Passive Optical Network Splitters: How PON Splitters Work

Passive Optical Network splitters distribute optical signals between an Optical Line Terminal and multiple subscriber connections without requiring electrical power. Consequently, they form a critical part of FTTH, GPON, XGS-PON and Passive Optical LAN networks because they allow one feeder fibre and one OLT port to serve multiple Optical Network Terminals.

However, every optical split introduces signal loss. Therefore, network designers must select the splitter technology, split ratio, location and overall architecture within the available optical power budget.

This guide explains how PON splitters work, compares PLC and FBT splitter technology, examines single-stage and cascaded designs, and outlines the installation and testing practices required to create a reliable passive optical network.

What Is a Passive Optical Network Splitter?

A Passive Optical Network splitter is a passive optical device that divides the light entering one fibre between two or more output fibres.

In the downstream direction, the splitter distributes the signal transmitted by the Optical Line Terminal, or OLT, to every connected Optical Network Terminal or Optical Network Unit. Meanwhile, in the upstream direction, the same device combines the optical signals transmitted by the subscriber terminals and directs them towards the OLT.

Importantly, an optical splitter contains no powered electronics. Furthermore, it does not inspect, route or modify the data passing through it. Instead, it distributes optical power between its output ports.

The broader collection of fibres, splitters, connectors, closures and other passive components between the OLT and subscriber terminals forms the Optical Distribution Network, or ODN. Accordingly, the ODN must be designed as a complete optical system rather than as a collection of unrelated components.

The ITU-T G.984.2 GPON standard describes an ODN as a network of optical fibres supplemented by power or wavelength splitters, filters and other passive optical devices.

How Does a PON Splitter Work?

A Passive Optical Network uses a point-to-multipoint topology.

First, the OLT transmits a downstream optical signal over a feeder fibre. The splitter then divides the available optical power between its output ports. Consequently, the same downstream transmission can reach multiple subscriber terminals.

Each ONT receives the downstream signal and extracts the information addressed to that subscriber. However, the optical splitter itself does not determine which subscriber receives particular data. Instead, the active network equipment manages addressing, encryption and subscriber access.

Upstream transmission requires additional coordination because several ONTs share the same optical route back to the OLT. Therefore, in time-division PON systems, the OLT allocates transmission windows so that each ONT sends data during its assigned time slot. As a result, multiple subscriber signals can share the upstream fibre without transmitting simultaneously.

The splitter performs only the passive optical distribution function. In contrast, the OLT and ONTs manage subscriber identification, bandwidth allocation, encryption and data transmission.

PON Splitters Work in Both Directions

An optical splitter is fundamentally bidirectional.

In the downstream direction, it operates as a one-to-many device:

• One OLT signal enters the splitter.
• The optical power divides between the output ports.
• Consequently, every connected ONT receives the downstream transmission.

In the upstream direction, it operates as a many-to-one combiner:

• Individual ONTs transmit during their allocated time slots.
• Their optical signals enter through the distribution ports.
• The splitter then combines those signals onto the feeder fibre leading back to the OLT.

For a symmetrical splitter, the fundamental splitting loss is broadly comparable in either direction. Nevertheless, measured end-to-end link loss can vary because of connectors, splices, fibre length, wavelength, component tolerances and the direction of testing.

What Is a PON Split Ratio?

The PON split ratio indicates the number of subscriber-side outputs served by one input fibre.

Common configurations include:

• 1:2
• 1:4
• 1:8
• 1:16
• 1:32
• 1:64

Higher split ratios allow an OLT port to serve more potential subscribers. However, dividing the available optical power between more outputs also increases insertion loss.

Therefore, designers must manage a fundamental trade-off:

• A higher split ratio can improve the utilisation of feeder fibres and OLT ports.
• In contrast, a lower split ratio preserves more optical power for transmission distance, connectors, splices and engineering margin.

For example, a 1:32 splitter allows one OLT port to serve twice as many subscriber connections as a 1:16 splitter. Nevertheless, the 1:32 arrangement introduces approximately 3 dB more theoretical splitting loss before excess component loss is considered.

The maximum split ratio that a network can support depends on the PON technology, optical class, transmission distance, component losses and network operator rules. Consequently, designers should never choose a split ratio solely according to the maximum number of subscribers.

How Much Loss Does an Optical Splitter Introduce?

