Electric Drive Technology at dspace Developing and Testing Electronic Control Units

Transcription

1 Electric Drive Technology at dspace Developing and Testing Electronic Control Units

2 Contents Electric Drive Technology 3 Developing with Rapid Control Prototyping 4 Use Cases 6 Success Stories 11 dspace Products MicroAutoBox II 17 AC Motor Control Solution 19 RapidPro Hardware 22 DS1103 PPC Controller Board 23 DS5203 FPGA Board 26 Battery Cell Voltage Measurement and Balancing 29 Testing with Hardware-in-the-Loop Simulation 30 Use Cases 32 dspace Products EMH (Electric Motor HIL) Solution 40 DS2655 FPGA Base Module 45 Programmable Generic Interface (PGI1) 47 Electronic Load Modules 48 Battery Cell Voltage Emulation 52 ASM Electric Components Model 54 XSG Electric Component Library 56 XSG Utils Library 57 JMAG-RT Parameterization Support 58

3 Electric Drive Technology Developing and Testing Electronic Control Units Electric drives have been used increasingly over the past few years. Not only in vehicles, but also in automation technologies and other applications. The reason: They have numerous advantages and wide-ranging application potential in many areas. Electric motors can be very small and fit almost anywhere. They have very high dynamics and provide high torque at lower rotational speed. Other advantages are improved energy savings due to power-on-demand, better controllability, and easier maintenance. Application Areas of Electric Motors In automotive applications, electric drives are being incorporated into numerous complex, basic, and safety-relevant vehicle functions. Some examples of automotive applications are: Electric motors are also in widespread use in the capital goods industry, medical engineering, and aerospace. Especially aerospace applications have high requirements regarding safety and reliability. Electric steering systems Powertrain actuators Starter-generator systems Electric vehicles Mild/full-hybrid systems Electric brake systems Electric window lifters Auxiliary aggregates: oil pumps, water pumps, etc. Other applications include: Wind energy converters Electric trains Printing machines Roller mills for cold and warm roll forming Zinc coating plants (headway and drive control) Artificial respirators Magnetic resonance tomography Bipedal walking robots 3

4 Developing with Rapid Control Prototyping Advantages of Rapid Control Prototyping To develop a new control strategy, you have to evaluate different approaches and functions. You need to concentrate completely on your function development and should not have to worry about the performance of the prototyping hardware. Ideally, you can optimize your function designs both on the test bench and in the actual vehicle or industrial drive system until they meet the requirements all without having to do any programming. Challenges of Developing Electric Motors The trend towards mechatronic integration means that electric motors are gaining popularity. However, their advantages are accompanied by a higher workload for integrating the additional control algorithms into the respective controller. The result is a more complex controller software, which usually leads to increased development times. This drawback can be countered by using model-based design along with rapid control prototyping (RCP) to accelerate design iterations of the control algorithm on the real object. RCP systems for electric motor control development need to meet specific requirements with regard to: Powerful system architecture Flexible I/O interfaces Dynamic power stages An intuitive software environment Controller New controller function Developed in Simulink/Stateflow Tests for: Electric motors Hybrid controllers EPS (electronic power steering)... 4

5 Rapid Control Prototyping Easy Workflow New functions are typically developed in MATLAB /Simulink /Stateflow. dspace Real-Time Interface (RTI) is the link between this development software and the dspace hardware. It automatically implements the MATLAB/Simulink/ Stateflow model on the dspace MicroAutoBox, the modular hardware installed in the AutoBox, or the stationary dspace Expansion Box. If function modifications are necessary during the tests, you can simply correct a function in Simulink and flash it to the hardware again. The dspace prototyping system substitutes for any controller and its connections to the vehicle or the controlled system during the development process. For further product information, please see: MicroAutoBox II, page 17 RapidPro, page 22 DS1103 PPC Controller Board, page 23 AC Motor Control Solution, page 19 DS5203 FPGA Board, page 26 Solutions for Electric Drive Applications dspace offers specialized products for the highly dynamic requirements of electric drives: The AC Motor Control Solution is installed in the AutoBox to control diverse AC motors. The AC Motor Control Solution upgrades MicroAutoBox II to a compact, flexible development system for electric motor control applications. The MicroAutoBox II acts as the central controller. Its high performance enables unrestricted testing of new functions. The RapidPro system can be used to extend the AutoBox or MicroAutoBox II. It offers specialized halfbridge and full-bridge modules which can deliver peak currents of up to 60 A for applications with electric drives and valves. The RapidPro module for the universal control of brushless electric motors provides special support for tasks such as electrifying auxiliary aggregates. The DS1103 PPC Controller Board is an all-rounder for rapid control prototyping that provides a real-time processor and comprehensive I/O. Design, simulation and analysis on a PC I/O connection via library Monitoring, tuning Implementation on real-time hardware Real-time simulation and verification in a real environment 5

