What Is Backlash in a Planetary Gearbox? Arcmin Explained with Real-World Impact Calculations

Precision planetary gearbox internal gear mesh — understanding backlash and arcminute specification

TECHNICAL KNOWLEDGE

What Is Backlash in a Planetary Gearbox?

Backlash is the single specification that separates a $200 servo reducer from a $1,200 one. Yet most datasheets list it as a number in arcminutes without explaining what that number means in real-world positioning error. This guide converts the abstraction into millimetres you can measure at the tool tip.

1′
= 1/60 OF A DEGREE
0.29 mm
1 ARCMIN @ 1 m RADIUS
≤3′
BEST AVAILABLE GRADE

Discuss Your Precision Requirement →

Backlash — The Gap Between Gear Teeth

Backlash in a planetary gearbox is the small angular dead zone between the input and output shafts. When the input shaft reverses direction, the output shaft does not respond instantly — it pauses for a fraction of a degree while the gear teeth cross the clearance gap to re-engage on the opposite tooth flank. That pause is backlash, and it is measured in arcminutes.

For a conveyor or fan that rotates in one direction only, backlash is functionally invisible. For a servo-driven robot arm or CNC axis that reverses direction hundreds of times per minute, backlash creates measurable positional error at every reversal point. This is why backlash is the most scrutinised parameter on every planetary gearbox datasheet — and why understanding what the number actually means in your specific application prevents both under-specification (leading to poor accuracy) and over-specification (leading to unnecessary cost).

Some clearance between gear teeth is structurally necessary. Without it, gears would bind under thermal expansion, lubricant film could not form between the contact surfaces, and assembly of the planetary gear train would be mechanically impossible. The engineering challenge is to make this clearance as small as possible without eliminating it entirely.

In a planetary gearbox, backlash accumulates from three sources: the clearance between sun gear and planet gear teeth, the clearance between planet gear and ring gear teeth, and the bearing play in the planet carrier. Each planetary stage adds its own backlash contribution, which is why multi-stage units have wider total backlash than single-stage units of the same precision grade.

Types of Planetary Gearbox 1 (1)

Three sources of backlash:
① Sun ↔ Planet tooth clearance
② Planet ↔ Ring tooth clearance
③ Planet carrier bearing play

What Is an Arcminute?

Backlash is measured in arcminutes (symbol: ′ or arcmin). One arcminute is 1/60 of a degree — a unit small enough to describe the precision of astronomical telescopes and surveying instruments. It became the standard measurement unit for planetary gearbox backlash because the angular errors involved are too small to express meaningfully in degrees.

UNIT CONVERSION REFERENCE
= 60 arcminutes
1′
= 0.01667°
1′
= 0.000291 radians
1′ @ 1 m
= 0.29 mm arc length

The last conversion — arcminutes to linear displacement — is the most practically useful. When you know the radius from the gearbox output shaft to the point where positioning accuracy matters (the tool tip, the robot end-effector, the laser focal point), you can convert backlash directly into millimetres of positional uncertainty.

From Arcmin to Millimetres — Real-World Positioning Impact

The formula to convert backlash (in arcminutes) to linear positional error (in millimetres) at a given arm radius is:

Error (mm) = Backlash (arcmin) × π ÷ 10,800 × Radius (mm)
Simplified: Error ≈ Backlash × 0.000291 × Radius

The table below pre-calculates the positional error for common backlash grades and working radii. Find your gearbox backlash specification in the left column and your application’s working radius across the top to read the expected positional uncertainty in millimetres.

Backlash @ 100 mm @ 200 mm @ 500 mm @ 1,000 mm Typical Series
≤3 arcmin 0.09 mm 0.17 mm 0.44 mm 0.87 mm EP-HAB
≤5 arcmin 0.15 mm 0.29 mm 0.73 mm 1.45 mm EP-PBL/PBF
≤8 arcmin 0.23 mm 0.47 mm 1.16 mm 2.33 mm EP-PL/PF, EP-WPBL
≤15 arcmin 0.44 mm 0.87 mm 2.18 mm 4.36 mm EP-PL/PF (2-stg)

Positional error = maximum arc-length displacement at the working radius during a torque direction reversal. Actual positioning repeatability also depends on servo loop stiffness, load compliance, and mechanical assembly tolerances.

