OD vs PD: which diameter do you actually need?

Every pulley measurement question starts the same way: you put calipers on the pulley and get one number, the outside diameter. Whether that number is useful depends entirely on what you are about to do with it. Speed ratios, belt length formulas and center-distance solves all operate on the pitch diameter — the diameter of the neutral line inside the belt as it sits in the groove. On a flat pulley or an idler that contacts the belt’s backside, the pitch diameter is simply the OD plus or minus the belt thickness; on a grooved V-belt sheave, the pitch line sits a defined distance below the OD called the datum offset, and PD = OD − 2 × offset.

The rule for deciding is short: if the number will enter a formula — ratio, RPM, belt length, center distance — it must be a pitch diameter. If the number will enter a purchase order or a fit check — does this pulley clear the guard, does it match the casting drawing — the outside diameter is the right figure. Mixing the two is the single most common pulley measurement error: using a 4.75 in OD B-pulley as “4.75 in” in a ratio calculation inflates the computed speed of that shaft by about 3.2%, because the true pitch diameter is 4.75 − 0.29 = 4.46 in.

Measuring technique matters as much as the arithmetic. Cast and machined pulleys have real runout; a sheave that averages 0.1 mm of eccentricity will read differently every quarter turn. Take at least three caliper readings at rotations spread around the pulley and average them. For pulleys too large to span with calipers, wrap the outer land with a strip of paper or a flexible tape, measure that circumference, and divide by π — this “PI tape” method is standard shop practice for sheaves above roughly 150 mm.

The datum offset table: OD to pitch diameter by section

The offsets below follow the RMA/ISO datum system used by the calculator engine on this site. The offset is the distance from the pulley’s outside diameter down to the belt’s pitch line — where the cord line of the belt sits when the belt is seated in the groove. Subtract twice the offset from your measured OD to get PD; add twice the offset to a known PD to predict the OD before you buy.

Two things in that table deserve a second look. First, the narrow sections (3V/5V/8V) carry the same offsets per width class as their metric SP cousins within about half a millimeter, but they are not interchangeable — a 5V belt in an SPB groove wedges differently and will ride at a slightly different effective radius. Second, the offset scales with belt height, not width: a deep C-section belt has its cord line 4.7 mm below the top, so a C-pulley measured at 250 mm OD has a pitch diameter of 240.6 mm, not the 250 mm a naive calculation would use. On a 3:1 ratio drive, that 3.8% error shows up directly as a speed error on the driven shaft.

For idlers and backside contact the offset works in the opposite direction. A flat backside idler of diameter D contacts the belt’s back face, so the belt’s pitch line sits outside the idler surface, at D + 2 × (belt height − datum offset). An A-section belt on a 60 mm flat idler, for example, has an effective pitch diameter of 60 + 2 × 5.2 = 70.4 mm. Getting this sign wrong on a tensioner idler throws off multi-pulley length calculations by the full circumference of the error, which is exactly why the multi-pulley calculator asks whether each wheel contacts the belt face or the backside.

Measuring a V-belt groove: top width, depth and angle

A groove check answers three questions in order: does the groove match the belt section, is it worn, and is it safe to reuse. Work through them with the same calipers you used for the OD.

Groove top width at the OD

Place the caliper jaw tips across the opening of the groove at its outermost edge. The reading should match the belt section’s nominal top width closely — 17 mm for a B-section groove, 9.5 mm for 3V, and so on, using the same table above. A reading more than about half a millimeter over nominal means the groove has been bell-mouthed by wear, and a worn groove lets the belt sink lower, effectively shrinking the pitch diameter while increasing slip under peak load. Bell-mouthing is also visible: worn grooves show a polished, concave wear band at the widest part of the opening and a step where the original land remains.

Groove depth

Rest a straight edge across the two lands and measure down to the groove root with a depth rod. The root must clear the belt’s inner corner radius — the belt bottom should never contact the groove root. As a practical rule the groove depth should exceed the belt section height by at least 2–3 mm; a C-section belt (14 mm tall) wants a groove deeper than about 17 mm. If a belt bottoms out, it loses its wedge action entirely, tension readings stop making sense, and the drive glides to a stop under load no matter how much tension you add.

Groove flank angle

V-groove flanks are cut at angles between roughly 32° and 38° included, depending on the section and the pulley’s datum diameter — smaller diameters use the narrower angles so the belt wedges fully despite curving around the sheave. Without a dedicated groove-angle gauge, the fastest reliable check is comparative: seat a known-good new belt of the correct section in the groove and look for daylight between the flanks and the belt sides. A belt that sits visibly low or rocks in the groove confirms either a worn (opened) angle or the wrong section. Dedicated gauge sets covering the standard 32/34/36/38° angles cost little and remove the guesswork on multi-groove sheaves, where worn grooves on one band silently overload the remaining belts.

Timing pulleys: teeth × pitch ÷ π

Timing pulleys invert the V-belt problem in a helpful way: instead of subtracting a wear-dependent offset, the pitch diameter is fixed exactly by the tooth count and pitch. The formula is PD = pitch × N ÷ π, where N is the number of teeth and pitch is the belt’s pitch (tooth-center to tooth-center) in consistent units. No offset table, no judgment — a 40-tooth HTD 8M pulley is exactly 101.86 mm pitch diameter, forever.

