Wood Lathe Speed Calculator: Safe RPM by Blank Diameter and Operation
Multiply blank diameter in inches by RPM and keep the product between 6,000 and 9,000 for general turning, 4,000 to 6,000 for roughing an out-of-round blank, and under 3,000 for anything cracked. A 12in bowl blank therefore roughs at about 400 RPM and finishes near 750. Below half its belt-range top speed, a variable-speed head has already given away half its usable power.
Take your blank diameter in inches and multiply it by the spindle speed. Keep that product between 6,000 and 9,000 for general turning, pull it back to 4,000-6,000 while a blank is still out of round, and stay under 3,000 for anything cracked, bark-edged or noticeably heavier on one side.
That single constant is the whole chart. A 12in bowl blank roughs at about 400 RPM, shapes near 580 and takes a finishing cut around 750. An 8in blank does the same jobs at roughly 625, 875 and 1,125.
What no lathe manual mentions is the second half of the problem. The dial says 400 RPM and the lathe honestly is turning 400 RPM — but if you got there by winding an electronic variable-speed head down while leaving the belt in its high range, you are now working with about a quarter of the horsepower on the badge. The speed is right and the torque has quietly gone.
The speed chart
Figures below are rounded to something you can actually dial. Where a small blank calls for a speed above 3,000 RPM the table caps it, because most home-shop lathes top out between 3,000 and 4,000 and there is nothing to gain from chasing the ceiling on a 1in spindle.
Blank diameter Roughing RPM Shaping RPM Finishing cut RPM Sanding RPM Ceiling for a cracked or out-of-balance blank 1 in 2,500 3,000 3,000 2,000 Not a concern at this size 2 in 2,400 3,000 3,000 2,000 1,500 3 in 1,650 2,300 3,000 1,350 1,000 4 in 1,250 1,750 2,250 1,000 750 6 in 830 1,150 1,500 670 500 8 in 625 875 1,125 500 375 10 in 500 700 900 400 300 12 in 415 580 750 335 250 14 in 355 500 640 285 215 16 in 310 435 560 250 190 18 in 275 390 500 220 165 20 in 250 350 450 200 150 24 in 210 290 375 165 125 Start in the roughing column, move right only once the blank runs true and quiet. If the blank has a check, a void or bark on one side, the last column is a hard limit and not a suggestion. Read the sanding column carefully, because it surprises people. Sanding speed is lower than finishing speed, not higher. Abrasive on wood at high surface speed generates heat that opens heat checks in end grain and glazes the paper in seconds.
Where those numbers come from
The underlying quantity is surface speed — how fast the wood passes the tool edge, in feet per minute. Surface speed equals pi times diameter in feet times RPM. A 12in blank at 750 RPM presents roughly 2,360 surface feet per minute at the rim, which sits comfortably in the band where a bowl gouge shears cleanly rather than tearing.
The diameter-times-RPM shortcut exists because dividing by 12 and multiplying by pi gives you a constant of roughly 0.26, so a product of 9,000 lands at about 2,350 SFPM. Turners have used the shortcut for decades because it is arithmetic you can do at the lathe with cold hands.
Two practical consequences fall out of it.
- Speed must come down as diameter goes up, and it comes down fast. Doubling the diameter halves the speed for the same surface rate.
- The rim of a 16in bowl at 400 RPM is travelling at about 1,675 SFPM. The centre is travelling at zero. That is why the same gouge that shears beautifully at the rim tears at the base, and why you change technique rather than speed as you work inward.
Out-of-balance blanks get their own ceiling
Centrifugal force rises with the square of the speed. Double the RPM on an unbalanced blank and the force trying to throw it off the lathe goes up fourfold, not twofold.
This is why the last column exists. A green blank with bark on one edge, a cracked burl, or anything mounted off centre wants the diameter-times-RPM product held near 3,000. On a 16in blank that is 190 RPM, which feels absurdly slow until the lathe walks across the floor at 400.
The other honest answer is bolt the lathe down or load the base. Most benchtop and mid-size lathes ship with a base weighing far less than the blank you are about to hang off it. Sandbags in the stand, a shelf of concrete blocks, or lag bolts into the floor do more for safe low-speed roughing than any spec on the machine.
Your dial reading is not your torque
Horsepower equals torque times RPM divided by 5,252. Hold that equation still and the consequence is unavoidable: if a motor is holding constant torque and you halve its speed, you have halved its power. Electronic variable-speed lathes run their motors in exactly that constant-torque mode below base speed, which is why the belt has ranges at all.
Dial speed as a share of that belt range's top Share of rated HP still available What it feels like at the gouge What to do 100 percent Full rated output Motor holds pitch through a heavy roughing cut Nothing 75 percent About 75 percent Barely perceptible drop on a deep cut Nothing 50 percent About 50 percent Motor sags on a catch instead of shrugging it off Fine for finishing cuts, marginal for roughing 33 percent About 33 percent A 1.5 HP lathe now behaves like a half-horse one Drop the belt a range and dial the speed back up 25 percent About 25 percent A moderate bite with a roughing gouge stops the spindle The belt has to come down 15 percent About 15 percent, plus cooling problems Bogs on the first pass, and the motor heats with no fan speed to shed it Never work here for more than a minute Work out roughly where your dial sits within its current belt range, then read across. If you are below half, move the belt rather than fighting the wood. The fix is free and takes ninety seconds. Every variable-speed lathe with a belt has two or three ranges precisely so you can get low RPM at high motor speed. Turners skip the belt change because opening the headstock is a nuisance, then conclude their lathe is underpowered. It is not underpowered. It is in the wrong gear.
Reeves drives fail differently
Mechanical variable speed — a Reeves drive, where sprung sheaves open and close to change effective pulley diameter — has a different derating shape. It does not lose torque the way a VFD does, but it loses grip.
Symptoms worth knowing. At the extreme low end the belt rides deep in one sheave and shallow in the other, contact area drops, and the belt slips under load rather than stalling the motor. That slip is silent from the operator position and shows up as a spindle that is turning slower than the indicator claims, sometimes by 15 percent or more.
Never change speed on a Reeves drive while the lathe is stopped. The sheaves need to be rotating to reposition, and forcing the handwheel against a stationary belt is the single most common way these drives get wrecked. A drive that has been abused this way develops a rough band in the middle of its range where the speed hunts on its own.
The vibration test beats any chart
Charts give you a starting point. The blank gives you the answer.
Set a coin on the lathe bed — a wide flat coin on edge is even better if you have the patience. Bring the speed up gradually through the roughing band. The moment the coin walks or falls, you have found the resonant speed of that blank on that machine at that mount, and you back off 15 to 20 percent from it.
Resonance is specific, not general. The same 12in blank can be smooth at 400 RPM, alarming at 480, and smooth again at 550. If a speed feels wrong, the correct response is to change it rather than to work through it, and going faster is sometimes the fix. Trust the coin over the table.
Trueness Index: what a lathe spec sheet leaves out
Three published lathe numbers routinely fail in a working shop, and they are the ones the Trueness Index scores.
Swing over bed is measured to the bed ways, so a 16in swing lathe genuinely takes a 16in blank — right up until the tool rest banjo is in the way, at which point the usable figure is smaller. Stated horsepower is measured at base speed, which as the derating table shows is not where bowl work happens. And variable-speed range is quoted end to end across all belt positions, so a lathe advertised as 50-3,200 RPM may only offer 50-400 in its low range and 500-3,200 in its high, with a gap in the middle nobody mentions.
Runout at the spindle nose is the fourth. Anything past 0.001in on the taper will show up as a chattering surface on a long spindle and gets blamed on technique for years.