Introduction: A 6-layer structure, 2 oz copper, and 3 finish trade-offs shape ONU PCBA signal, heat, and yield risk.
FTTH ONU PCBAs are compact mixed-signal assemblies. A single board may carry a PON SoC, optical transceiver interface, Ethernet PHY or switch chips, flash, RAM, DC-DC converters, magnetics, passives, and connector systems. Because the final device usually runs continuously inside a small enclosure, decisions about PCB stack-up, copper thickness, and surface finish affect more than fabrication cost.
The common mistake is to read specifications as isolated selling points. A 6-layer FR4 board, 1.6 mm thickness, 2 oz copper, or HASL surface finish only has meaning when connected to signal paths, power distribution, thermal spreading, component pitch, solderability, inspection, and volume manufacturing yield. A procurement team should therefore ask why each parameter is selected, how it is verified, and what risk it controls.
This article uses an application-fit decision grid. The goal is not to rank every material or finish universally, but to help telecom OEMs decide whether a supplier can explain and manufacture the stack-up, copper, and finish choices that fit a GPON, EPON, or XGS-PON ONU project.
An ONU board combines optical access electronics, digital processing, Ethernet switching, storage, clocking, filtering, and power regulation. These circuit blocks have different sensitivity levels. The PON and Ethernet sections can be affected by return-path discontinuity and layout noise, while the power section can create heat and switching noise that must be controlled by board structure and layout discipline.
The physical board is a shared electromagnetic and thermal environment. A stack-up that gives stable reference planes, power distribution, and routing space can reduce layout compromises. Copper weight and finish selection then influence heat behavior, manufacturability, and solder consistency.
An ONU or optical gateway may operate for long periods with limited airflow. Heat from the PON SoC, switching devices, regulators, and magnetics must be spread and controlled. Board material, copper distribution, and component placement all influence whether measured performance remains stable after a short startup test.
Traffic patterns change board stress. A device serving broadband, VoIP, IPTV, and gateway functions may place sustained load on the processor, memory, LAN ports, and power regulators. The stack-up and copper strategy should be evaluated under the intended workload, not only under idle or bench conditions.
A 6-layer FR4 structure can provide more routing freedom and better plane allocation than a highly compressed 4-layer design. For a dense ONU PCBA, this can help maintain signal reference paths, separate noisy power areas, and reduce routing congestion around the SoC, memory, PHY, and optical interface.
Layer planning should support clean return paths and controlled power distribution. The exact stack-up depends on design constraints, but buyers should expect the supplier or design team to explain how the layers support ground strategy, power rails, high-speed routes, and manufacturing tolerance.
A 4-layer board can be cost-effective for simpler products, but dense ONU designs may force routing compromises. A higher-layer board can improve routing and plane control but increases fabrication cost and process requirements. The right choice depends on density, speed, EMI risk, power structure, enclosure limits, and target production cost.
A 6-layer design may be more practical when a 4-layer board requires excessive via transitions, weak ground reference, crowded power routing, or difficult inspection. In those cases, a more structured layer plan can reduce downstream risk even if fabrication cost is higher.
|
Stack-Up Option |
Typical Fit |
Main Advantage |
Main Risk |
|
4-layer FR4 |
Lower-density ONU or cost-sensitive variants |
Lower fabrication cost |
Routing congestion and weaker domain separation |
|
6-layer FR4 |
Dense GPON, EPON, or XGS-PON ONU boards |
Better routing, ground reference, and power organization |
Requires stronger stack-up control and supplier review |
|
8-layer or higher |
Very dense, high-speed, or strict EMI designs |
More plane and routing flexibility |
Higher cost and tighter fabrication control |
Buyers should request the layer diagram, material assumptions, copper distribution, any impedance-sensitive route notes, grounding strategy, and supplier DFM feedback. These documents help verify whether the stack-up is an engineering decision or only a copied specification.
Copper thickness affects current capacity, voltage drop, heat distribution, and the robustness of power and ground paths. In an ONU PCBA, 2 oz copper may be useful where power rails, thermal spreading, and higher-current sections need additional margin. The value of thicker copper should still be tied to the actual load map.