Reading example: A robot arm with an EP-PBL/PBF reducer (≤5 arcmin) and a 500 mm arm length will produce no more than 0.73 mm of positional uncertainty at the end-effector during direction reversals. If this exceeds the application tolerance, upgrade to the EP-HAB series (≤3 arcmin) to reduce the error to 0.44 mm at the same radius.

How Is Backlash Measured in a Planetary Gearbox?

Backlash is measured by locking the input shaft stationary and applying a controlled torque to the output shaft in alternating clockwise and counter-clockwise directions. The angular displacement of the output shaft between the two contact positions is the total backlash, measured in arcminutes.

The measurement protocol matters. Industry standard practice (used by Korea Ever-Power and most reputable manufacturers) specifies:

Test load: 2% of the rated output torque applied in both directions
Measurement instrument: Precision angle encoder or dial indicator at the output flange face
Condition: Gearbox at room temperature (20±5°C) with standard factory lubrication
Direction: Measured at multiple angular positions of the output shaft; the maximum reading is reported

Some manufacturers test at higher loads (5% or 10% of rated torque), which compresses the backlash and produces a flattering but misleading lower number. Always verify the test load percentage when comparing backlash specifications between different brands.

Korea Ever-Power backlash testing centre measuring planetary gearbox precision with angle encoder

⚠ Specification trap: A datasheet that states “≤5 arcmin backlash” without specifying the test load is incomplete. If that same unit were tested at 2% of rated torque (the industry norm), its actual backlash might measure 7–8 arcmin. Always ask: “At what percentage of rated torque was the backlash measured?”

Choosing the Right Backlash Grade for Your Application

Not every application needs the tightest backlash available. Over-specifying precision is the most common source of unnecessary cost in servo reducer procurement. A ≤3 arcmin unit typically costs 3–5 times more than a ≤8 arcmin unit of the same frame size. The guide below maps common applications to the appropriate backlash grade.

APPLICATION → BACKLASH GRADE MAPPING
≤3′
Ultra-Precision
CNC rotary tables
Robot base joints
Optical stages
Chip shooters
≤5′
High Precision
Semiconductor
Laser cutting
AOI inspection
Medical devices
≤8′
Standard Precision
CNC feed axes
Packaging lines
Printing presses
AGV drives
≤15′
Economy
Conveyors
Winding machines
Mixers
Non-positioning

Korea Ever-Power covers all four tiers: the EP-PBL/PBF high precision series delivers ≤5 arcmin with IP65 sealing, while the EP-PL/PF standard series provides ≤8 arcmin at the most competitive cost per N.m of output torque. For the most demanding applications, the EP-HAB flagship reaches ≤3 arcmin with 50 N.m/arcmin torsional stiffness.

Precision planetary gearbox applications requiring different backlash grades — CNC, robotics, packaging

Does Backlash Increase Over the Gearbox Lifespan?

Yes — but the rate of increase depends heavily on gear quality, load level, operating temperature, and lubrication maintenance. In a well-made planetary gearbox with case-hardened steel gears (HRC 58–62), backlash growth over the rated 2,000-hour service life is typically less than 1 arcmin under proper loading conditions. Over an extended service life of 10,000+ hours at derated torque (80% of rated), cumulative backlash growth may reach 2–3 arcmin.

This is one of the key advantages of a planetary design over a worm drive. In a worm reducer, the softer bronze wheel wears measurably faster against the hardened steel worm, and backlash can double from the original specification within 5,000–8,000 operating hours. A planetary gear train with all-hardened-steel components maintains its original precision far longer, making it the preferred choice for applications where consistent accuracy over years of service is essential.

