LED PCB Board: How to Choose the Right Architecture

An LED PCB board is a PCB that provides electrical connections for LED emitters and their support circuitry, as well as the means for mechanical mounting and heat transfer from the entire assembly (LEDs + circuit board). The term LED PCB refers to an application, not a material used in the manufacture of the board. An LED PCB can be manufactured from FR-4, a metal core, flexible material, ceramic, or a hybrid construction. The choice of which type of PCB to use will depend on the electrical load and the complete thermal path to ambient.

While the PCB offers some ability to spread heat from an LED package, it cannot remove heat from a product by itself. Heat from an LED package may exit the PCB directly via convection and radiation or may continue to be transferred through the interface of the PCB to the housing or heat sink of the product and then to the surrounding air. This distinction results in a change from “What type of material is best?” to “What type of PCB architecture closes all electrical, thermal, and mechanical paths?”

What Is an LED PCB Board?

An LED PCB board is a PCB designed specifically for LED emitters and their associated circuitry. The PCB provides all of the electrical connections and mechanical mounting support and forms one of the thermal paths for the complete assembly (LEDs + circuit board). The final assembly is referred to as a populated assembly or PCBA once the LEDs and other components have been mounted on the bare PCB.

The term LED PCB can be used for a low-current indicator board, a concentrated lighting module, or a flexible strip. These three types of LED PCBs do not require the same type of substrate or stackup since their current levels, heat densities, routing paths, shapes, and mounting conditions will differ. Referring to all three types of PCB as aluminum PCBs is misleading since each requires a unique set of engineering decisions that relate to the product design.

The efficiency, service life, and optical output of an LED are determined by the selected emitter and its operating conditions, not by the PCB. The PCB’s role is to support current regulation, spread heat, maintain the mechanical and optical alignment of the emitters, and allow for a repeatable method of assembly.

Functions can move between boards

The functions needed by LED products typically include emitters, current control, electrical connections, protection for the emitters, and sometimes digital or analog control. These functions may share one PCB. Some products use the driver and control circuitry on a separate PCB. The arrangement or type of board will affect the location of heat sources, physical connections, service access, and the way the product is tested.

The term LED PCB does not define a complete lamp, display, luminaire, or power supply. The term only identifies the board-level part of the overall product assembly (lamp, display, or power supply). The characteristics of the total product will be determined by the properties of the optics, housing, driver, wire harness, cooling hardware, and ambient environment.

Is an LED PCB Board Always a Metal-Core Board?

No, the term “LED PCB” does not specifically mean that the PCB has a metal core. The metal core is an option for many LED products and should be compared to many other types of materials, including FR-4, flex, ceramic, and hybrid materials. Each of these materials should be compared based on the defined heat flow, routing, isolation, mechanical contact, assembly, and cooling for an LED system.

When metal core becomes a serious candidate

If concentrated heat is generated by the emitters and there is an established mechanical connection to the housing or heat sink, then a MCPCB is a good candidate to consider. If a dielectric material is used between copper and the base, the material will have to conduct heat and provide electrical isolation, so that will have to be considered. The mounting contact, interface material, and the next cooling surface would still all be part of the same thermal path.

Cree LED’s Thermal Management of XLamp LEDs discusses the PCB, TIM, heat sink, and ambient as separate thermal segments. The note from ams OSRAM, Thermal Management of LED Light Sources, uses the same thermal separation in its architectural approach.

Start With Current, Not the Substrate

The substrate should be determined once the exact LEDs and current-control method have been identified. The arrangement of the LEDs in series or parallel will determine the voltage required, branch current, fault characteristics, and the points at which electrical power will become heat. Making a substrate material decision before completing the thermal loss map may result in selecting the wrong thermal material to solve the problem.

An indicator LED could be operated through an appropriate limiting method. A high-brightness LED string would likely require a dedicated constant-current driver. Texas Instruments has established this relationship with its LM3405AEVAL design, where a buck regulator drives the series LED string. While the document illustrates the basic architecture, the component values and operating limits apply only to this device and its documentation.

