An Epicyclic / Planetary Gear Train is a coaxial gear arrangement where one or more planet gears orbit around a central sun gear while meshing with an outer ring gear. Power flows through three concentric members — sun, carrier, and ring — and locking any one of them sets a different ratio between the other two. This packs high reduction ratios into a compact, coaxial envelope while splitting load across multiple planet meshes. You see it in every automatic transmission, wind turbine gearbox, and cordless drill on the market.
Epicyclic / Planetary Gear Train Interactive Calculator
Vary sun, planet, and ring tooth counts plus sun speed to see the fixed-ring reduction ratio and carrier output motion.
Equation Used
For the fixed-ring configuration, the sun gear drives the planets and the carrier becomes the output. The reduction ratio is set by the ring-to-sun tooth ratio: larger ring gears or smaller sun gears create more reduction. A valid simple planetary set also satisfies N_r = N_s + 2N_p, so the calculator reports planet tooth fit error.
- Ring gear is fixed and the sun gear is the input.
- Carrier is the output and rotates in the same direction as the sun.
- Ideal kinematics are used; friction, tooth losses, backlash, and deflection are ignored.
- Planet tooth fit is checked using N_r = N_s + 2*N_p.
How the Epicyclic / Planetary Gear Train Works
The Epicyclic / Planetary Gear Train, also called the Planetary Gear Train in automotive and aerospace work, runs three coaxial members at the same time: a central sun gear, an outer internal-tooth ring gear (the annulus), and a planet carrier holding 3 to 5 planet gears that mesh with both. The fundamental trick is that the planets orbit around the sun while spinning on their own pins. Lock the ring and drive the sun, the carrier turns slowly in the same direction. Lock the carrier and drive the sun, the ring turns backwards. Lock the sun and drive the ring, the carrier turns the same way at reduced speed. That gives you three discrete ratios from one set of gears.
Why build it this way? Two reasons. First, load sharing. With 3 planets, each tooth mesh carries roughly a third of the input torque, so you get the torque density of a much larger parallel-shaft gearbox in a fraction of the volume. Second, the input and output shafts sit on the same axis, which matters in places like a wheel hub or a turbine nacelle where you cannot afford an offset.
Tolerances bite hard here. Planet pin position must hold within about 0.02 mm true position on a typical 80 mm carrier — go looser and one planet will carry most of the load while the others coast, the loaded mesh wears, then the load shifts and you get a ratcheting noise under torque reversal. Backlash also stacks: sun-to-planet plus planet-to-ring backlash adds, so a sloppy build that feels fine static will hunt audibly under reversing load. The most common failure modes you see in the field are planet bearing fatigue (needle rollers in the planet bore), ring-gear pitting from misalignment, and carrier deflection under torque that knocks the planets off-parallel and chews the tooth ends.
Key Components
- Sun Gear: Central external-tooth gear on the main axis. Usually the high-speed input in reduction service. Tooth count typically 12 to 30; below 12 you get undercut, above 30 the package grows fast for a given ratio.
- Planet Gears: External-tooth gears, typically 3 to 5 of them, that mesh with both sun and ring. Each planet rotates on a needle bearing or bushing inside its bore. Planet-to-planet tooth-count match must be within 1 tooth difference for proper assembly into the ring.
- Planet Carrier: Rigid spider that holds the planet pins on a shared pitch circle. Pin true position must hold within about 0.02 mm; carrier torsional stiffness controls planet load sharing under torque.
- Ring Gear (Annulus): Internal-tooth gear surrounding the planets. Tooth count equals sun teeth plus 2× planet teeth. Often the fixed reaction member in cordless-drill and wheel-hub gearboxes.
- Planet Bearings: Needle rollers or plain bushings inside each planet bore. They see the highest cyclic load in the train and are the typical wear-out point — L10 life under 5,000 hours is normal in heavy-duty service.
Industries That Rely on the Epicyclic / Planetary Gear Train
Anywhere you need a high reduction ratio in a coaxial package, an Epicyclic Train shows up. The same arrangement appears under names like High-Speed Epicyclic Train in turbofan reduction gearboxes, Planetary gear transmission in automatic gearboxes, and Epicyclic Gear in clockwork and watch movements. The geometry is identical — what differs is the operating regime: tip speed, torque density, and which member is held fixed.
- Automotive: ZF 8HP automatic transmission uses 4 planetary gear sets in series with clutches and brakes to deliver 8 forward ratios from a 5.0:1 spread.
- Aerospace: Pratt & Whitney PW1100G geared turbofan uses a 3:1 High-Speed Epicyclic Train between the fan and low-pressure turbine to let each spin at its optimum speed.
- Wind Energy: Vestas V90 nacelle gearbox combines a planetary first stage with two parallel-shaft helical stages to step 16 RPM rotor speed up to 1,500 RPM generator speed.
- Power Tools: Milwaukee M18 Fuel cordless drill uses a 3-stage Planetary Gear Train inside the head to reach roughly 60:1 reduction in a 38 mm diameter housing.