Factors that accelerate backlash growth
◆ Operating above rated torque consistently
◆ Frequent high-impact direction reversals
◆ Contaminated or degraded lubricant
◆ Excessive operating temperature (>90°C ambient)
◆ Misalignment between motor and reducer
Factors that preserve original precision
◆ Operating at ≤80% of rated torque
◆ Maintaining ambient temperature within spec
◆ Using the manufacturer’s specified lubricant grade
◆ Ensuring correct motor-to-reducer concentricity
◆ Periodic re-lubrication per maintenance schedule

How Manufacturers Achieve Low Backlash

The difference between a ≤15 arcmin economy unit and a ≤3 arcmin ultra-precision unit is not just tighter quality control — it requires fundamentally different manufacturing processes at every stage. Understanding these processes helps explain the cost differential between precision grades.

1

Gear tooth accuracy grade
Standard units use DIN/JIS Class 6–7 gear teeth (hobbed and shaved). Precision units require Class 4–5 (profile-ground after hardening). The grinding step removes heat-treatment distortion and achieves tooth profile errors below 3–5 μm, which directly reduces the clearance between mating teeth.
2

Matched planet gear sets
In a standard unit, planet gears are assembled from general production stock. In a precision unit, all three or four planet gears are individually measured and assembled as a matched set with total pitch variation below 2 μm, ensuring equal load sharing and minimum cumulative backlash.
3

Bearing preloading
Radial play in the planet carrier bearings contributes directly to output shaft backlash. Precision units use preloaded angular-contact or tapered-roller bearings that eliminate radial play, at the cost of slightly higher friction and bearing heat. This preloading alone can reduce total backlash by 1–2 arcmin compared to standard bearings with clearance.
4

Housing bore concentricity
The internal bore of the gearbox housing must be concentric with the output shaft centreline to within a few micrometres. In precision units, the housing bore is finish-ground in a single clamping operation to eliminate setup error between bearing seats. Any eccentricity here translates directly into uneven planet gear loading and increased effective backlash.
5

100% individual verification
Standard units may be batch-tested with statistical sampling. Precision units are 100% individually tested under load, with the measured backlash value recorded on a unit-specific test certificate. Any unit exceeding the specification is rejected — not regraded to a lower tier. This is the final quality gate that ensures the datasheet specification is met on every delivered unit.

Korea Ever-Power engineering team supporting planetary gearbox backlash specification and selection

When Backlash Does Not Affect Your Application

Not every servo drive application is sensitive to backlash in a planetary gearbox. Understanding when backlash is irrelevant prevents you from over-specifying and overpaying for precision that delivers no functional improvement to your machine.

Many engineers default to specifying the tightest available backlash grade on every planetary gearbox they purchase, reasoning that “lower backlash is always better.” In absolute engineering terms this is true, but in cost-benefit terms it is often wasteful. A ≤3 arcmin unit on a continuous-rotation conveyor drive delivers zero positioning advantage over a ≤15 arcmin unit — because the conveyor never reverses direction. The premium paid for that unnecessary precision is pure waste.

Backlash is irrelevant when:
1 The output rotates in one direction only — conveyors, fans, pumps, winding machines. Backlash only manifests during direction reversals; if there are no reversals, backlash has zero effect on performance.
2 The load is spring-loaded or gravity-loaded in one direction — vertical Z-axis drives under constant downward force keep the gear teeth engaged on one flank at all times, effectively eliminating backlash from the positioning equation.
3 The positional tolerance is much larger than the backlash effect — if your application tolerance is ±5 mm at the working radius and the backlash produces only ±0.5 mm of uncertainty, the gearbox precision is not the limiting factor. Specifying tighter backlash in this case provides zero functional benefit.
4 The servo controller compensates electronically — advanced servo amplifiers with backlash compensation algorithms can partially correct for known backlash values by pre-commanding an offset at each direction reversal. This does not eliminate backlash physically but can reduce its effect on positioning accuracy by 50–70% in well-tuned systems.

