Use the tool first to get an immediate fit decision, then use the report layers to verify assumptions, understand risk boundaries, and choose the next engineering action.
Input your torque, ratio, duty, and backlash target to get a first-pass fit decision.
No result yet.
Run the checker to see stage recommendation, estimated efficiency, and next action.
Published families consistently map low ratio to 1-stage and high ratio to 3-stage (for example, PGII lists 3-10, 15-100, and 120-1000). Start with ratio architecture before micro-optimizing backlash.
Public planetary catalogs often show 20,000 h service life, but those values are tied to catalog test assumptions (load, duty, lubrication, temperature). Treat them as screening anchors only.
Planetary pages often quote backlash bands, while strain-wave vendors emphasize lost motion (e.g., measured at +/- 4% rated torque). Direct number-to-number comparison can be misleading.
Bevel stages are commonly published at 96-98% per stage, while worm stages with roughly i > 10 are explicitly flagged as lower-efficiency and startup-friction-sensitive.
ISO 6336 and AGMA rating methods are explicit that gear-rating formulae do not guarantee assembled drivetrain performance. Final release still needs full system validation.
When ratio, backlash, or torque demand crosses tool boundaries, the correct next step is engineering review with dimensional, stiffness, thermal, and life verification.
Numbers below are used as screening anchors. They are not universal guarantees and must be confirmed against target frame/model.
| Metric | Published Range / Example | Why It Matters | Source Family |
|---|---|---|---|
| Typical stage ratio windows | 1-stage: 3-10 | 2-stage: 10-100 | 3-stage: 100+ (PGII example: 3-10, 15-100, 120-1000) | Primary architecture split for quick sizing | APEX PGII series |
| Planetary catalog lifecycle envelope | Service life 20,000 h, -25 to +90 C, IP65 | Published life and protection values are conditional boundaries | Neugart PSFN technical data |
| Planetary catalog lifecycle envelope (alt source) | Service life 20,000 h, 0 to +90 C, IP65 | Cross-vendor check reduces single-source bias | APEX PII/PGII catalog |
| Strain-wave precision condition | Lost motion < 1 arcmin at +/- 4% rated torque; ratio 50:1-160:1 | Metric definition differs from common planetary backlash classes | Harmonic Drive technology + CSF-GH page |
| Cycloidal RV published robustness signal | Backlash/lost motion 1 arcmin; momentary torque up to 5x rated; life 6,000 h | Useful for shock-heavy alternatives, but application class differs | Nabtesco product guide |
| Bevel vs worm efficiency contrast | Bevel: 96-98% per stage (1:1 to 1:10); worm i > 10 lower efficiency | Topology choice changes thermal load and motor sizing | NORD bevel/worm engineering notes |
| Worm non-reversing boundary | Static locking mainly at i = 64+; startup efficiency penalty explicitly noted | Self-locking potential often trades off with duty efficiency | Bonfiglioli VF-W catalog |
This round focuses on evidence quality upgrades rather than rewriting structure. Each row tracks a specific gap and repair state.
| Gap | Why It Was Weak | Enhancement in Stage1b | Status |
|---|---|---|---|
| Service-factor explanation had weak standard boundary | Earlier version relied on heuristic narrative without formal rating-method scope. | Added ISO 6336 and AGMA scope limits and clarified screening-only intent. | Closed in this round |
| Lifecycle boundary was under-specified | No clear reminder that temperature/IP/lubrication assumptions gate life claims. | Added Neugart/APEX condition envelopes with explicit values and dates. | Closed in this round |
| Cross-topology comparison lacked measurement normalization | Backlash and lost motion were mixed without noting different measurement conditions. | Added Harmonic/Nabtesco measurement-context rows and applicability warnings. | Closed in this round |
| Commercial benchmark certainty was overstated | Cross-brand, same-duty price/life data are not publicly standardized. | Added explicit uncertainty table and minimum executable next-step path. | Open as public-data limitation |
These are explicit conditions that determine whether a published number is transferable to your project.
