3:15PM Using design by optimization for reducing the weight of a SiC switching cell [#678] Mylene Delhommais, Gnimdu Dadanema, Yvan Avenas, Jean-Luc Schanen, Francois Costa and Christian Vollaire, G2ELab, France; Satie, France; Ampere, France
This paper uses the design by optimization method to determine the best combination of passives, EMC filter, heatsink, taking into account the components constraints, in order to reduce the weight of a SiC switching cell, implemented in a simple buck converter. All elements and associated restrictions are described using analytical expressions, and the optimizer performs the reduction of the objective function (weight) respecting all constraints. The models and the optimization strategy are described, and the optimization results are presented and discussed, in comparison with a conventional Si switching cell.
Active Power Filters
Thursday, September 22, 2:00PM-3:40PM, Room: 101B, Chair: Luis Moran, Andy Knight
2:00PM Multilevel Nine-Leg Converter Universal Active Power Filter [#1174] Phelipe Leal Serafim Rodrigues, Cursino Brandao Jacobina and Mauricio Beltrao de Rossiter Correa, DEE UFCG, Brazil
This paper presents a multilevel universal active power filter (UAPF) topology with nine legs. The proposed system is a combination of two series and one shunt active power filters with a connection series-shunt-series. It is suitable to improve the harmonic distortion in the grid currents and a voltage compensation. The filter is composed of three three-phase converters resulting in nine legs sharing a single DC-link. The complete control system, including the PWM techniques of the proposed topology for different transformers turn ratios, is developed. Simulation and experimental results are shown for validation purposes.
2:25PM Central Control and Distributed Protection of the DSBC and DSCC Modular Multilevel Converters [#1539] Andre Hillers, Hao Tu and Juergen Biela, Laboratory for High Power Electronic Systems, Switzerland; North Carolina State University, United States
The benefits of modular multilevel converters come at a price: Their high level of complexity demands for an intricate design methodology and a sophisticated control structure. This is especially true for converters that need to be able to recover from short-circuits, grid faults, communication faults or module faults. In order to build a reliable system, it is thus important to understand the effect that the control structure has on the converter design. When protecting each module against overvoltages and overcurrents individually, the performance of the central controller is no longer immediately safety critical. This allows to communicate the switching times and measurements with a daisy-chain-like communication architecture that does not require point-to-point connections from the central controller to every module. The presented approach is verified by time-domain simulations and validated on a downscaled hardware prototype system.
2:50PM Mitigating the Effect of Series Capacitance Unbalance on the Voltage Reduction Capability of an Auxiliary CSI used as Switching Ripple Active Filter [#1545] Savvas Papadopoulos, Mohamed Rashed, Christian Klumpner and Pat Wheeler, University of Nottingham, United Kingdom
The use of series connected capacitors for high voltage applications has been proven to be beneficial for voltage stress reduction across power semiconductors. In a 3-phase grid any asymmetry in the value of the series capacitance may lead to significant variations in the voltage seen across the low voltage converter. This paper investigates the effects of an unbalanced set of series connected capacitors used to reduce the voltage stress across a three phase current source inverter (CSI) used as an active power filter and proposes a method to minimize the impact of unbalance on CSI voltage ratings. It is shown that through a proposed solution which adjusts the level of inverse sequence current component in the series capacitors, the reduced CSI voltage stress can be maintained for large capacitor unbalance and validated by simulation and experimental results.
3:15PM A New Control Method of Suppressing DC-Capacitor Voltage Ripples Caused by Third-Order Harmonic Compensation in Three-Phase Active Power Filters [#1375] Tomoyuki Mannen, Issei Fukasawa and Hideaki Fujita, Tokyo Institute of Technology, Japan
This paper proposes a new control method suitable for active power filters to reduce the dc capacitor voltage ripples associated with the third-order harmonic current compensation. The proposed method superimposes a negative-sequence fundamental current on the compensating current to cancel out the active power ripple caused by the third-order harmonic current. As a result, the proposed method has the capability to eliminate the dc capacitor voltage ripple at double the source frequency. Experimental results obtained by a 10-kW three-phase diode rectifier verify the validity of the proposed method. The proposed method exhibits a good suppression performance of the dc capacitor voltage ripple at double the source frequency.
