Showing papers by "Dragan Maksimovic published in 2013"

Architectures and Control of Submodule Integrated DC–DC Converters for Photovoltaic Applications

Abstract: This paper describes photovoltaic (PV) module architectures with parallel-connected submodule-integrated dc-dc converters (subMICs) that improve efficiency of energy capture in the presence of partial shading or other mismatch conditions. The subMICs are bidirectional isolated dc-dc converters capable of injecting or subtracting currents to balance the module substring voltages. When no mismatches are present, the subMICs are simply shut down, resulting in zero insertion losses. It is shown that the objective of minimum subMIC power processing can be solved as a linear programming problem. A simple close-to-optimal distributed control approach is presented that allows autonomous subMIC control without the need for a central controller or any communication among the subMICs. Furthermore, the proposed control approach is well suited for an isolated-port architecture, which yields additional practical advantages including reduced subMIC power and voltage ratings. The architectures and the control approach are validated by simulations and experimental results using three bidirectional flyback subMICs attached to a standard 180-W, 72-cell PV module, yielding greater than 98% module-level power processing efficiency for a mismatch less than 25%.

Mitigation of Solar Irradiance Intermittency in Photovoltaic Power Systems With Integrated Electric-Vehicle Charging Functionality

Abstract: Mitigation of the variability in output power of renewable generators such as solar photovoltaic (PV) systems is a growing concern as these generators reach higher penetrations on electric grids. Furthermore, increased penetration of electric vehicle (EV) loads presents a challenge for distribution feeders. This paper presents a system where a bidirectional, highly efficient, dc-dc EV charger is placed between the high-voltage dc bus of a PV inverter and the EV battery. The system partially alleviates feeder overloading by providing fast charging for the EV battery from the PV system. In addition, the charger is capable of diverting fast changes in PV power output to the battery, thereby reducing the rate of change of inverter output power to a level below the ramp rate of existing grid resources. The paper addresses sizing of the charger and energy storage based on the PV system rating, the desired maximum ramp rate, and site solar irradiation characteristics, including geographic dispersion of PV arrays. Analysis suggests that small amounts of energy storage can accomplish large reductions in output power ramp rate. Experimental results are shown for a 10 kW, 98% efficient dc-dc charger based on bidirectional four-phase zero-voltage-switching converter.

Design and Control for High Efficiency in High Step-Down Dual Active Bridge Converters Operating at High Switching Frequency

Abstract: A control scheme is developed to maximize efficiency over a wide range of loads for a dual active bridge converter. A simple control circuit using only phase-shift modulation is proposed which considers both the converter conversion ratio and switching dead times in order to maintain high efficiency in the presence of varying loads. To demonstrate feasibility of the proposed control method, experimental results are presented for a 150-to-12 V, 120-W, 1-MHz prototype converter which has 97.4% peak efficiency and maintains greater than 90% efficiency over a load range between 20 and 120 W.

Active power control of photovoltaic power systems

Abstract: Historically, electric power system operators have seen photovoltaic (PV) power systems as potential sources of problems due to intermittency and lack of controllability. However, the flexibility of power electronic inverters allows PV to provide grid-friendly features including volt-VAR control, ramp-rate control, high-frequency power curtailment, and event ride-through. These technologies offer power quality improvements and enable wider penetrations of PV systems. Commercially available smart PV inverters can further provide frequency down-regulation by curtailing power, but they are unable to provide true frequency regulation through active power control (APC) because they are unable to increase power on command. The development of inverters capable of APC for primary and secondary frequency regulation without the need for energy storage has the potential to transform the way grid operators view high PV penetration levels. This paper provides a brief tutorial on APC of PV inverters, summarizes state of the art research in the area, and suggests future research directions.

Smart DC Power Management System Based on Software-Configurable Power Modules

Abstract: DC bus-based systems are envisioned as an enabling technology to integrate renewable energy sources, energy storage devices, and a variety of loads in a number of power management and distribution scenarios. The system design and integration challenges include accommodating heterogeneous components, the wide variability of operating conditions, and system stability issues arising from dynamical interactions between the components. This paper proposes a flexible, smart dc power management architecture based on identical digitally controlled bidirectional dc–dc modules that can be software configured to enable simple system design, exceptional system flexibility, and optimization of the use of available resources. Design of the reconfigurable digital control infrastructure of such versatile dc power system is discussed in detail from its system-level description to the low-level design of the digital compensators. Stability analysis of the dc bus voltage is also discussed, proving the robustness of the power architecture from a theoretical standpoint. The proposed approach is then demonstrated on an experimental dc power management system consisting of several 500 W, bidirectional dc–dc modules operating from a 24-V dc bus.

