Reflecting on the recent Research Workshop held at Tyndall National Institute, it is clear that there is a relentless drive towards miniaturization of power management systems.

Our annual MCCI Research workshop seeks to bring industry and international experts together on a specific research theme, to identify gaps in knowledge and opportunities for further research. Our aim is to create a mandate for MCCI researchers to pursue industry relevant research. The theme this year was “Research Challenges in Power Conversion Circuits”.

We were delighted to welcome input from our Industry partners, Dr Basil AlMukhtar (ON Semiconductor) and Tue Trong Vu (ICERGI), academic collaborators Dr. Robert Pilawa, University of Illinois at Urbana-Champaign, Prof. Aleksandar Prodic, University of Toronto, and Prof. Bruno Allard, Head of Ampere Lab, INSA de Lyon, in addition to contributions from our own research leaders Seamus O’Driscoll, MCCI and Prof. Cian O’Mathuna, Tyndall National Institute.

“It’s good to see academic researchers are in line with Industry demand of compact higher power density solutions”, observed Dr Basil AlMukhtar. “We need to break power converters operating frequency legacy. Higher levels of integration, with power management, all passives and energy storage elements in one system on chip solution demands 20MHz to 50MHz operating frequency. Multilevel and series switch cap converters show great potential in reducing transistor stress levels, but the capacitors involved use considerable die area, depending on load levels. This limits the feasibility of this approach for compact, multiphase, multi-rail power solutions”.

Magnetics, New Converter Topologies, and HF Switching

For many years, Tyndall National Institute has been at the forefront of fundamental research on high-frequency integrated magnetics, with the results obtained by the teams in Tyndall among the very best on-chip integrated inductors in the world. Still, for many power electronics applications, those inductor values are just too small, despite being in some cases an order of magnitude larger than those presented by other research institutions. The research challenge here then, quite simply, is that power electronics requires much larger inductor values, than most other microelectronics applications.

This conundrum has improved dramatically in the last 4-5 years however. Research on integrated magnetics has been constantly improving the characteristics of inductors (increasing their values) while on the other hand, new topologies and high frequency switching based solutions have significantly reduced the requirements for inductors. So, these two trajectories, one on the increased inductance values and the other on reduced requirements for the inductance values, have been converging and the realisation at the MCCI Research Workshop was that we may now be observing the first evidence of the intersection of the two trends.

On circuit topologies, hybrid and multi-level converters require much smaller inductance values than conventional solutions, and on the device side, high frequency switching provides just that. High frequency switching can be obtained both through the reduction of voltage stress and/or through the utilization of wide bandgap semiconductor devices such as GaN and SiC. One such informative presentation on the day was the GaNonCMOS project funded by EU Horizon2020, where the research proposes a new GaN switch and monolithic bridge structures in the 25V and 100V range. These are intended to enable breakthrough performance in power electronic regulators for a number of next generation applications such as multi-node computing, 48V automotive, aerospace and energy.

Digital (Mixed-Signal) Control and Data Converters

Digital and mixed signal control solutions for emerging applications are becoming a necessity. This is being driven by increased complexity of converters and the need for advanced power management control algorithms, both on the converter and system level. Key elements in building those controllers will be pulse-width modulators & data converters. Our research team have strong experience in this technology and are exceptionally well placed, close to system and application level researchers, to create custom ADC and DAC solutions.

Trends

There is a relentless drive towards miniaturization of power management systems, and integrated solutions will win the day. PMICs are becoming a bottle neck in future minimization of devices and the reduction of weight/size has become the key priority for many applications, including mobile electronics, automotive, and avionics.

The trend toward compact and highly efficient power supplies is well underway in mobile applications where customer demand is strong and persistent. While in higher power applications including medical/industrial, renewable energy, etc., the evolution happens at a slower pace due to cost and reliability constraints.

Multi-level converters, including switched capacitor types, open the door to new power density and efficiency benchmarks for power supplies. However, there are control and driver challenges associated with usage of multiples switching devices. So being able to integrate power magnetics, capacitors, and active switches in the same substrate is key, enabling a significant reduction in cost, complexity and consequently realizing the deployment of multilevel converters in a broader context.

New solutions for power management will require more advanced packaging for a number of reasons. Key among which are a requirement for reduced size, for dealing with parasitics, which are much more dominant at higher frequencies, as well as dealing with potential thermal issues. The possibility of packaging the entire converter on a single IC, for those reasons has outstanding value.

Application Driven Research


Internet of Things (IoT) and the expected exponential increase number of sensors and wearable electronics are making power harvesting very attractive. A significant barrier to the widespread adoption of power harvesting however is the need to up-convert and regulate the energy harvested to levels more commonly used to drive the circuits in the application. The possibility of creating energy harvesting sensors, followed by ultra-low power electronic solutions for regulating the harvested unregulated energy is also seen to be of extremely high value.

As with all emerging and disruptive technologies, it is important to carefully define the value proposition in each application case, for example integrated capacitor technologies vs existing off-chip solutions. When comparing different possible solutions the use of “relevant” Figures of Merit is more important to understand, and may need to be redefined, than the use of sometimes artificial academic figures of merit.

Addressing the Challenges

Significant challenges remain to realising many of the solutions discussed at our Workshop, chief among those being standardised, cost-effective integration processes, the manufacturability of solutions, reliability, and addressing customer needs. Collaboration is the key to addressing some of the greatest challenges, as the expertise & knowledge will never reside in one location. The immediate and obvious areas identified in which collaboration will have the greatest impact are in the trial of existing Tyndall ALN Capacitors, Switched Capacitor Research Prototypes, ALD capacitors in Hybrid Switched Capacitor PWRSoC point of load, advanced gate driver and advanced integrated Hybrid point of load research.

MCCI Roadmap

Our proposed Integrated Power Systems Roadmap was strongly validated by all three US Research Leaders in this field and by our Irish Industry presenters on the day, which we are very pleased about. Our unique position of being part of the wider Tyndall research community gives us the competitive edge to take a world-leading role in low-power, power management. The audacious goal of making bulky power management systems “invisible” is within reach, with many of the “lego blocks” and “know-how” almost there.

Summary
I must again acknowledge the contributions both on the day, and in the subsequent discussions of all the speakers and attendees at the Workshop. I believe this is a technology facing significant disruption in the coming years, from higher integration enabled by emerging solutions for higher switching frequencies, new circuit topologies, integrated magnetics and multi-level switched cap approaches.

A strong mandate was achieved to pursue the goal of miniaturization of power management systems, and specifically the control and regulation circuits required for next generation power electronics systems. True disruption however means threatening your existing product lines and your past investments. Breakthrough products disrupt competitors, sometimes whole markets. I am looking forward to the follow up discussions with our industry partners, on how best to fund and support this research, as innovation is not just about the idea or identifying the opportunity, it’s about the execution.

Donnacha O’Riordan