An ideal two-way split divides the available optical power equally and creates approximately 3.01 dB of theoretical splitting loss at each output.

Furthermore, every doubling of the output count adds approximately another 3.01 dB of ideal loss:

Split ratio	Ideal splitting loss
1:2	3.01 dB
1:4	6.02 dB
1:8	9.03 dB
1:16	12.04 dB
1:32	15.05 dB
1:64	18.06 dB

However, these figures represent ideal mathematical loss only.

In practice, a real splitter introduces additional excess loss through its waveguides, optical couplings, manufacturing tolerances, fibre pigtails and connectors. Therefore, the manufacturer’s specified maximum insertion loss will be higher than the theoretical figures shown above.

Consequently, network designers must use the manufacturer’s guaranteed maximum specifications when calculating the optical power budget. Theoretical splitting loss can help explain the relationship between split ratios, but it should not replace the certified component data.

Splitter Insertion Loss and Uniformity

Insertion loss measures the reduction in optical power between the splitter input and a selected output port.

However, total insertion loss is not the only important performance characteristic. Designers should also examine:

• Uniformity: The difference between the highest-loss and lowest-loss output ports.
• Return loss: The relationship between transmitted power and the light reflected towards the source.
• Directivity: The level of isolation between output ports.
• Operating wavelength range: The wavelengths across which the splitter meets its specifications.
• Polarisation-dependent loss: Variations in loss caused by the optical signal’s polarisation.
• Temperature stability: The ability to maintain performance across the stated environmental range.
• Port-to-port consistency: The repeatability of performance across all outputs.

Poor uniformity can create unequal receive levels between subscribers. For example, one ONT may operate comfortably within its acceptable power range, while another sits close to its receiver threshold despite having a similar route length.

Therefore, designers should review both maximum insertion loss and port uniformity before approving a splitter for a PON deployment.

PLC Splitters

Planar Lightwave Circuit splitters use optical waveguides formed on a substrate, commonly through silica-based technology. The waveguide structure divides the incoming light between the required number of output ports.

PLC splitter technology generally provides:

• Even splitting across multiple output ports
• Broad operating wavelength performance
• Good port-to-port uniformity
• Compact packaging
• Support for higher split counts
• Stable performance across a wider temperature range
• Suitability for GPON, EPON and XGS-PON networks

Moreover, PLC technology can integrate a large number of output channels into a relatively compact component. Consequently, balanced PLC splitters have become the usual choice for modern FTTH and Passive Optical LAN deployments.

The Fiber Optic Association’s optical splitter guide explains that PLC splitters generally provide stronger loss and wavelength uniformity than FBT devices and can support a greater number of splits.

Nevertheless, the term “PLC splitter” does not guarantee that every product provides identical performance. Therefore, buyers should still compare the maximum insertion loss, uniformity, return loss, operating temperature, wavelength range and manufacturing test results.

FBT Splitters

Fused Biconical Taper splitters are manufactured by twisting, heating and stretching two or more optical fibres until their optical fields couple.

FBT technology can remain useful where the design requires:

• A small number of output ports
• An unequal optical split
• Operation across a limited wavelength range
• A straightforward optical coupler
• A cost-sensitive specialised application

However, FBT splitters generally provide less uniform performance across broad wavelength ranges. Furthermore, they become less practical as the required number of output ports increases.

Therefore, describing FBT technology simply as an older or cheaper version of PLC technology can be misleading. Instead, each construction method has different optical characteristics, advantages and limitations.

For example, an FBT coupler may suit a specialised unequal-tap application. In contrast, a balanced PLC splitter will generally provide a more suitable solution for a modern high-port-count PON deployment.

Balanced and Unbalanced Splitters

A balanced splitter distributes approximately equal optical power to every output port.

For example, a balanced 1:4 splitter directs approximately one-quarter of the incoming power towards each output before excess loss is considered. As a result, each subscriber path begins with broadly comparable splitting loss.

An unbalanced splitter deliberately directs different proportions of optical power to its outputs. For instance, a device may allocate 90% of the power to one path and 10% to another.

Consequently, unbalanced splitters can support specialised tap or bus architectures. However, designers must calculate the power available at every downstream point carefully because each tap removes optical power from the continuing route.

