The Five-Step Selection Framework at a Glance
A precizni planetarni mjenjač sits directly between your servo motor and the machine load. Every mismatch in that interface — torque, inertia, configuration, or IP rating — is amplified through every cycle the machine runs. The five-step process below is the minimum rigorous approach. Steps 1 and 2 are where most early failures originate; Steps 4 and 5 are where installation problems begin.
Step 1 — Define Your Load Profile and Duty Cycle
Most engineers start a planetarni mjenjač selection by asking what the rated continuous torque of their servo motor is, and then directly match a gearbox to that number. That approach is incomplete. What the gearbox must actually survive is the full shape of the torque demand over time — not just the average.
Before calculating a single number, document the following four elements of your load profile:
The torque the load demands during sustained steady-state operation. For a robot arm at constant velocity, this is the gravitational torque plus friction. This value sets the thermal sizing floor.
The maximum torque demanded during acceleration, deceleration, or impact. For servo axes with fast positioning cycles, this is often 2–4× continuous torque. The gearbox instant stop rating must exceed this.
IEC and DIN standards classify shock loads into three levels. Light shock (uniform conveyor belt) applies SF=1.0–1.25. Moderate shock (indexing table with direction reversals) applies SF=1.5–2.0. Heavy shock (impact press, robot collision stop) applies SF=2.0–2.5.
The percentage of each cycle during which the motor applies torque. A 60% duty cycle with a 5-second period means 3 seconds on, 2 seconds off. This determines thermal load on the gearbox and lubricant, especially in sealed lifetime-lubricated units.
| Application Type | Shock Class | Typical ED% | Recommended SF |
|---|---|---|---|
| Single-direction conveyor, fan, pump | Light | 80–100% | 1.0–1.25 |
| AGV drive wheel, packaging line servo axis | Light–Moderate | 50–80% | 1.25–1.5 |
| CNC rotary axis, indexing table, robot arm joint | Moderate | 30–60% | 1.5–2.0 |
| Press line transfer, collision-rated robot axis | Moderate–Heavy | 20–50% | 2.0–2.5 |
| Servo press main drive, heavy-impact transfer | Heavy | <30% | 2.5+ |
Step 2 — Calculate Required Output Torque with Service Factor (The Step Most Engineers Skip)
The service factor (SF) is not a bureaucratic safety margin added by cautious engineers. It accounts for three real physical phenomena that a simple rated-torque calculation cannot capture: load variations that are faster than the servo’s closed-loop response, thermal effects on lubricant film strength under varying duty cycles, and duty cycle asymmetries between acceleration and deceleration phases that create cumulative bearing fatigue loads exceeding what steady-state continuous torque implies.
Skipping the service factor is the single most common cause of early-life gearbox failure in servo automation systems, responsible for approximately 40% of premature failures in high-cycle servo applications.
Worked Example — Automotive Transfer Robot J2 Arm Axis
A Korean automotive body-shop supplier needs a servo gearbox for a 6-axis transfer robot’s J2 (large-arm) joint. The servo motor is a 1.5 kW unit rated at 3,000 rpm. The machine cycle involves rapid positioning with direction reversals (Moderate–Heavy shock class). Service factor selected: SF = 2.0.
Without SF, the engineer selects a gearbox rated for 71.9 N·m — a unit in the EP-ZDE-60 range. At the actual peak torque during emergency braking (estimated 2× continuous = 143.8 N·m), the gearbox operates at 200% of its rated load every time the servo triggers an emergency stop. After a few thousand such events, planet gear flank pitting initiates. Backlash grows. By month eight the axis develops oscillation and a full gearbox replacement is required. This is not a hypothetical — it is the documented failure pattern of the Korean Tier-1 case referenced in the introduction.
Step 3 — Gear Ratio Selection and Inertia Matching
The gear ratio of a servo planetarni mjenjač determines two things simultaneously: the output shaft speed and the reflected inertia of the load as seen by the motor. Getting the torque right but misjudging inertia means your servo drive will struggle to tune correctly — and may oscillate, overshoot, or trigger overcurrent faults under rapid acceleration even with a mechanically adequate gearbox.
The table below shows how a change in gear ratio transforms the same load inertia into dramatically different reflected values at the motor shaft. This is why ratio selection is not just a speed calculation — it is the primary lever for matching servo motor to mechanical load.
| Gear Ratio i | Stage | J_reflected (kg·m²) * | Inertia Ratio | Servo Tuning Status |
|---|---|---|---|---|
| 3:1 | 1 | 0.00222 | 2.2 : 1 | ✅ Ideal |
| 5:1 | 1 | 0.000800 | 0.8 : 1 | ✅ Good |
| 10:1 | 1 | 0.000200 | 0.2 : 1 | ⚠️ Over-geared, slow response |
| 20:1 | 2 | 0.000050 | 0.05 : 1 | ❌ Torque underutilised, poor response |
* Example: J_load = 0.02 kg·m², J_motor = 0.001 kg·m². Actual values depend on your specific load geometry and motor specification.
