Kimmo Rauma

Open-Loop Adaptive Filter for Power Electronics Applications

An open-loop adaptive filter, which can be used in various power electronics control systems, is presented. The purpose of the filter is to give as fast a step response as possible while retaining good noise canceling properties. The proposed filter consists of a first-order infinite-impulse-response filter, the coefficients of which are adapted according to the changes in the input signal. In particular, good performance is obtained if the filter must suppress a low-frequency narrow-band noise. The performance of the filter is verified through step response simulations and with a laboratory prototype of a welding machine. The control system of the prototype is implemented on a field-programmable gate array. The proposed filter is compared to two optimum linear filters: (1) the moving average filter and (2) a filter whose step response and white noise gain are both optimized. The step response simulations show that, with the proposed filter, the step response time can be considerably decreased compared to these filters. The prototype also shows that, when using the proposed filter, the step response of the welding current can be made better than that of the original analog control.

Verification of frequency converter with small DC-link capacitor

Traditionally, large DC-link capacitors have been used in voltage source inverters. By using MPPF-type capacitor the line current harmonic content is improved and the unreliable electrolytic capacitor can be omitted. In this paper, the dynamic behavior of an inverter with small DC-link capacitor has been studied by simulations and laboratory tests. A new modulating principle, named differential space vector pulse width modulation (DSVPWM), can deal with heavily fluctuating DC-link voltage and produce correct voltage vector with high accuracy. The performance of the DSVPWM has been verified with simulations

Overmodulation in Voltage Source Inverter with Small DC-link Capacitor

Overmodulation (OM) is used to increase the voltage output of the PWM controlled frequency converter. Full inverter voltage utilization is important because of the cost and output power improvement perspectives. The voltage source inverter (VSI) with high overmodulation performance is less sensitive to inverter DC-link voltage disturbances. DC-link voltage drop may result in unintentional entrance to the Overmodulation region. DC-link drop is often due to line voltage sag or fault conditions. When small DC-link capacitor is used, voltage drops due to full-wave rectifier bridge are present all the time. A high performance overmodulation method improves the drive performance under such conditions

DTC driven single phase fed voltage source inverter with small dc-link capacitor

In voltage source inverters (VSI) relatively large de-link capacitors are used to provide stable dc-link voltage and energy storage. However electrolytic capacitors are large, heavy and expensive. In many cases the lifetime of the electrolytic capacitor is the main factor that limits the lifetime of the frequency converter. Large de-link capacitor causes high line current peaks during dc-link capacitor load. Power factor correction circuits are traditionally used to correct line current quality in single phase supplied VSI. To overcome the problem of the line current quality and big electrolytic capacitor, a relatively small film capacitor (MPPF, metallized polypropylene film capacitor) is selected. While overcoming the problem of the line current waveform and the life time of the electrolytic capacitor the new problem is arisen in the form of the heavily fluctuating dc-link voltage. This leads to heavily fluctuating torque. Many industrial dynamically low performance applications, such as pumps and fans, are high inertia loads. Thus even heavily fluctuating torque can be tolerated. DTC motor control is used to maintain as smooth torque behavior as possible.

Electric drive emulator using dSPACE real time platform for VHDL verification

The use of programmable logic devices e.g. field programmable gate arrays (FPGA) in the motor control devices increase the complexity of the testing and verification. Simulation and verification can be linked together using Simulink, HDL-simulation library and dSPACE real time platform. VHDL co-simulations with Simulink model allows control algorithms to be simulated together with the rest of the system model. dSPACE real time platform allows Simulink models to be executed in real time. dSPACE can be used to emulate inverter and motor in the verification phase where control algorithms run in FPGA circuit and control the whole emulated electric drive system.

Control of an inverter output active du/dt filtering method

Frequency converters are used to increase the controllability and efficiency of electric motor drives. Induction motors may experience an insulation failure caused by an overvoltage at the motor terminals. The overvoltage is caused by an impedance mismatch between the motor terminals and the motor cable. Fast pulse rise times compared with the propagation delay in the motor cable will make the terminal voltage exceed the DC link voltage level. Passive filtering techniques for two-level inverters have been developed to cancel out the overvoltage. In this paper, a control method for active du/dt filtering to reduce motor terminal overvoltage is introduced. A filter topology and the basis for setting the filter component values are presented. The size of the passive filter used in this filtering method is smaller than in traditional passive du/dt filters. Functional filtering can be achieved with a smaller filter by controlling the filter output voltage rise time with an accurately timed inverter output voltage. The control of the system is implemented on a field programmable gate array (FPGA) to provide precise gate drive signals with high time resolution.

New Bus Structure for Programmable Logic Devices Controlling Power Electronics

Development of PLDs (programmable logic device) like large FPGA (field programmable gate array) circuits allow those to be used as modern control platforms for power electronics. The increase of the computational power makes it possible to implement a whole control system inside one chip. This single chip solution needs efficient design methodology to be developed for system implementation. In this paper a new architecture for an FPGA circuit in power electronics control is presented. Motor drive system controlled by one million gate Virtex-II FPGA circuit is presented as a test case

FPGA based dead-time compensation for PWM inverters

Dead-times of power switches causes significant errors in some electric drive applications. Compensation of the error can be done with a small amount of discrete components and small programmable logic circuit. Larger field programmable gate array (FPGA) circuit allows integration of the modulator and compensation logic in the same circuit resulting a very compact and cheap solution. This paper presents one solution to implement a dead-time compensation logic with a cheap and exact voltage feedback and a simple logic that can be implemented in a FPGA or application specific integrated circuits (ASIC). Full test system and measurement results are presented

FPGA implementation of DSVPWM modulator

In this paper a novel field programmable gate array (FPGA) implementation of differential space vector pulse width modulator (DSVPWM) is presented. FPGA circuit was chosen because DSVPWM requires high computational power and its algorithms can be easily parallelized. The use of traditional sequential processor would need very fast and expensive hardware compared to the FPGA implementation. DSVPWM enables more accurate flux linkage formation than traditional modulators. DSVPWM is completely new modulation method and this paper presents its first implementation

Using Switching Function in Preliminary EMI-analysis of a Switching Power Supply

Electromagnetic interference (EMI) is usually difficult to predict and the interference problems may dramatically lengthen the time-to-market of a new power supply product. The possibility of predicting the EMI-spectrum from the switching function of a switching power supply by means of simulation is analyzed in this paper. Simulated spectra are compared to the measurements made with a spread-spectrum DC-DC converter