Most conventional FTTH tree architectures use balanced PLC splitters. Nevertheless, unbalanced devices can provide an effective solution where the network topology requires controlled optical tapping rather than equal point-to-multipoint distribution.

Where Are PON Splitters Installed?

PON splitters sit within the Optical Distribution Network between the OLT and subscriber ONTs. However, their physical location varies according to the network architecture.

Centralised Splitter Architecture

In a centralised design, the main splitter sits at a central office, exchange, fibre distribution hub or another controlled distribution location.

This approach can provide:

• Centralised access to splitter ports
• Easier port management
• Simplified testing from a common location
• Improved visibility of OLT-to-subscriber assignments
• Straightforward reassignment of distribution fibres

Moreover, concentrating splitters in an accessible location can make inspection, cleaning and replacement easier. However, the architecture may require a larger number of distribution fibres leaving the central splitter location.

Consequently, designers must balance operational accessibility against cable-count and pathway requirements.

Distributed Splitter Architecture

A distributed architecture positions the splitter closer to subscribers. For example, the splitter may sit within an external cabinet, pit, closure, pole-mounted enclosure or local building distribution point.

Potential advantages include:

• Reduced distribution-fibre requirements
• Efficient servicing of geographically grouped customers
• Smaller local cable counts after the split
• Flexible deployment across neighbourhoods and campuses
• Closer alignment between splitter capacity and local demand

However, external splitter locations require suitable environmental protection. Furthermore, technicians need safe and practical access for installation, testing and maintenance.

Therefore, the enclosure must protect the splitter, pigtails, connectors and fibre-management system from water, dust, mechanical stress, uncontrolled bending and temperature extremes.

Multiple-Dwelling Units

Apartment buildings and other Multiple-Dwelling Units may place splitters in:

• Basement communications rooms
• Main distribution frames
• Building entrance facilities
• Floor distributors
• Riser cupboards
• Secure telecommunications enclosures

However, the most convenient splitter location may not always provide the best whole-of-life network outcome. Therefore, designers must also consider riser capacity, fire requirements, access control, subscriber identification and the pathway between the splitter and each tenancy.

Moreover, building owners and network operators need clear responsibility for access, maintenance and record keeping.

Single-Stage and Cascaded Splitter Designs

A single-stage design uses one splitter between the OLT and subscriber.

For example, a 1:32 splitter can connect one feeder fibre to as many as 32 distribution fibres. Consequently, the complete split occurs at one physical point within the ODN.

In contrast, a cascaded design uses two or more splitter stages. For instance, a 1:4 first-stage splitter may feed four 1:8 second-stage splitters, producing a total effective ratio of 1:32.

Advantages of a Single-Stage Split

A single-stage architecture can provide:

• Fewer optical components
• Lower accumulated excess loss
• Simpler optical-loss calculations
• Easier fault isolation
• Fewer closures or cabinets containing splitters
• More straightforward network documentation

Additionally, technicians can assess the principal split from one location. Therefore, single-stage designs can simplify maintenance where pathway and fibre-count requirements permit their use.

Advantages of Cascaded Splitting

Cascaded splitting can provide:

• Greater flexibility across dispersed service areas
• Smaller second-stage splitters close to subscriber groups
• More efficient use of distribution cables in some topologies
• Staged network expansion
• Reduced initial deployment in low-take-up areas

However, every splitter stage introduces additional excess loss. Furthermore, each stage creates another potential fault, maintenance and documentation point.

Consequently, a single-stage splitter is not automatically superior in every project. Instead, the correct architecture depends on subscriber density, fibre availability, pathway topology, expected take-up, maintenance strategy and the available optical power budget.

Calculating the Optical Power Budget

The optical power budget represents the maximum total loss the PON system can tolerate while maintaining reliable communication.

A complete loss calculation should include:

• Splitter insertion loss
• Fibre attenuation
• Connector losses
• Fusion-splice losses
• Mechanical-splice losses, where applicable
• Wavelength-dependent loss
• Engineering and ageing margin
• Repair allowance
• Coexistence filters or wavelength components
• Equipment transmitter and receiver limits

The splitter usually represents the largest single source of loss in the ODN. Nevertheless, designers should not overlook the cumulative effect of smaller losses.

For example, several patching points, field splices and repair closures can consume a meaningful proportion of the remaining power budget after the main split. Additionally, connector contamination or an undocumented bend can move a marginal link outside its acceptable operating range.

