IEEE Educational Videos on Power Electronics

Frequency Response Debug

Jorge Vega — Sat, 14 Feb 2026 00:00:00 +0000

This video explores the frequency response behavior of a switching power converter under varying line and load conditions. While most operating points produced clean, predictable responses—featuring a 20 dB/decade gain slope crossing 0 dB and approximately 65° of phase margin—the low-line, maximum-load corner yielded only noise. Investigation revealed that at this corner the controller’s compensation (COMP) pin had reached its 5 V compliance limit, indicating the control loop was fully saturated. In this state, the converter was effectively running open-loop because no additional duty cycle was available to respond to further load or line changes. The result is a loss of meaningful frequency response data and a clear demonstration of how controller headroom directly affects stability measurements. This experiment underscores the importance of monitoring COMP pin behavior when troubleshooting unexpected Bode plot results.

Stability and Compensation: Simple Average Model

Robert Bolanos, Jorge Vega — Sat, 14 Feb 2026 00:00:00 +0000

A pole-zero cancellation method is presented to compensate switching power converters. The method models the plant as a single pole model and the error amplifier with a zero and high frequency pole. The pole-zero cancellation matches the plant’s pole with an error amplifier zero and cancels the ESR zero with a high-frequency pole. This produces a clean 0 dB crossover at 20 dB/decade for stable operation. The approach is quick and predictable, requiring only basic calculations from known component values. It avoids complex modeling by treating all downstream circuitry as a single voltage-controlled current source. The method is well suited for early-stage design, quick prototypes, or educational purposes.

Introduction to AI in Power Electronics

Xinze Li — Fri, 20 Feb 2026 00:00:00 +0000

This video introduces the transformative role of Artificial Intelligence (AI) in power electronics design. This episode outlines the vision of AI as a co-designer, not a replacement, empowering engineers to navigate the complex, nonlinear design space of converters more efficiently and creatively.

Through the scope of designing a flyback converter, the episode contrasts traditional, iterative engineering workflows with a generative AI-driven paradigm enabled by tools like PE-GPT. Viewers witness how natural language prompts can generate complete converter designs, simulate circuit behavior, and guide intelligent component selection.

The episode also previews the full PELSTube series, covering meta-heuristic optimization, machine learning-based modeling, reinforcement learning for adaptive control, and physics-informed ML. Expert insights from leading researchers highlight cutting-edge applications of AI across magnetic design, power module layout, and design automation.

Core Concepts of AI in Power Electronics Design

Xinze Li — Fri, 20 Feb 2026 00:00:00 +0000

Continuing from the previous introductory video, this session introduces the fundamental concepts of AI and their applications in power electronics design. It sequentially covers key AI techniques—meta-heuristic algorithms, machine learning, and generative AI engineering—each contributing toward the vision of autonomous power electronics design, or PE Design 4.0. To reinforce these concepts, case studies on AI-augmented synchronous buck converter design are presented, offering practical insights and deepening understanding.

Stability of Wireless Power Receivers

Kerui Li — Tue, 06 Jan 2026 00:00:00 +0000

This video presents a study focused on the stability of wireless power receivers. The author identifies two instability factors and explains their underlying mechanisms. Building upon this understanding, the study then proposes practical solutions to mitigate these stability issues.

Distributed Predictive Cooperative Control of Smart Power Converters in Microgrids

Oluleke Babayomi — Tue, 06 Jan 2026 00:00:00 +0000

Microgrids with power converter-based distributed energy resources are modular electrical networks that make our power systems more sustainable, reliable, and resilient. However, traditional model predictive control of power converters suffers from parameter uncertainties and limited distributed energy resource coordination, while existing solutions inadequately address measurement noise and dynamic response limitations. This talk presents novel control strategies for smart power converters in microgrids to enhance sustainability, reliability, and resilience. Two main contributions are presented.

First, a model-free predictive control approach using a multifrequency extended state observer (ESO) framework that estimates sub-frequency harmonics of lumped disturbances, achieving 55% better measurement noise suppression than conventional ESOs without compromising disturbance rejection or parameter robustness.

Second, distributed cooperative secondary control with virtual synchronous generator inertia emulation and novel virtual capacitance algorithms for accurate reactive power sharing, delivering 60% improvement over state-of-the-art methods despite line parameter mismatches. Real-time hardware-in-the-loop tests and hardware experiments validate performance across islanded and grid-connected microgrid scenarios.

Multi-Level Flying Capacitor Buck Converters with Digital-Predictive Current-Mode Control

Giovanni Bonanno — Tue, 06 Jan 2026 00:00:00 +0000

This talk presents the main results of my PhD research on multi-level flying capacitor buck converters with digital predictive current-mode control, focusing on their application in modern automotive electronics. The increasing demand for compact, high-efficiency, and high-performance DC-DC converters in vehicles has driven the exploration of advanced converter topologies. Multi-level flying capacitor architectures offer significant advantages in terms of efficiency, dynamic response, volume, and converter footprint reduction. Unfortunately, their practical deployment is challenged by the need to maintain stable and balanced flying capacitor voltages.

To address this, I developed a unified analytical methodology for the closed-loop stability analysis of flying capacitor voltages under digital predictive current-mode control. This approach fills a gap in the literature, providing closed-form solutions for both three-level and generic n-level converters. The proposed methodology is more general and may be used for a wide range of digital controllers, enabling precise prediction of steady-state and dynamic behavior.

Additionally, I introduced two novel digital corrective controllers that enhance the speed of corrective actions and expand the available control bandwidth without increasing the switching frequency. These controllers, validated through extensive simulation and experimental testing, demonstrate robust self-balancing properties and improved dynamic performance. The experimental results, obtained using custom prototypes, confirm the theoretical predictions and highlight the practical feasibility of the proposed solutions.

In summary, this research delivers new analytical tools and control strategies for multi-level flying capacitor converters, advancing their reliability and performance in automotive and other demanding applications. Given the ongoing trend toward electrification and digitalization in transportation and industrial systems, these innovations have the potential to significantly influence the next generation of power electronic converters, enabling more efficient, compact, and reliable solutions for future smart mobility and energy management platforms.