This paper presents a hypothetical Turing test in power electronics, leveraging structured computer vision as a step towards domain-specific artificial general intelligence (AGI). To illustrate the key principles of such a power electronics Turing test, we developed PowerVision, a computer vision framework designed to teach machines to understand schematic drawings. PowerVision comprises four key components: 1) ComponentNet: an image database for component recognition; 2) CircuitNet: an image database for schematic recognition; 3) NetlistMaker: a schematic recognition tool that converts human-readable schematics into netlists for SPICE simulations; and 4) NetlistClassifier: a circuit classification tool that can categorize different power electronics circuits based on machine-generated netlists. The PowerVision platform can facilitate the learning of power electronics fundamental principles by large-scale AGI models through human-accessible information including texts, schematics, computer simulations, and experimental results, ultimately enabling machines to comprehend power electronics.

Circuit models for multiphase coupled inductors are summarized, compared, and unified. Multiwinding magnetic structures are classified into parallel-coupled structures and series-coupled structures. For parallel-coupled structures used for multiphase inductors, the relationships between a) inductance-matrix models, b) extended cantilever models, c) magnetic-circuit models, d) multiwinding transformer models, e) gyrator-capacitor models, and f) inductance-dual models are examined and discussed. These models represent identical physical relationships in the multiphase coupled inductors, but emphasize different physical aspects and offer distinct design insights. The circuit duality between the series-coupled structure and the parallel-coupled structure is explored. Design equations for interleaved multiphase buck converters based on these models are streamlined and summarized, and a simplified equation showing the relationships between current ripple with and without coupling is presented. The models and design equations are verified through theoretical derivation, SPICE simulation, and experimental measurements.

The future electric grid is supported by a vast number of smart inverters interfacing with distributed energy resources at the edge. These inverters’ dynamics are typically characterized as impedances, which are crucial for ensuring grid stability and resiliency. However, the physical implementation of these inverters may change significantly from inverters to inverters and may be kept confidential. Existing analytical impedance models require a complete and precise understanding of system parameters. They can hardly capture the complete electrical behaviors when the inverters are performing complex functions. Online impedance measurements for many inverters across multiple operating points are not scalable. To address these issues, we present InvNet, a machine learning framework to systematically evaluate the effectiveness of data-driven methods for modeling inverter impedance patterns across a wide operation range, even with limited impedance data. Leveraging transfer learning, the InvNet can extrapolate from physics-based models to real-world ones and from one inverter to another with very limited data. This framework demonstrates machine  learning as a powerful tool for modeling and analyzing black-box characteristics of grid-tied inverter systems that cannot be accurately described by traditional analytical methods, such as inverters under model predictive control. Comprehensive evaluations were conducted to verify the effectiveness of the InvNet in various scenarios. All data and models were open-sourced.

This paper investigates the modeling, analysis, and design methods for balancing flying capacitor multilevel (FCML) converters using coupled inductors. Coupled inductors introduce multi-resonant dynamics to multiphase FCML converters. It synergizes with FCML converters by reducing inductor current ripple, reducing switch stress, and providing a strong and fast voltage and current balancing mechanism. Coupled inductor balancing is shown to be robust against many types of disturbances. A piece-wise linear (PWL) model and an energy-balance-based model are developed to show that coupled inductor balancing is effective in intrinsically unbalanced converters. Analytical and numerical methods are used to evaluate the effect of initial conditions, circuit parameters, and losses on the balancing behavior, and to show how coupled inductors in many cases balance the flying capacitors. The models and key principles are verified with simulations and experimental results.

This paper presents the MagNet-AI platform as an online platform to demonstrate the “Neural Network as Datasheet” concept for B–H loop modeling and material recommendation of power magnetics across wide operation range. Instead of directly presenting the measured characteristics of magnetic core materials as time sequences, we employ a neural network to capture the B–H loop mapping relationships of magnetic materials under different excitation waveforms at different temperatures and dc-bias. Both LSTM and Transformer based neural network models are developed, verified, and compared. The training and inference process of the neural network are fully automated to minimize the impact of human error. The neural network can be used to rapidly predict B–H loops under different operating conditions, compare materials, and recommend materials for design. Neural networks are also effective in compressing the information contained in the raw database, and enable rapid material evaluation and comparison.

This paper presents a multistack switched-capacitor point-of-load (MSC-PoL) voltage regulation module (VRM) with coupled magnetics for ultrahigh-current chiplet systems. In the MSC-PoL architecture, the stacked switched-capacitor cells split the high input voltage into several intermediate voltage rails, which are loaded with the switched-inductor cells to achieve soft charging and voltage regulation. Automatic capacitor voltage balancing and inductor current sharing are realized during the soft charging process. Many inductors of the switched-inductor cells are coupled into one and operated in interleaving to reduce the inductor current ripple and boost the transient speed. A 48-to-1-V/450-A VRM containing two MSC-PoL modules is built and tested, leveraging high voltage GaN devices for the front-end and high current Silicon devices for the back-end. Two ladder-structured coupled inductor designs are developed and compared, one of which installs a leakage magnetic plate to adjust the leakage inductance for lower current ripple. Featuring 3D stacked packaging, the entire power stage, gate drivers, and bootstrap circuits of one MSC-PoL module are enclosed into a 1/16-brick/0.31-in3/6-mm-thick package. The peak and the full-load efficiencies as well as the full-load power density (including both gate loss and size) of the MSC-PoL prototype with and without using the leakage plate are 91.7% and 89.5%, 85.8% and 85.6%, and 621 W/in3 and 724 W/in3, respectively. The 6-mm-thick MSC-PoL converter can be embedded into the chiplet or CPU socket, enabling power-supply-in-package (PwrSiP) for extreme efficiency, density, and control bandwidth.