Therefore, the loss budget should include a realistic engineering margin rather than being designed exactly to the equipment’s theoretical limit. Moreover, commissioning results should be compared with both the calculated design loss and the applicable acceptance limits.

GPON and XGS-PON Splitters

GPON and XGS-PON both use point-to-multipoint optical access infrastructure. Therefore, suitable parts of an existing ODN may sometimes remain in service when an operator changes or adds active network equipment.

GPON commonly provides nominal line rates of approximately:

• 2.488 Gbit/s downstream
• 1.244 Gbit/s upstream

In contrast, XGS-PON provides nominal symmetric transmission of approximately 10 Gbit/s downstream and upstream. The ITU-T G.9807.1 XGS-PON recommendation defines the physical-layer requirements for this 10-gigabit symmetric system.

However, operators should not automatically assume that every existing passive splitter is suitable for a newer PON generation. Instead, they must confirm:

• Operating wavelength range
• Maximum insertion loss
• Return-loss performance
• Directivity
• Connector type and condition
• Complete ODN power budget
• Existing network documentation
• Network operator design requirements

Importantly, the splitter does not determine the data rate. Instead, it must transmit the required wavelengths while keeping the complete ODN within the applicable optical-loss limits.

Consequently, upgrading GPON to XGS-PON involves more than changing the OLT and subscriber equipment. The existing splitter network, fibres, connectors, closures and test records must also support the intended service.

Do AWGs Replace Splitters in WDM-PON?

Arrayed Waveguide Gratings and conventional optical power splitters perform different functions.

A conventional power splitter distributes supported wavelengths across its output ports. In contrast, an AWG separates or routes individual wavelength channels towards specific ports.

Therefore:

• A power-split PON uses optical power division.
• A wavelength-routed PON may use an AWG to direct individual wavelengths.
• A hybrid network may contain splitters, wavelength filters and other passive optical components.

Consequently, an AWG is not simply a more efficient general-purpose replacement for a PLC splitter. Instead, it performs a wavelength-routing function within a different network architecture.

Designers must therefore select components according to the network’s transmission and wavelength plan rather than treating power splitters and AWGs as interchangeable devices.

PON Splitter Connector Types

PON splitter modules may use connectorised inputs and outputs or unterminated pigtails intended for fusion splicing.

SC/APC connectors are common in optical-access networks because their angled physical-contact end face can provide strong return-loss performance. However, connector selection ultimately depends on the network operator, equipment interface, enclosure design and approved installation practices.

Splitter package options include:

• Bare fibre devices
• Miniature steel-tube modules
• ABS-box modules
• Connectorised cassettes
• Rack-mounted panels
• Closure-mounted modules
• LGX-style modules
• Purpose-built distribution assemblies

Importantly, every mating point must use compatible connector types. APC and UPC connectors must not be mated together, even if their outer connector bodies appear mechanically compatible.

Furthermore, technicians should verify both the connector body and ferrule polish before making a connection. Otherwise, an incorrect mating may introduce substantial loss and potentially damage the connector end faces.

Cable Selection Around the Splitter Network

Although the splitter is a defining passive component, the surrounding fibre cable determines how reliably the ODN survives installation and long-term operation.

External feeder and distribution routes may require protection against:

• Water penetration
• Pulling tension
• Crush loads
• UV exposure
• Tight or congested ducts
• Repeated temperature changes
• Rodent or termite risks
• Handling in pits and closures

Therefore, cable selection should reflect the entire route and installation environment.

Compact loose-tube cable can provide a practical option for ducts and congested external pathways. Moreover, its reduced diameter can improve pathway utilisation where available conduit space is limited.

The Anderson Corporation Mini Loose Tube Fibre Optic Cable Authority Hub explains the cable construction, mechanical requirements and environmental considerations involved in selecting compact fibre cable for demanding telecommunications and infrastructure pathways.

However, designers must still verify every project-specific requirement. In particular, carrier-network projects may require nominated products, approvals, standards or operator-specific qualification.

Where Pre-Terminated Fibre Can Support PON Deployments

Not every connection associated with a PON deployment must be terminated in the field.

Where route lengths and interfaces can be determined accurately, pre-terminated fibre optic cable systems can support connections between communications rooms, distribution cabinets, equipment racks and other controlled network locations.