| Boundary Topic | Published Condition | Decision Impact | Source |
|---|---|---|---|
| ISO 6336 method boundary | Validated around pressure angle 15-25 deg, helix angle <=30 deg, contact ratio 1.0-2.5; outside this range needs confirmation. | Do not use quick-screen outputs as full certification for atypical geometry or vibration-heavy systems. | ISO 6336-1:2019 |
| AGMA rating method boundary | AGMA 2101 scope states rating formulas compare designs but do not assure assembled gearbox-system performance. | A "go" result still needs full drivetrain integration checks before final release. | ANSI/AGMA 2101-D04 scope |
| Planetary catalog life assumptions (Neugart) | PSFN publishes 20,000 h life with defined efficiency, backlash, temperature, and IP boundaries. | Treat catalog life as context-dependent, not a blanket field-life promise. | Neugart PSFN technical data |
| Planetary catalog life assumptions (APEX) | APEX PII/PGII publishes 20,000 h and 0 to +90 C with IP65 and synthetic grease assumptions. | Thermal and lubrication mismatch can invalidate nominal efficiency/life expectations. | APEX PII/PGII catalog |
| Strain-wave precision metric definition | Harmonic Drive cites zero-backlash concept, but numeric precision is reported as lost motion (<1 arcmin) at +/- 4% rated torque. | Normalize measurement definitions before comparing to planetary backlash classes. | Harmonic Drive technology note |
| Cycloidal robustness envelope | Nabtesco RV guide publishes 1 arcmin class and high shock tolerance (up to 5x rated torque) with 6,000 h rated life lines. | Strong for shock resistance, but lifecycle and application context differ from planetary catalogs. | Nabtesco product guide |
| Worm reversibility and startup behavior | Worm efficiency/reversibility depends on ratio and static/dynamic conditions; i=64+ often used for locking behavior. | If application needs frequent starts or high duty, verify startup-loss penalty early. | Bonfiglioli VF-W + NORD notes |
| Scenario | Good Fit Signal | Not-Fit Warning | Decision Note |
|---|---|---|---|
| Servo indexing, moderate ratio, medium precision | Planetary usually fits quickly | If ultra-tight repeatability is <1-2 arcmin without upgrade budget | Start with 1-2 stage screening and reduced-backlash options. |
| High ratio with compact envelope constraints | Planetary works if thermal margin is controlled | If duty and starts force oversized frames beyond envelope | Check stage split and derating before locking flange geometry. |
| Continuous heavy shock load | Possible with high service factor allowance | If procurement assumes nominal torque without load class margin | Service factor and torsional stiffness verification become mandatory. |
| Extreme zero-backlash positioning | Possible only with premium reduced-backlash classes | If project budget only supports standard backlash classes | Compare planetary premium classes vs harmonic/cycloidal alternatives. |
The tool combines boundary validation, stage architecture rules, duty-based service factor, and backlash class screening.
Boundary states intentionally stop overconfidence and route you to manual engineering review.
| Step | Logic | Output |
|---|---|---|
| Input normalization | Validate torque, ratio, duty, starts/hour, and backlash target; reject non-physical boundaries (e.g., >24 duty hours/day). | Clean numeric input with explicit fail-fast errors |
| Stage architecture selection | Auto-map ratio to 1/2/3-stage unless the user explicitly overrides stage mode. | Candidate stage count and baseline ratio corridor |
| Torque and service-factor checkpoint | Estimate output torque from motor torque × ratio × stage efficiency; scale with peak factor and duty-derived service factor. | Rated torque checkpoint for shortlist filtering |
| Backlash boundary screening | Compare target backlash to stage baseline bands; flag review/no-go when target is tighter than baseline class. | Go / review / boundary status with explicit next action |
This comparison is decision-oriented. Unknown/partial values are explicitly marked instead of guessed.
| Option | Strength | Tradeoff | Data Confidence | Typical Fit |
|---|---|---|---|---|
| Planetary gearbox | Balanced torque density and industrial availability | Backlash usually not the absolute minimum without reduced-backlash options | Multi-source numeric data available (ratios, efficiency, life envelopes) | General automation axes and OEM machines |
| Harmonic drive | Very low lost-motion class and high precision repeatability | Lost-motion metric is not directly equivalent to planetary backlash labels | Official ratio and lost-motion condition available; cross-brand lifecycle comparables are partial | Ultra-precision compact joints |
| Cycloidal reducer | Strong shock-load tolerance and low backlash class in robot-duty lines | Published life and load assumptions are often application-family-specific | Structured public tables exist but vary by series and duty | High-load robotic or indexing systems |
| Worm gearbox | Simple architecture and cost accessibility | Efficiency drops at high ratios and startup friction can penalize duty cycles | Good public non-reversing guidance; precision-class comparables are limited | Cost-sensitive, lower precision duties |
This table adds reproducible numeric anchors and explicitly states where metric definitions are not directly comparable.
