10:00am - 10:20am
Power Electronics: 1
A new topology for DC-DC converter with enhanced current multiplication
Electronic DC to DC converters divide into those that use transformers for impedance transformation and those that use resonant transfer between capacitors and inductors. In the latter, an inductance is used as intermediate storage that is energized at one voltage and deenergized at another. Some circuits combine the two approaches in one circuit such as opening a current carrying primary of a transformer. In general, the resonant transfer type such as the buck or buck boost circuits come in several basic variations and are inherently limited in their impedance transformation. For higher transformation ratios multiple stages are used which reduce the efficiency of the total system.
A circuit termed the “Meatgrinder” was developed in the 1980s to enhance the energy transfer between an energy storage inductor and an uncoupled inductive load by breaking the energy storage inductor into coupled sections interspersed with intermediate switches and a prescribed sequential switching scheme. The arrangement increased the total energy transfer efficiency with the number of sections. Significantly, a variation of this circuit termed the “Ringer” and developed principally for energizing electromagnetic guns which used only two sections; and furthermore, used the ringing of the source capacitor to perform the opening switch function and thereby eliminating the opening switch altogether. This Ringer circuit has now been modified to energize various PWM inverters for motor control and related. Applications with substantially increased impedance transformation and efficiency and with more relaxed switching requirement.
Converter switch requirements include both opening and closing switches. Opening switches are affected by the ever-present leakage inductance whose inductive kick must be suppressed with capacitors and resistors to protect the switch from over voltage surges. Alternative approaches are to open the switch only when the current is zero. This often goes counter to the function we would like the switch to perform. the circuit described in this paper combines the switch opening of the inductive circuit with both enhanced current multiplication and efficiency and substantially reduced switch requirement.
The circuit perceived application is predominantly in compact high efficiency converters requiring large impedance ratio transformations.
 B. W. Williams, "Basic DC-to-DC Converters," IEEE Trans. Power Electron., vol. 23, no. 1, pp. 387,401, Jan. 2008.
 Hua, G., Leu, C.S., Jiang, Y., et al.: ‘Novel zero-voltage-transition PWM converters’, IEEE Trans. Power Electron., 1994, 9, (2), pp. 213–219
 O. Zucker, J. Wyatt and K. Lindner, "The meat grinder: Theoretical and practical limitations," in IEEE Transactions on Magnetics, vol. 20, no. 2, pp. 391-394, March 1984, doi: 10.1109/TMAG.1984.1063078.
10:20am - 10:40am
Power Electronics: 2
In-depth Analysis of current-fed resonant Full-bridge Converter Application for high voltage DC Power Supply
1Bern University of Applied Sciences, Switzerland; 2Industrial X-Ray, Comet AG, Switzerland
In-depth analysis of current-fed resonant full-bridge converter application for high voltage DC power supply
In , current –fed full-bridge converter with resonant switching and constant on-time control for high–voltage (HV) application is introduced and demonstrated on 1kV/600W prototype converter with single rectifier output stage. While taking into account major parasitic components of HV transformer, theoretical analysis and demonstrator design in  does not cover the following aspects:
- influence of multistage multiplier and its component characteristics on converter performance and design for high output voltages e.g. 100kV
- impact of current-source implementation on the overall converter characteristics
- sensitivity of converter design to variation and uncertainties of HV transformer model parameters
- implementation aspects of current-fed converter.
This paper provides in-depth analysis of the aspects listed above. Using analytical and simulation approach, non-negligible effect of the multistage multiplier on the converter design is shown. Specifically, presence of significant HV diode junction capacitance leads to reduction of converter operating range. It is also shown that, in case of thin winding wire and multiple insulation mediums in the winding, analytical approach for HV-transformer capacitance calculation  does not provide estimates that are accurate enough. As an alternative, measurement-based approach, introduced and tested for two-winding transformers in , is pursued for parasitic parameter extraction and associated challenges are shown. Design trade-offs for current source implementation (in this case, buck converter) vs full-bridge converter on-time selection are shown. Specifically, it is demonstrated that time shift between constant on-time of full-bridge converter and buck converter significantly impacts current source output ripple and must be considered during the design. Finally, selected simulation results are compared with measurements on oil-insulated 150kV/300W high-voltage power supply prototype, discrepancies discussed and opportunities for future work identified.