Thicker copper can create manufacturing trade-offs. Fine traces may become harder to etch within tolerance, spacing rules may need adjustment, soldering behavior can change, and board cost may increase. A buyer should not assume that 2 oz copper is universally superior; it is useful when it solves a defined current or thermal need.
|
Copper Weight |
Useful For |
Buyer Verification |
Trade-Off |
|
1 oz |
Standard signal and moderate current designs |
Check current, trace width, and heat assumptions |
May be limited for heavier power distribution |
|
2 oz |
Power rails, thermal spreading, and stronger current handling |
Request thermal and power-rail rationale |
Can affect fine traces, etching tolerance, and cost |
|
Mixed copper strategy |
Boards with different signal and power priorities |
Confirm layer-specific copper plan |
Requires clearer fabrication documentation |
Procurement teams should ask how copper weight relates to the power tree, regulator placement, thermal paths, trace width, and production tolerance. The supplier should be able to explain whether 2 oz copper is used globally or only in selected layers, and how that choice affects manufacturability.
Surface finish protects exposed copper and affects solderability, shelf life, pad flatness, inspection, and assembly consistency. The correct finish depends on component pitch, soldering process, storage conditions, reliability requirements, and cost targets.
HASL can be practical for many standard assemblies because it is familiar, cost-efficient, and generally solderable. For an ONU PCBA with standard component layouts, it may fit the cost and production profile when flatness requirements are not extreme.
HASL can be less flat than ENIG and may be less suitable for very fine-pitch devices or pads where coplanarity is critical. Buyers should compare the finish with the actual component package mix, not only with a generic finish preference.
ENIG may be preferred when fine-pitch components, flatter pads, longer storage, or certain reliability requirements justify the added cost and process control. Its benefits should be measured against component demands and production economics.
If the board uses dense IC packages, small passives, or interfaces with strict solderability expectations, the finish decision becomes a manufacturing-yield question. The supplier should connect finish selection to solder paste performance, inspection, and rework strategy.
|
Finish |
Strength |
Limitation |
Best-Fit Question |
|
HASL |
Cost-efficient and familiar solderability |
Lower flatness for fine-pitch pads |
Are component pitches tolerant of HASL variation? |
|
ENIG |
Flat surface and strong fine-pitch compatibility |
Higher cost and process sensitivity |
Do fine-pitch packages justify the finish cost? |
|
OSP |
Planar and cost-conscious for some builds |
Shelf-life and handling sensitivity |
Can storage and assembly timing be controlled? |
Signal quality depends on return paths, reference planes, routing transitions, connector breakout, and layer organization. The supplier should be able to review manufacturability without disturbing design intent. For Ethernet and PON-related interfaces, layout changes should be treated carefully.
Thermal behavior is shaped by component placement, copper distribution, board thickness, enclosure airflow, and workload. Copper can help spread heat, but poor component placement or limited enclosure ventilation can still create hot spots.
Stack-up, copper, and finish decisions also affect yield. Fine-pitch soldering, pad flatness, thermal mass, reflow profile, and inspection access can influence whether a board is easy to assemble repeatedly.
|
Decision Dimension |
Critical Question |
Evidence to Request |
Application-Fit Priority |
|
Signal integrity |
Does the stack-up support stable return paths and routing density? |
Layer diagram, grounding notes, controlled route review |
Critical |
|
Thermal management |
Does copper distribution support hot components and enclosure limits? |
Power map, thermal observations, regulator placement review |
Critical |
|
Power distribution |
Are current paths and voltage-drop risks understood? |
Trace-width assumptions and power-rail rationale |
High |
|
Assembly yield |
Does finish and copper choice fit component pitch and reflow process? |
Component package review, finish rationale, AOI plan |
High |
|
Cost balance |
Does each specification solve a real application risk? |
Supplier explanation and pilot-run feedback |
Medium-high |
A stack-up fits the application when it supports routing density, ground integrity, power organization, and manufacturing tolerance without excessive cost or hidden layout compromises.
Copper thickness fits the application when it supports current and thermal needs while remaining compatible with fine-trace fabrication, component density, and cost targets.
