Gear Train

← Back to Engineering Library

A Gear Train is a set of two or more meshing gears arranged on parallel or angled shafts to transmit rotational power between them. Each mesh trades speed for torque in inverse proportion to the tooth count ratio, so a small driver turning a large driven gear slows the output and multiplies torque. We use Gear Trains to match a motor's natural speed to a load's required speed and force — anything from a 100:1 reduction in a winch drive down to the 4-stage epicyclic train inside a Bosch cordless drill turning at 400 RPM under load.

Gear Train Interactive Calculator

Vary the tooth counts and input speed of a 2-stage compound gear train to see stage ratios, total reduction, and output RPM.

Stage 1
--
Stage 2
--
Total Ratio
--
Output Speed
--

Equation Used

stage1 = N2 / N1; stage2 = N4 / N3; total ratio = stage1 * stage2; output RPM = input RPM / total ratio

Each mesh ratio is the driven gear tooth count divided by the driver gear tooth count. In a compound train, the ratios multiply, so the output speed equals input speed divided by the total reduction ratio.

  • Spur gears are ideal with no slip.
  • Intermediate driven gear and second-stage driver are fixed on the same shaft.
  • Efficiency losses are ignored, so this calculator sizes speed ratio only.
Watch the Gear Train in motion
Video: Gear train and rack by Nguyen Duc Thang (thang010146) on YouTube. Used here to complement the diagram below.
Compound Gear Train Diagram A 2-stage compound spur gear train showing ratio multiplication. Input Pinion 16T 1440 RPM Intermediate Shaft 64T + 18T fixed Output Gear 72T 90 RPM Stage 1 4:1 Stage 2 4:1 ● Orange = Driver gears ● Gray = Driven gears
Compound Gear Train Diagram.

Inside the Gear Train

A Gear Train works by passing rotation through successive meshes, where each pair of teeth in contact applies a tangential force at the pitch line. The pitch diameter — not the outside diameter — is what sets the ratio, and the ratio at each mesh equals the driven gear's tooth count divided by the driver's tooth count. Stack two or more reductions on a single intermediate shaft and you have a compound gear train, where ratios multiply rather than add. A 5:1 stage feeding a 4:1 stage gives 20:1 overall, in a package far smaller than a single 20:1 mesh would need.

The geometry has to be right or the train will whine, wear, or seize. Centre distance between two spur gears must equal the sum of their pitch radii within roughly 0.05 mm for a module-1 gear — too tight and you bind the teeth and overheat the lubricant, too loose and backlash opens up and the train rattles under load reversal. Tooth profile matters too: standard 20° pressure angle involute teeth tolerate small centre-distance errors gracefully, but a 14.5° profile is far less forgiving. If you notice pitting on the drive flank within the first 100 hours of run-in, the most common cause is misalignment between the two shafts — parallelism error above about 0.02 mm per 100 mm of face width concentrates load on one end of the tooth and chews it off.

What about the idler gear? An idler sits between driver and driven without changing the overall ratio — its tooth count cancels out — but it reverses output direction and bridges centre distances that would otherwise need a huge driven gear. You see this in lathe headstocks, automotive reverse gears, and printing-press feed trains. The catch is that idlers see load on both flanks each revolution, so their fatigue life is roughly half that of a single-flank gear at the same load.

Key Components

  • Driver Gear (Pinion): The input gear connected to the power source. Tooth count typically ranges from 12 to 20 for a reduction stage — below 12 teeth on a 20° pressure angle spur gear you start getting undercut at the root, which weakens the tooth by 15-25%.
  • Driven Gear: The output gear that receives torque from the driver. Its tooth count divided by the driver's tooth count gives the stage ratio. Pitch diameter must be machined to within roughly 0.025 mm of nominal for a quiet mesh at module 1-2.
  • Idler Gear: A free-spinning gear between driver and driven that reverses rotation direction without affecting overall ratio. Sees alternating flank loads each revolution, so case-hardened steel (58-62 HRC) is standard for any cycle count above 10⁶.
  • Intermediate Shaft: Carries a compound gear pair — large driven gear and small driver pinion fixed together — to multiply ratios within a compact envelope. Shaft deflection under load must stay below 0.01 mm at the gear face or the mesh pattern shifts toward the tooth tip.
  • Bearings: Support each shaft so the gears can mesh without wobble. Radial play above 0.05 mm in a 20 mm bore deep-groove ball bearing throws the centre distance off enough to spike noise by 6-8 dB.
  • Housing: Sets and holds the centre distances between shafts. Bore-to-bore tolerance of ±0.02 mm is standard for AGMA Q10 quality gears — looser than that and the rated load capacity drops sharply.