Parallel branches need a sharing strategy

Paralleled LEDs do not always share the same current. The Texas Instruments article LED Current Regulation with Ballast Resistors shows how a small voltage difference can result in the parallel branches having unequal current through them. A difference in bin, a difference in junction temperature, and resistance in the electrical path all contribute to this variation in the parallel branch currents. Ballast resistors can serve to balance the current in parallel LED branches, but they will not serve as a universal solution for all drivers and power levels.

A current plan should be defined, thermal losses mapped, thermal load estimated, and PCB and cooling design options compared. When the regulator is near the emitters, the PCB may still be heated by the regulator, even if the LED calculation did not indicate excessive heat.

Verify before release:
Part number; anode and cathode orientation; mapping from symbol to pad; current limit.

Which Board Architecture Fits the Thermal and Mechanical Job?

There is no universal wattage threshold to validate an LED PCB architecture. Each candidate must be reviewed to see how the candidate will meet the requirements of the heat path, routing within the PCB, electrical isolation, mechanical characteristics, assembly methods, and cooling interfaces.

When the heat density, copper spreading, via strategy, board area, and external cooling path support the design, FR-4 is a viable option. However, designing with an MCPCB does not necessarily conclude the analysis; for instance, a poor interface between the board and its enclosure may dominate system performance and thermal transfer regardless of how efficiently the board transfers heat through its layers.

Although flexible PCBs address shape and packaging issues, bending the board alters the mechanical characteristics of the board as well as the available area for heat spreading. A board that has only been bent once during installation may not have the same requirements for copper fatigue as a board subjected to repetitive dynamic motion. While ceramic or hybrid stack-ups can be used to address localized heating and thermal isolation, the geometry of these structures, along with the assembly conditions and ambient temperature, ultimately impacts the final product.

LED PCB Design Risks

Voltage Drop in Long LED Boards

Voltage drop results from resistance in copper traces, connectors located along the length of the board, and cut or splice points. The effect of long LED boards losing voltage may manifest itself as lower branch current, uneven brightness of the LEDs, a color shift in LED devices sensitive to current, or excessive heat being generated in close proximity to narrow copper traces.

LED PCB design requires evaluating the entire current loop rather than only considering copper weight; in general, wider conductors, shorter branches, multiple feed points distributed across the board, and separate return paths will help, but the best choice will depend on the length of the conductors, the current, acceptable temperature rise, and the driver’s regulation range.

Thermal Vias in FR-4 LED Boards

Thermal vias in FR-4 LED boards are plated holes used with copper thermal pads. These vias can work with copper planes to move heat away from exposed or bottom thermal pads. While the number of vias may affect heat transfer away from the soldered thermal pad, the via hole size, copper plating, via spacing, copper connection, board thickness, and the next available heat-spreading surface all should be considered when defining how well the thermal pad will transfer heat.

The placement of vias too close to a solderable thermal pad may draw solder away from the thermal pad during soldering and thus may result in voiding; design and fabrication controls for the openings in the PCB are critical when designing with vias.

Optical Uniformity Across LED Arrays

To achieve optical uniformity from the LED devices used in an LED array, one must consider additional factors besides the nominal output of each LED; LED bin, branch current, junction temperature, package orientation, placement accuracy, solder thickness, board flatness, LED pitch, and the distance to any optical device must all be evaluated. The assembled board should be validated using the intended optics during the intended operating mode of the board, rather than using individual LED measurements.

Solder Mask Performance

The performance of a solder mask can be specified by its material type and performance characteristics. The white solder mask (Taiyo America PSR-4000 LEW1) is specified to have reflectance and colour-retention properties. While this data sheet provides some basis for specifying a mask based on both material type and performance characteristics, it does not demonstrate that all white solder masks have the same reflectance or that all white solder masks will perform the same under all assembly profiles. The most useful specification of a solder mask’s performance is how much reflectance is retained after the solder mask has been assembled and subjected to the expected operating environment over its expected life.

Warpage in Long LED Boards

When designing long LED boards (light bars), large panel boards, or boards that must contact a flat housing surface, warpage must be taken into account. Warpage can have an adverse effect on solder-joint coplanarity, LED-to-lens alignment, and mounting pressure, as well as thermal-interface thickness. Factors contributing to warpage include copper thickness imbalance, board thickness, panel geometry, and thermal exposure during assembly. Sometimes a board may pass electrical testing but will produce optical nonuniformity in the field or create an inadequate thermal interface.