| Option | Numeric Signal | Limit / Counterexample | Decision Use | Source Family |
|---|---|---|---|---|
| Planetary (precision class) | Typical published windows cluster around 3-10 (1-stage), 10-100 (2-stage), and 100+ (3-stage); PGII example family shows 3-10, 15-100, 120-1000. | Life and efficiency are catalog-condition values, not guaranteed field outcomes. | Best first-pass architecture when ratio + torque density + compactness all matter. | APEX + Neugart |
| Harmonic / strain-wave | Servo gearheads commonly 50:1-160:1; lost motion <1 arcmin is reported under specified torque test conditions. | Lost motion and backlash are not interchangeable metrics without test-condition normalization. | Use when ultra-tight positioning dominates and budget accepts higher precision topology cost. | Harmonic Drive |
| Cycloidal (RV class) | Published lines include backlash/lost motion around 1 arcmin with momentary torque envelopes up to 5x rated. | Many public data sets are robotics-oriented; direct transfer to all industrial duties needs validation. | Use when shock-load resilience and torsional robustness dominate. | Nabtesco |
| Worm and bevel right-angle options | Bevel stages are published around 96-98% per stage; worm high-ratio stages are explicitly flagged lower efficiency with possible self-locking behavior. | Worm reversibility depends on ratio, friction, and vibration context; static locking is not universal. | Use worm when passive holding is valuable; use bevel when duty efficiency is critical. | NORD + Bonfiglioli |
Risks are grouped by misuse, cost, and scenario mismatch so each has an executable mitigation.
| Risk | Trigger | Impact | Mitigation |
|---|---|---|---|
| Precision misfit risk | Backlash target defined late or ignored during ratio selection | Positioning error and rework after pilot | Freeze backlash budget at concept stage and validate class before PO. |
| Thermal/rated torque risk | Nominal torque used without duty and service-factor uplift | Overheating or premature wear under real cycle load | Apply duty/load-class factor and run conservative scenario before shortlist. |
| Metric mismatch risk | Backlash, lost motion, and repeatability numbers compared without matching test conditions | False precision ranking and wrong topology lock-in | Normalize measurement conditions and request test method notes in RFQ. |
| Layout interface risk | Inline vs right-angle chosen by packaging only | Mounting, stiffness, and maintenance tradeoffs discovered too late | Review interface, stiffness, and access constraints together in pre-RFQ checklist. |
| Procurement expectation risk | Tool result interpreted as final model release | Commercial commitment before engineering closure | Treat quick screen as gate-0 only; require engineering sign-off for final BOM. |
| Price benchmark overconfidence risk | Assuming cross-brand same-duty price/performance is publicly comparable from open catalogs | Premature sourcing decision and hidden lifecycle cost | Run controlled RFQ template with identical duty profile; mark open-web pricing as non-decision evidence. |
No forced conclusions are made for gaps without reliable public data. Each row includes a minimal next action.
| Topic | Current Status | Why Uncertain | Minimum Next Step |
|---|---|---|---|
| Cross-brand normalized price benchmark | No reliable public normalized dataset found in this round | Open sources report model-level specs, but rarely disclose same-duty transactional price with identical commercial terms. | Collect at least 3 vendor RFQs under one unified duty template before final cost ranking. |
| Cross-topology same-duty lifecycle benchmark | Public evidence remains partial | Published life values (e.g., 20,000 h vs 6,000 h) use different series contexts and rating assumptions. | Request fatigue/life assumptions and derating logic per vendor, then compare under a common duty-cycle sheet. |
| Scenario | Premise | Process | Outcome |
|---|---|---|---|
| Pick-and-place axis retrofit | Need 10:1 ratio, moderate duty, <=8 arcmin target, compact inline envelope | Tool yields 1-stage or 2-stage candidate depending on torque margin. Team runs conservative mode to test thermal headroom. | Shortlist narrowed to two frame sizes; engineering review requested for flange fit before sample PO. |
| High-cycle packaging line | Starts/hour spikes during indexing; procurement initially used nominal torque only | Service-factor uplift in tool changes rated torque checkpoint and pushes one frame size up. | Avoided undersized selection and reduced pilot failure risk at commissioning. |
| Ultra-tight motion budget project | Backlash target below standard class while ratio requirement stays high | Tool flags review/no-go boundary and recommends reduced-backlash class or alternative topology review. | Project avoids false certainty and starts topology comparison earlier. |
Is this tool a final model selector?
No. It is a fast screening layer for architecture and risk gating. Final selection still requires engineering confirmation.
Why does stage count matter so much?
Stage count shifts both efficiency and backlash baseline. It changes thermal margin and achievable precision class at the same time.
Can I force stage count manually?
Yes. Manual stage mode is useful for what-if checks, but forced settings can violate ratio/backlash feasibility.
What if my ratio is outside 3 to 1000?
The checker enters boundary mode and asks for manual engineering sizing to avoid false confidence.
How should I interpret backlash numbers?
Use backlash as a motion-budget input, not a marketing label. Arcmin values must be tied to endpoint displacement tolerance.
Why can a low backlash target trigger review even when torque is feasible?
