
A fiber optic splitter network design is ready for procurement only when its topology, shortest and longest optical paths, active-equipment limits, splitter interfaces, host equipment, acceptance method and document revision agree. A topology sketch becomes a release design only after assumptions, supplier inputs, test limits and as-built records have defined owners.
Use this guide to control that handoff. For ratio selection, see the GPON splitter selection guide. For a worked loss-budget method, use the fiber loss-budget guide. This page owns the process that connects those decisions to the RFQ, acceptance plan and as-built record.
Already have a topology? Send the revision, endpoint models and shortest/longest routes for a configuration gap review before turning the drawing into an RFQ.
Quick answer
Design a splitter network by freezing the equipment and service basis, mapping route constraints, selecting splitter placement by zone, checking the shortest and longest paths plus capacity, mapping every splitter to its host, and releasing one controlled document set. Issue the RFQ from that revision, control supplier deviations, then reconcile component tests and installed-path records into the as-built design.
Use this eight-step fiber optic splitter network design process:
- Freeze the service, endpoint and migration basis.
- Map route, zone and environmental constraints.
- Choose centralized, cascaded or hybrid splitter placement.
- Check the shortest path, longest path and service capacity separately.
- Match each splitter configuration to its host equipment and interfaces.
- Build one revision-controlled design and acceptance pack.
- Issue a comparable RFQ and control supplier deviations.
- Reconcile component tests and installed-path records into the as-built design.
Buyer summary
Before procurement, require one controlled revision that identifies the endpoint models and optical class, shortest and longest routes, every splitter stage, host equipment, connector interfaces, test method, acceptance limits and unresolved assumptions. Compare suppliers only against that same revision and evidence scope. A lower unit price is not equivalent if the offer changes the splitter package, omits drawings or tests, substitutes interfaces or leaves delivery documents undefined. Hold award until optical limits, physical fit, sample purpose, deviation handling, quantity, destination, lead time and commercial basis are explicit.
Key release conclusions
– A topology sketch is not a procurement baseline; the evidence source and owner of each open field must be recorded.
– The shortest path, longest path and service capacity are separate release gates.
– A splitter model must be checked with its host enclosure, interfaces and project test method, not as an isolated component.
– Supplier quotations are comparable only when configuration, evidence, approval stages and commercial basis match.
The release rule: do not procure from a topology sketch
Put the design on HOLD when any of these items is still implicit:
- the OLT and ONT/ONU models or implemented optical class;
- the shortest and longest physical paths;
- every splitter stage, coexistence device, connector interface and splice allowance;
- the splitter port map and the host enclosure at each node;
- the source of each optical and mechanical limit;
- the test method, wavelengths, reference method and acceptance limit;
- the revision that suppliers, installers and reviewers are using;
- the owner of unresolved assumptions or deviations.
Until every open item has an owner and an evidence source, the drawing is not an RFQ baseline.
1. Freeze the design basis
Start with a design-basis register. It prevents a planning assumption from becoming a purchase specification simply because it appeared in an early drawing.
| Design-basis field | What to record | Evidence owner |
|---|---|---|
| Service and capacity | Subscriber or endpoint forecast, service profiles, busy-hour assumptions, protection requirement | Network planning |
| Active equipment | OLT and ONT/ONU models, implemented optical class, transmitter and receiver limits, supported split and distance constraints | Active-equipment owner |
| Routes | Shortest and longest paths by zone, fiber sections, existing assets, inaccessible or uncertain segments | Outside-plant or building design |
| Passive architecture | Splitter stages, nominal ratios, location IDs, coexistence devices and test access | ODN designer |
| Environment | Indoor/outdoor boundary, enclosure responsibility, temperature, water, dust, UV, fire and access requirements | Site/civil/building owner |
| Operations | Activation, fault isolation, restoration, spare policy, labeling and record system | Operations |
| Commercial boundary | Quantity, phase, destination, schedule, required evidence and permitted alternatives | Procurement |
For G-PON, verify the implemented configuration against ITU-T G.984.2. For XGS-PON, use the in-force ITU-T G.9807.1 and applicable amendments. The standard identifies system requirements; it does not prove that a named OLT/ONT combination, splitter model or field path has been approved.
