Dongsheng Zhao

Bio: Dongsheng Zhao is an academic researcher from Delft University of Technology. The author has contributed to research in topics: EMI & Active filter. The author has an hindex of 10, co-authored 32 publications receiving 310 citations. Previous affiliations of Dongsheng Zhao include Eindhoven University of Technology.

Characterization of a Wideband Digitizer for Power Measurements up to 1 MHz

Abstract: A two-channel high-speed digitizer has been extensively characterized in the frequency range of 50 Hz-1 MHz. Measurements involved alternating-current (ac) flatness, phase, linearity, input impedance, and the effects of direct-current offsets, temperature, and an internal self-calibration routine. A digital antialiasing filter inside the digitizer negatively affects ac flatness in the low-frequency region. An inverse compensation filter has been designed and applied, which improves frequency response with a factor of 25-60 and makes it flat within 25 μV/V up to 100 kHz and within 100 μV/V up to 1 MHz. The phase difference between the two channels can be modeled by a constant time delay between the two channels, which for the 2- Vpp range equals (250 ± 30) ps. The overall results of the characterization indicate that the digitizer can be applied in wideband power measurements under practical circumstances with (k = 1) uncertainty contributions of not more than 70 and 400 μW/VA at 10 kHz and 1 MHz, respectively. This uncertainty excludes loading effects, which are significant at frequencies above 100 kHz. At low frequencies, up to three times lower uncertainty are achieved when the digitizer is calibrated at the signal level and temperature at which it is subsequently used.

A Multistep Approach for Accurate Permittivity Measurements of Liquids Using a Transmission Line Method

Abstract: High-precision permittivity measurements of liquids are essential for specifying the specific absorption rate (SAR) of radio-frequency power absorbed by the human tissue. At Van Swinden Laboratorium (VSL), a commercial unit is used for these measurements, which uses a combination of transmission and reflection methods. The permittivity of unknown liquids is derived from the measured different scattering parameters of a transverse electromagnetic cell when the cell is empty and when it is filled with liquid. The flat slab structure of the cell makes it convenient for liquid filling and cleaning. The method is an efficient complementary approach for permittivity measurements compared with more complex setups in other national metrology institutes, but for reaching a low uncertainty, some inherent disadvantages need to be overcome. In this paper, a multistep approach is proposed using a combination of curve fitting, nonlinear progressive technique, and improved transmission line models in order to improve the uncertainty of the permittivity measurements. This approach is applicable even without a priori knowledge of the liquid properties. Using this approach, the useful frequency range of the commercial unit is successfully extended from 200 MHz-3 GHz to at least 100 MHz-4 GHz, and the accuracy is significantly improved to the level where it becomes comparable to more primary methods. The results obtained with this new method agree very well with the reference data from the literature.

Principal Component Compression Method for Covariance Matrices Used for Uncertainty Propagation

Abstract: We investigate a principal component analysis approach for compressing the covariance matrices derived from real-time and sampling oscilloscope measurements. The objective of reducing the data storage requirements to scale proportional to the trace length ${n}$ rather than ${n}^{2}$ is achieved, making the approach practical for representing results and uncertainties in either the time or frequency domain. Simulation results indicate that the covariance matrices can be represented in a compact form with negligible error. Mathematical manipulation of the compressed matrix can be achieved without the need to reconstruct the full covariance matrix. We have demonstrated compression of data sets containing up to 10 000 complex frequency components.

A new behavioural model for performance evaluation of common mode chokes

Abstract: A behavioural model for performance evaluation of a common mode choke (CMC) is proposed: the topology and methods used to model the different properties of the CMC are presented. It combines an accurate model of the complex permeability, of the saturation, of leakage inductances and of parasitic capacitances. Inputs needed from the designer are measured values of currents, voltages and impedance of the actual system, and parameters available in literature and/or from manufacturers for the modelled CMC. The behavioural model is essential in simulation of EMI effects.