Who Uses the Gear Train

Gear Trains turn up wherever a prime mover and a load disagree about speed and torque, which is to say almost everywhere mechanical power moves. The choice between a simple two-gear train, a compound multi-stage layout, and an epicyclic (planetary) arrangement comes down to ratio, package size, and concentricity of input and output shafts. Spur gears dominate when noise allows, helical gears take over when noise doesn't, and bevel gears bridge angled shafts. You would be amazed how often a single Gear Train design problem decides the entire form factor of a finished product.

  • Automotive: The 8-speed ZF 8HP automatic transmission used in BMW, Jaguar, and Ram trucks — four planetary gear trains plus clutches give ratios from 4.71:1 in 1st to 0.67:1 in 8th.
  • Power Tools: Bosch GSR 18V-110 cordless drill — a 3-stage planetary Gear Train reduces motor output from 21,000 RPM down to 400-1,800 RPM at the chuck, with 110 Nm peak torque.
  • Wind Energy: Vestas V90 turbine gearbox — a compound train of one planetary and two helical stages converts 16 RPM rotor input to 1,500 RPM generator input at roughly 100:1.
  • Robotics: Harmonic Drive strain-wave reducers used in FANUC and KUKA robot wrists — though strain-wave is technically a different family, the input bevel and helical Gear Train upstream is what lets the wrist motor sit clear of the joint.
  • Marine: ZF 220A marine reduction gearbox on small commercial trawlers — a 2.0:1 to 2.5:1 single-reduction Gear Train lets a 2,500 RPM diesel turn a 1,000 RPM propeller efficiently.
  • Machine Tools: Hardinge HLV-H toolroom lathe headstock — a back-gear compound Gear Train provides 8 spindle speeds from 125 to 3,000 RPM through engaged and disengaged ratios.
  • Clocks and Instruments: Mechanical wristwatch trains — typically a 4-wheel train stepping the mainspring barrel down to the 1 Hz escape wheel through ratios totalling roughly 7,200:1.

The Formula Behind the Gear Train

The overall Gear Train ratio tells you how much speed reduction and torque multiplication you actually get from input shaft to output shaft. At the low end of useful ratios — say 2:1 to 5:1 in a single-stage spur train — you get clean efficiency around 97-98% per mesh and almost no thermal management problems. The sweet spot for a compound 2-stage train is roughly 15:1 to 50:1, where you balance package size, tooth count, and bearing load. Push past 100:1 in a single planetary stage and you start running into tooth-count constraints (sun and ring need integer relationships), heat from friction in deep reductions, and efficiency drops below 85% per stage. The formula below is what you use to choose stage counts and tooth counts before you commit to a housing.

itotal = (N2 / N1) × (N4 / N3) × ... × (Nn / Nn-1)

Variables

Symbol Meaning Unit (SI) Unit (Imperial)
itotal Overall gear ratio from input to output dimensionless dimensionless
N1, N3, ... Tooth count of each driver gear in the train teeth teeth
N2, N4, ... Tooth count of each driven gear in the train teeth teeth
ωout Output angular speed = ω<sub>in</sub> / i<sub>total</sub> rad/s or RPM RPM
Tout Output torque = T<sub>in</sub> × i<sub>total</sub> × η N·m lbf·ft
η Combined mesh efficiency (typically 0.97-0.99 per spur stage) dimensionless dimensionless

Worked Example: Gear Train in a peat-cutter conveyor drive

Spec the 2-stage compound Gear Train for a peat-cutter belt conveyor at a Bord na Móna bog operation in County Offaly. The conveyor needs roughly 45 RPM at the head pulley to run the belt at 0.7 m/s, and you have a 1,440 RPM 4-pole 3 kW SEW-Eurodrive motor sitting on the frame. The conveyor pulls between 60 Nm at light load and 180 Nm at peak when wet peat slabs hit the belt all at once. Pick the tooth counts and verify the output speed and torque across the operating range.