Treat migration and coexistence as separate architecture decisions
A splitter with a broad operating band does not by itself make an ODN ready for GPON/XGS-PON coexistence. ITU-T G.9805 covers coexistence methods that may use an external coexistence element, an integrated multi-PON module or a splitter-based method. The chosen equipment, filtering/isolation, added loss, wavelength plan and operating class still need project-specific verification.
Record the intended migration state in the design basis:
- current PON generation and endpoint population;
- target generation and migration sequence;
- coexistence method and component responsibility;
- added passive losses and required isolation evidence;
- legacy component records that are missing or uncertain;
- retest and rollback conditions.
2. Map routes and zones before choosing splitter placement
Density is one input, not a topology rule. Map take rate, available feeder and distribution fibers, cabinet or riser access, restoration practice, optical-path limits, construction phases and the migration plan for each zone.
If the reader first needs the network-layer context, use the ODN architecture guide and FTTH architecture overview. This article starts where those concept pages stop: converting the selected architecture into a controlled splitter design.
| Zone question | Centralized placement may fit when… | Cascaded or distributed placement may fit when… |
|---|---|---|
| Fiber availability | Distribution fibers and pathways can support home-run branches | A smaller downstream fiber count materially reduces route constraints |
| Access and testing | One controlled hub is accessible for patching and fault isolation | Remote nodes are accessible and their records can be maintained reliably |
| Take rate and phasing | Ports can be allocated flexibly from a common location | Phased zones need local activation without building all downstream fibers at once |
| Restoration | Central replacement and rerouting are operational priorities | Localized faults and shorter downstream restoration paths are more important |
| Optical checks | One-stage model limits close on every path | Each cascade path closes with all stages and interfaces included |
| Space and host equipment | A hub can accept the splitter, adapters, routing and labels | Field terminals or building boxes can safely host the required modules and fiber management |
A mixed design is valid when the evidence differs by zone—for example, a controlled central hub in one area and an accessible downstream node where feeder pathways are constrained. Record the approved reason at zone level; a network-wide slogan is not a design decision.
Adapt the same control method to FTTH, FTTB and passive optical LAN
| Context | Design emphasis | Evidence that should reach the release pack |
|---|---|---|
| FTTH | Subscriber take rate, feeder/distribution/drop boundaries, terminal capacity, activation and restoration | Zone map, terminal schedule, shortest/longest paths, spare policy and address-to-port mapping |
| FTTB/MDU | Building entry, riser or floor access, indoor requirements, ONT/ONU placement and responsibility for the in-building network | Building topology, pathway and fire requirements, equipment-room/riser interfaces and floor/tenant mapping |
| Passive optical LAN | Vendor/system architecture, work-area endpoint count, riser/horizontal pathways, redundancy and operational model | Approved system architecture, endpoint schedule, splitter locations, pathway record and equipment-specific loss/capacity limits |
Passive optical LAN designs vary by system. Do not assume that a neighborhood FTTH splitter plan can be copied into a campus simply because both use passive splitters.
Keep cable construction in the design basis without turning this page into a cable-selection guide. Use the ADSS cable selection page for a self-supporting aerial segment, the FTTH drop-cable page for the subscriber segment, and the armored/direct-burial approval guide where burial or mechanical exposure controls the construction. Equipment, panel and test-interface jumpers belong with the fiber patch-cord owner. A data-center link is relevant only when the project has a documented campus or data-center PON boundary.
Keep fiber and cable requirements tied to the route
The G-PON and XGS-PON paths covered here use single-mode fiber. ITU-T G.652 and ITU-T G.657 define fiber and cable attributes; they do not approve a finished cable for a route. A project may use G.652.D in feeder or distribution sections and specify G.657.A1 or G.657.A2 where tighter routing is required, but the exact category, splice compatibility and finished-cable limits must be confirmed. A G.657 macrobending test condition is not the installation or long-term bend radius of the finished cable.