- Robotics: Nabtesco RV-series cycloidal-planetary reducers drive the base joints of FANUC and ABB industrial robot arms at ratios from 30:1 to 200:1 with under 1 arc-min of backlash.
- Horology: Tourbillon escapement carriages in high-end Patek Philippe and Greubel Forsey movements run on miniature Epicyclic Gear Train frames to average out gravity errors in the balance wheel.
The Formula Behind the Epicyclic / Planetary Gear Train
The single equation you need to size any Epicyclic / Planetary Gear Train is the Willis equation, which relates the speeds of the three members through the fixed-carrier ratio. The fixed-carrier ratio R₀ is set the moment you pick tooth counts on sun and ring — it does not change with operating speed. What changes across the operating range is which member you hold and how that maps the input speed to output speed. At the low end of the typical range (small sun, large ring), you get reductions of 3:1 to 4:1 per stage with high torque density. At the nominal sweet spot of about 5:1 per stage, planet size, sun strength, and ring wall thickness all balance. Push beyond 8:1 in a single stage and the sun gets so small it undercuts, planet bearings shrink to where they cannot carry the torque, and you end up cascading two stages anyway.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| ωsun | Angular speed of the sun gear | rad/s | RPM |
| ωcarrier | Angular speed of the planet carrier | rad/s | RPM |
| ωring | Angular speed of the ring gear (annulus) | rad/s | RPM |
| Zsun | Number of teeth on the sun gear | teeth | teeth |
| Zring | Number of teeth on the ring gear | teeth | teeth |
Worked Example: Epicyclic / Planetary Gear Train in an autonomous warehouse robot wheel hub drive
Sizing the in-hub Planetary Gear Train for the drive wheel of a Locus Robotics LocusBot-style autonomous mobile robot. The wheel diameter is 200 mm, the brushless DC motor delivers nominal 3,000 RPM at the input, and the target ground speed is 1.5 m/s. You picked a single-stage Epicyclic train with a fixed ring (a standard configuration for hub drives, sometimes called Epicyclic train (form 2) in older textbooks). Sun has 18 teeth, planets have 27 teeth, ring has 72 teeth.
Given
- Zsun = 18 teeth
- Zplanet = 27 teeth
- Zring = 72 teeth
- ωsun (nominal) = 3000 RPM
- Dwheel = 200 mm
Solution
Step 1 — with the ring held fixed (ωring = 0), the reduction ratio from sun to carrier follows directly from the Willis equation:
Step 2 — at nominal motor speed of 3,000 RPM, the carrier speed (and therefore wheel speed) is:
Step 3 — convert to ground speed using wheel circumference:
That is well above the 1.5 m/s target, so in practice the motor controller throttles the sun to about 720 RPM to hit cruise. At the low end of the duty range — pallet pick-up at 0.3 m/s — the sun runs at roughly 144 RPM:
That is slow enough for the bot to nose up to a pick station without slamming. At the high end, when the controller commands full sprint at the 3,000 RPM nameplate, you'd theoretically get 6.28 m/s — but in a typical warehouse the path-planning ceiling is about 2.0 m/s for safety, so the gearbox almost never sees its design redline. The sweet spot for this gear set sits around 720 RPM input / 1.5 m/s ground speed, where planet bearing PV product stays well below the bushing manufacturer's limit and the ring gear sees about 40% of its allowable Hertzian stress.
Result
The nominal reduction is 5. 0:1, giving 600 RPM at the wheel and 6.28 m/s top speed at full motor RPM. In practice the bot cruises at about 1.5 m/s, which puts the sun at roughly 720 RPM and lands the gear set in its low-stress sweet spot. At pallet-pickup speeds near 0.3 m/s the train barely loads, while at the 6.28 m/s theoretical maximum the planet needle bearings would exceed their 5,000-hour L10 life inside a year of two-shift operation. If your measured wheel speed comes in 5–10% low, suspect: (1) ring gear shifted in the housing because the anti-rotation lugs are a slip fit instead of a 0.05 mm interference, letting the ring back-drive slightly under load, (2) planet pin true position out beyond 0.02 mm causing one planet to grab and the others to lag through the lash, or (3) carrier-to-output spline wear adding 0.5–1° of compliance that shows up as missed encoder counts.