Where Does the Heat Go After It Reaches the PCB?

Heat leaves the LED junction but must continue further through the package, solder point, and into the PCB. After the heat has entered the PCB, part of the heat may escape from the PCB through convection and radiation off the board surface, and part of the heat will pass through the contact or thermal interface into the product enclosure or heat sink before reaching ambient air.

Interfaces belong in the model

Cree has modelled the thermal paths from the LED to the ambient environment as combinations of series and parallel thermal resistances. This is especially important for systems that use multiple emitters, as board-level thermal spreading is not the only source of thermal conduction. Contact area, interface material, mounting pressure, the geometry of the housing, convection from the heat sink, and ambient conditions will affect what occurs next.

ams OSRAM links the temperature of the LED with its light output, forward voltage, wavelength, and lifetime. The temperature of the LED cannot be treated as a separate reliability factor, and it will affect both electrical and optical behaviour of the LED while in use. The allowable limit and conversion method should be derived from the requirements of the LED, including the entire system, not from a generalised web table.

Material conductivity alone may give false impressions as to how a material behaves under specified conditions. The assembled thermal path includes several factors: thickness, area, interfaces, and boundary conditions. An assembled board segment could potentially show a dramatically low thermal resistance, but if there is inadequate contact between the board and the housing, that segment will have little effect on the overall performance. The prototype should be tested in the same configuration as the production units.

Should the LED Driver and Emitters Share One Board?

Using a single board for both the LED driver and emitters can reduce the number of interconnects and package volume and ultimately shorten the driver-to-LED electrical path. Placing the driver directly on the same PCB as the emitters moves driver loss closer to the LEDs, thereby potentially creating a temperature gradient between the two components. Alternatively, there may be cases where separating the driver from the emitters can reduce some of the thermal coupling but adds connectors, wiring, footprint, and assembly work, as well as additional test interfaces.

Treat the driver as a heat source

The closer the driver is to the emitters, the greater the potential impact of the driver on the emitter junction temperature. The components (switching devices, current regulators, resistors, and control components) that are placed onto the same thermal structure as the emitters will also add to the cumulative effect of losses generated by those components on the overall thermal setup. Similar to the situation of having the driver next to the emitter, the local copper areas, required spacing, and testing and validation data will all change as a result of the placement of the driver.

A split architecture can give the emitter board a simpler thermal task and move control electronics into a different environment. Connectors and cables add resistance, assembly operations, possible failure points, and space. They can also improve test access or field replacement, depending on the product.

Compare both layouts by locating heat sources, tracing current paths, checking routing or interference constraints, identifying power and control connectors, and confirming that each board can be measured under realistic operating conditions.

LED PCB Assembly Methods

Surface-Mount LED Assembly

Polarity, rotation, pad mapping, solder-paste volume, and coplanarity of packages with exposed thermal pads must all be controlled when working with surface-mount LED assemblies. Stencil and reflow methods help avoid excessive voiding or starvation of the electrical joints in packages with thermal pads, and placement errors can create an offset from the optical center of the LED. Moisture sensitivity, handling, pick-and-place tooling capabilities, and reflow profile parameters must follow the LED manufacturer’s specifications.

Through-Hole LED Assembly

Through-hole LEDs add mechanical lead retention but are subject to height and alignment variables. Anode and cathode orientation, lead forming, stand-off height, insertion force, and the position of the optical center are all variables that must remain consistent throughout the assembly process. During wave, selective, or manual soldering, the thermal conditions applied to each package and lead may differ. If the product is expected to endure vibration or be mounted to a panel, check for mechanical support and lead stress instead of assuming that the strength of the through-hole connection is sufficient.

