The University of Wisconsin-Platteville

Transcription

1 Embedded Motor Drive Development Platform for Undergraduate Education By: Nicholas, Advisor Dr. Xiaomin Kou This research and development lead to the creation of an Embedded Motor Drive Prototyping station for the development and study of energy efficient motor drives. This development platform uses the Texas Instruments TMS320C2808 DSP, and employs Targeted Code Generation using MatLab, Simulink, and the Target Support Package TC2 for TIC2000 microcontrollers. In addition to the development of the prototyping station, several academic tools are also created to aid students in reaching their desired learning outcomes. Introduction Energy efficiency is a key aspect of modern motor drive design. Motors account for approximately 60% of all electric loads in the United States. To decrease electric energy usage, it is necessary to develop and teach energy efficient design concepts at the undergraduate level. The embedded motor drive development platform is a tool students can use to develop and analyze energy efficient power systems. This development platform integrates the most recent digital signal processing technology and state-of-the-art design methods. These technologies not only give students the experience of working with the tools they will use in their post education careers, but also allow for the development of the next generation of energy efficient motor drives. Included in this motor drive development platform are the necessary tools for students to get started. Along with a video file describing a brief overview of the development platform, tutorials have also been created to aid in the development of individual student projects. I. PROJECT DEVELOPMENT A. Goals and Features The goal of this embedded motor drive development platform is to create a testing bench for use by students of all levels in a range of engineering courses, particularly Electrical Engineering. The most recent digital signal processing, DSP, technologies, along with advanced power conversion hardware are integrated into this platform. Also, a graphical programming interface, with hardware-in-loop design processes, is used to reduce the student learning curve, and allow for more rapid development and analysis of innovative, efficient motor drive solutions. To aid in the design of this embedded development platform, it is necessary to focus on the development of a particular application, in this case, an Embedded Three-Phase Brushless The Big M, Vol V,

2 DC Motor Drive [1]. This type of motor drive is chosen for the fact that three phase BLDC motors are an advanced motor topology which offer higher power densities and increased control capabilities, making them especially suitable for automated processes. B. Overview of the Embedded Motor Drive The embedded motor drive used here consists of three major components, a Three-Phase Brushless DC Motor, Power Converter, and a Microcontroller. A diagram of the embedded motor drive is shown in Fig. 1. Fig. 1. Diagram of the Embedded Three Phase Brushless DC Motor Drive. The three-phase BLDC motor is unique in that the magnetic field is formed by permanent magnets on the rotor, and power can be supplied to very specific regions of the stator. As power is applied to a certain section of the stator, a magnetic field is created. This field attracts the opposite polarity magnet on the rotor, making the rotor rotate. Turning on these sections of the stator in a specific and controlled sequence allows for the controlled commutation of the The Big M, Vol V,

3 rotor. Many BLDC motors come equipped with Hall Effect sensors [2] which sense the polarity of the magnetic field within the vicinity of the sensor. Typically, three of these sensors are located 120 apart around the stator. As the rotor turns, these sensors output logic 1 or 0, depending on the position of the rotor. These Hall sensors are used to determine which section of the stator has power supplied to it, and, depending on the desired direction of rotation, will determine which section of the motor to apply power to next, such that rotation will occur. Also, BLDC motors can come equipped with a position encoder [3]. The encoder consists of a circular sheet with slits evenly spaced around a circle and attached to the rotor. As the rotor rotates, a light is shined through these slits onto a sensor, which then outputs a series of pulses as the rotor turns. By counting these pulses, the speed and position of the rotor can be determined. The power converter is the power processing hardware. This section interfaces between the DC power supply and the three-phase motor using a set of controllable switches. Controlling which switches are on or off will determine which section of the stator has power applied to it. To create smooth rotation of the rotor, a microcontroller is used to sense and interpret signals received from the hall sensors and encoder, and determines the next state of the each switch. This is what makes the drive Embedded ; all of the control algorithms and switching signals are implemented on a single digital signal processing microchip [4]. C. Prototype Development Platform This embedded motor drive development platform integrates some of the most recent technologies available. The BLDC motor used is the MBE.171.E500 [5]. This motor has a maximum speed of 4000rpm, 62.5mNm nominal Torque, and 500 slit encoder. When used with a Quadrature encoder, position resolution of 0.18 can be achieved. The DMC550 [6] power converter is used as the power conversion hardware. This is a professional product available from Spectrum Digital. In the design of this development platform, it is determined that using a professional product offers two distinct advantages. Number one, it allows increased focus on the development of the specific tools for student use, and number two, it gives students exposure to professional quality products which they can study further, especially interfacing techniques and protection circuitry which they might not otherwise be exposed to. The Big M, Vol V,