Given

  • ωin = 1440 RPM
  • ωout, target = 45 RPM
  • Tin, peak = 19.9 N·m
  • N1 (stage 1 driver) = 16 teeth
  • N2 (stage 1 driven) = 90 teeth
  • N3 (stage 2 driver) = 18 teeth
  • N4 (stage 2 driven) = 102 teeth
  • η (per stage) = 0.98 dimensionless

Solution

Step 1 — compute each stage ratio, then multiply for the overall ratio:

itotal = (90 / 16) × (102 / 18) = 5.625 × 5.667 = 31.875

Step 2 — at nominal motor speed of 1,440 RPM, divide input speed by the ratio:

ωout, nom = 1440 / 31.875 = 45.18 RPM

That lands within 0.4% of the 45 RPM target — close enough that belt speed will be 0.703 m/s rather than the design 0.700 m/s, which a peat conveyor will not notice.

Step 3 — compute output torque at the nominal input torque (3 kW at 1,440 RPM = 19.9 N·m):

Tout, nom = 19.9 × 31.875 × (0.98)2 = 609 N·m

Step 4 — at the low end of the operating range, when the belt is empty and the motor is loaded to maybe 30% of rated torque (≈6 N·m input), the output sits around 184 N·m. The conveyor barely notices — it is essentially loafing, and gear teeth see only run-in level loads. Bearing life at this duty point easily clears 50,000 hours.

Step 5 — at the high end, when wet peat slabs slam onto the belt and demand 180 N·m at the head pulley, you need 180 / (31.875 × 0.96) ≈ 5.88 N·m at the motor, which is only about 30% of rated motor torque. The motor handles the spike with margin, but the gear teeth see roughly 3× their nominal tangential load for the duration of the slug — this is where AGMA bending fatigue becomes the limit, not the average load.

Result

The Gear Train delivers 45. 18 RPM and 609 N·m at the head pulley under nominal motor torque — bang on target for the 0.7 m/s belt speed the cutting head expects. Across the operating range you swing from a sleepy 184 N·m at empty-belt running to a brief 575 N·m+ output during wet-slab impact loads, and the sweet spot for tooth life sits at around 60% of rated torque where bending stress stays below the AGMA fatigue limit. If you measure 42 RPM instead of the predicted 45 RPM under load, the most likely causes are: (1) belt creep on a worn drive pulley lagging — check the lagging thickness, anything below 8 mm on a 250 mm pulley loses grip when wet; (2) motor slip exceeding nameplate 2.8% because the supply voltage has sagged below 380 V at the bog's distribution end; or (3) coupling slip between motor and stage-1 pinion if you used a taper-lock bushing torqued below the SKF spec of 30 N·m for a 1610 series.

Choosing the Gear Train: Pros and Cons

A Gear Train is one of three main choices when you need to change speed and torque between two shafts. Belt-and-pulley drives are cheaper and quieter but slip and stretch. Chain drives handle dirt and shock loads but need lubrication and tensioning. Compare on the dimensions that actually drive the design decision.

Property Gear Train Belt and Pulley Drive Chain Drive
Efficiency per stage 97-99% (spur), 94-98% (helical) 95-98% (V-belt), 98-99% (timing belt) 96-98%
Maximum practical ratio per stage 6:1 (spur), 10:1 (worm), 200:1 (planetary multi-stage) 8:1 typical, 15:1 with idler 7:1 typical
Speed accuracy Exact integer ratio, no slip 0.5-2% slip under load (V-belt) 0.1-0.5% chordal action variation
Load capacity at 100 mm centre distance High — up to 50 kW continuous Medium — typically up to 15 kW High — up to 30 kW continuous
Maintenance interval Oil change every 5,000-10,000 hours Belt inspection every 500 hours, replacement every 3,000-5,000 Lubrication every 50-100 hours, chain replacement every 5,000-10,000
Tolerance to misalignment Tight — 0.02 mm/100 mm parallelism Forgiving — up to 1° angular offset Forgiving — up to 0.5° angular offset
Initial cost (3 kW, 30:1 reduction) $$ high — machined housing, hardened gears $ low — pulleys, belt, simple bracket $$ medium — sprockets, chain, tensioner
Acoustic noise at 1 m 65-80 dB(A) (spur), 55-70 dB(A) (helical) 55-65 dB(A) 70-85 dB(A)

Frequently Asked Questions About Gear Train

The textbook calculation assumes ideal mesh efficiency around 0.98 per stage, but real-world losses stack faster than people expect. Bearing drag, oil churning, and seal friction each remove 1-2% on top of the tooth-mesh loss — so a 2-stage train delivering 0.96² = 0.92 from gears alone, minus another 2-3% from bearings and oil, lands you at the 89-93% range you are seeing.

