The pulley speed formula, and which diameter goes into it

For an open belt run between two pulleys, the belt moves at the same surface speed everywhere. Surface speed equals circumference times rotational speed: v = π · D · rpm. Setting the driver’s and driven pulley’s surface speeds equal and canceling π gives the working formula:

rpm₂ = rpm₁ × D₁ ÷ D₂

The ratio version is just as useful in the other direction: D₂ = D₁ × rpm₁ ÷ rpm₂ — that is the “Solve driven diameter” mode above, and it is how you go from a target speed to an orderable part. Both forms assume no slip; real V-belt drives creep about 1–2%, so treat computed speeds as the no-load figure and expect the tachometer to read a hair lower under power.

The single biggest source of error in this formula is feeding it the wrong diameter. As covered in the pulley measurement guide, the formulas want pitch diameters. For a B-section sheave the pitch diameter is the measured OD minus 7.4 mm (0.29 in); for A it is OD − 5.6 mm; for 3V, OD − 5.6 mm as well. Using OD instead of PD skews each pulley’s contribution by roughly 3–6%, and the errors partially cancel in the ratio — which is exactly why the mistake survives so long unnoticed. It stops canceling the moment the two pulleys carry different belt sections, or the moment you feed the same wrong diameters into a belt-length calculation.

Why the bigger pulley always turns slower — and what you gain

Speed and torque trade off along the belt. Power in ≈ power out (minus a few percent of loss), and power is torque times rotational speed, so whatever factor the RPM drops by, the available torque rises by nearly the same factor. A 6 in pulley driven by a 2 in driver at 1750 RPM turns at 583 RPM and delivers roughly 3× the motor’s torque at that shaft (less slip and bearing losses). This is why the reduction stage sits close to the load: you want the slow, strong shaft at the workpiece, not the fast, weak one.

The belt itself does not care which pulley is bigger — it travels at one speed. That speed is set by whichever pulley it is wrapped around, and both pulleys see the same belt speed. This is the quantity the ceiling check in the calculator watches: a belt is limited by how fast material can flex in and out of its grooves and by centrifugal loading, not by RPM. A 2 in pulley at 3450 RPM and a 12 in pulley at 575 RPM put exactly the same 1806 ft/min through the belt — the same flexing duty per second, the same centrifugal load, the same life expectancy.

There is a floor on how small the driver can be, too. Every belt section has a minimum recommended pitch diameter: below it, the belt bends too sharply, runs hot, and sheds life rapidly. Classical A-section wants roughly 2.6 in (66 mm) PD and B-section roughly 3.7 in (94 mm) as round practical floors, with narrow sections tolerating somewhat smaller drivers. If the solved driver diameter comes out below your section’s floor, the fix is a larger driver and larger driven pulley at the same ratio, not a smaller driver.

The belt-speed ceiling: 30 m/s for V-belts

The engine used across this site — including the homepage calculator and the multi-pulley solver — flags any V-belt drive whose belt surface speed exceeds 30 m/s (about 5,905 ft/min). The limit exists because centrifugal force grows with the square of belt speed: at the ceiling, the belt’s own mass flings it outward hard enough to visibly unload the wedging contact in the groove. Power capacity does not fall off politely past this point; it drops steeply, and the first symptom is usually slipping under a load the drive handled comfortably at lower speed.

Practical implications by scenario: a 3450 RPM motor on a 2 in driver runs the belt at 1806 ft/min (9.2 m/s) — comfortable. The same motor needs only a 6.4 in driver to breach 5,900 ft/min, and small pulleys spinning at very high RPM (small engine mods, spindle upgrades) hit the ceiling long before anything else in the drive looks wrong. If the calculator above shows the ceiling breached, your options in order of preference are: reduce driver RPM at the source, move the reduction into a jackshaft with two modest stages, or switch to a belt type rated for the speed. Timing belts and flat belts have different (generally higher) speed envelopes, but their own limits still apply — “the calculator said 40 m/s is fine for the geometry” never means it is fine for the belt.

Quick-reference: common motors × common pulley pairs

The two tables below are pure arithmetic from the formula — driven RPM = motor RPM × D₁ ÷ D₂ — so they are exact for the diameters listed and pitch-diameter-correct if you treat the diameters as PD. They exist so you can sanity-check an off-the-shelf sheave purchase in ten seconds without opening the calculator.

Belt surface speed from a single wheel (sfpm)

Belt grinders and contact-wheel drives read better in surface speed than RPM, because the abrasive or belt rating is specified in sfpm. The table below converts contact-wheel diameter and motor RPM directly — sfpm = π × D × rpm ÷ 12 — so a 2 in, 2.5 in or 4 in contact wheel on a standard motor can be compared at a glance.

Read the pair of tables together and the design pattern falls out: doubling wheel diameter doubles belt speed at fixed RPM, and doubling motor RPM does the same. To double belt speed without breaching the belt ceiling, you have exactly those two levers — or the compound route described next, which changes neither and instead trades torque for speed.

Jackshafts and compound ratios

When one stage cannot bridge the gap — either the ratio is too large for a single pair of sheaves, or the single-stage belt speed breaches the ceiling — a jackshaft splits the reduction. Stage ratios multiply: a shaft carrying an 8 in driven pulley fed by a 2 in motor pulley reduces 4:1; the same shaft then driving a 3 in pulley that feeds a 6 in final pulley reduces 2:1; the total is 8:1. A 1750 RPM motor ends at 219 RPM, and each stage’s belt speed stays modest even though the compound ratio would have required an impractically huge final pulley to achieve in one step.

Two rules keep compound drives honest. First, compute each stage’s belt speed separately with the stage’s driver RPM — the second stage’s driver spins at 437 RPM in the example above, so its belt speed is a quarter of the first stage’s even with similar diameters. Second, keep every driver pulley above its section’s minimum diameter floor; compound layouts tempt builders into tiny intermediate drivers, and those are the pulleys that cook belts. Check the tension side of a new jackshaft layout with the belt tension calculator, because two short spans generally need firmer static tension than one long span carrying the same power.

A worked end-to-end example ties the whole page together — the bandsaw from the measuring guide. Its motor pulley measured 2.83 in PD (A-section) and the arbor pulley 5.33 in PD (B-section). Ratio = 5.33 ÷ 2.83 = 1.88:1; driven RPM = 1750 ÷ 1.88 = 929 RPM; belt speed = π × 2.83 × 1750 ÷ 12 = 1296 ft/min (6.6 m/s) — comfortably inside the ceiling, with the A-section driver above its 2.6 in minimum. Every number here came from calipers, the offset table, and the two formulas on this page. That is the entire workflow: measure, convert to PD, apply the ratio, verify belt speed.

For length-side verification of any drive you size here — will a stock belt length actually close the loop at your center distance — run the diameters through the multi-pulley belt length calculator or, for simple two-pulley drives, the open-belt mode on the homepage belt calculator. Both hubs and guides on the site cross-reference: the calculators hub lists every solver, and the guides index collects the measurement methods behind the numbers.

Frequently asked questions