2:00PM Design of Dual Purpose No Voltage Combined Windings for Bearingless Motors [#1524] Eric Severson, Robert Nilssen, Tore Undeland and Ned Mohan, University of Minnesota, United States; NTNU, Norway
A winding design approach is proposed to create a single motor winding which is able to produce both radial force and torque. This approach can be used to design new bearingless motors as well as to transform conventional motor designs into bearingless motors by simply modifying the winding end- connections. The resulting winding has two sets of terminal connections: one for torque and one for suspension. The suspension terminals experience no motional-EMF when the rotor is centered, which means that the suspension drive can have a low voltage rating and that rotor vibrations can be passively dampened by simply short-circuiting the suspension terminals. Bearingless motors that use these so-called ``dual purpose no voltage windings" can be designed to have higher torque density and lower losses associated with the magnetic suspension operation than traditional bearingless motors which utilize separate torque and suspension windings. It will be shown that many popular winding designs, including fractional- slot and concentrated windings, can be realized as dual purpose no voltage windings. The proposed approach applies to traditional p +- 1 pole-pair bearingless motors as well as the bearingless consequent-pole and ac homopolar motors. Fractional-slot motor winding theory is used to derive the new winding requirements and a generalized design procedure; example designs are explored through finite element analysis and experimental results from a hardware prototype of a bearingless ac homopolar motor.
2:25PM Synchronous Generator Field Excitation Via Capacitive Coupling Through a Journal Bearing [#629] Jiejian Dai Dai, Skyler Hagen, Daniel Ludois and Ian Brown, University of Wisconsin -Madison, United States; Illinois Institute of Technology, United States
Wound field synchronous generators (WFSG) are the standard for back-up and utility scale power generation. Rotor field current and prime mover speed are the only control parameters required to regulate power conversion in a generator application. Maintenance costs may be minimized by adopting non- contact or brushless technologies to replace sliding slip ring connections. This paper presents a brushless excitation approach using ceramic insulated sleeve or journal bearings with oil lubrication to form capacitively coupled slip rings, in contrast to more traditional inductive brushless exciters and rotary transformers. This capacitive power transfer CPT approach exhibits advantages including low weight, low volume and has a relatively simple construction using off-the-self components. Analysis, design and prototype construction of the CPT system are presented. Experimental results demonstrate that 1.7nF of capacitive couplingtransfers 340W to the rotor field winding of a 10kW 208V WFSG. Voltage regulation of a WFSG is demonstrated during steady state and 1 per unit load step changes yielding a NEMA-MG1 class G2 rating.
2:50PM Development of Stator-Magnetless Linear Synchronous Motor for Sensorless Control [#314] Makino Shogo, Kakihara Masanobu, Takase Yoshiyasu, Takaki Mamoru, Shikayama Toru, Ohto Motomichi, Higuchi Tsuyoshi and Abe Takashi, Yaskawa Electric Corporation, Japan; Nagasaki University, Japan
Sensorless control technique is desired in applications of a long stroke linear synchronous motor (LSM) without using a linear scale. This paper discusses the development of a stator-magnetless LSM (no magnet mounted on the stator of LSM) for sensorless control technique that includes a high speed position estimation algorithm based on magnetic hysteresis phenomenon. This paper presents the new structure of a flux-switching LSM to achieve a high saliency ratio by designing a cutout which causes magnetic saturation in the armature core. The effect of sub-teeth to reduce cogging thrust is also presented. Furthermore, this paper also reports the analytical and experimental characteristics of inductance, thrust, cogging thrust, and sensorless control drive using a prototype.