Because standard stage classes may not meet the target. Precision class upgrades or topology alternatives can be required.
Why do balanced and conservative modes differ?
Conservative mode increases service-factor pressure and lowers efficiency assumptions to expose risk earlier.
What is the most common sizing mistake?
Locking frame size on nominal torque without duty-cycle and starts-per-hour uplift.
What should be included in RFQ after using the checker?
Include torque profile, ratio, backlash target, duty hours/day, starts/hour, motor model, and layout constraints.
When should we compare against harmonic or cycloidal?
When target backlash is significantly tighter than standard planetary classes or when stiffness/shock priorities dominate.
Can right-angle and inline be treated as interchangeable?
Not safely. They often differ in envelope, interface behavior, and torque packaging constraints.
How do we reduce project risk after a go result?
Run conservative-mode confirmation, complete interface checks, and require engineering sign-off before purchase release.
Source-backed fields are listed below with checkpoint date. Any non-source value is explicitly treated as heuristic.
| Source | Checkpoint Date | Data Used | Link |
|---|---|---|---|
| ISO 6336-1:2019 (official scope page) | Snapshot checked: 2026-05-06 | Method scope limits and non-applicability boundaries | https://www.iso.org/standard/63819.html |
| ANSI/AGMA 2101-D04 scope page | Snapshot checked: 2026-05-06 | Design-rating scope and "not assurance of assembled system" statement | https://members.agma.org/ItemDetail?Category=STANDARDS&iProductCode=2101_D04 |
| AGMA technical publications catalog (member PDF) | Snapshot checked: 2026-05-06 | Current ANSI AGMA 2101-E25 listing and abstract | https://members.agma.org/common/Uploaded%20files/__AGMA%20Publications%20Catalog.pdf |
| APEX Dynamics PGII product page | Snapshot checked: 2026-05-06 | Stage ratio sets and 20,000 h-life context | https://www.apexdyna.nl/en/products/industrial-planetary-gearboxes/pgii-series |
| APEX Dynamics PII/PGII catalog PDF | Snapshot checked: 2026-05-06 | Efficiency, service life, operating temperature, IP65, ratio grids | https://apexdynamicsusa.com/pub/media/sebwite/productdownloads/p/i/pii_piir-eng_1_6.pdf |
| Neugart PSFN product page | Snapshot checked: 2026-05-06 | Efficiency, temperature, IP65, backlash classes, ratio lists | https://www.neugart.com/en-us/gearboxes/precision-gearboxes/psfn |
| Neugart PSFN 2025 chapter PDF | Snapshot checked: 2026-05-06 | 20,000 h service life and technical boundary fields | https://cdn.neugart.com/fileadmin/user_upload/Downloads/Catalog_Chapters/11_2025/PSFN/PSFN_2025_11_EN.pdf |
| Harmonic Drive technology page | Snapshot checked: 2026-05-06 | Zero-backlash claim and <1 arcmin lost motion measurement condition (+/-4% rated torque) | https://staging.harmonicdrive.net/technology |
| Harmonic Drive CSF-GH product page | Snapshot checked: 2026-05-06 | Typical ratio range 50:1 to 160:1 for servo gearheads | https://www.harmonicdrive.net/products/servo-mount-gearheads/harmonic-drive/csf-gh |
| Nabtesco RV-E page | Snapshot checked: 2026-05-06 | Backlash/lost-motion class statements (<1 arcmin) | https://precision.nabtesco.com/en/products/detail/RV-E |
| Nabtesco product guide PDF | Snapshot checked: 2026-05-06 | Backlash/lost-motion tables, rated service life rows, shock/load capacity lines | https://precision.nabtesco.com/en/download/pdf/Product_guide_en.pdf |
| NORD bevel vs worm engineering blog | Snapshot checked: 2026-05-06 | Bevel efficiency 96-98%/stage and worm high-ratio efficiency caveats | https://www.nord.com/en/nord-group/current-events/blog/angled-gear-units-bevel-gear-or-worm.jsp |
| NORD bevel gear unit blog | Snapshot checked: 2026-05-06 | Bevel stage ratio 1:1 to 1:10 and non-self-locking boundary | https://www.nord.com/en/nord-group/current-events/blog/bevel-gear-units.jsp |
| Bonfiglioli VF-W catalog | Snapshot checked: 2026-05-06 | Worm non-reversing boundary (i=64+), dynamic/static efficiency, startup caveat | https://www.bonfiglioli.com/_default_upload_bucket/BR_CAT_VF-W_IE2-IE3_ENG_R11_5_1.pdf |
After screening on this page, continue with adjacent decision modules based on your project stage.
Inquiry Email