Factory termination can provide:

• Consistent connector preparation
• Reduced field-termination time
• Documented insertion-loss testing
• Faster connection and commissioning
• Less dependence on field-polishing conditions
• Repeatable breakout and labelling arrangements

Additionally, a pre-terminated assembly can reduce the amount of specialised termination work required on site. Consequently, it may support faster deployment across buildings, campuses and other defined network environments.

Nevertheless, the assembly must be designed for the physical pathway. Before manufacture, the installer should confirm:

• Required cable length
• Connector type
• Fibre count
• Pulling direction
• Protective pulling-eye requirements
• Breakout length
• Conduit size and occupancy
• Minimum bend radius
• Equipment and enclosure interfaces

The Anderson Corporation Pre-Terminated Fibre Optic Cable Authority Hub provides further guidance on system specification, pathway planning, factory testing, installation and commissioning.

Common PON Splitter Faults

Because a PON branches into multiple subscriber paths, faults can be more difficult to isolate than faults in a straightforward point-to-point link.

Contaminated Connectors

Dust, oil and other contamination can increase insertion loss and reflectance. Furthermore, mating a contaminated connector can transfer debris or permanently damage both connector end faces.

Therefore, technicians should inspect, clean and reinspect connector end faces before making a connection.

Damaged Splitter Ports

A damaged connector, pigtail or internal waveguide may cause excessive loss on one port or across several outputs.

Consequently, comparing loss measurements across the output ports can help determine whether a fault affects one distribution path or the splitter itself.

Poor Fusion Splices

Incorrect cleaving, contamination, poorly maintained equipment or unsuitable splice protection can increase loss or weaken the joint mechanically.

Therefore, technicians should verify suspect splices and ensure that closures protect them against strain, moisture and handling damage.

Macrobends

A macrobend occurs when optical fibre bends more tightly than its permitted radius. Moreover, the resulting loss may become more severe at longer wavelengths.

Macrobends may occur in:

• Splice trays
• Splitter cassettes
• Cabinets
• Congested pits
• Building-entry points
• Fibre storage loops
• Over-tight cable ties

Therefore, fibre-management hardware must maintain the specified bend radius throughout installation and subsequent maintenance.

Cross-Connections and Documentation Errors

Incorrect port assignments can disconnect one customer while appearing to create a fault elsewhere in the network.

Consequently, clear port identification, accurate route records and current subscriber documentation are essential.

Water or Enclosure Damage

A correctly manufactured splitter may still fail if its enclosure allows water ingress, mechanical loading or uncontrolled fibre movement.

Therefore, external closures and cabinets must suit the installation environment. Additionally, they should provide appropriate cable sealing, strain relief, fibre storage and bend-radius control.

Testing a PON Splitter Network

Testing should confirm that every subscriber path remains within the design loss budget.

However, technicians must use the correct equipment, wavelengths and test procedures. Otherwise, the results may not represent the performance of the operational network.

Insertion-Loss Testing

A calibrated optical light source and power meter can measure end-to-end loss through the ODN.

First, the technician must establish a reference using the nominated test method. The network should then be tested at the appropriate wavelengths. Finally, the measured results should be compared with the calculated loss budget and project acceptance limits.

The Fiber Optic Association’s splitter-testing guidance explains how splitters and couplers can be assessed using established fibre-optic loss-testing methods.

Importantly, technicians should retain the results for each subscriber path. As a result, future maintenance teams gain a reliable commissioning baseline for comparison.

OTDR Testing

An Optical Time-Domain Reflectometer can help locate:

• Fibre breaks
• Reflective connector events
• High-loss splices
• Macrobends
• Unexpected route distances
• Faults before or after a splitter

However, splitters make OTDR traces more difficult to interpret. The splitter creates a large loss event and, when testing from the OLT side, may combine the effects of several downstream branches.

Therefore, testing from the subscriber side can help isolate an individual branch. Additionally, correct launch fibres, receive fibres and PON-aware testing equipment may improve fault identification.

Technicians working on an active PON must use equipment and procedures designed for live-network testing. Furthermore, they should never connect unsuitable test equipment to an active optical system.

