… So let’s challenge old design concepts.
As Martin Cotton begins to enjoy retirement (it seems that’s his word for embarking on complex wireless projects he’s always wanted to explore), I have been invited to take over this spot and share my perspectives on the substrates industry. You may know I have recently become technology ambassador for Ventec after starting my career as a research physicist and spending many years in the PCB business, most recently as chairman of the EIPC.
As a technology ambassador, my role begins with encouraging designers and fabricators to think creatively at the substrate level. By taking advantage of critical innovations happening here, it’s possible to deliver new products that deliver much better functionality, form factor and reliability than their predecessors. This presents tremendous opportunities to leapfrog competitors technically by challenging accepted practices and creating new products that really stand out.
Consider the automotive industry. We only need to look at what’s coming toward us to appreciate the tremendous impact LED lighting is having on the cars we drive. While the logo will always express the brand’s identity, the emotion in the lamps now expresses the attitude at least as strongly. Think of those daytime running lights, which are on at any time of day or night. We respond to what these tell us about the vehicle long before we see the badge.
Clearly LEDs have opened a new universe of styling opportunities for car designers, who now have much more freedom to experiment with the shape and size of lighting units. They can take advantage of a larger light-emitting surface, while at the same time reducing the overall volume of the assembly; as more and more features such as structural crumple zones, mandatory safety equipment, and electronic driver-assistance systems compete with passenger and luggage space within the vehicle, every cubic centimeter saved at the corners is highly valuable. Of course, LEDs bring other advantages too, such as greater reliability – lights are now expected to run maintenance-free for the lifetime of the vehicle – and the fact that headlamps can now be brighter and whiter, to better illuminate the road ahead for safer night-time driving.
The LED emitters themselves continue to improve at a rapid rate. While Moore’s law has famously described progression of large-scale integration (LSI) chips, the LED business goes by a similar rule of thumb: Haitz’s law states the cost per lumen falls by a factor of 10 every decade, while the luminous flux per LED package increases by a factor of 20 for a given wavelength of light.
Designers can take advantage of this to simultaneously raise the performance and increase the cost-effectiveness of automotive lighting. The fact remains, however, that although LEDs are far more efficient than their incandescent predecessors, between 60% and 85% of the energy supplied is emitted as heat, not light. This heat must be managed, and this can slow the rate of progress as far as performance, reliability, and overall size of the light units is concerned.
At the beginning of this automotive lighting revolution, designers established a model for thermal management that many have not yet challenged or changed. It combines the conventional LED chip-on-board (COB) module with a large finned heatsink to dissipate heat from the emitter. A thermal interface material (TIM) interposed between the two has done its best to transfer heat from the board into the heatsink. We must remember the thermal conductivity of the TIM is only good relative to the air that it displaces between the two surfaces. Its conductivity is quite low relative to the metal heatsink, and hence dominates the assembly’s overall thermal performance.
In addition to imposing performance limitations, the TIM represents a manufacturing process that takes time and adds supply-chain challenges. The TIM must be procured, stored, supplied to the factory floor at the right time, and applied either by machine or manually.
You could say this approach to designing automotive lamps has become a habit. By thinking creatively at the substrate level, designers can now kick that habit by using insulated metal substrate (IMS) as both the mechanical support and heatsink for the assembly. No TIM is required, so we can eliminate both the associated performance barrier and process overheads.
Sounds great, but what’s the catch? Designers need to understand that the IMS thermal-transfer properties are not the only parameters that matter. The mismatch in coefficient of thermal expansion (CTE) between the LED package, copper circuit layer, dielectric layer, and aluminum baseplate can be enough to cause stress-induced fracturing of electrical interconnects. At Ventec, we have seen CTE-related failures in solder joints and in circuit-layer traces close to LED packages.
It is possible to overcome these effects through careful design. For example, by combining a special low-CTE aluminum with a dielectric of reduced elastic modulus, we have been able to reduce no-light failures by nearly 40% in thermal cycling tests.
Moreover, the dielectric formula can be engineered for the exact thermal conductivity needed to maintain the emitter junction at the desired operating temperature, within the desired maximum size and weight. The running temperature is critical because operating outside the desired range causes noticeable changes in both the flux output and spectral content of the light, which are obviously undesirable.
We can now see adopting new approaches to design, and engineering solutions to the challenges they bring, is needed to satisfy the automotive industry’s stringent demands for performance, quality, and reliability, therefore ensuring the continued success of LED-lighting technology in this important market.