OM3, OM4 and OM5 are multimode categories covered by IEC 60793-2-10. They are not substitutes for the single-mode PON path. Include them only when the project contains a separate premises, industrial or data-center optical link, then design that link under its own interface and cabling requirements.
| Route or interface decision | Relationship that must remain clear | Release evidence |
|---|---|---|
| Aerial self-supporting section | ADSS is an all-dielectric self-supporting cable family; span, loading, sag/tension, hardware and the electrical environment control selection | Route/loading data, cable and hardware drawings, installation method and applicable IEC 60794-4-20 or owner specification |
| Duct section | Duct placement does not define the cable by itself; pulling tension, bend limits, crush, water exposure, duct fill and access still need project limits | Route survey, pulling method and model-matched mechanical/environmental data |
| Direct-burial or high-risk section | Direct burial is an installation condition; armor is one construction option, not a synonym. Metallic elements may introduce bonding and grounding decisions | Burial environment, crush/water/rodent requirements, construction drawing and bonding/grounding review where applicable |
| Subscriber drop and building entry | Drop-cable construction must match outdoor exposure, entry transition, indoor pathway, termination and the authority having jurisdiction | Approved route, entry detail, local fire/building requirement and finished-cable data |
| Loose-tube or tight-buffered construction | These terms describe fiber protection and cable construction; neither one proves aerial, duct, burial, indoor or fire suitability | Exact cable construction plus route-specific qualification |
| Jacket and environmental claims | PE, LSZH, UV resistance, water blocking, flame spread, smoke, halogen and rodent protection are separate properties | A requirement-to-evidence matrix with the exact test, classification, edition and cable model |
| Fiber count and spare capacity | Fiber count follows service, restoration and growth decisions; spare fibers, splitter ports and pathway space are not interchangeable | Capacity forecast, restoration policy and labeled spare schedule |
IEC 60794-1-1:2023 provides generic optical-cable property categories, while product-family standards and project specifications provide the applicable detail. For an indoor/outdoor transition, IEC 60794-6:2020 treats outdoor robustness, indoor fire performance, bend behavior, UV and humidity as connected but distinct requirements. An LSZH material statement alone does not prove flame spread, smoke performance, UV resistance, water resistance, rodent protection or compliance with a country’s building rules.
Regional differences should enter the design as named requirements, not stereotypes. Record the authority having jurisdiction, operator specification, climate/loading data, indoor fire classification, language/marking requirements, document set and acceptance rules for the actual destination. For a premises cabling scope, ISO/IEC 11801-1 or the contractually required TIA-568 provisions may apply; neither replaces the PON equipment class, optical-path or ODN acceptance checks.
Localize the procurement pack without treating a region as a specification
Use the region only to prompt questions. The contract, authority having jurisdiction, network owner and actual route decide the requirement.
| Destination market | Environment and installation questions to confirm | Standards and document questions to confirm | Commercial and delivery questions to confirm |
|---|---|---|---|
| North America | Aerial, underground, building or mixed route; pole/duct ownership; loading, fire and access conditions | Operator/AHJ requirements, contractual standards and editions, labeling language, test-record format | Ship-to site, freight basis, phased deliveries, approval lead time and required import documents |
| Europe | Country, indoor/outdoor boundary, pathway, fire classification and environmental exposure | Applicable national/EU requirements, cited EN/IEC/ISO editions, declaration/certificate scope and language | Destination, Incoterm, packaging/recycling obligations if contractual, document approval and delivery window |
| Middle East | Temperature, solar exposure, dust, saline/coastal exposure, flooding risk, route access and enclosure responsibility where applicable | Utility/operator specification, authority approvals, document language, inspection/test witnessing and exact certificate scope | Port/site delivery, approval cycle, packing protection, partial shipment and site-storage constraints |
| Southeast Asia | Humidity, heavy rain or flood exposure, UV, aerial/duct/building route, dense access constraints and storage conditions where applicable | Country/operator requirements, fire rules for indoor sections, document language and local approval evidence | Customs documents, multi-site drops, monsoon or access scheduling where relevant, packaging and storage handoff |
| Latin America | Aerial/duct/building route, pole access, UV, rain, coastal or high-altitude conditions where applicable | Country/operator specification, contractual IEC/ITU/TIA references, Spanish/Portuguese documentation and test format as required | Importer-of-record responsibility, inland delivery, phased quantities, labels and approval timing |
| Africa | Urban/rural route, aerial/duct/direct-burial sections, UV, heat, dust, rain/flood or long storage/transport exposure where applicable | Country/operator/financier specification, required standard editions, inspection, language and traceability documents | Port and inland route, customs responsibility, split deliveries, storage protection, spares and remote-site handoff |
For any destination, keep the RFQ and quotation on the same Incoterm, currency, packing unit, shipment mode, destination, approval sequence and evidence scope. Treat OEM/ODM, private labeling, bulk capacity, certificates, samples and fixed lead times as supplier-specific fields that require written confirmation; the region alone proves none of them.