 R.Y. Chen ; R.L. Lin ; T.J. Liang ; J.F. Chen ; K.C. Tseng, “Current-fed full-bridge boost converter with zero current switching for high voltage applications” in IEEE 2005 Industrial Applications Conference, 2005, pp.2000-2006
 L. Dalessandro ; F. Cavalcante ; J. Kolar, ”Self-Capacitance of High-Voltage Transformers”, in IEEE Transactions on Power Electronics, 2007, vol22, pp.2081-2092
 F. Blache ; J.P. Keradec ; B. Cogitore, “Stray capacitances of two winding transformers: equivalent circuit, measurements, calculation and lowering” in IEEE Industry Applications Society Annual Meeting, 1994, pp.1211-1217
10:40am - 11:00am
Power Electronics: 3
Study of a Medium Voltage AC/DC Testbed Employing Electrochemical Energy Storage and Power Electronic Regulation
1University of Texas at Arlington (UTA), Electrical Engineering Department, 416 Yates Street, Rm. 518, Arlington, TX 76019 USA; 2Naval Surface Warfare Center – Philadelphia Division, 5001 S Broad St, Philadelphia, PA 19112 USA
Electrical power system architectures that are more intellingent are needed to meet the ever changing power requirements in both civilian and defense applications, respectively. Countless migrogrid topologies are being designed, modeled, and experimentally studied everyday to identify and hopefully overcome the many challenges that arise when distributed power sources and loads are interconnected and controlled autonomously. The Pulsed Power and Energy Laboratory (PPEL) at the University of Texas at Arlington (UTA) has developed a real-world testbed for the purpose of investigating control topologies that can be used to reliably control the power electronic distribution system when transient loads are sourced such that power quality is maintained. The testbed, operating at power levels in excess of 300 kW, utilizes distributed AC and DC power sources and loads operating at 480 VAC, 4160 VAC, 1 kVDC, 6 kVDC, and 12 kVDC, respectively. The testbed is being extended utilizing a hardware in the loop (HIL) simulator. The presentation will discuss the design of the testbed, the test plan methodology, and some results collected so far.
11:00am - 11:20am
Power Electronics: 4
Analysis of Resonant Behavior of Voltage Multiplier
Spellman High Voltage Electronics Corp., United States of America
By “resonant” behavior, in broad terms, we mean influence of the multiplier on the current waveforms of the HV transformer. The multiplier is seen by the transformer as an impedance, usually, a capacitive one. For a linearized circuitry, the multiplier can be substituted by a lumped capacitance, Cadd, connected to the transformer secondary terminals. Such an approach was actively investigated and formalized in the early years of Cockroft-Walton type multiplier developments. It is being equally actively and eagerly reinvented now, many decades later.
Parasitic capacitances Cp’s of the multiplier include geometrical capacitances, which are, in general, linear, and nonlinear capacitances of the semiconductors. Early researchers and those who followed them were fully aware of the complex nature of the parasitics. Schemes, of mostly inductive compensation, were suggested to alleviate Cp’s negative impact, including improved voltage sharing, or smaller “compression”, between the stages.
Usually, Cp’s are most prominent at no-load or at a light load. They were mostly how treated this way in the past. It was shown that even at no-load the output voltage would be lower than the expected 2NV0, where V0 is the peak transformer output voltage, and N is the number of cascades (for a classic CW multiplier). However, even under load, Cp’s are important, especially if the transformer is fed by a square- or other waveform containing higher harmonics. Cp’s are an unalienable part of the converter.
An interplay between the multiplier construction capacitances, Cm, and its Cp’s results in complex current waveforms. It is shown both theoretically and experimentally that at a low duty cycle, the transformer windings can be subjected to a “backlash” from the multiplier Cp’s, which generates spurious currents in the primary winding. Another non-trivial consequence is that with low Cm/Cp ratio, the leakage inductance is compensated more effectively, and higher power can be made, at a price of a larger compression. Yet another peculiarity is that even with very high Cm/Cp ratio, low Cm values result in the current waveforms that are more resonant, under heavy load, with N exceeding certain number of stages. This report concentrates on a classical CW multiplier. Special attention is given to analyzing compression caused by parasitics at no-load.