The servo drive’s velocity feedback loop Kv gain is effectively limited. The axis responds sluggishly to velocity commands and overshoots on position stops. Increasing the proportional gain to compensate causes mechanical resonance — a problem software alone cannot fully solve because it originates in the physics of the drivetrain inertia mismatch.
For ratios in this range, a single planetary stage (EP-ZDE/ZDF/ZDWE/ZDWF, 1-stage) provides 96% efficiency (inline) or 94% efficiency (right-angle input). This is the preferred range for high-dynamic servo axes — CNC feed axes, laser cutting heads, and pick-and-place robots — where both inertia ratio and efficiency matter equally.
Two-stage units are appropriate when output speed must be very low (<200 rpm) at rated motor speed. Efficiency drops to 94% (inline) or 92% (right-angle). Acceptable for AGV drive wheels, pallet changers, and solar trackers where efficiency loss is less critical than the high ratio for torque multiplication. Backlash is slightly wider than single-stage.
Step 4 — Choose the Right Configuration (Inline vs Right-Angle, Round vs Square Flange)
The Korea Ever-Power EP series of precision planetary gearboxes offers four physical configurations across five product lines. Each solves a specific combination of installation constraints. This is a structural decision — not a performance preference — driven by your machine geometry and available machine shop operations.
| Serija | Motor Input | Output Flange | Max Torque | IP adresa | Best For |
|---|---|---|---|---|---|
| EP-ZDE | Inline | Round Φ | 800 N·m | IP54 | Standard precision servo axes — CNC, robot, laser cutter |
| EP-ZDF | Inline | Square □ | 800 N·m | IP54 | Plate-mount frames — no boring needed |
| EP-ZDWE | 90° bevel | Round Φ | 800 N·m | IP54 | 30–50% shorter axial depth — compact machine heads |
| EP-ZDWF | 90° bevel | Square □ | 800 N·m | IP54 | AGV/AMR low-profile chassis, welded frames |
| EP-ZDS | Inline | Square □ | 1,800 N·m | IP65 | Heavy robot joints, press drives, food processing, washdown |
Right-angle input efficiency trade-off (ZDWE/ZDWF): The 90° bevel gear input stage adds approximately 2% efficiency loss compared to an inline unit of the same frame size. For a 750 W servo motor running 16 hours per day, this equates to approximately 15 W additional heat generation — negligible for most applications. For continuous 24/7 high-power operation, verify thermal budget using the formula: P_heat = P_input × (1 − η), where η = 0.92 for ZDWE/ZDWF two-stage.
Step 5 — Motor Interface Verification: The 12-Point Checklist
A precision planetary gear reducer correctly sized for torque, ratio, and configuration can still fail in service within weeks if the motor-to-gearbox interface is improperly specified. Interface errors typically manifest as elevated vibration, early input bearing failure, and in severe cases, input shaft coupling fracture. This 12-point checklist covers every dimension of the motor-gearbox interface that must be verified before order placement.
Backlash Specification — Matching Precision Grade to Application Requirement
Once torque, ratio, and configuration are confirmed, verify that the backlash specification of the selected precision planetary gearbox is appropriate for your positioning accuracy requirement. Backlash is the angular play at the output shaft when the input direction reverses — measured in arcminutes (arcmin), where 1 arcmin = 1/60th of a degree.
Do not over-specify backlash. A unit with <1 arcmin backlash may cost 3–5 times more than a <8 arcmin unit of the same frame size, with no measurable performance benefit in applications that position in a single direction or where the servo closed-loop compensates for the backlash contribution. Match the specification to the actual requirement:
Three Sizing Errors That Lead Directly to Early Failure
The most frequent error. A gearbox rated at the calculated steady-state output torque appears to match on paper. At the first emergency stop or direction reversal under full load, the actual torque spikes to 2–3× continuous. Without SF, the unit is operating at 200–300% of its design point. After several thousand such events, planet gear surface fatigue initiates and backlash begins to grow rapidly.
When load inertia reflected to the motor exceeds five times the motor rotor inertia, the servo velocity loop becomes difficult to tune. Engineers who push the proportional gain up to compensate create mechanical resonance — a problem that manifests as axis oscillation, audible vibration, and ultimately early planet carrier bearing fatigue from cyclic overload at the resonant frequency. Software filters help but cannot fully resolve the underlying mechanical mismatch.
An IP54-rated planetarni mjenjač resists water splashing from any direction — but it does not protect against a direct water jet. Korean food-processing facilities under HACCP protocols apply high-pressure hose washing to all machine surfaces including gearboxes. Over 6–18 months, even IP54-rated lip seals degrade under repeated chemical cleaning cycles. Water ingress emulsifies the lifetime lubricant, destroying the grease film and dramatically accelerating bearing wear. The gearbox housing temperature rises, the noise increases, and the rated 20,000-hour lifespan may be achieved in under 5,000 hours.
Selection Summary and Next Steps
Korea Ever-Power’s application engineering team provides gearbox selection support — including service factor verification, inertia ratio calculation, and motor interface confirmation — in Korean and English for Korean OEM manufacturers. Provide your servo motor model, load parameters, and installation constraints to receive a complete selection recommendation at no charge.
Urednik: Cxm