3. Centralized vs cascaded splitter network design: choose placement by zone
For each zone, record the candidate architectures and why one was approved. A compact decision record should include:
- candidate split stages and locations;
- required feeder, distribution and drop fibers;
- shortest and longest path implications;
- host-enclosure and access requirements;
- activation, testing and restoration workflow;
- capacity and take-rate implications;
- change cost if the take rate or route plan changes;
- unresolved evidence and the named owner;
- approved option, approver and revision.
Avoid rules such as “dense always means centralized” or “long routes always mean cascaded.” A topology that reduces one cable count can add a second splitter stage, more interfaces, more field records and a different restoration burden. The design record should show which tradeoff the network owner accepted.
4. How to check a splitter network design: shortest path, longest path and capacity
“Below the maximum budget” is not a complete release test. Check every relevant direction and wavelength using the implemented equipment configuration, and check both ends of the permitted optical-path-loss window.
Longest path: build an upper-bound loss
Include, as applicable:
- maximum insertion loss for every selected splitter stage;
- maximum cabled-fiber attenuation at the project wavelengths;
- connector and splice allowances defined by the project;
- coexistence/WDM devices and other passive components;
- a design reserve defined by the network owner.
Use the selected model’s maximum insertion loss, not ideal split loss or a category average. Keep design reserve as a separate, owner-approved entry that states what uncertainty it covers.
Shortest path: check minimum loss and receiver limits
A short branch or low split can fall below the permitted minimum loss or approach a receiver-overload condition. Estimate the lowest credible installed loss using the actual endpoint models and configuration. If attenuation is required, use an approved engineering rule and document the component and test impact.
Capacity: keep optical and service checks separate
An optical calculation does not prove that the PON port meets service requirements. Record endpoint count, service profiles, busy-hour assumptions, OLT limits, protection policy and growth state separately. The final split ratio must pass both the optical and capacity decisions.
Use this release table for each PON tree or building segment:
| Gate | Shortest path | Longest path | Controlling evidence | Status |
|---|---|---|---|---|
| Endpoint models / implemented class | Enter model and class | Enter model and class | Approved equipment configuration | HOLD until approved |
| Physical route and fiber | Enter route and fiber basis | Enter route and fiber basis | Route/as-built record and cable requirement | HOLD until traced |
| Splitter stages and maximum IL | Enter every stage and model limit | Enter every stage and model limit | Model-matched datasheet | HOLD until model matched |
| Other passive interfaces | Enter all included interfaces | Enter all included interfaces | Interface schedule and project allowances | HOLD until complete |
| Minimum / upper-bound physical loss | Enter calculated minimum | Enter calculated upper bound | Calculation revision | HOLD until checked |
| Design reserve / headroom | Enter owner-approved reserve | Enter remaining headroom | Network-owner rule | HOLD until approved |
| Capacity and service result | Enter endpoint/service result | Enter endpoint/service result | OLT configuration and traffic model | HOLD until both pass |
The loss-budget guide provides the calculation method. This table is the release record that links the result to equipment, route and revision.