11:20am - 11:40am
Power Electronics: 5
Compact Magnetron Power Supply for Industrial Heating Applications
University of Nottingham, United Kingdom
This paper proposes a high-power Magnetron power supply for industrial heating applications with a rated power in excess of 120 kW and an operating voltage over 20 kV. With the aid of silicon carbide MOSFETs, the switching frequency of the power supply can reach 100 kHz, enabling a physical footprint reduction in which is advantageous in some target heating applications, such as those undertaken offshore. The converter modelling analysis and the combined voltage and current control methodology are presented. A non-linear Magnetron type load emulator circuit for the simulation work is also introduced.
The magnetron is an RF vacuum tube capable of generating microwaves. In industrial heating applications, the use of microwaves is much more efficient than traditional approaches using fossil fuels. One such application where the footprint of the heating system is critical is in offshore waste material treatment, where the cost of physical space is very high. Recent advances in the development of wide bandgap semiconductors and magnetic materials enable the potential for operating at a higher switching frequency potentially leading to a size reduction in the overall microwave generation system.
Due to the non-linearity of the magnetron load, a resistor is not sufficient to represent its characteristic in simulation environments, and hence an emulator circuit has been constructed. This provides a better approximation to the voltage-current curve for different working regions of the magnetron.
The proposed converter structure is a single active bridge (SAB). An AC input is rectified to produce a DC bus which forms the source of an H-Bridge circuit. This is connected to a high voltage transformer-rectifier unit (HVTRU) via an inductor for current control. The output of the transformer is then rectified and filtered before being connected to the anode and cathode of the Magnetron. Further details on the topology will be given in the final paper.
The SAB can operate under both continuous and discontinuous current conduction modes (CCM and DCM), and hence the transfer functions should be derived separately. Also, due to the requirement of the magnetron load, a voltage controller needs to be used when increasing the voltage before the tube conducts current and a current controller is required once the tube is conducting. Since the voltage controller works in DCM for start-up and the current controller works in both CCM and DCM, 3 different control approaches are required for the full operation of the system. The controllers are tuned to ensure that the output power of the RF tube can be controlled with an acceptable response and speed, limiting the possibility of arcs and being able to respond quickly to limit input power into the magnetron if one occurs, extending lifetime.
In the final paper, full details of the system design will be presented. The control design challenges and solutions will be presented and simulation results used to validate the operation of the power supply.
11:40am - 12:00pm
Power Electronics: 6
Impulse Generator for Simulating Lightning-Induced Pulse Transients for Airborne Equipment Test
Chung-Ang University, Korea, Republic of (South Korea)
This paper deals with design and implementation of the test equipment capable of simulating various types of lightning-induced pulse transients required by an international standard which defines the test condition and procedures of airborne equipment. The impulse waveforms required by the standard are composed of dozens of types based on various criteria such as voltage and current level, shape, rising/falling time, and repetition rate. Therefore, existing commercialized test facilities use a method of charging a storage capacitor using high-voltage capacitor charging power supply (CCPS), and configuring PFN circuit with the capacitor to generate a required test waveform. However, according to various conditions as mentioned earlier, multiple PFN circuits are designed and the storage capacitors of each PFN have a wide capacitance range from several nF to hundreds of μF; so the CCPS is also manufactured in two or more units, and the method of selecting and replacing each CCPS and PFN according to the test condition is currently used. However, in this study, 11 PFN circuits for each waveform and a single high-voltage CCPS that can cover a wide range of load capacitances ranging from 1 nF to 150 μF with maximum ratings of 12 kV charging voltage, 20 kHz repetition rate, and 245 kW peak power were designed and developed. Since the CCPS should be able to charge the load capacitor with a relatively large capacitance in a short time while considering the precision of the charging voltage for the loads of several nF level, the proposed single CCPS is designed as a dual-converter configuration consisting of a high-power (HP) and low-power (LP) converter. Each converter is designed based on the modified LCC resonant converter for high-efficient operation, and is capable of being operated independently or in parallel depending on the type of test waveform. In addition, considering the required level and precision of charging voltage for each load, the HP converter is designed to automatically stop generating the gate signal when the charging voltage reaches approximately 7 kV. Finally, a single CCPS was developed in a volume of 29.3 L, with design considerations such as high voltage isolation and development of high voltage and high frequency transformers. Through the experiments using the developed CCPS and each PFN load, it was confirmed that all test waveforms defined in the standard could be generated.