LED Lighting PCB: Materials, Thermal Design and Assembly

The Board Is the Thermal Design

An LED is a semiconductor, and like every semiconductor it degrades faster when it runs hot. That single fact makes the circuit board carry more responsibility than in almost any other lighting product. The board has to hold the emitter, deliver the current and, above all, move heat away from a device that converts only part of its input into light. The rest becomes heat concentrated in a footprint of a few square millimetres.

A light emitting diode is also more sensitive to temperature than most components. A rule of thumb used across the lighting industry is that every ten degrees Celsius of additional junction temperature roughly halves the expected life. Before that, the symptoms appear as colour shift and reduced output: the luminaire dims over time, and the light that remains no longer matches the colour it was specified for. Good thermal design is therefore not a reliability detail bolted on at the end. It is the design.

What an LED Lighting Board Has to Do

The board performs three jobs at once. It mounts the emitters in a defined pattern, it carries the current from the driver to each emitter, and it conducts heat out of the package and into whatever structure the luminaire provides.

The electrical duty is heavier than on a signal board. A single high power emitter may draw several hundred milliamps, and a board in a streetlight or a floodlight carries strings of them. Trace widths, copper weight and the resistance of the finished path all matter, because voltage drop along a long string shows up as uneven brightness between the emitters at each end. The layout therefore has to be designed for current first and cosmetics second.

Types of LED Board

• Single sided FR-4. One copper layer, simple and cheap, used where the emitters are low power and the board is not the main thermal path. Common in indicator and low output decorative products.
• Double sided FR-4. More routing flexibility and better current distribution, used for mid power arrays and where the driver circuitry shares the board.
• Metal core board. An aluminium or copper base under a thin dielectric. This is the mainstream choice for high power lighting, because heat passes through a conductor rather than an insulator.
• Flexible board. Polyimide based, used for curved fixtures, strip lighting and shapes that cannot be formed from a rigid panel.

Our notes on metal core PCB construction explain how the dielectric layer in the middle of that list is built and what it costs thermally.

Substrate and Copper Weight

The dielectric layer in a metal core board is the compromise at the heart of the design. It has to insulate, because the base is conductive, and it has to conduct heat, because that is its purpose. Typical filled dielectrics reach thermal conductivities of about 1.5 W per metre kelvin and above, and thicker dielectric means higher breakdown voltage but higher thermal resistance. The design has to satisfy the safety requirement without over-insulating the emitters. The choice between an aluminium PCB base and a copper base then follows the power level, since copper spreads heat better and costs more.

Copper weight follows the current. One ounce copper is normal for standard lighting boards. Two to three ounces is used where currents are high or where the copper is doing double duty as a heat spreader. Heavier copper lowers resistance along the string, reduces volt drop, spreads heat sideways from each pad and makes the temperature across the array more even. It also costs more and sets a minimum feature size, so it is specified where it earns its place rather than by default.

The Thermal Path

Heat leaves an emitter through a chain: the die, the package, the solder joint, the copper pad, the dielectric, the metal base and finally the heatsink or enclosure. Each interface adds resistance, and the largest practical gains come from the weakest link rather than from the last one.

Two techniques are used to shorten the chain. Thermal vias placed under the emitter pad carry heat through a multilayer board into a plane below, which is useful when the board is not a metal core design. Direct conduction into a metal base is used when it is, because the dielectric is thin and the area is large. In both cases the solder joint under the pad has to be void free; a void under the thermal pad is a local insulator sitting exactly where heat is trying to leave. The wider thermal programme around these choices is covered in our notes on thermal management.

Layout Rules for Lighting Boards

Several rules distinguish a lighting board from a general purpose circuit board.

• Widen the high current runs and keep them short, because resistance in the copper becomes uneven brightness across the array.
• Keep the layout symmetrical. Series strings that share a driver should have the same trace length and the same copper, so each emitter sees the same current.
• Space emitters evenly and avoid clustering, because a group of hot packages next to each other raises the local ambient for all of them.
• Keep the driver away from the emitters. Driver losses and LED heat should not share the same area of board.
• Respect creepage and clearance. Mains fed luminaires have to pass safety requirements, and the metal base is part of that consideration.
• Avoid sharp corners and narrow necks in current carrying traces, which create local heating and reduce reliability.

How the Board Is Made

A metal core or FR-4 lighting board is fabricated with the usual imaging, etching and finishing steps, with the dielectric lamination added when a metal base is used. Surface finish is chosen for solderability and for the environment: hot air solder levelling, electroless nickel immersion gold and organic solderability preservative are all common, and a finish that survives the luminaire environment matters as much as the assembly process.

Assembly is then a surface mount process: stencil printing, component placement, controlled reflow and automated optical inspection. LED assembly brings two extra constraints. The reflow profile has to stay inside the emitter manufacturer limits, including for the lens and the internal materials, and the optical surfaces have to stay clean, because flux residue or handling marks on a lens become visible light loss. These requirements are the same ones that govern any SMT PCB assembly run, applied with tighter limits.

Driving the Array

Emitters are connected in series, in parallel, or in a mixture of the two, and the choice has a strong effect on reliability. Series strings are preferred because every emitter in the string carries exactly the same current, so brightness and temperature stay even, at the cost of a higher drive voltage.

Parallel groups carry a specific risk. If two emitters share a current path, the one that is slightly hotter has a lower forward voltage and takes more of the current, which makes it hotter still. That runaway is prevented by driving with a constant current source, keeping parallel groups small, matching emitters from the same bin, and making sure the thermal path is the same for each device in the group.

Applications and Cost

Lighting boards appear in domestic and commercial luminaires such as bulbs, panels and downlights; in automotive lamps and interior ambient lighting; and in industrial and outdoor products such as streetlights, floodlights and high bay fittings. Each has a different balance between cost, colour quality and thermal headroom.

As a 2026 reference, an FR-4 lighting board in medium volume falls roughly between 50 cents and two dollars per piece. An aluminium based board runs from about 1.50 to five dollars, and a high power metal core board from around four to ten dollars or more. Material, copper weight, layer count, size and quantity drive the spread, which is why a quotation against a specific board is more useful than a price list. Our LED PCB overview covers the same territory from the fabrication side.

Frequently Asked Questions

How long does an LED lighting board last? With a properly designed thermal path, light engines are commonly rated for 50,000 to 100,000 hours, but that figure depends entirely on the junction temperature the board maintains.

Is an aluminium board always better than FR-4? For high power lighting, yes, because the metal base is the thermal path. For low power products FR-4 is adequate and costs less.

How much copper is needed? One ounce suits standard boards. Two to three ounces is used for high current arrays and where the copper has to spread heat.

Why do emitters fail in parallel groups? Uneven current sharing causes thermal runaway. Constant current drive and small, matched groups prevent it.

What is the most common manufacturing defect? Voids in the solder joint under the thermal pad, which directly raise junction temperature.

Conclusion

An LED lighting board looks simple: an array of emitters on a panel. In practice it is a thermal design with electrical constraints attached. Choose the substrate from the power level, use enough copper to keep the current even and the heat spread, give every emitter a short thermal path with a void free joint, lay the strings out symmetrically and keep the driver separate. Do those things and the luminaire keeps its colour and its output for its rated life. Skip them and the board becomes the reason the product fails in the field.