If the inputs are spread across drawings, spreadsheets and emails, request the editable splitter design-review worksheet. Attach the current topology revision and identify the source of each equipment and passive-component limit; BWNFiber can then flag missing configuration fields rather than infer them.
5. Map the splitter component to its host equipment
IEC 60875-1:2024 defines generic requirements for passive, non-wavelength-selective branching devices. That component scope does not approve the complete installed node. The splitter and its host enclosure need separate, connected records.
| Interface | Splitter record | Host/system record |
|---|---|---|
| Optical function | 1×N or 2×N, equal/unequal port map, maximum IL, uniformity, PDL, return loss/directivity where specified, wavelength conditions | System path-loss window and coexistence architecture |
| Physical fit | Module dimensions, cable exits, pigtail construction and length, connector/adaptor interface | Tray, cassette, LGX, cabinet, closure or rack compatibility; routing and access clearance |
| Environment | Component operating/storage limits and documented construction | Sealing, UV, fire, water, dust, impact and mounting responsibility of the installed assembly |
| Identification | Model, revision, input/output labels, port map and lot/serial rules | Node ID, rack/slot/tray position, address/service mapping and record-system identifier |
| Maintenance | Cleaning, inspection and replacement requirements | Safe access, spare policy, test access, restoration method and change authorization |
A blockless tube, ABS module or cassette name does not prove outdoor protection, host compatibility or fire performance. Use the dimensional drawing and exact installed assembly. Browse published splitter configurations only after the architecture and interface schedule are defined.
For a project-selection review, send the host drawing, routing limits, connector interfaces and required document set. The useful output at this stage is a list of candidate configuration families and unresolved approvals—not a generic claim that one package fits every cabinet or closure.
6. Build one controlled design pack
A controlled pack keeps the calculation basis, commercial baseline and field return records on the same identifiers:
- design-basis register and open-assumption log;
- zone and route drawing with revision and length source;
- approved topology and splitter-location schedule;
- OLT/ONT configuration and PON-tree capacity record;
- shortest/longest-path worksheets with value sources;
- splitter schedule with model or performance requirement, port map and host;
- connector, adapter and splice schedule;
- fiber and cable schedule by route;
- labeling and record-system convention;
- acceptance plan with method, wavelengths, reference method, limits and sample level;
- deviation, substitution and engineering-change log;
- as-built and test-record return requirements.
Each document needs an owner, revision, date and approval state. Cross-reference the same node and port identifiers throughout. A revised splitter location that does not update the budget worksheet, splice schedule and RFQ is not a controlled change.
Reopen the right checks when the design changes
Do not restart the entire design blindly, and do not update only the drawing. Use the change trigger to identify every dependent check and record that must be reopened.
| Change trigger | Checks to reopen | Records to revise | Approval owner |
|---|---|---|---|
| OLT/ONT model, optical class or service profile | Shortest/longest paths, wavelengths, reach, receiver limits and PON-tree capacity | Equipment register, path worksheets, topology and acceptance limits | Active-equipment and network-design owners |
| Splitter ratio, stage, model or location | Every affected path, port allocation, host fit, access and restoration method | Topology, splitter/port map, splice schedule, RFQ and test plan | ODN designer |
| Route, measured length, cable or construction method | Attenuation, passive-event count, bend/tension limits, environment and spare capacity | Route drawing, cable schedule, path worksheets and installation method | OSP or building-pathway owner |
| Connector, splice or coexistence interface | Passive losses, reflectance/return-loss requirements where specified, wavelength path and test access | Interface schedule, path worksheets, cleaning/inspection plan and test plan | ODN and active-equipment owners |
| Demand, take rate or service forecast | PON-tree capacity, port utilization, phasing, spare fibers/ports and restoration policy | Forecast, port-allocation schedule, quantity basis and RFQ | Planning and procurement owners |
| Supplier substitution or package change | Model-specific optical limits, physical fit, environment, labeling, evidence and traceability | Deviation log, approved drawing/sample record, purchase baseline and inspection plan | Technical approver and procurement owner |
The approved change record should name the affected nodes and paths, not merely state “equivalent.” If a dependency cannot be rechecked from controlled evidence, return the design to HOLD until the responsible owner closes it.