Maximizing lithium ion vehicle battery life through optimized partial charging

Abstract: The limited lifetime, high cost, and large size of current lithium ion batteries are some of the primary obstacles to wider adoption of electric vehicles and plug-in hybrid electric vehicles. Simulations presented in this paper predict that Li-ion battery life can be extended through intelligent charging, especially when predictions of next-day energy needs are used to charge the battery only as needed. As-needed charging minimizes battery degradation by minimizing time spent at high state-of-charge. Preliminary results presented here indicate that the battery of a vehicle used for daily commuting and short errands could see its useable life extended by up to 150 % over unoptimized charging.

Simple Digital Pulse Width Modulator Under 100 ps Resolution Using General-Purpose FPGAs

Abstract: This letter describes a very simple implementation of a digital pulse width modulator (DPWM) under 100 ps resolution in low-cost field-programmable gate arrays (FPGAs). The implementation is based on internal carry chains and logic resources which are present in most FPGA families. The proposed approach does not require manual routing or placement, consumes few hardware resources, and does not rely heavily on specialized phase-locked loop or clock management resources. A 50-MHz switching frequency DPWM with 60-ps resolution and a 1-MHz switching frequency DPWM with 90-ps resolution are experimentally demonstrated, with monotonicity and excellent linearity.

Inherent volt-second balancing of magnetic devices in zero-voltage switched power converters

Abstract: Small mismatches in inductor-applied volt-seconds may arise in power converters due to asymmetries in circuit parasitics or modulation waveforms. These small mismatches can have significant impact on circuit operation, including the saturation of magnetic components, loss of regulation, and decrease in converter efficiency. Various auxiliary circuits and control methods have been developed to prevent volt-second imbalances from being applied to magnetic components. In this work, an inherent feedback specific to Zero-Voltage Switched (ZVS) converters is examined which automatically compensates for volt-second mismatch. A closed-form linearized relation between volt-second mismatch and inductor current offset is derived. This relation is then verified through simulation and experimental results using two prototype circuits comprised of an inductively loaded full-bridge and a dual active bridge (DAB) converter.

Nonlinear control design for the photovoltaic isolated-port architecture with submodule integrated converters

Abstract: Differential power processing (DPP) isolated-port architecture with submodule integrated converters (subMICs) enables improved energy capture in photovoltaic (PV) power systems. This paper describes a control scheme for flyback subMICs operating in discontinuous conduction mode (DCM). The duty-cycle is driven in proportion to a voltage error, with no further processing, thus eliminating the need to measure or to control the current explicitly. The approach is well suited for very simple, analog PWM controller implementation, but the resulting control loops become nonlinear. The system stability is demonstrated using the Lyapunov stability theorem. Experimentally measured module-level efficiency is greater than 95% for a 72-cell PV module having 3 substrings with solar irradiation of 80%, 60% and 40%, respectively.

A cell-level photovoltaic model for high-granularity simulations

Abstract: This paper presents a cell-level simulation model of photovoltaic (PV) systems that is capable of reproducing partial-shading and other mismatch effects at high granularity. The nonlinear diode voltage-current characteristic is stored in a piecewise-linear model, and the stored values are employed to compute the Jacobian matrices very efficiently. As a result, the model can be solved with low computational effort, so that simulations at cell-level granularity and over long time spans can be carried out efficiently. The model is well suited for evaluations of energy capture, including the effects of partial shading or other mismatches, in large PV systems over long periods of time. Comparisons of simulation and experimental results are provided for three representative PV systems.

Balancing, filtering and/or controlling series-connected cells

Abstract: A balancing circuit for a plurality of series connected cells or substrings of cells is provided. In one implementation, the balancing circuit includes a plurality of primary ports; an isolated secondary port; and one or more DC-DC converters connected between the primary ports and the isolated secondary port. Each DC-DC converter includes at least one power switch. The DC-DC converters are configured to adjust a primary port current received at one or more of the plurality of primary ports based upon a difference between a voltage at the one of the primary ports and a reference voltage. Also provided are an electrical power system including such as balancing circuit and a method of balancing a plurality of electric cell substrings using such a balancing circuit.

A branch and bound algorithm for high-granularity PV simulations with power limited SubMICs

Abstract: Differential power processing (DPP) architectures employ low power rated dc-dc converters to mitigate, without insertion losses, mismatches in series strings of photovoltaic (PV) cells. This paper presents tools for simulation of PV systems based on the isolated-port submodule integrated converter (subMIC) DPP architecture. While the ideal operation with full power rated subMICs is relatively easy to solve using a combination of numerical and analytical methods, the solution with partial power rated (current limited) subMICs is more challenging. A heuristic branch and bound algorithm is proposed as an effective method to deal with nonlinearities introduced by current limited subMICs. The approach is well suited for high-granularity simulations of large PV systems over long periods of time, as illustrated by experimental and simulations results.