Selecting a PON Splitter

Before specifying a PON splitter, confirm:

• PON technology
• Required split ratio
• Single-stage or cascaded architecture
• Balanced or unbalanced output
• Operating wavelength range
• Maximum insertion loss
• Port uniformity
• Return loss
• Directivity
• Connector type
• Package format
• Pigtail length and fibre type
• Operating temperature
• Environmental protection
• Applicable standards
• Network operator approval requirements

Price should not be the only deciding factor. After all, a splitter sits at a branching point where one component failure can affect multiple subscribers.

Therefore, consistent optical performance, environmental suitability, traceability and documented factory testing may provide considerably more value than a small reduction in purchase cost.

Moreover, designers should confirm that the selected package fits the intended closure, cabinet or rack. Otherwise, poor fibre management or unsuitable environmental protection can undermine the performance of an otherwise compliant splitter.

Frequently Asked Questions About Passive Optical Network Splitters

Does a PON splitter require power?

No. A PON splitter contains no powered electronics and divides or combines optical power passively.

However, the OLT and subscriber ONTs still require electrical power.

Does an optical splitter amplify the signal?

No. A passive optical splitter cannot amplify light.

Instead, it divides the available optical power between its output ports and introduces insertion loss. Therefore, the network must provide sufficient transmitter power and receiver sensitivity to accommodate the complete ODN loss.

What is the most common PON split ratio?

Ratios such as 1:8, 1:16 and 1:32 are widely used. Additionally, some network designs support 1:64 or other configurations.

However, the appropriate ratio depends on the PON technology, optical class, distance, architecture and network operator requirements.

Is every PON splitter a PLC splitter?

No. Optical splitters can use PLC, FBT or other specialised technologies.

Nevertheless, balanced PLC splitters are commonly selected for modern FTTH networks because they provide good uniformity, broad wavelength performance and support for higher output counts.

Is a 1:32 splitter always better than a 1:16 splitter?

No. A 1:32 splitter serves more potential subscribers, but it also introduces greater optical loss.

In contrast, a 1:16 design preserves more optical power. Consequently, it may provide additional margin for longer routes, more connection points or future repairs.

Can GPON and XGS-PON use the same splitter network?

Potentially, but the existing splitter and complete ODN must support the required wavelengths, loss limits and performance specifications.

Therefore, operators must assess the entire network rather than assuming compatibility simply because the splitter is passive.

Are splitter losses identical in both directions?

A passive symmetrical splitter has broadly comparable splitting behaviour in either direction.

However, measured end-to-end results may differ because of connectors, splices, wavelengths, fibre characteristics and test methods.

Can an OTDR test through a PON splitter?

Yes, but the splitter creates a substantial loss event and can make downstream traces difficult to interpret.

Therefore, technicians may need to test from the subscriber side, use appropriate launch fibres and apply PON-specific testing procedures.

What is the difference between a splitter and an AWG?

A conventional optical splitter divides optical power between its outputs. In contrast, an Arrayed Waveguide Grating separates or routes individual wavelength channels.

Consequently, the two devices serve different functions and should not be treated as interchangeable components.

What connectors are commonly used with PON splitters?

SC/APC connectors are commonly used in optical-access networks because they provide strong return-loss performance.

However, the correct connector depends on the equipment, network operator and enclosure design. Furthermore, APC and UPC connector end faces must not be mated together.

Conclusion – Passive Optical Network

Passive Optical Network splitters make scalable point-to-multipoint fibre access possible by allowing one OLT port and feeder fibre to serve multiple subscriber terminals.

PLC splitters generally provide the broad wavelength performance, output uniformity and higher split counts required by modern GPON and XGS-PON networks. Meanwhile, FBT devices remain useful for selected low-port-count, restricted-wavelength and unequal-split applications.

However, successful PON design involves more than choosing a splitter ratio. Designers must also calculate the complete optical power budget, select suitable cable and connectors, protect every component against environmental and mechanical damage, and document every subscriber path.

Furthermore, installers must inspect connector end faces, manage fibre bend radius, protect splices and verify every completed route through appropriate testing. Consequently, reliable performance depends on the quality of the complete Optical Distribution Network rather than the splitter alone.

For further guidance on compact external distribution cable, visit the Anderson Corporation Mini Loose Tube Fibre Optic Cable Authority Hub. Alternatively, for factory-terminated connections between defined network locations, explore the Pre-Terminated Fibre Optic Cable Authority Hub.