3:15PM Ultralightweight Motor Design Using Electromagnetic Resonance Coupling [#201] Kazuto Sakai and Yuta Sugasawa, Toyo University, Japan
To produce an ultra-lightweight motor, we propose a technology that converts electrical energy between the stator and rotor using magnetic coupling. Magnetic coupling causes electromagnetic resonance between the respective multiphase windings of the stator and rotor. Electromagnetic resonance coupling technology allows electric motors to convert energy without magnetic cores, significantly reducing their overall weight. In this study, we proposed a motor design based on magnetic resonance coupling (MRC) and described its operating principles and characteristics. A model of the proposed MRC motor was analyzed using magnetic analysis to verify its rotational energy conversion and understand its fundamental characteristics. Our results confirmed that the proposed MRC motor without a magnetic core is capable of converting electromagnetic energy between the stator and rotor and producing sufficient operating torque.
Magnetic Gears
Thursday, September 22, 2:00PM-3:40PM, Room: 102B, Chair: Siavash Pakdelian, Kan Akatsu
2:00PM A Novel Reluctance Magnetic Gear for High Speed Motor [#1667] Kohei Aiso and Kan Akatsu, Shibaura institute of technology, Japan; Shibaura Institute of Technology, Japan
In the industrial field, the drive system integrated the high speed motor and the gear is required to effectively utilize the limited space. However, mechanical gears often require their lubrication and cooling, they generate whilst noise and vibration. To overcome these problems, the magnetic gears have been expected. Magnetic gears achieve the system with maintenance-free, low noise and low vibration characteristics by contact-less power transformation. However, conventional magnetic gears are not suitable system for high speed motors because the mechanical strength is weak due to permanent magnets of the high speed rotor. Adding that, the magnet eddy current loss in the high speed rotor seriously decreases the gear efficiency in high speed drive. This paper presents a novel reluctance magnetic gear for the super high speed motor. The high speed rotor of the proposed magnetic gear is constructed by only iron core. Therefore, the structure is very simple and robust and it is possible to rotate in the high speed region. Moreover, the magnetic gear achieves high efficiency in the high speed region because the magnet eddy current loss in the high speed rotor is not generated because of having no magnets. The downsizing of the system and high efficiency power transfer are realized by applying the reluctance magnetic gear.
2:25PM Analysis of a Magnetically Geared Lead Screw [#1277] Mojtaba Bahrami Kouhshahi and Jonathan Bird, Portland State University, United States
Linear magnetic gearboxes (LMG) and magnetic lead screws (MLS) have been shown to be capable of operating at significantly higher volumetric force densities than traditional electromagnetic linear actuators (ELA). However in both such devices the linear translator must be made of magnet material and therefore if the stroke length of the translator is long the cost of the MLS and LMG will become prohibitively high. In this paper a magnetically geared lead screw (MGLS) is investigated and its performance capability is compared with the LMG and MLS. The advantage of the MGLS is that the translator does not contain magnets.
2:50PM Design Comparison of NdFeB and Ferrite Radial Flux Magnetic Gears [#1358] Matthew Johnson, Matthew C. Gardner and Hamid A. Toliyat, Texas A and M University, United States
Magnetic gears promise the benefits of mechanical gears with added advantages from contactless power transfer. Although most literature focuses on minimizing the size of magnetic gears, their material costs must also be reduced to achieve economic feasibility. This work compares the active material costs of NdFeB and ferrite radial flux magnetic gears with surface permanent magnets through a parametric 2D and 3D finite element analysis (FEA) study. Differences in optimal design trends such as pole count and magnet thicknesses are illustrated for the two materials. The results demonstrate that, for most historical price combination scenarios, NdFeB gear designs are capable of achieving lower active material costs than ferrite gear designs, and they are always capable of achieving much higher torque densities. Based on the selected design constraints, relative to a nominal ferrite cost of $10/kg, NdFeB must cost more than $90/kg before ferrite is cost competitive. Additionally, contour plots are provided to show the impact of material price rate variation on the cost break points.
