Why is power densification an important goal for converter designers? The power conversion circuits that supply power-hungry systems like data center servers and the increasingly intelligent vehicles on our roads need to be able to handle more power, but in a smaller footprint. more reduced. As simple as that.
As we push more of these systems, they must do more work in the same or less time. By definition, that means more power. But space is at a premium, both in data centers and in vehicles. Building larger circuits to handle the extra power is usually not an option. In fact, there is pressure to significantly reduce size while increasing energy efficiency and power. Hence, the densification of energy is a primary objective for designers. Added to this is increased efficiency to mitigate the thermal challenges of heat dissipation. It is also important for conserving energy, as the world is increasingly dependent on generating electricity from renewable sources.
The broadband effect in power system design
Among the options available to help achieve power densification, broadband semiconductors (WBGs) have quickly achieved widespread adoption. Before the pandemic, for example, EV upstarts were the main proponents of WBG semiconductors in the automotive space. Now the tables have turned, and established brands are moving quickly to ensure comparable performance in their upcoming all-electric product ranges.
WBG technologies - silicon carbide (SiC) in particular, as well as gallium nitride (GaN) and others - make it possible to significantly increase the efficiency of power conversion, especially when switching at a frequency significantly higher than the range suitable for their silicon counterparts. They can also operate reliably at higher temperatures, alleviating thermal management issues and can reduce the size, weight, and complexity of any refrigeration system.
Faster switching also allows much smaller circuits to handle the same or more power. Specifically, operating the switch at a higher frequency allows smaller associated components, such as capacitors and inductors, to manage and smooth the flow of power in the input and output circuits. This is already widely known, although the lower values of capacitance and inductance required are only part of the story.
Typical switching frequencies of ordinary silicon power semiconductor-based converters have been in the range of a few tens of kilohertz: say, 30-80kHz. At these frequencies, polypropylene capacitors are suitable and widely used, as they have been proven, are reliable and, above all, cost-effective. However, above this frequency range, parasitic effects cause excessive resistive losses and self-heating.
More material science
We have worked with most of the major power electronics teams to develop new prototype converters around SiC power transistors. Investigating the new demands that these power switches place on support circuits allowed us to develop our KC-LINK ceramic capacitors, based on a proprietary high-voltage C0G dielectric that ensures extremely low Effective Series Resistance (ESR) and very low thermal resistance. They can operate with minimal loss at frequencies in the low megahertz range and handle very high ripple currents with no change in capacitance against DC voltage. Capacitance is also extremely stable over temperature. The ability to work up to 150°C allows mounting close to fast switching semiconductors in high power density applications. Available series offer voltage ratings from 500V to 2000V to cover a wide range of applications, including use with 400V and 800V electric vehicle battery systems.
We have also developed Transient Liquid Phase Sintering (TLPS), a solderless interconnect technology that enables the construction of high-capacity lead-free MLCC cells that take up little space and take advantage of the temperature stability of the Class I C0G dielectric to address power applications. high power that can reach temperatures of 150 °C and more without refrigeration.
On the other hand, WBG penetration in data center server applications is typically based on GaN technology. Typical switching frequencies have been stuck at around 300 kHz for many years. This has increased with the advent of GaN, although it is still around 900kHz. In this case, we find that the performance of the magnetic components is the main limiting factor. Inductors have two loss mechanisms, comprising resistive losses due to winding, as well as power losses experienced in the form of heating of the ferrite or metal compound core. The ideal is to minimize core losses without compromising its magnetic permeability, which is the basis of its ability to resist current changes within the circuit and to store energy in the magnetic field.
Another challenge for materials science is one that our teams have embraced and are ready to announce a solution for. While retaining high magnetic permeability, this new material is optimized for lowest loss in the 1-5MHz frequency range to allow increased switching frequencies of GaN-based converters. As in a SiC converter, increasing the switching frequency allows the values of capacitance and inductance to be reduced, which ultimately results in a higher power density.
Increasing the switching frequency of the power supply has other advantages. The transient decoupling capacitance required to protect critical parts, such as the main processor, can be greatly reduced. Historically, these capacitors have been either tantalum or aluminum polymer. Reducing the decoupling capacitance dependency allows a small array of Class II MLCCs, such as the X5R, X6S, or X7R devices, to be placed directly next to the processor. The next goal we are working on is to integrate aluminum polymer decoupling capacitors on the chip holder inside the package, so that they work together with the on-chip silicon capacitors. This could overcome the decoupling problems faced by processor designers today and allow for higher converter frequencies; possibly up to 10MHz and more in the future. It could take about five more years of engineering effort.
We've also found that performance improvements in one part of a system can hit a dead end, causing designers to look more closely at other parts of the system for further improvement. Our materials department formulated U2J ceramic dielectric specifically to help develop early switched tank converters. With custom inductor geometries added to the mix to reduce magnetic core loss, these converters unleashed a dramatic increase in 48V to 12V conversion efficiency in data center server distributed power systems.
These converters define the upper limit -for now- in terms of 48V-12V conversion efficiency. When that limit was reached, attention shifted to point-of-load (POL) converters. Here, high-performance processors and FPGAs operate with a combination of low digital supply voltage and high clock frequencies that cause the current demand to change rapidly, reaching a high peak value. The multi-phase voltage regulators commonly used to power these ICs force designers to trade off transient response with ripple current. Transient response is limited as all phases need time to settle in sequence. In addition, these polyphase regulators prevent power densification, since it is not possible to reduce the width of the inductor and maintain mechanical stability. Four-terminal, double-wound inductors have enabled the development of the transducer voltage regulator (TLVR) in which all phases respond simultaneously for faster transient response. Pulse Electronics, part of the Yageo Group, is a leader in TLVR inductors.
WBG and noise emissions
The fast switching transitions of WBG semiconductors create an unwanted challenge for designers: electrical noise emissions, or EMI/EMC. To address this design challenge and bring converters and inverters into compliance, KEMET's Magnetism group has developed nanocrystalline core materials for use in EMI common mode chokes that deliver broadband performance in a smaller package.
What the future brings us
The developments we are seeing, including advanced materials, new circuit topologies, and new demands for capacitors and inductors, are highly interrelated. Together, they enable continued progress in the effort to increase energy efficiency and power density. Who knows when, or even if, we will reach a limit beyond which no further improvement is possible?
Author Peter A. Blaise, Senior Director of Application Engineering at KEMET, a YAGEO company.