COB LED Assembly

COB LED assembly uses bare LED dies directly attached to the substrate, connected by either wire-bond or flip-chip interconnection followed by phosphor and encapsulation processes. Die attachment, surface cleanliness of the substrate and LED dies, bond quality, coating uniformity, and local heat flux should be controlled as primary variables in the design. Because of the high local concentration of heat, substrate and housing interfaces may be thermally sensitive. When an LED die or bond structure is exposed, opportunities for rework or redesign can be limited. The requirements for COB LEDs must conform with those specified by the manufacturer for the COB device and manufacturing process rather than ordinary SMD rules.

LED PCB Applications

LED Strip Lights

Flexible and rigid LED strips are composed of repeated LED sections that use long power and return paths. Voltage drop can result in decreased current and apparent brightness variations along the length of the LED strip, but is especially noticeable across connectors, cut points, or narrow copper conductor transitions. Bend zones must be designed so that the copper does not fatigue and LEDs and solder joints are kept away from areas of concentrated strain. Adhesive mounting or the use of an aluminum channel affects the heat path; installation conditions must be considered during thermal validation.

A strip powered only from one end may pass a short bench test but show visibly different brightness between the beginning and the end when fully installed. Wider conductors or shorter segments may reduce the voltage drop, but there is no universal feed interval.

Automotive Lighting

When creating automotive LED circuit boards, the arrangement of the emitters on the circuit board must align with the lens and reflector of the lamp housing. However, the design of the PCB must also account for other factors (vibration, humidity, thermal cycling, changes in supply voltage, etc.). The mounting holes, connector location, and thermal contact area in the lamp housing need to be established before the PCB is designed; otherwise, this could affect how the PCB will fit into the lamp housing once all these components are placed. Coatings, seals, and potting can protect the assembly but may also change heat transfer and material compatibility. Validation should use the final housing and optical geometry.

Even if the replacement PCB has the same electrical load requirements as the original PCB, it may not have the same lens focal position or contact the lamp housing evenly. A poorly shaped beam may result, or there may be localized hot spots even if the bare PCB functions correctly.

Outdoor Lighting

Streetlights, floodlights, and architectural lights that use LED circuit boards typically have a sealed or semi-sealed enclosure. Reduced air movement, moisture ingress, corrosion, surge exposure, thermal-interface aging, and housing flatness can affect long-term performance.

Coating and potting materials used in outdoor lighting should not interfere with the LED package or obstruct intended heat-transfer surfaces. Enclosure assemblies should be tested for temperature and optical output after achieving stable operating conditions, rather than only testing the circuit boards on the bench.

Display Backlights

A display or backlight assembly must be designed to maintain consistent LED pitch and orientation, branch current, bin selection, and distance from the LED package to the diffuser or light guide assembly. Variations in the placement or solder heights of LEDs across an array may result in highly visible hot spots. If a backlight board warps, it can create regional differences in brightness due to a change in the optical gap. To verify uniformity, spatial luminance or color uniformity should be measured using the assembled optical stack rather than only using individual isolated LEDs.

Medical Lighting

The lighting systems used in medical and inspection applications often have requirements for the controlled nature of the CCT, CRI, spectral output, as well as the amount of stable illumination that can be maintained over long periods of time. The cleanliness of PCBs, the presence of flux residue on the surface, thermal drift as the unit approaches its operating temperature, and the repeatability of the current used will all affect the optical consistency of the source. The acceptance limits for some devices, however, will depend on the device applications. The PCB does not establish medical performance by itself. Thus, to evaluate performance, the assembled light source needs to be evaluated in conjunction with its associated optics, operating modes, and defined measurement geometries.

Machine-Vision Lighting

Machine-vision-type boards often use ring, bar-style, dome, coaxial, or backlight-type arrangements to provide a repeatable illumination source for the camera. Both the LED positioning and datum (reference plane) of the machine-vision board will directly affect how the machine-vision lighting will illuminate the camera’s field of view. Since strobe operation during machine vision will create high peak currents for very short duty cycles, monitoring average power consumption only will not necessarily provide an accurate indication of how much transient electrical and thermal stress is generated during this operation. To accurately assess how the machine-vision illumination system performs, synchronization, pulse width of the strobe operation, uniformity of illumination, and stability of the emitted wavelength should be examined with the actual camera exposure and target distance.