3:15PM Power Transferring of Magnetic-Geared Permanent Magnet Machines [#364] Leilei Wu, Ronghai Qu, Dawei Li and Yuting Gao, Huazhong University of Science and Technology, China
Due to the high torque density and sharing similar operating principles with vernier machines, the magnetic-geared permanent magnet machines have been gaining more attention recently. This paper focuses on power transferring feature of a magnetic geared machine among the stator, modulation layer and rotor by analytical methods. Analytical equations of the back electromotive force (back EMF), torque and power are derived. It is found that the power expression is a function of gear ratio between modulation layer and rotor. In addition, the influence of the parameters on the back EMF and torque are also discussed. Finally, the analytical results are verified by the finite-element analysis (FEA), and the good agreements have been achieved.
2:00PM Robust Control of an Open-Ended Induction Motor Drive With a Floating Capacitor Bridge over a Wide Speed Range [#1629] Michele Mengoni, Albino Amerise, Luca Zarri, Angelo Tani, Giovanni Serra and Domenico Casadei, University of Bologna, Italy
An electric drive for high-speed applications is analyzed in this paper. The drive consists of a dual two-level inverter with a floating bridge, fed by a single voltage source, and a 3-phase induction motor with open-ended stator windings. The floating bridge compensates the reactive power of the motor, so that the main inverter operates at unity power factor and fully exploits its current capability. The constant power speed range of the motor can be significantly extended depending on the DC-link voltage of the floating inverter. The details of the control system are examined and the feasibility of an electric drive is experimentally assessed.
2:25PM High speed operation of permanent magnet machine position sensorless drive using discretized EMF estimator [#596] Shih-Chin Yang and Guan-Ren Chen, National Taiwan University, Taiwan
This paper proposes a permanent magnet (PM) machine position sensorless drive for high speed (>10-krpm) applications. A fully discretized back electromotive force (EMF) estimator is designed to overcome several high speed rotor position estimation issues, including the voltage error due to pulse-width- modulation (PWM), dq inductance cross-coupling effect and the deviation between the estimated EMF and actual EMF resulting from the digital implementation. To overcome these problems, an observer-based EMF estimator is proposed with two implementation considerations. They are i) a discretized dq current observer to minimize the inductance cross-coupling effect on the estimated EMF voltages, and ii) a PWM model considering the EMF cross-coupling effect to obtain actual EMF voltages. A high speed PM machine is experimentally tested to verify all proposed sensorless drive methods.
2:50PM DC Voltage Regulated PWM Inverter for High-Speed Electrical Drives [#1507] Vito Giuseppe Monopoli, Maria Concetta Poliseno, Maria Chiara Stomati and Francesco Cupertino, Politecnico di Bari, Italy; GE Avio srl, Italy
This paper presents a study on two converter topologies suitable for driving high-speed three-phase electrical machines. In such applications the use of standard inverters with Si-based IGBTs may lead to high switching losses due to the elevated fundamental frequencies of high-speed machines. The use of wide band-gap devices allows reducing losses at high switching frequencies but problems related to EMI and accelerated motor insulation aging may occur. In this work a DC-DC converter is introduced to limit the inverter switching frequency at motor terminals without reducing the sampling frequency of the motor control scheme. This scheme allows to reduce EMI phenomena, and insulation stress on the motor with positive impact also on converter total losses at very high fundamental frequencies. Experimental results are also presented.
3:15PM Variable Time Step Control with Synchronous PWM in Low Frequency Modulation Index for AC Machine Drive [#881] Sungho Jung, Jaeyong Park, Euihoon Chung and Jung-Ik Ha, Seoul National University, Korea (South)
This paper proposes a variable time step controller with synchronous PWM. Since a frequency modulation index is low in an ultra-high speed or high power drive system, the modulated voltage by a fixed time step control includes large harmonic components. The optimal PWM method is one of the solution to minimize the harmonic components, but it needs many offline calculations and its structure is complicated. The proposed variable time step control has small calculation burden and it does not need offline calculations. Furthermore, the harmonic components of the modulated voltage is small because the frequency modulation index is kept as integer value in all speed conditions. Also, the transition strategy between synchronous PWM methods is proposed and it is verified with simulation and experimental results.