6 Rapid Control Prototyping / Use Cases Use Cases Developing Control Functions in Electric or Hybrid Electric Vehicles Task The main task in electric or hybrid electric powertrain development is to design the overall control strategy. The control functions are spread over a distributed network of electronic control units. An additional task is to integrate the control strategy into these distributed ECU networks. Challenge To develop the optimal ECU algorithms for electric motors, you need to test various different control strategies. You therefore need a development system that acts as a substitute for the future central hybrid controller. The prototyping system has to offer various interfaces and should be usable in-vehicle. Solution During function prototyping, a dspace MicroAutoBox II takes the place of the central hybrid controller. It offers comprehensive bus interfaces and, with its compact and robust design, it can be used in-vehicle. The new functions developed with Simulink are implemented on the MicroAutoBox II with dspace Real-Time Interface (RTI). ControlDesk Next Generation Powertrain CAN Hybrid CAN Hybrid Powertrain ECU RTI Engine ECU Battery Management System E-Motor ECU Transmission ECU High Voltage Battery Combustion Engine Electric Motor Inverter Transmission to Drive Shaft (Vehicle Dynamics) The dspace MicroAutoBox II acts as the central hybrid ECU during the development of new functions. 6

7 Rapid Control Prototyping / Use Cases Controlling Li-ion Cell Voltages During Prototyping Task One reason for the combustion engine s great success in the 20th century is gasoline s high energy density. While one liter of gasoline can run for many kilometers, a modern battery of the same mass or volume takes an electric vehicle only a fraction of the distance. As this comparison clearly shows, developing powerful, high-density batteries with a maximum realizable capacity is key to the breakthrough of electrical vehicles. Challenge Li-ion batteries have to be constantly monitored and controlled because the usable voltage range of a Li-ion cell is limited to several 100 mv. The further the voltage moves out of this ideal range, the more the life span of the cell is impaired. In extreme cases, the cell can even be destroyed. Instances of battery fires in telephones, laptops, and last but by no means least, in planes, emphasize just how important it is to monitor the battery state. To maximize the battery s overall capacity it is necessary to keep all cells on the same level of charge. The modular structure of the dspace battery management system allows tailor-made confi gurations of up to around 200 cells and can also be installed directly in a vehicle. Solution To benefi t fully from the high energy density of Li-ion batteries, the state of charge of the individual cells must be monitored precisely. dspace has developed a RCP battery management system (p. 29) that performs this task throughout the development process, from the first model to in-vehicle testing. Its main focus is on measuring and controlling Li-ion batteries. The system is modular and can be assembled to create confi gurations of between 6 and approx. 200 cells. It can also be installed directly in a vehicle. The BMS modules are connected via Ethernet with a dspace prototyping system such as MicroAutoBox II. dspace PGI1 Sensor Board Board with balancing resistances Cut-off device Battery Ethernet dspace system platform R F Isolation watchdog Ground Electrical isolation Enclosure Isolation fault The isolation concept of the dspace battery management system makes it safe to use high battery voltages. 7

8 Rapid Control Prototyping / Use Cases Robotics Task Rapid prototyping for robotic applications requires flexible and fast interfaces, especially fast encoder interfaces that are easy to access from the real-time Simulink model. Challenge The functions of the robotic position controller have to be performed. In the example below, the controller board replaces the position controller. The prototyping hardware should also allow easy parameter modification for convenient design optimization. Solution The real-time system picks up the robot s six incremental encoder signals to determine the current robot position. Then this data is compared with the reference values. Afterwards, the DS1103 calculates the control algorithm and sends the controller output for example, data on positions and velocities back to the robot. Calculating a robotics control algorithm on a DS1103 PPC Controller Board. Further Processing Potential All reference values are calculated in real time, even for inverse kinematics with highly nonlinear functions. External sensors such as axis-force momentum sensors can be included. Trajectory planning and running advanced algorithms for collision avoidance are also very convenient with the DS1103 PPC Controller Board. 8

9 Rapid Control Prototyping / Use Cases In-Vehicle Prototyping Task The task is to develop and verify control strategies and distributed ECU functions in an electric vehicle (EV) or hybrid electric vehicle (HEV) "on the road". Challenge To develop and verify algorithms for an EV/HEV ECU network, you need a fl exible and in-vehicle capable development system. Universal I/O interfaces, support for common bus systems, and the ability to flexibly hook up the electric motor power stages are also necessary. Solution dspace offers a fl exible development environment for in-vehicle prototyping of EVs and HEVs. The MicroAutoBox provides convenient support of common bus interfaces (CAN, LIN, FlexRay) for high connectivity. You can also use the PGI1 (p. 47) with the MicroAutoBox to interface (via the TwinSync protocol) to various LTI power stages such as the LTI ServoOne. This combination provides high flexibility with regard to the power stages, as you are able to connect various electric motors with different power ranges exactly as required. MicroAutoBox LTi ServoOne PGI1 CAN LVDS Link HV Battery 9

