbers on the item type, not on the machine. A 30 oz tumbler and a wine bottle need different entries and you'll be swapping between them all week.
Reading the arc test
For a target line of 100 mm with 12,800 steps per rotation entered, here's what each measurement means and what to type in next.
Measured line length What it means Corrected steps per rotation Or, if steps are right, corrected diameter (from 73 mm entered) Visible error on a 100 mm design 96 mm Machine under-rotated by 4% 13,333 70.1 mm 4 mm too narrow 98 mm Under-rotated by 2% 13,061 71.5 mm 2 mm too narrow 100 mm Correct 12,800 — no change 73.0 mm — no change None 102 mm Over-rotated by 2% 12,549 74.5 mm 2 mm too wide 104 mm Over-rotated by 4% 12,308 75.9 mm 4 mm too wide 110 mm Over-rotated by 10% 11,636 80.3 mm 10 mm too wide, obvious to the eye Scale your own numbers the same way: new steps = old steps x (target ÷ measured). Do the full-rotation test first, or you'll be applying this correction to the wrong parameter. A 2% error is roughly the point at which a customer notices a circular logo has gone slightly oval. A 4% error is unmistakable. It's worth getting this right once per vessel type and keeping the numbers.
Fit table for the vessels people actually engrave
Dimensions vary by brand, so these are the bands you'll find across common stock rather than any one manufacturer's spec. Measure your own before you cut.
Vessel Diameter at the engraving band Height / usable band Taper Best rotary Watch out for 12 oz slim can cooler 62-66 mm 120 mm, band 80 mm None Roller Powder coat can slip on polished rollers — add a rubber band 20 oz skinny tumbler 70-75 mm at lip, 60-65 mm at base 215 mm, band 120 mm 1.5-2 degrees Chuck, or roller with a shim Measure at the band — the lip figure is 10 mm out 30 oz tumbler 88-95 mm at lip, 68-72 mm at base 200 mm, band 110 mm 2.5-3 degrees Chuck Too much taper for a roller to hold level 40 oz tumbler with handle 90-100 mm 280 mm, band 140 mm 2-3 degrees Chuck with the handle clocked out of the beam path Handle mass makes it rotate unevenly; slow the speed Conical pint glass 80-90 mm rim, 60-68 mm base 150 mm, band 90 mm 3-4 degrees Chuck on the base Thin glass chips — reduce power, add damp paper Stemless wine glass 75-85 mm 120 mm, band 70 mm Curved wall, not a straight taper Chuck, engraving only the flattest 40 mm Curvature takes the surface out of focus fast 750 ml wine bottle 74-78 mm at the body 300-330 mm, band 100 mm None on the body Roller with tailstock, or chuck on the neck Weight; keep the speed down and support the far end 330 ml beer bottle 60-62 mm 230 mm, band 90 mm None Roller Embossed logos break focus — check for raised glass 16 oz mason jar 85-88 mm over the threads 170 mm, band 90 mm None Roller or chuck Thread ridge lifts the item on a roller — sit it on the smooth section Pen or vape body 8-16 mm 140 mm, band 60 mm Slight Collet or chuck with soft jaws Too small for roller shafts — it drops between them Ring 15-25 mm inside 6-10 mm band None Ring cone in a chuck Curvature over a narrow band; a short focal lens helps Use the diameter column as your software input, not the manufacturer's advertised capacity. Anything in this table with more than about a degree of taper is a chuck job. Focus on a curve
The top of a cylinder is a single line, and everything either side of it curves away. On a 73 mm tumbler, a 40 mm wide engraving drops roughly 2.8 mm below focus at its edges — enough to visibly lighten the ends of a design that's crisp in the middle.
Three ways out, in order of how much they cost you. Keep the engraving band narrow, ideally under 30 mm on anything below 80 mm diameter. Fit a longer focal length lens, because the deeper focus window absorbs the curvature. Or split a wide design into segments and rotate between them, which is fiddly but produces the cleanest result on wraparound artwork.
Set focus to the top of the workpiece, not to the bed, and re-set it every time you change vessel diameter. That last point catches everyone at least once.
What goes wrong the first time
- Bed not lowered. The rotary raises the workpiece 60-100 mm, and the gantry hits it on the first move. Check clearance manually before you press start.
- Y axis not disabled. On most controllers the rotary replaces the Y motor, and if the gantry tries to move Y at the same time the design shears diagonally.
- Design longer than the circumference. Software will happily send a 250 mm design onto a 73 mm tumbler and wrap it over itself two and a half times.
- Workpiece slipping on the rollers. Powder-coated and glossy surfaces slide under acceleration. Slow the acceleration setting, not just the speed.
- Seam positioned under the beam. Rotate the vessel so the moulding seam or the printed logo sits away from the engraving band.
- Forgetting to reverse the direction. Half of all first attempts come out mirrored because the rotary spins the opposite way to the software's assumption. Test on a can cooler.
When a rotary is the wrong tool entirely
Flat-sided items don't need one. A hip flask, a rectangular tin or a lighter engraves better clamped flat on the bed, and a rotary just introduces a rotation error you didn't need.
Very short bands on very large diameters are another case. If you're marking a 40 mm logo on a 150 mm diameter drum, the surface across that band is nearly flat and a jig plus a longer lens will beat a rotary on both time and quality.
And if you're doing one item, once, the twenty minutes of setup and calibration will exceed the engraving time by a wide margin. That's fine — just go in knowing it.