Electronic identification can support record operations, but it is a separate layer; see intelligent ODN management rather than duplicating that architecture here.
7. Convert the network design into a comparable fiber splitter RFQ
A price comparison is valid only when every bidder quotes the same topology revision, interface definition, test scope and evidence request.
Move through the procurement chain without skipping gates
| Buyer stage | Decision to close | Evidence or action before moving on |
|---|---|---|
| 1. Recognize the problem | Is the current topology only conceptual, or release-ready? | Run the HOLD checklist and assign owners to missing inputs |
| 2. Understand product/configuration types | Which split arrangement, package and interface families remain possible? | Use the selection and product-owner pages without choosing a model prematurely |
| 3. Define the application | FTTH, FTTB or passive optical LAN; aerial, duct, buried or building pathway? | Approve route, environment, access and host constraints by zone |
| 4. Confirm technical requirements | Which optical, mechanical, interface and evidence limits control the purchase? | Freeze endpoint models, path extremes, splitter stages, host and acceptance method |
| 5. Compare suppliers | Are all offers technically equivalent and based on the same revision? | Normalize drawings, model limits, inclusions, exclusions and deviations |
| 6. Compare price and delivery | Does price cover the same configuration, documents, packaging and approval stages? | Compare price, MOQ, lead time and freight only after technical parity |
| 7. Request documents or a sample | Which uncertainty requires paper review, drawing approval or physical verification? | Request model-matched datasheets/drawings; use a pre-production sample when fit, labels or record format warrant it |
| 8. Verify quality evidence | Can the offered evidence be traced to the approved model, lot and test conditions? | Agree inspection, component test, installed-path test and exception records |
| 9. Start an inquiry | Can the supplier review the request without guessing? | Send the RFQ input set below and identify every TBC field |
| 10. Release the order/project discussion | Have technical and commercial deviations been approved? | Freeze the purchase baseline, approval stages, change authority and as-built return requirements |
This sequence does not require every project to buy a sample. It requires the buyer to choose the evidence step that closes the actual risk before bulk release.
| RFQ field | Required buyer input | Supplier response to normalize |
|---|---|---|
| Reference design | Topology revision, node schedule and quantities | Accepted revision and listed exceptions |
| Splitter configuration | 1×N/2×N, port map, ratio, package, fiber, pigtails, connectors and labels | Exact model/drawing and unresolved options |
| Optical limits | Project wavelengths, maximum IL and other required parameters/test conditions | Model limits and test-record format tied to the offer |
| Host interface | Enclosure/tray/cassette/rack reference, dimensions and routing | Fit confirmation, drawing and supplied mounting parts |
| Evidence | Datasheet, drawing, sample plan, test data, traceability and contract documents | Available document list, scope, timing and exclusions |
| Commercial basis | Quantity, destination, schedule, packaging and approval stages | Price, MOQ, lead time and freight basis for the same configuration |
| Change control | Permitted alternatives and approval authority | Written deviation or substitution process |
Clarify these RFQ responses before award
- the quotation cites an obsolete or unnamed topology revision;
- a product code is offered without the drawing or port map needed to verify the host interface;
- the loss value is typical or nominal when the design check requires a model-specific maximum;
- “compatible package” is stated without dimensions, cable exits, mounting parts or routing confirmation;
- the proposed test report is not tied to the offered model, lot or agreed test conditions;
- price or lead time assumes a different connector, pigtail, package or evidence set.
Treat any such difference as a deviation, not a purchasing shortcut. Reopen the affected interface and path checks, then reject the alternative or issue a new approved revision. Where physical fit, labeling or the test-record format remains open, approve the drawing and, when the project warrants it, a pre-production sample before bulk release.
BWNFiber’s role on this page is configuration review before quotation. BWNFiber can review a submitted topology and configuration for gaps, compare it with published splitter families and identify fields that still need buyer approval. Final ODN approval remains with the network owner or its authorized designer.