Power Assemblies
Thursday, September 22, 2:00PM-3:40PM, Room: 102A, Chair: Jelena Popovic, Yvan Avenas
2:00PM Implementation and Performance of a Current Sensor for Laminated Bus Bar [#721] Yoshikazu Kuwabara, Keiji Wada, Jean-Michel Guichon, Jean-Luc Schanen and James Roudet, Tokyo Metropolitan University, Japan; G2Elab, Univ. Grenoble Alps, France
This paper presents a current sensor embedded in a laminated bus bar. The proposed sensor is based on the Rogowski coil principle and it can detect the magnetic flux corresponding to a derivative waveform of the current within the bus bars. The frequency characteristics of the current sensor including the parasitic parameters depending on the circuit geometry are discussed. After the design of the current sensor using a PEEC-based software, the proposed sensor was implemented using four- layers and three PCBs. It is confirmed that the verification of the proposed current sensor embedded in the laminated bus bar can be performed by the measurement of turn- on and -off switching waveforms using a buck converter circuit, rated at 50 V and 4 A.
2:25PM Busbar Design for SiC-Based H-Bridge PEBB using 1.7 kV, 400 A SiC MOSFETs Operating at 100 kHz [#1640] Niloofar Rashidi Mehrabadi, Igor Cvetkovic, Jun Wang, Rolando Burgos and Dushan Boroyevich, CPES- Virginia Tech, United States
This paper presents a systematic study of the busbar design and optimization for SiC-based H-bridge power electronics building block (PEBB) used in high-frequency and high-power applications. Step-by-step guidelines are presented in which the design considerations and analysis are given. This paper presents a double-sided busbar concept to create a compact PEBB design with improved thermal and switching performance, which result from having double-side cooling and symmetric minimized current commutation loop inductances, respectively. The proposed concept is verified experimentally by evaluating the high-speed switching performance of the PEBB up to 400 A.
2:50PM Ultra-low Inductance Design for a GaN HEMT Based 3L-ANPC Inverter [#548] Emre Gurpinar, Francesco Iannuzzo, Yongheng Yang, Alberto Castellazzi and Frede Blaabjerg, University of Nottingham, United Kingdom; Aalborg University, Denmark
In this paper, an ultra-low inductance power cell design for a 3L-ANPC based on 650 V GaN HEMT devices is presented. The 3L-ANPC topology with GaN HEMT devices and the selected modulation scheme suitable for wide-bandgap (WBG) devices are presented. The commutation loops, which are the main contributors to voltage overshoots and increase of switching losses, are discussed. The ultra-low inductance power cell design based on a four layer PCB with the aim to maximise the switching performance of GaN HEMTs is explained. Gate driver design for GaN HEMT devices is presented. Common-mode behaviours based on SPICE model of the converter is analysed. Experimental results on the designed 3L-ANPC with the output power of up to 1 kW are presented, which verifies the performance of the proposed design in terms of ultra-low inductance.
3:15PM Layout Study of Contactless Magnetoresistor Current Sensor for High Frequency Converters [#1374] Mehrdad Biglarbegian, Shahriar Jalal Nibir, Hamidreza Jafarian, Johan Enslin and Babak Parkhideh, University of North Carolina at Charlotte, United States; Energy Production and Infrastructure Center, United States
In this paper, we present a new technique to unify and intensify the magnetic fields that results in higher performance of Anisotropic Magneto- Resistive (AMR) current sensors and consequently develop a closed loop controller for 100W synchronous GaN buck converter at 1 MHz. The closed loop operation of high switching frequency converters at high power has always been a big challenge due to lack of access to current information. The proposed method that also intensifies the magnetic fields through the sensor, significantly improves the bandwidth limits, and reduces electromagnetic interference (EMI) on AMR sensors, making them applicable for high switching frequency and high current power electronics converters. After verifying uniform distribution concept through simulation, we also implemented a prototype of AMR current sensors onto Printed Circuit Board (PCB) for verification of the concept at high frequency converter. We then present the design procedure and associated challenges of an integrated analogue peak current controller for creating the closed loop operation of a GaN buck converter at high switching frequency.