Backlash Comparison Across Korea Ever-Power Series

Korea Ever-Power offers five inline and right-angle planetary gearbox series spanning the full backlash range from ≤3 arcmin (ultra-precision) to ≤16 arcmin (standard right-angle). The table below consolidates the backlash specification for each series at single-stage and two-stage configurations, along with the output direction and IP protection grade, so you can identify the correct product family at a glance.

Series 1-Stage 2-Stage Output IP Tier
EP-HAB ≤3′ ≤5′ Inline 65 Flagship
EP-PBL/PBF ≤5′ ≤8′ Inline 65 Precision
EP-WPBL/WPBF ≤8′ ≤10′ 90° 65 Precision
EP-PL/PF ≤8′ ≤12′ Inline 54 Standard
EP-WPL/WPF ≤13–16′ ≤16–20′ 90° 54 Standard

Korea Ever-Power precision gear grinding workshop producing planetary gearbox components to achieve specified backlash grades

Frequently Asked Questions

▶ What does ≤5 arcmin backlash mean in practical terms?
Five arcminutes equals 0.083 degrees. At a 200 mm output arm radius, this produces a maximum positional uncertainty of 0.29 mm at the tool tip during torque direction reversals. At 500 mm radius, the uncertainty grows to 0.73 mm. For most servo-driven precision applications — laser processing, wafer handling, dispensing, inspection — this is well within the required tolerance envelope.
▶ Is “zero backlash” possible in a planetary gearbox?
Technically no. Some clearance between gear teeth is physically necessary for lubrication, thermal expansion, and assembly. However, specialised designs using split planet gears with spring preloading can achieve backlash below 1 arcmin, which is marketed as “zero backlash” or “backlash-free.” These units cost significantly more and are reserved for ultra-precision applications such as semiconductor lithography and optical alignment stages. For most industrial applications, ≤3 arcmin is considered the practical precision ceiling.
▶ Why does a two-stage unit have more backlash than a single-stage?
Each planetary stage contributes its own backlash from tooth clearances and bearing play. In a two-stage unit, the backlash from both stages adds together at the output. For example, if each stage has 4 arcmin of individual backlash, the total output backlash is approximately 8 arcmin. This is why precision-critical applications prefer to use the minimum number of stages needed for the required ratio — and why the EP-HAB flagship series is limited to 1–2 stages only, ensuring the ≤3 arcmin specification is met consistently.
▶ At what load is backlash tested?
The industry standard is to measure backlash at 2% of the rated output torque. This low test load ensures the measurement captures the true mechanical dead zone without compressing the tooth flanks. Some manufacturers test at higher loads (5–10%), which reduces the measured backlash and makes the specification appear better than it is. When comparing backlash across brands, always confirm the test load percentage to ensure you are comparing equivalent measurements.
▶ How much does tighter backlash cost?
Pricing scales roughly as follows: a ≤8 arcmin standard unit is the baseline. A ≤5 arcmin high precision unit at the same frame size and ratio typically costs 1.5–2 times the standard unit. A ≤3 arcmin ultra-precision unit costs 3–5 times the standard. The cost increase reflects tighter manufacturing tolerances, post-hardening gear grinding, matched planet gear sets, and 100% individual backlash verification at the factory.
▶ Does Korea Ever-Power provide individual backlash test certificates?
Yes. Every EP-PBL/PBF and EP-HAB unit undergoes 100% individual backlash verification at our Ansan-si test centre. An individual test certificate recording the measured backlash value, test load, and inspector identification is available on request at no additional charge. For the EP-PL/PF standard series, batch test certificates are standard; individual certificates are available on request for an additional fee.

Not Sure Which Backlash Grade You Need?

Send us your working radius, positional tolerance requirement, and application description. Our engineers will calculate the required backlash grade and recommend the most cost-effective series within one business day.

Get a Free Backlash Grade Recommendation →

Editor: Cxm