4 Spectrum Digital s F2808eZdsp [7] is the controller of choice for this test platform. This is a DSP control board with all of the necessary circuitry to fully utilize TI s TMS320C2808 DSP [8]. This is an attractive product because it has a very high operating frequency, 100MHz, and allows for the simultaneous control of two independent Three-Phase Motors. In addition, this DSP uses advanced hardware peripheral units to reduce board size and increase microcontroller flexibility. These peripheral units [9] are actual hardware circuits inside the chip which perform specific functions, including pulse width modulation, PWM, generation, capturing of encoder position and speed information, and handling communication with other devices. The epwm, enhanced Pulse Width Modulation [10], peripherals are useful in the generation of switching commands, and the eqep, enhanced Quadrature Encoder Pulse [11], is useful in obtaining data required for calculating position and speed information. A photo of the Embedded Motor Drive Development Platform prototype is shown in Fig. 2. BLDC Motor Microcontroller Power Converter IDC Connectors +3V Power Supply Bread Boards Fig. 2. Photograph of the Embedded Motor Drive development platform prototype. Notice in Fig. 2, the addition of breadboards and Insulation Displacement Connectors, IDCs, to access all of the communication signals for analytic purposes. Also, due to The Big M, Vol V,

5 the frequency limitations of the DMC550 encoder interfacing circuits, it is necessary to connect the motor s encoder output directly to the ezdsp through a 4.7kΩ pull-up resistor with a +3V power supply. Caution must be used when connecting external circuits to the ezdsp, as voltages greater than 3.3V can damage the device. D. Microcontroller Programming Tools One of the goals of this development platform is to provide the ability to rapidly develop the control algorithms for the embedded motor drives [12]. To do this, a method of graphical programming is used. The MatLab, Simulink Target Support Package TC2 [13] is a code generation toolbox for the Simulink modeling environment which provides the ability to graphically program the TI C2000 series DSPs. This toolbox provides block diagrams specific to the peripheral hardware units and control capabilities of the DSP used in this platform. Dialog boxes and drop down menus greatly simplify the configuration and programming of the DSP. To download the program to the DSP board, Simulink actually generates the C language program files and links with Texas Instruments Code Composer Studio, which then compiles the C-files and downloads them to the DSP. The entire process of building the C-files, linking with CCS, compiling, and downloading to the DSP, takes less than a minute. An example of programming an epwm peripheral using the Simulink block diagram is shown in Fig. 3. Fig. 3. Example of programming the ezdsp using Simulink block diagrams. The Big M, Vol V,

6 Simulink was chosen as the code generation software package of choice for multiple reasons. The Help menu in Simulink is an excellent source of information during the model debugging process. If there is an error in the program, Simulink will tell the user what the problem is and how to fix it. Simulink can also be used to program other TI DSPs, including the C6000 series. Another useful aspect of Simulink is that it provides the option of communicating with the DSP to provide real-time information during actual program runtime. This allows for the hardware-in-loop design process to be fully utilized in the development of control algorithms; controller performance can be tuned in real time. Furthermore, Simulink is already incorporated into the current curriculum at this university, meaning students will not have to learn an entirely new program environment. For these reasons, MatLab, Simulink is the ideal choice of graphical programming software. II. ACADEMIC TOOLS A. Tutorials This development platform is meant to be used by students of all levels in a multitude of engineering courses. For this reason, it is necessary to have specific academic tools available, for students to use, which provide the fundamental basis for understanding. In this effort, tutorials and video files have been created for students to use as an educational guide. Tutorial 1 ensures that the student has the development platform correctly configured to use with their host PC. It walks them through step-by-step, to make sure all software components are correctly installed, and that the microcontroller properly communicates with the host PC. Tests are presented to ensure both the ezdsp and DMC550 work properly, and also provide an understanding of the interfacing requirements for each board. This tutorial also explores the implementation of simple algorithms on the DSP, such that the student becomes familiar with the Simulink graphical programming environment and understands how the various components are integral to the overall performance of the embedded system. Tutorial 2 provides the understanding of how to configure the ezdsp through Simulink to control the operation of the six independent switches in the power converter. The pulse-widthmodulation and dead-time injection capabilities of the C2808 epwm peripherals are explored in detail, and a step-by-step procedure for configuring the peripheral s operation is presented. The Big M, Vol V,