The practical work is in identifying the pitch system. Count the teeth first, then measure the pitch itself with calipers across, say, ten tooth crests from crest-center to crest-center and divide by ten. Compare against the common systems: 2 mm (GT2), 3 mm (GT3, HTD 3M), 5 mm (GT5, HTD 5M), 8 mm (HTD 8M), 14 mm (HTD 14M), and the imperial series 0.080 in (MXL), 0.200 in (XL), 0.375 in (L) and 0.500 in (H). Ten-tooth averaging matters because individual teeth on small pulleys fall under 2 mm — single-tooth caliper reads drown in jaw-width error.

Two measurements beyond teeth and pitch complete a timing-pulley check. First, the flange-to-flange width must accept the belt width plus a little lateral float — a common rebuild error is substituting a pulley of the right pitch but one width class too narrow, which forces the belt against a flange and sheds teeth. Second, the tooth profile: HTD curvilinear teeth will not seat properly in a trapezoidal (AT or imperial-profile) pulley even when the pitch matches exactly, and vice versa. The mismatch usually runs — the drive skips under load — so confirm the profile family by the tooth shape visually: rounded crests indicate HTD/GT, flat-topped trapezoids indicate classical profiles.

Bore, keyway and mounting checks

The bore decides whether the pulley fits the shaft; the keyway decides whether it transmits torque. Measure the bore with an inside caliper or a small bore gauge at two depths (near each face), rotating the gauge 90° between readings. Cast pulleys are commonly slightly oval, and a bore that reads 25.00 mm at the mouth may read 25.15 mm mid-depth — enough to make a press fit feel like a slip fit during assembly. Record the largest consistent reading; that is the dimension the shaft must engage.

Keyways are specified as width × depth pairs keyed to the shaft diameter. In the inch series, common square-key pairings run as follows (ANSI B17.1 style):

Use the table as a cross-check, not a substitute for measurement: keyway width across the flats tells you the shaft size the pulley was bored for even when the bore itself is corroded or bushed. Note also whether the keyway runs through both sides of the bore (double-keyway pulleys accept either shaft orientation) and check set screws — a pulley relying on set screws alone over a plain keyway walks axially under reversing loads. Taper-bushed pulleys replace all of this with a split bushing and screws: measure the bushing series number stamped on its face instead of the raw bore, because the bore of the bushing, not the hub, defines the fit.

Working backwards: identifying the pulley from the belt

The most common real-world scenario is inverted: the pulley is buried in a machine, seized, or long gone, and all you hold is the belt. Belt measurements then drive everything. The belt’s top width identifies the section, the section fixes the datum offset, and the length designation constrains the pitch-diameter and center-distance combination the replacement pulleys must satisfy.

The identification chain is the mirror of this guide’s order: top width → section (13 mm is A, 17 mm is B, 9.5 mm is 3V, 9.7 mm is SPZ, and the rest follow the same table); length marking → pitch length (A45 means 45 in inside length, 3V250 means 25.0 in pitch, SPB2240 means 2240 mm pitch); then a solve of the belt-length equation for the unknown pulley sizes at your measured center distance. The full method, including how to measure each belt type, lives in the companion guide How to Measure a Belt, and the arithmetic solve — belt length fixed, diameters unknown — is the reverse mode of the homepage belt calculator.

That final check is fast with the pulley RPM & ratio calculator: enter both pitch diameters and one known shaft speed, and confirm the driven speed lands where the machine spec says it should. For drives where belt tension matters as much as geometry — short centers, high torque, or any serpentine path with a spring idler — pair the replacement with a measurement plan from the site’s tension tool so the new pulley is not installed at the wrong static deflection.

Tools, tolerances and a worked example

The complete kit is short: a 150 mm digital caliper (0.01 mm resolution) for OD, groove width and bore; a depth rod or a straight edge plus rule for groove depth; a PI tape or flexible steel tape for large sheaves; and optionally a groove-angle gauge set. Add a bore gauge or a set of telescoping gauges if you service machinery regularly — caliper-inside reads compress in worn bores and hide up to 0.1 mm of clearance, which is the difference between a snug and a knocking fit.

Assemble the whole guide into one worked example. A used sheave comes off a bandsaw; calipers give an OD of 5.62 in averaged over four readings, the groove top width reads 17.1 mm, and the groove is deep enough that a B-belt’s inner corners never touch root. The top width identifies a B-section groove, so the datum offset is 3.7 mm per side: PD = 5.62 × 25.4 − 7.4 = 135.4 mm = 5.33 in. Suppose the motor pulley measured 3.05 in OD in an A-groove: PD = 3.05 × 25.4 − 5.6 = 71.87 mm = 2.83 in. The ratio is 5.33 ÷ 2.83 = 1.88, so a 1750 RPM motor drives the saw arbor at about 929 RPM. Those are the numbers that enter a belt-length or RPM calculation — and note how far both sit from the raw caliper readings. Every one of the errors in this example (using OD directly: −3.2% on one pulley, −1.9% on the other; combined ratio error ≈ 1.3%) is exactly the class of mistake the offset table above exists to prevent.

Tolerances worth recording when you write the numbers down: OD to ±0.1 mm for shop work (±0.05 mm if the result feeds a matched-speed drive), groove top width to ±0.25 mm, pitch diameter effectively exact for timing pulleys once teeth and pitch are confirmed, and bore to the shaft tolerance class the machine specifies. Log measurements with the pulley, not just in a notebook — a paint-pen mark of “PD 135.4 / B / 1.375 bore” on the hub saves the next teardown.

For a wider view of every measurement and calculation tool on the site — including the drives covered here plus length, tension and ratio solvers — browse the full calculators hub, and for the other measurement guides in this series see the guides index. Both hubs stay current as new tools and guides are added, so they are the fastest way back to any page in the series.

Frequently asked questions