10 Rapid Control Prototyping / Use Cases Developing Electric Motor Control Algorithms Task The task is to develop control functions for all types of electric motors: Asynchronous motors Brushless DC (BLDC) motors Permanent magnet synchronous motors (PMSM) Challenge Fast current and voltage measurements are required and diverse position encoders have to be connected. Solution AC Motor Control Solution The ACMC Solution, based on the MicroAutoBox or the DS5202 FPGA Base Board mounted in an AutoBox, is ideal for fast current/voltage measurement, connecting diverse position encoders, and controlling AC motors. The MicroAutoBox and the AutoBox can be installed in-vehicle and connected to the electric motor. If installed in a dspace Expansion Box, the ACMC Solution can also control an industrial electric drive application. You can use the AC Motor Control Solution together with the dspace RapidPro system to control PMSM and BLDC motors. Various piggyback modules can be plugged onto the DS5202 to provide specialized, comprehensive I/O functionality, with the control algorithms running on a DS1005 or DS1006 processor board. PHS Bus DS1005/ DS1006 Control signals or Current signals DS5202 EV1048 Piggyback module used inside MicroAutoBox II Resolver, SSI, EnDat Hall / Encoder RapidPro Power Unit RTI Blockset Motor The ACMC solution offers the I/O interfaces required for devel oping control strategies for various AC motors such as BLDC motors. 10

11 Rapid Control Prototyping / Success Stories Success Stories E-Motion: Motion Control Algorithms for Electric Vehicles Research Focus at Fujimoto Research Laboratory The Fujimoto Research Laboratory at Yokohama National University in Japan investigates electric vehicles, focusing particularly on methods of electric drive technology. The laboratory is working on a type of drive known as an in-wheel motor, and is also studying the safety aspects of electric vehicles on slippery road surfaces. Research is being conducted on attitude control methods that employ yaw rate control, using this yaw moment to prevent spinning and drifting when turning. Development Objective: A Yaw-Stable Vehicle An electric motor goes straight from zero to its maximum torque. Thus, uncontrolled torque requests can result in immediate loss of static friction, which results in vehicle oversteer during extreme cornering. To detect the beginnings of oversteer, the vehicle s yaw rate has to be determined. The yaw rate is the angular velocity with which a vehicle rotates around its vertical axis. If external effects push a yaw-stable vehicle off course, in the ideal case it returns to a straight path without the driver having to steer. Test Drive with dspace AutoBox To test the control algorithms in prac tical test drives, the FPEV 2-Kanon test vehicle was equipped with a dspace AutoBox containing a DS1103 PPC Controller Board that was responsible for computing the algorithms. A control system modeled with MATLAB /Simulink was loaded to the AutoBox. The AutoBox drives the electric motors via converters. The angular velocity, the torque, the acceleration and the yaw rate are available as analog signals. Motor for SBW Acceleration sensor Yaw-rate sensor Steering angle sensor * f a x,a y Controller AutoBox DS1103 ( ),T real Li-ion battery 15Vx10 150V 300V Chopper T* R Inverter L Inverter R Motor L Motor Motor for SBW Effectiveness of the dspace AutoBox To make full use of the advantages of electric motors, the control algorithms have to be calculated extremely fast. The short sample time of the DS1103 PPC Controller Boards and its low latencies during I/O access meant that the algorithms could be executed in real time. Since the hardware has such extremely fast response times, the algorithms behaved as expected. r * Confi guration of the vehicle control system. 11