7 Students can use this tutorial as a supplement to their course work, in which they learn about various PWM characteristics, techniques, and applications. Tutorial 3 explores the development and control of an embedded motor drive. In this tutorial, students will learn how to perform the proper tests to determine the characteristics of the motor hall sensors and position encoder. This information is necessary in the design of the motor drive. A configuration of the DSP s epwm and eqep peripherals is presented for the three-phase BLDC motor and power converter present on this development platform. By following this tutorial, a student will successfully design their own embedded motor drive. B. Configurable Brushless DC drive with Closed Loop Speed Control For students with less knowledge of engineering and drives, a configurable Brushless DC motor drive is available for them to implement on the development platform. The Simulink block diagram of this configurable drive is shown in Fig. 4. Fig 4. Simulink block diagram of the configurable three-phase brushless DC motor drive with closed-loop speed control. The Big M, Vol V,

8 This motor drive has configurable parameters which students can adjust. This allows the student to interact with the embedded system and provides the opportunity to get hands on experience with the latest development tools available. They can perform all of the same tests and analyses as the more advanced students, but without having to design their own motor drive. They can see physical devices working and gain an understanding of the development process. For entry level students, this type of experience can be very motivational and can inspire them for future projects. The configurable BLDC drive is useful for senior students as well. Because the system is already developed, engineering students can focus further on a specific segment of the efficient motor drive design. A controls engineer can use the existing drive and concentrate their time and effort on the design and analysis of their particular control strategies. The pre-built drive can also be used in conjunction with particular classes. Professors can use the development board to emphasize the curriculum in their courses, while students can verify those theories and concepts which they learn in class. III. PROJECT OUTCOMES AND CONCLUSIONS This research and development was successful in creating an embedded motor drive development platform. With this platform, professors can emphasize theories and concepts taught in their courses, and students can verify those theories and concepts while getting the hands on experience of working with the most advanced embedded motor drive technologies. With the developed academic tools accompanying this platform, students will be able to focus more on any particular aspect of the motor drive design. The graphical programming and hardware-in-loop design process, provided through the MatLab, Simulink TC2000 toolbox, reduces the learning curve and allows for further advances in the efficiency of the motor drive. They can modify and analyze existing drives and conduct performance analyses, or they can create their own motor drive using the tutorials as guidance. Together, these hardware, software, and academic tools provide an excellent resource for both students and faculty in the study and design of embedded motor drive systems. The key aspects and achievements of this embedded motor drive development platform have been presented at the UW-Platteville Poster Day, EMS Engineering Expo, and PURF Spring Luncheon. The Big M, Vol V,

9 ACKNOWLEDGMENTS This research and development was funded by the UW-Platteville PURF grant, as well as Dr. Xiaomin Kou s Opportunity Fund. A special thanks goes to the UW-P PURF committee for providing us this opportunity, as well as to Dr. Kou for all of his knowledge and support. REFERENCES [1] Bose, Bimal K.. Power Electronics and Motor Drives: Advances and Trends. Academic Press, [2] Gilbert, Joe. "Technical Advances in Hall-Effect Sensing." Allegro MicroSystems. 20 Jan 2009 < [3] "Encoder Basics." Danaher Precision Systems. 30 Apr 2009 < [4] Peckol, James K.. Embedded Systems: A Contemporary Design Tool. Hamilton Printing, [5] "Brushless Motors." Inteligent Control Technology Co., Ltd.. 09 Sep 2008 < [6] A WENWEN MCU Laboratory. 09 Sep 2008 < [7] "Spectrum Digital Support." Spectrum Digital. 30 Apr 2009 < [8] "TMS320F2808.pdf." Texas Instruments. 30 Apr 2009 < [9] "SPRU566h.pdf." Texas Instruments. 30 Apr 2009 < [10] "SPRU791e.pdf." Texas Instruments. 09 Sep 2008 < [11] "SPRU790d.pdf." Texas Instruments. 09 Sep 2008 < [12] Duma, R., P. Dobra, M. Abrudean, and M. Dobra. "Rapid Prototyping of Control Systems using Embedded Target for TI C2000 DSP." IEEE Xplore 2007 Web.13 Nov [13] "DSP - Digital Signal Processing and Communications - MatLab." MathWorks. 30 Apr 2009 < The Big M, Vol V,