High frequency synchronous Buck converter using GaN-on-SiC HEMTs

Abstract: GaN-on-SiC technology has been widely used in radiofrequency (RF) power amplifiers to achieve high power density at very high frequencies due to the enhanced power handling capabilities provided by the SiC substrate and very high transition frequencies of GaN High Electron Mobility Transistors (HEMTs). Along with other well-known advantages of GaN, such as high breakdown voltage and high temperature operation, these characteristics make GaN-on-SiC technology suitable for power electronics applications. This paper explores the use of GaN-on-SiC depletion-mode HEMTs in switched-mode power converters, targeting high efficiencies at switching frequencies in the tens of Megahertz range. The fundamental limits of the technology are analyzed, and design principles and guidelines are given to exploit the capabilities of the devices. Results are presented for a 10 W, 20 V synchronous Buck converter prototype operating at 10 to 40 MHz. Measured efficiency peaks above 96% at 10 MHz and remains above 90% over a wide range of operating conditions. To demonstrate the applicability of the technology, the Buck converter switching at 20 MHz is used as a supply modulator to track a 3 MHz envelope signal with greater than 90% average efficiency.

RFPA supply modulator using wide-bandwidth linear amplifier with a GaN HEMT output stage

Abstract: This paper presents an envelope tracking supply modulator for a high efficiency radiofrequency power amplifier (RFPA). A wide bandwidth linear amplifier using a GaN high electron mobility transistor (HEMT) as a source follower output stage is described. A low bandwidth, high efficiency current controlled Buck converter is used to assist the linear amplifier and improve the system efficiency. Steps towards integration of the supply modulator and the RFPA are given through inclusion of the modulator output stage with an RFPA on a microwave monolithic integrated circuit (MMIC). Simulation and experimental results are presented for a 150 MHz bandwidth, 9 V to 27 V output voltage modulator supplying a 10 GHz RFPA.

Loss modeling and optimization for monolithic implementation of the three-level buck converter

Abstract: This paper presents parameter extraction based loss modeling to reduce the high-order design space for the three-level buck converter optimized for monolithic implementation. Loss models derived from simulation extracted parameters are presented to reduce model complexity while maintaining accurate loss predictions. The design space is reduced further by analysis of converter characteristics. An optimization approach is applied to select the inductor, the switching frequency, and the sizes of the power devices and the gate drive stages. The loss model and the optimization approach are validated for a 3.7-to-1.15 V, 4 MHz, 2 A converter through detailed circuit simulations in a 0.18 μm CMOS process.

State-machine realization of second-order sliding-mode control for synchronous buck DC-DC converters

Abstract: The paper presents a second-order sliding-mode (SOSM) control approach for synchronous buck dc-dc converters. The SOSM controller is implemented digitally, using a simple state machine, which does not require current sensing or an integral term. A hysteresis is introduced to control the switching frequency. The SOSM controller results in fast transient responses without overshoots, and is robust against parameter uncertainties. The proposed approach is verified by experimental results on a 1.25 V, 10 A prototype.

Comparison of reverse recovery behavior of silicon and wide bandgap diodes in high frequency power converters

Abstract: The nature of reverse recovery losses is examined in hard-switched and soft-switched converters, using silicon (Si), silicon carbide (SiC), or gallium nitride (GaN) devices. A loss model and experimental results with a prototype 150-to-400 V, 150 W, boost converter operated at switching frequencies between 500 kHz and 2 MHz are used to characterize and quantify losses related to diode reverse recovery. It is found that reverse-recovery related losses with Si diodes cannot be neglected even when the converter is soft switched, with zero-current switching of the diode and zero-voltage switching of the transistor. The switching losses with SiC or GaN diodes are substantially smaller in all cases considered, and can be reduced to negligible values when the converter is soft switching.

On/off control of a modular DC-DC converter based on active-clamp LLC modules

Abstract: This paper studies on/off control of a dc-dc system that contains multiple active-clamp LLC resonant modules operating in parallel. The architecture and the on/off control method yield high overall efficiency and fast transient responses. Two converter models are introduced for different purposes: (1) a high-frequency model to determine a gate timing sequence that enables fast on/off control, and (2) an averaged model to assist in the system control loop design. The models and the on/off control method are verified by experiments on a 1 MHz, 24 V-to-3.3 V, 5.5 W active-clamp LLC module prototype. The performance of PI and PID system controllers is analyzed and compared by simulations of a twenty-module system.

Active clamp LLC resonant converter for point-of-load applications

Abstract: This paper describes an active clamp LLC resonant converter intended for point-of-load (POL) power supplies. The active clamp LLC converter inherits advantages of the standard LLC resonant converter including zero voltage switching (ZVS) of primary side MOSFETs, zero current switching (ZCS) of synchronous rectifier MOSFETs and 50% duty cycle operation of all switching devices. The active clamp addresses the voltage oscillation across the rectifier devices caused by transformer secondary-side leakage inductances and MOSFET output capacitors by clamping voltage to twice the output voltage while reducing the output capacitor current ripple. The output voltage can be regulated by simple on/off control, and the converter is well suited for a multiple-parallel-module configuration. The converter module operation, modeling, and design approach are verified on two 1 MHz, 24 V-to-3.3 V, 6 W experimental prototypes.