When the RFQ basis is stable, request a configuration-matched quotation or sample review. Include the approved revision, quantity, destination, required delivery date, document/test scope and the specific uncertainty the sample must resolve. Price, MOQ, lead time, model fit and sample availability remain order-specific until confirmed.
8. Test, accept and reconcile the as-built splitter network
Component acceptance and installed-path acceptance answer different questions.
| Acceptance layer | Typical question | Record |
|---|---|---|
| Document review | Is the offered item the approved model and revision? | Datasheet, drawing, deviation log |
| Incoming/component check | Do identity, interfaces and measured component results meet the purchase specification? | Inspection and model/lot-linked test record |
| Installation inspection | Is the node mounted, routed, cleaned, labeled and protected as designed? | Installation checklist and photographs where authorized |
| Installed-path test | Does the complete path meet the project method and limits? | Insertion-loss/power and OTDR records as specified by the owner |
| As-built reconciliation | Do routes, splices, ports, equipment and exceptions match the controlled records? | Approved as-built revision and closed exception log |
OTDR does not replace all other tests, and splitter-containing paths need a defined method and access plan. State the instruments, wavelengths, direction, launch/receive arrangements, reference method, trace naming, limits and exception workflow instead of requesting a generic “OTDR report.”
Connector inspection, component attenuation and installed-path testing also answer different questions. IEC 61300-3-35:2022 explicitly treats visual connector inspection as additional to—not a replacement for—attenuation or return-loss measurement. IEC 61300-3-4:2023 addresses attenuation measurement of optical components. The purchase specification still has to name which parameter, method, conditions and limits apply to the offered splitter and to the installed path.
Control field handoffs, not just the final test
| Handoff | Risk to close | Required record |
|---|---|---|
| Receipt and storage | Wrong model or revision, damaged packaging, contaminated interfaces, lost traceability or storage outside documented conditions | Receipt inspection, model/lot identity, deviation record and storage/handling requirement |
| Installation | Unapproved routing, exceeded cable or pigtail limits, inaccessible splitter host, uncontrolled connector handling or labels that do not match the port map | Approved method, route/node inspection, connector-cleanliness control and photographs where authorized |
| Activation and test | Wrong wavelength, direction, reference method, instrument setting, filename or limit | Approved test plan, instrument identity, native records and exception disposition |
| Maintenance handover | Unknown spare ports/fibers, blocked access, undocumented replacement or no link between alarms and physical records | As-built node/port map, spare schedule, access method, replacement rule and change owner |
Transport and storage limits come from the exact product and packaging documentation. Do not convert a generic cable or splitter handling practice into a universal temperature, bend, stacking or reel rule.
Design headroom is also a controlled record. Name spare splitter ports, spare fibers, spare pathway capacity and reserved identifiers explicitly. Do not apply a universal spare percentage. Unused outputs do not remove the optical loss of the selected total split; protect and identify them according to the approved assembly and operating practice.
Close acceptance by returning measured routes, approved substitutions, field constraints and restoration findings to the design basis. The next phase should inherit the approved as-built record, not the assumptions in the tender drawing.
Frequently asked design questions
What should trigger a splitter network redesign?
Reopen the design when the equipment class, splitter stage or model, coexistence method, route length, passive interface count, host enclosure, endpoint forecast or acceptance method changes. The owner should identify which path, capacity, fit and document checks the change affects before approving a new revision.
What is the difference between centralized and cascaded splitting?
Centralized splitting places a split stage at a common hub; cascaded splitting distributes two or more stages across network nodes. Compare fiber counts, optical interfaces, access, testing, restoration, phasing and record complexity. Density alone is not a sufficient rule.
Does a 1×32 splitter always fit a GPON design?
No. The approved ratio depends on the implemented OLT/ONT configuration, shortest and longest paths, model-specific splitter loss, other passive losses, capacity policy and project reserve. A ratio label does not close those decisions.
How much design margin should a PON use?
There is no universal number for every network. The network owner should define the reserve, its purpose and the conditions it covers. Keep the reserve separate from physical attenuation and show the remaining headroom.
Why must the shortest optical path be checked?