UV or IR Modules

UV and IR LED modules can concentrate a substantial amount of heat within a small area and can produce a significant amount of output that cannot be seen with the unaided eye. Both the emitted wavelength and radiant output will vary with temperature, and unsuitable solder masks, adhesives, connectors, and encapsulants may degrade as a result of UV exposure. Define optical alignment, current control, compatibility with the LED module’s package, and safe measurement limits when using UV or IR emissions. The validation of UV and IR LED modules should use calibrated detectors that are appropriate for the wavelengths of operation and the application distance.

Indicator Panels

Indicator and control panels may have lower power output than other light sources; however, they still require current limiting, polarity control, ESD protection, and consistent brightness across the various indicator channels. In addition to creating lower power output per channel, multiplexing or scanning of displays also imposes duty-cycle and peak-current limitations, and the densely packaged driver circuits that run multiple indicators may produce more heat than the indicators themselves. FR-4 material is typically adequate for the design of indicator panels, provided local losses of the indicators are accounted for in the design, as well as conditions at the connector, the enclosure temperature experienced by the indicator, and whether the indicator will be easily visible through the final panel.

Turn Design Assumptions Into Verification Evidence

The specification of a particular material, stackup, or heat sink represents an assumption on the part of the designer with respect to the intended application and conditions of the component. Verification testing verifies that the design assumption was correct through the identification of the item to be measured; the determination of the position of the item; an acknowledgment of the operating conditions; the specification of the acceptance limit; and the provision of results. Without these fields, temperature measurements and inspection results do not provide any information regarding whether the design meets the intended application and conditions.

Board temperature is not automatically junction temperature

The junction temperature remains undefined until the thermal characteristics of an LED package are measured using the specified solder-point measurement method.

This thermal measurement model connects the solder point of a supported Cree LED package with the manufacturer’s junction-temperature estimate.

Measurements taken from a location other than the specified solder point cannot be used as an alternative.

When documenting the measurements, the LED part number, measurement location, ambient conditions, time taken to reach thermal stability, drive current being used, and the basis of calculations should be documented.

Without this data, a reviewer reading it at a later date will not be able to properly determine what the reported temperature means.

Inspection methods should be connected to a failure mode.

You should be able to determine the polarity and position of the LED from first-article inspection.

You can validate the current being drawn by the LED using electrical measurement and use optical testing to verify optical output and color.

A single AOI, ICT, functional, or thermal check does not validate the entire product; each inspection method must also establish an acceptance criterion that corresponds to the assumption it is testing.

Prepare the Inputs for Design or Manufacturing Review

To identify conflicts before a substrate name is treated as an approved solution, document the electrical architecture, thermal environment, mechanical interface, material and assembly constraints, and validation criteria.

Electrical definition

• The exact MPNs for the LED and the driver.
• The schematic or net intent, series and parallel arrangement, and current limits for the configuration.
• The input range and operating modes for the driver, protection requirements, and estimated component losses.

Thermal and mechanical definition

• The ambient operating range and duty cycle for the LED configuration, optical targets, and package-specific junction limits.
• The board outline, mounting and contact-area features, flatness requirements, and enclosure geometry.
• The TIM, housing or heat-sink concept, airflow, isolation, and touch-temperature constraints, if applicable.

Manufacturing and validation definition

• The materials required for the PCB containing the LEDs and routing or layer requirements.
• The required BOM, Gerber or ODB++ data, pick-and-place data, and drawings.
• The acceptance criteria for polarity, placement, electrical current, temperature, optical output, and functional requirements.
• The prototype operating conditions and records required to validate results against design assumptions.

Record the LED bin code, chromaticity window, luminous-flux tolerance, dimming characteristics, startup transient, surge protection, connector pinout, cable-length restrictions, coating restrictions, optical-alignment datum, thermal-sensor location, thermocouple attachment point, stabilization criterion, airflow state, enclosure state, measurement uncertainty, and pass/fail calculation.

Conclusion

The ideal LED PCB board architecture is not determined solely by the material with the highest conductivity number. The best product architecture is achieved by using a system approach to close electrical and thermal paths from junction to ambient under stated operating conditions.

Define current and losses first, then compare PCB candidates based on the real mechanical interface, and finally verify temperature, current, optical output, and assembly results.