This paper applies machine learning to modeling power magnetics. We first introduce an open-source database - MagNet - which hosts a large amount of experimentally measured excitation data for many materials across a variety of operating conditions. The processes for data acquisition and data quality control are explained. We then demonstrate a few neural network-based power magnetics modeling tools for modeling core losses and B–H loops. Machine learning allows the many factors that may influence the magnetic characteristics being modeled in a unified framework, and provides insights to quantify the complexity of magnetic characteristics and reduce the size of the measurement data required to build a precise model. Neural network models are found effective in compressing the measurement data and predicting the behaviors of magnetic materials such as the core loss and the B–H loop. The behaviors of a typical power magnetic material (e.g., TDK N87) across a wide range of operating conditions (e.g., temperature, waveform, dc-bias) can be well described by a small-scale neural network (200 KB) which is 10,000 times smaller than the raw measured time-series data (2 GB), paving the way toward using neural networks as an interactive datasheet to assist magnetics design.

This paper motivates the development of sophisti- cated data-driven models for power magnetic material charac- teristics. Core losses and hysteresis loops are critical information in the design process of power magnetics, yet the physics behind them is not fully understood. Both losses and hysteresis loops change for each magnetic material, are highly nonlinear, and depend heavily on the electrical operating conditions (e.g., wave- form, frequency, amplitude, dc bias), the mechanical properties (e.g. pressure, vibration), as well as temperature and geometry of the magnetic components. Understanding the complexity of these factors is important for the development of accurate models and their applicability and limitations. Existing studies on power magnetics are usually developed based on a small amount of data and do not reveal the full magnetic behavior across a wide range of operating conditions. In this paper, based on a recently developed large-scale open-source database – MagNet – the core losses and hysteresis loops of Mn-Zn ferrites are analyzed over a wide range of amplitudes, frequencies, waveform shapes, dc bias levels, and temperatures, to quantify the complexity of modeling magnetic core losses, amplitude permeability, and hysteresis loops and provide guidelines for modeling power magnetics with data- driven methods.

This paper presents a 48 V–1 V merged-two-stage hybrid-switched-capacitor converter with a Linear Extendable Group Operated Point-of-Load (LEGO-PoL) architecture for ultra-high-current microprocessors, featuring 3-D stacked packaging and coupled inductors for miniaturized size and vertical power delivery. The architecture is highly modular and scalable. The switched capacitor circuits are connected in series on the input side to split the high input voltage into multiple stacked voltage domains. The multiphase buck circuits are connected in parallel to distribute the high output current into multiple parallel current paths. It leverages the advantages of switched capacitor circuits and multiphase buck circuits to achieve soft charging, current sharing, and voltage balancing. The inductors of the multiphase buck converters are used as current sources to soft-charge and soft-switch the switched-capacitor circuits, and the switched-capacitor circuits are utilized to ensure current sharing among the multiphase buck circuits. A 780 A vertical stacked CPU voltage regulator with a peak efficiency of 91.1% and a full load efficiency of 79.2% at an output voltage of 1 V with liquid cooling is built and tested. This is the first demonstration of a 48 V–1 V CPU voltage regulator to achieve over 1 A/mm2 current density and the first to achieve 1,000 W/in3 power density. It regulates output voltage between 0.8 V and 1.5 V through the entire 780 A current range.

Flying capacitor voltage balancing is critical for the performance of flying capacitor multilevel (FCML) converters. This paper investigates the intrinsic capacitor voltage balancing of multiphase FCML converters with coupled inductors. It is shown that the coupled inductor provides flying capacitor voltage balancing that minimizes steady-state imbalances due to periodic disturbances compared to converters with uncoupled inductors. A dynamic model of natural balancing of the converter is derived and used to estimate the time required for the flying capacitors to settle from an initial imbalance. The theoretical predictions are verified with analytical derivations, SPICE simulations, and experimental results.

This letter investigates the ripple reduction mechanisms of pulsed-width-modulated (PWM) coupled inductors based on the principles of multiphase interleaving. By adopting circuit duality and superposition analysis, the ripple reduction ratio of PWM coupled inductors is, for the first time, presented as a function of the ripple reduction ratio of multiphase interleaving and the coupling coefficient of the coupled inductor. The analysis simplifies the assumptions used in previous study, and results in concise and intuitive description of the ripple reduction mechanisms of coupled inductors. The results provide useful guidelines for designing interleaved and coupled PWM converters.