Investigation of Hybrid EMI Filters for Common-Mode EMI Suppression in a Motor Drive System

Abstract: This paper begins with an analysis of the common-mode (CM) noise in a motor drive system. Based on the developed CM noise model, two cancellation techniques, CM noise voltage cancellation and CM noise current cancellation, are discussed. The constraints and impedance requirements for these two cancellation methods are investigated. An active filter with a feedforward current cancellation technique is proposed, implemented, and tested, and techniques to improve the performance of active filters are explored. It is found that due to the limitations of speed, power loss, and gain bandwidth of active filters, active electromagnetic interference (EMI) filters are not good at suppressing high di/dt or high amplitude noise current. Hybrid filters that include a passive filter and an active filter are proposed to overcome the shortcomings of active filters. Hybrid EMI filters are investigated based on the impedance requirements and frequency responses between the passive and active filters. The experiments show that the proposed active filter can greatly reduce noise by up to 50 dB at low frequencies (LFs), and therefore, the corner frequency of the passive filter can be increased considerably; as a result, the CM inductance of the passive filter is greatly reduced. The power loss of the proposed active EMI filter can be well-controlled in the experiments.

Multilevel inverters for low-power application

Abstract: Multilevel inverters are an attractive solution in the medium-voltage and high-power applications. However in the low-power range also it can be a better solution compared to two-level inverters, if MOSFETs are used as devices switching in the order of 100 kHz. The effect of clamping diodes in the diode-clamped multilevel inverters play an important role in determining its efficiency. Power loss introduced by the reverse recovery of MOSFET body diode prohibits the use of MOSFET in hard-switched inverter legs. A technique of avoiding reverse recovery loss of MOSFET body diode in a three-level neutral point clamped inverter is suggested. The use of multilevel inverters topology enables operation at high switching frequency without sacrificing efficiency. High switching frequency of operation reduces the output filter requirement, which in turn helps in reducing the size of the inverter. This study elaborates the trade-off analysis to quantify the suitability of multilevel inverters in the low-power applications. Advantages of using a MOSFET-based three-level diode-clamped inverter for a PM motor drive and UPS systems are discussed.

Comparison and Reduction of Conducted EMI in SiC JFET and Si IGBT-Based Motor Drives

Abstract: This paper compares the conducted electromagnetic interference (EMI) in Si insulated gate bipolar transistor (IGBT) and silicon carbide (SiC) junction field-effect transistor (JFET) based motor drives. Two inverters, with the SiC and Si transistors, respectively, are built using the same circuit layout and investigated accordingly. Their conducted EMI levels are compared under the conditions of without filter and with traditional common mode (CM) filters. Reasons of the exhibited different noise emissions are analyzed. To verify the discussions, two inverters are tested in the CM. This allows for the identification and analysis of their maximized CM and differential mode (DM) wave shapes. It is shown that the excited parasitic oscillations during the switching transients are magnified more in the SiC JFET inverter, which is the main cause of the noise difference. Lastly, improved filtering solutions are proposed, which effectively suppressed the increased high-frequency noise due to the faster SiC switching speed.

Investigation of Conducted EMI in SiC JFET Inverters Using Separated Heat Sinks

Abstract: This paper systematically investigates the conducted electromagnetic interference (EMI) using separated heat sinks for a silicon carbide (SiC) JFET inverter for motor drives. The inverter circuit layout is implemented with discrete SiC JFETs attached on top of the heat sink, which creates extensive capacitive couplings and moreover increases parasitic oscillations. To minimize the influence, the solution of using separated heat sinks is proposed. For better common mode performance, the high-side heat sink is grounded to avoid fast dv/dts that occur between the drain of the lower switch and the low-side heat sink. For better differential mode performance, the RC snubber circuit and ferrite beads are used to damp parasitic oscillations. Two 2.2 kW inverter prototypes, with six discrete SiC JFETs on one common heat sink and separated heat sinks, respectively, are built using the same circuit layout. Their EMI spectra are compared under unfiltered and filtered conditions. The experiments show that the separate heat sinks inverter system exhibits significantly reduced EMI. Last, three improved solutions are proposed, which effectively suppresses the emitted EMI to the target level.