The shortest path can fall below a specified minimum loss or approach a receiver-overload condition. Verify the actual transmitter and receiver limits for the implemented equipment configuration.
Is a broadband PLC splitter automatically ready for XGS-PON migration?
No. Migration may require a verified coexistence architecture, compatible active equipment, wavelength/isolation controls, added-loss checks and records for the installed splitter and other passive components.
What belongs in a splitter RFQ?
Send the topology revision, 1×N/2×N arrangement, port map, ratio, package, host, fiber and pigtails, connector interfaces, optical limits and test conditions, document pack, quantity, destination, schedule and change-control rules.
Which tests should be required at acceptance?
Define separate document, component, installation and installed-path checks. Name the method, wavelengths, direction, reference method, sample level, limits, instruments, record format and response to a failed result.
How much spare capacity should the design include?
Set ports, fibers, pathways and identifiers from the demand scenario, restoration plan and owner policy. Record the chosen quantities and triggers; do not copy a universal percentage.
When should the design be frozen?
Freeze the technical baseline before issuing a comparable RFQ, then control every approved change. Construction and acceptance may update the as-built revision, but they should not silently overwrite the design basis.
Where should optical splitters be placed in an FTTH network?
Choose the location by zone after comparing fiber availability, path limits, cabinet or closure access, take rate, activation workflow, fault isolation, restoration and future migration. A centralized hub, distributed terminal or hybrid plan is valid only when those tradeoffs are recorded and every affected path closes.
What documents should accompany a fiber splitter network design?
Include the design-basis register, topology and route revision, splitter-location and port maps, shortest/longest-path worksheets, active-equipment record, component and host-interface schedules, splice/connector schedules, acceptance plan, RFQ evidence list, deviation log and as-built return requirements.
Request a configuration gap review
Use the final CTA when the design has enough evidence for a supplier to respond without silently filling gaps. Send BWNFiber:
- topology and splitter-location revision;
- OLT and ONT/ONU models or implemented class;
- shortest and longest routes by zone;
- splitter stages, ratios and current model references;
- host-enclosure dimensions and interface constraints;
- project wavelengths, optical limits and test method;
- required drawings, test records, labels and traceability;
- quantity, destination, schedule and unresolved fields.
Request a fiber optic splitter network design gap review. The first useful outcome is a configuration-gap list tied to the submitted revision; a quotation or sample decision follows only after the open technical and commercial fields are resolved. Document availability, model fit, price, MOQ, lead time and test scope require confirmation for the selected configuration and order.
Technical scope and sources
- ITU-T G.984.2: G-PON physical media dependent layer specification.
- ITU-T G.9807.1: XGS-PON system and optical requirements, including in-force amendments.
- ITU-T G.9805: coexistence of PON systems.
- ITU-T G.652 (08/2024): characteristics of single-mode optical fiber and cable.
- ITU-T G.657 (08/2024): characteristics of bending-loss-insensitive single-mode optical fiber and cable.
- IEC 60875-1:2024: generic specification for passive, non-wavelength-selective branching devices.
- IEC 60794-1-1:2023: generic optical-fiber-cable property requirements.
- IEC 60794-4-20:2018: ADSS cable construction, performance, installation and environmental scope.
- IEC 60794-6:2020: indoor/outdoor optical-fiber-cable scope and boundary conditions.
- IEC 61300-3-4:2023: attenuation measurement for optical components.
- IEC 61300-3-35:2022: connector end-face inspection and its boundary relative to optical performance measurement.
Standards define system or component requirements; they do not certify a particular BWNFiber product or approve a project design. Use current, model-matched documents and the network owner’s approved engineering rules.
Prepared by the BWNFiber Technical Content Team. A named optical-network SME must approve the technical release. This guide supports design review and procurement preparation; it does not replace project-specific engineering approval.
Last technically reviewed: 30 September 2026.
Disclaimer: Planning methods, standards references and configuration fields are for design review and procurement preparation. They are not product guarantees or project approval. Verify the current equipment specifications, model-matched documents, field conditions and network-owner requirements before release.



Related product category: Bare Fiber PLC Splitters.