Quick diagnostic: spin the output shaft by hand with the input disconnected. If you feel notable resistance beyond bearing pre-load, your seal is over-tight or the oil viscosity is too high. ISO VG 220 in a small reducer running at 30 °C ambient will eat 4-5% by itself — drop to VG 150 and you recover most of it.

An idler sees double-flank loading every revolution. The drive flank pushes the driven gear, and the same teeth take a reaction load on the back flank from the driver. So you are running both flanks at full load, which roughly halves AGMA pitting fatigue life compared to a single-flank gear at the same torque.

Two fixes worth considering: bump the idler one face-width size larger than the calculation suggests (gives you 30-40% more life), or use a case-hardened idler at 58-62 HRC even if your driver and driven are through-hardened. The hardness differential helps the idler outlive the meshes either side.

Pick planetary when you need concentric input and output, high torque density, or the smallest possible package. A 3-stage planetary at 50:1 fits inside a cylinder roughly 100 mm diameter by 150 mm long for a 1 kW motor; the equivalent compound parallel-shaft reducer is closer to 200 × 200 × 300 mm because each stage needs its own pair of bearing bores.

Pick parallel-shaft compound when cost matters, repair access matters, or you need an offset output that lets you mount a brake or encoder on the back of the motor. Compound parallel reducers are also more efficient at high ratios — typically 92-94% at 50:1 versus 88-91% for a 3-stage planetary with the same gear quality.

The rule of thumb is ±0.025 mm per module size for AGMA Q10 quality. For a module-2 gear, that is ±0.05 mm centre-distance tolerance. Tighter than ±0.025 mm and you risk binding when the housing thermally expands — cast iron grows about 0.011 mm per metre per °C, so a 100 mm centre distance shifts 0.022 mm over a 20 °C temperature rise.

Looser than ±0.075 mm and backlash exceeds 0.15 mm at the pitch line, which is where you start hearing audible clatter on load reversal. If you cannot machine the bores precisely, use eccentric bushings on one shaft and dial the centre distance in at assembly — that is how Hardinge sets up its lathe headstock back-gear meshes.

Heat in a Gear Train comes from three sources: tooth-mesh friction, oil churning, and bearing drag. If your housing is hitting 90 °C+ after 30 minutes on a duty that should sit at 60 °C, the oil level is almost always wrong — too high, and the gears churn the oil into foam that shears at every mesh. The standard is to fill to the centreline of the lowest gear, never above.

Second cause: the motor is fine but the gearbox is undersized for thermal capacity, not torque capacity. A reducer rated for 3 kW mechanical might only dissipate 1.5 kW of heat continuously through its housing. If your duty cycle is constant rather than the intermittent assumed at rating, you need either a bigger housing, forced cooling, or a synthetic oil like Mobil SHC 630 that tolerates 110 °C without breaking down.

What you are seeing is accumulated backlash. Each mesh contributes 0.05-0.15 mm of backlash at the pitch line, and across 4 stages that compounds at the output. For a 7,200:1 watch-style train, output free play can reach several degrees of rotation before any input torque actually moves the load.

The fix depends on the application. For instruments, use anti-backlash split gears with a torsion spring between the halves — that pre-loads each mesh against one flank only. For positioning duty, switch to ground gears at AGMA Q12 or better, which cuts per-mesh backlash to 0.02 mm. For absolute zero-backlash duty, you cannot do it with a conventional Gear Train at all — you need a strain-wave (harmonic) reducer or a cycloidal drive.

References & Further Reading

  • Wikipedia contributors. Gear train. Wikipedia

Building or designing a mechanism like this?

Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.

← Back to Mechanisms Index
Share This Article
Tags: