Industrial planetary gearboxes now cover ratio spans from roughly 3:1 to 4,000:1, torque ratings from sub-100 N·m servo-class units up to 2,600,000 N·m heavy-industry units, and input power windows from 0.4 kW to 12,934 kW [S3]. Three numbers — torque (Nm), ratio (i), and service factor (Sf) — decide 80% of a spec; everything else is refinement.
This guide is written for buyers and engineers who have already moved past “which brand is best” and now need a defensible selection logic: load profile, ratio architecture, backlash class, lubrication, and radial-load rating. For background on the underlying gear geometry, the planetary reducer encyclopedia entry is a useful primer, and the cycloidal reducer reference explains where twin-cyclo alternatives win or lose against a true planetary stage.
Load, Torque and Service-Factor Sizing
Selection starts with the equivalent torque T₂ₑq = T₂ · Sf, where Sf compounds shock, daily operating hours, and driven-machine class per AGMA 2001 / ISO 6336 conventions; for an 8-16 h/day crusher or mixer under heavy shock, Sf commonly lands between 1.75 and 2.5 [S1][S3]. Compare T₂ₑq against the reducer's *rated* (not *peak* or *emergency-stop*) torque on the manufacturer's torque-speed curve, otherwise the unit runs hot and bearing fatigue arrives before the calculated L₁₀ life.
Three concrete bands dominate 2026 sourcing. Compact coaxial planetary units (Bonfiglioli S-series class) deliver 10 kNm and above for slewing, crane, and excavator duty at ratios of 23:1 to 67:1 [S1]. Mid-range JRP-class industrial units span nine to 36 frame sizes, 25:1 to 4,000:1 ratios, 0.4–12,934 kW input, and 2,600,000 N·m output torque [S3]. Economy spur-planetary hybrids (FHTW class) sit below both in price and rating, targeted at OEM light-cycle production where backstop-class rigidity is not required [S4].
Ratio Architecture and Backlash Class
Backlash separates market tiers cleanly: standard backlash for general industrial drives (≈ 8-15 arc-min), reduced backlash (≈ 3-5 arc-min) for indexing and packaging, and precision backlash (under 1 arc-min) reserved for cycloidal and strain-wave alternatives such as Nidec's ERP series, which lists sub-1 arc-min backlash and 20-30 arc-sec repeatability [S6].
For high-ratio precision under 160:1, harmonic-strain-wave units (CSG/CSF-2UH) at ratios of 50, 80, 100, 120 and 160:1 remain the dominant compact-precision choice [S2]. For ratios above 160:1 in precision applications, an RV reducer (planetary + second cycloidal stage) is usually a better match than a stacked three-stage planetary; the harmonic reducer reference and RV reducer reference document the trade-offs in detail.
Frame, Mounting and Radial-Load Rating
Frame size dictates output-shaft-bore diameter, which sets the maximum allowable radial load Fr at the shaft midpoint. A practical rule: a 30 mm bore in-line planetary is typically rated for 4-6 kN Fr at the midpoint; a 90 mm bore jumps to 18-28 kN, and a 150 mm+ heavy-industrial bore exceeds 60 kN, with the exact number read off the manufacturer's Fr-vs-overhung-distance chart, not estimated. Hollow-shaft coaxial designs (S-series class) simplify mounting by sliding the gearbox directly onto the driven shaft, eliminating a coupling and a baseplate while keeping the 10 kNm torque rating [S1].
For right-angle packaging or vertical-shaft pump drives, helical-bevel and right-angle planetary units (Nidec right-angle planetary class) re-route the input 90° using a separate spiral-bevel stage, which adds 5-8% to the BOM cost but removes the need for an external shaft/belt turn. Cycloidal right-angle alternatives (ERP series) offer 613-3,185 Nm acceleration torque across four frame sizes with hollow-shaft pass-through for cabling or ball screws [S6].
Lubrication, Thermal Limits and Cycloidal Comparison
Most industrial planetary reducers run splash or forced oil at temperatures up to 80-90 °C sump, with synthetic PAO gear oil (ISO VG 220-460) as the factory fill. Above 20 kNm or 95 °C sump, switch to a water-cooled heat-exchanger or external fan; for ambient dust (cement, mining, steel-mill), specify a sealed labyrinth + breather + oil-sight-glass package. In cyclic shock applications — slewing, crane, and track-drive — verify the manufacturer's e-stop (5x rated torque) capacity, not just the continuous rating, since most premature failures originate in 0.5-2 s shock events that bypass service factor [S1][S6].
When comparing planetary against a cycloidal alternative, line the selection criteria up directly: planetary wins on efficiency (one stage ≈ 97%, two stages ≈ 94-96%) and on torque density per kg; cycloidal wins on shock tolerance (5x rated torque e-stop is a design baseline, not a derate) and on zero-backlash repeatability (20-30 arc-sec) [S6]. For mixer, agitator, conveyor, and extruder duty, planetary is the default; for indexing tables, paint robots, and high-shock packaging, cycloidal or strain-wave is usually cheaper in the long run.
Who It Is For — and Who It Is Not
Planetary reducers fit buyers needing high torque density in a compact envelope, ratios above 10:1, and a continuous-duty efficiency profile. Heavy-industry users — crane builders, port slewing drives, wind pitch/yaw, mining conveyors, and large mixers — are the primary beneficiaries of the 1-2.6 MN·m class units [S1][S3]. On the light end, OEM machine builders running multi-axis servo systems benefit from the linear guide compatibility and standard backlash in the 8-15 arc-min class, where stacked-planetary is the cost-effective default.
A planetary reducer is the wrong pick for ultra-precision direct-drive applications under 1 arc-min repeatability, where harmonic or cycloidal stages dominate; for sub-1 kW sub-Nm applications where a simple spur or worm gear suffices; or for highly corrosive, wash-down food/pharmaceutical duty where a stainless-housed servo gearbox is the proper fit. It is also rarely the right pick for sub-1 kNm cost-driven OEM cycles under $200 unit cost, where a bevel or worm pair finishes the job. The crossed-roller guide compatibility and standard bearing stacks can also constrain the radial-load number, so always read the Fr chart against your application.
Sourcing, Standards and 2026 Supply Reality
Three sourcing realities define 2026 procurement. First, European and Japanese brands (Bonfiglioli, Nidec, Sumitomo, SEW) hold the 10 kNm+ class with documented AGMA/ISO 6336 torque curves and 12-24 month warranty; expect 8-14 week lead times [S1][S6]. Second, Chinese heavy-industrial suppliers (JIE-class manufacturers) cover the 25-4,000 ratio and 2,600,000 N·m class with 4-8 week lead times and roughly 30-50% cost advantage on like-for-like torque; verify the AGMA or ISO compliance certificate and the cyclic-rating curve before placing a critical-service order [S3]. Third, mid-range economy spur-planetary hybrids (FHTW, FHT-class) are appropriate for OEM light-cycle production where the buyer accepts a shorter design life in exchange for cost; do not specify them for heavy-shock or 24/7 duty [S4].
For buyers evaluating large-frame slewing and crane-class drives, a parallel look at the slewing bearing supplier landscape and the overhead bridge crane sizing logic helps keep the gearbox, bearing, and crane-duty-class decisions consistent — mismatches between a 2.6 MN·m reducer and a light-duty bearing stack are the most common field failure I see.
Verify the unit comes with a documented AGMA 2001 / ISO 6336 service-factor curve, an L₁₀ bearing-life calculation at your Fr and Fa, and a thermal-rating chart at your ambient and duty cycle. Reject any quote that ships only a catalog torque number with no overload curve. Cross-check the brand's e-stop rating — a 5x rated-torque e-stop capacity is a 2026 baseline for slewing, crane, and track-drive duty [S1][S6]. Watch for next-node signals: AGMA 2001 / ISO 6336 revisions are tracked by every major manufacturer, and a published update to the cyclic-shock derating table inside the next 12 months would shift several of the Sf ranges cited above.