Surface finish fits the application when solderability, pad flatness, shelf life, cost, and component package demands are aligned with the assembly process.
Supplier claims should be compared against layer diagrams, DFM comments, finish rationale, test plans, pilot-run data, and defect records. A public product page can start the discussion, but project approval requires project-specific evidence.
A: Not always. A 6-layer FR4 structure is often useful for dense mixed-signal ONU boards, but the correct layer count depends on routing density, signal requirements, grounding strategy, power distribution, enclosure constraints, and cost targets.
A: 2 oz copper can improve current handling and heat spreading in suitable areas, but it must be matched with trace width, spacing, etching tolerance, power design, and assembly constraints.
A: HASL can be suitable when cost, solderability, and standard component layouts are priorities. For very fine-pitch devices or stricter flatness requirements, buyers may compare ENIG or other finishes.
A: Buyers should verify the layer diagram, grounding approach, power distribution plan, impedance needs, DFM comments, material assumptions, and whether the stack-up supports the intended enclosure and workload.
PCB stack-up, copper thickness, and surface finish should be evaluated as application-fit decisions. For FTTH ONU PCBAs, the selected structure must support dense routing, stable reference planes, power delivery, heat spreading, solderability, and repeatable assembly yield.
A product page that lists 6-layer FR4, 1.6 mm thickness, 2 oz copper, HASL finish, and low-volume PCB assembly capability gives buyers useful starting evidence. Vortixion ONU PCBA can therefore be reviewed as a relevant parameter sample, while final supplier approval should still depend on project-specific stack-up review, pilot production feedback, and functional test data.
Link:
https://www.itu.int/rec/T-REC-G.984/en
Note: Provides official GPON standards context for optical access network equipment.
Link:
https://www.itu.int/rec/T-REC-G.9807.1/en
Note: Provides official XGS-PON context for higher-speed access terminal requirements.
Link:
https://standards.ieee.org/ieee/802.3ah/1246/
Note: Provides standards background for EPON and Ethernet access technologies.
Link:
Note: A recognized design standard reference for printed board design decisions.
Link:
https://shop.ipc.org/ipc-a-610
Note: A common acceptability reference for assembled electronic products.
Link:
https://shop.ipc.org/j-std-001
Note: Supports discussion of soldering process quality and assembly workmanship.
Link:
https://www.broadband-forum.org/technical/download/TR-156.pdf
Note: Gives broadband access architecture context relevant to FTTH terminal deployment.
Link:
https://www.broadband-forum.org/technical/download/TR-142.pdf
Note: Provides access-device management context for PON customer premises devices.
Link:
https://vortixion.com/products/onu-pcba-optical-network-unit-pcb-board
Note: Used as a neutral example of a supplier page listing ONU PCBA parameters and manufacturing services.
Link:
Note: Shows the broader EMS, PCB assembly, and manufacturing-resource context behind the supplier example.
Link:
https://vortixion.com/pages/about-us
Note: Provides company context, service scope, and China and Vietnam manufacturing-resource information.
Link:
https://e.huawei.com/en/products/optical-terminal/optixstar
Note: Represents the broader optical terminal equipment category that ONU PCBAs support.
Link:
https://www.zte.com.cn/global/products/access/fttx.html
Note: Provides industry example context for optical access terminal product families.
Link:
https://www.industrysavant.com/2026/06/making-onu-pcbas-production-ready-for.html
Note: Mandatory user-provided reference discussing ONU PCBA production readiness and Vortixion engineering context.
Link:
https://www.protoexpress.com/blog/pcb-surface-finish-types/
Note: Useful background on surface finish choices and assembly trade-offs.
Link:
https://resources.altium.com/p/pcb-stackup-design
Note: Provides practical background on stack-up planning and signal integrity considerations.
Link:
https://www.ti.com/lit/an/szza009/szza009.pdf
Note: Supports discussion of return paths, grounding, and high-speed layout sensitivity.
This post was reproduced from: https://www.industrysavant.com/2026/06/pcb-stack-up-copper-thickness-and.html