Choosing the Epicyclic / Planetary Gear Train: Pros and Cons
An Epicyclic Gear is not the only way to get high reduction in a small box. Cycloidal drives and harmonic (strain wave) gearboxes compete in the same space, and each has a different sweet spot on cost, backlash, and torque density. Pick by what your application actually demands.
| Property | Epicyclic / Planetary Gear Train | Cycloidal Drive | Harmonic (Strain Wave) Drive |
|---|---|---|---|
| Typical single-stage ratio | 3:1 to 8:1 | 10:1 to 100:1 | 30:1 to 320:1 |
| Backlash | 3–15 arc-min standard, 1 arc-min precision grade | <1 arc-min | <0.5 arc-min |
| Torque density | High — 3–5 planets share load | Very high — multiple lobe contacts | Moderate — flexspline limits peak torque |
| Cost (relative, mass-produced) | 1.0× (baseline) | 2–3× a comparable planetary | 4–6× a comparable planetary |
| Efficiency per stage | 96–98% | 90–94% | 70–85% |
| Typical service life | 20,000+ hours in industrial use | 10,000–15,000 hours | 8,000–10,000 hours |
| Best application fit | Automotive, power tools, wind, mobile robotics | Industrial robot joints, heavy-duty positioning | Semicon, precision robotics, optics stages |
Frequently Asked Questions About Epicyclic / Planetary Gear Train
That is a planet-pass frequency resonance. Three planets sweeping past a reference tooth at carrier speed × 3 generate a tonal excitation, and if that frequency lines up with a torsional mode of the housing, output shaft, or motor stator mount, you get a narrow-band whine. Sweep the input speed and you'll hear it appear over maybe a 50 RPM window then disappear.
The fix is almost never re-cutting gears. Either shift the housing mode (add a stiffening rib or a damper mass), shift the excitation frequency (go to 4 planets if the tooth-count assembly condition allows it), or detune the motor mount with a softer isolator. Helical planets with 15–20° helix angle also smear the excitation across a wider band and usually kill the tone outright.
No. That equation gets you geometric meshing, but you also need the assembly condition: (Zsun + Zring) must be divisible by the number of planets. With 18-sun / 72-ring and 3 planets, 90/3 = 30 — fine. Try 19-sun / 73-ring with 3 planets, you get 92/3 = 30.67 and the third planet physically will not drop into mesh. You would have to file teeth or shift one planet by a fraction of a pitch, both of which destroy load sharing.
If you want the freedom to pick any tooth count, use unequally-spaced planets — but then you give up the noise-cancellation benefit of equal spacing and the gearbox gets noticeably louder.
Yes. They are the same mechanism — sun, planets, carrier, ring — just named differently in different industries. Automotive and aerospace engineers usually say Planetary Gear Train or Planetary gear transmission. Mechanism textbooks and the British engineering tradition say Epicyclic Gear or Epicyclic Train. A High-Speed Epicyclic Train is the same hardware, optimised for tip speeds above about 100 m/s as you'd find in a turbofan reduction gearbox.
You almost always cascade. A single 25:1 planetary needs a sun small enough (around 8 teeth) that you'd undercut badly and the sun shaft becomes the weakest link in the whole drive. Two 5:1 stages keep every sun in the 15–25 tooth range where bending strength is healthy, the planets stay big enough to take real bearings, and overall efficiency lands at about 0.97² ≈ 94% — better than the single-stage equivalent once you account for the small-sun losses.
The exception is when axial length is brutally constrained, like inside a 50 mm cordless-drill head. Then you accept the single-stage compromise and use case-hardened pin bushings instead of needle bearings to survive the load.
Check the planet bearing preload and grease condition first. Needle rollers in a planet bore run hot under cyclic load, and the original lithium grease can churn out from between the rollers within 200–500 hours of heavy duty. Once the rollers run dry-ish, friction torque inside the planets eats input power before it ever reaches the carrier, and you measure that as a 5–10% drop in output torque for the same input current.
A second cause is carrier deflection. If the carrier is a stamped sheet-metal cage rather than a forged one, it bends elastically under torque and tilts the planet pins by 0.1–0.3°. The planet teeth then contact only at one end of the face width, peak Hertzian stress doubles, and you get edge pitting that masquerades as a torque-loss issue when really it is a stiffness issue.
The fixed-carrier ratio R₀ = −Zring/Zsun is the speed ratio between sun and ring as you would observe it from a reference frame rotating with the carrier. You care because it is the one number that stays constant no matter which member you hold or drive — it lets you compute every possible operating mode of the same gear set from a single equation.
In practice this matters when you build a differential or a power-split device, like the Toyota Hybrid Synergy Drive, where the carrier spins at a varying speed determined by the road and the engine, and you need to know the ring (motor-generator) speed for any combination of inputs. Without the fixed-carrier ratio you'd be solving the kinematics from scratch for every operating point.
Some lash is normal — 5–15 arc-min total at the output is typical for a commercial-grade planetary. Notchy is different. Notchiness usually means the planets are not equally loaded, so as you reverse, one planet takes up its lash, then the second, then the third, and you feel three discrete steps instead of one smooth take-up.
Root cause is almost always planet pin position error or planet tooth-thickness variation between the three planets. Measure pin true position on the carrier — if any pin is more than 0.03 mm off the nominal pitch circle, that's your problem. The fix is either a precision-ground carrier or selective assembly where you measure each planet's effective tooth thickness and pair them within 5 µm.
References & Further Reading
- Wikipedia contributors. Epicyclic gearing. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