12 Rapid Control Prototyping / Success Stories Younicos: New Energy Purely Regenerative Energy Supply An autonomous, CO 2-neutral power supply based on regenerative energies for remote areas islands or villages that are far away from the main power grid: That s what Younicous is planning and developing. The first project is for the island of Graciosa in the Azores, where 70-90% of the required energy could come from the sun and the wind, and the remaining 10-30% could be generated from locally produced biofuels. A 3-megawatt sodium sulfur battery as electricity storage to compensate for large supply fluctuations, and the island will be completely independent of fossil fuels. Developing the Converter Control The battery converter control has two main components: a real-time controller and a communication system. To find the optimum control for the converter, Younicos uses rapid prototying to test different voltage and frequency control algorithms that were designed in MATLAB /Simulink. For the actual tests, the AC Motor Control Solution from dspace was used. This consists of a DS1005 Processor Board and DS5202 FPGA Base Board with a piggyback module. The algorithms are implemented on the DS1005 by means of the dspace Real-Time Inter face (RTI), and then executed on the board. The DS5202 provides the necessary I/O connection between the processor board and the converter. If any changes are made to an algorithm, they can quickly be transferred from MATLAB/Simulink to the DS1005 by using RTI. Simulating Consumption, Wind, and Sun For simulating wind turbines and solar power plants Younicos implemented and executed their own simulation models on several dspace DS1005 PPC Boards. Real wind and sun data measured on the island of La Graciosa provides the input parameters for ascertaining the currently available power. This available power is then compared with a consumption profile that represents the island population s energy requirements throughout the day. Converters then perform energy distribution. Each battery is coupled to the simulated supply grid via a converter. The load on the grid is represented by another converter that runs through a scaled load profile of the island. Carrying out the Project In August 2012, Younicos and the local power supplier signed agreements on power input to the electricity grid and on the price of the electricity the commercial base of the project. The construction of the photovoltaic plant, wind park and battery storage is expected to be completed at the end of 2014, when the entire system will go into operation. This solar power system feeds an autonomous charging station for electric vehicles. 12

13 Rapid Control Prototyping / Success Stories MAGNA STEYR: Hybrid Drive MAGNA STEYR and its cooperation partners integrated new hybrid components in a vehicle and implemented a control system using a dspace prototyping system (Micro- AutoBox plus RapidPro). The hybrid demo vehicle HySUV (Mercedes M-class) with a dspace prototyping system as the central drivetrain control has made the hybrid drive a reality. MAGNA STEYR and its partners use the demo vehicle as a platform for further optimization of driving behavior, consumption, and emissions. Drive Systems of the Future MAGNA STEYR worked with MAGNA POWERTRAIN and Siemens VDO to develop modular hybrid drive systems, taking into account the research fi ndings from K-net KFZ, the competence network for Vehicle Drives of the Future. With the support of the OEMs, hybrid components developed by MAGNA are integrated in the drivetrains of prototype vehicles to investigate the optimization potential of the consumption, dynamics, and emissions. The control system and the cross-linking of new components in the drivetrain are implemented with the dspace prototyping system (MicroAutoBox plus RapidPro) on the basis of a central hybrid drive strategy. MAGNA STEYR has put this into operation in the hybrid demo vehicle HySUV. The automatic transmission and transfer case of a Mercedes ML350 were replaced by an automated manual transmission and MAGNA s E4WD module consisting of 2 electric drives and clutches. A full hybrid drivetrain with electrical all-wheel drive was implemented in this way. A lithium-ion battery system, developed by MAGNA STEYR, provides energy storage. Prototyping Hardware and Function Development The control software comprises the functions and interfaces of the entire torque path in the drivetrain. The objective was to control all the components of the hybrid drivetrain with just one prototyping system. In addition to their standard software development platform MicroAuto- Box, MAGNA STEYR decided to use the RapidPro system to effi ciently realize the broad range of signal conditioning and power stages. Its flexibility provided by software- and hardware-configurable signal I/O proved to be an advantage, particularly in early phases of prototype development, when the sensor and actuator systems are not yet completely defined. After the function software had been successfully implemented and tested, MAGNA STEYR entered the test drive phase, with the objective of further optimization. dspace MicroAutoBox dspace RapidPro Function System manager Diagnostics CAN I/O LVDS Actuator driver Sensor I/O Hardware diagnostics Measurement CAN Engine CAN Hybrid CAN Drivetrain CAN Converters 1/2 Gateway ECU Battery management system Air conditioning Gateway ESP Electrical machines 1/2 Combustion engine High-voltage battery system Air conditioning compressor Cooling circuits 1-3 AMT E4WD HMI System architecture: The dspace prototyping system net - Hardware connections Bus connections High-voltage component 12-V component worked in the vehicle. 13

14 Rapid Control Prototyping / Success Stories Ohio State University: Control of a Power-Split Hybrid-Electric SUV As part of this competition, Ohio State University (OSU) engineering students developed an HEV that is powered by a combination of a turbocharged diesel engine, a highvoltage, belted starter-alternator (BSA) and an AC induction type traction electric machine. In this confi guration, the rear and front drive systems are coupled through-the-road. Control Implementation Using the MicroAutoBox Prior to the actual implementation, OSU tested the performance of its control strategy using custom-designed vehicle simulation tools developed in the MATLAB /Simulink environment. After initial testing, the control strategy was implemented on the MicroAutoBox system via dspace s Real-Time Interface and the RTI CAN Blockset. MicroAutoBox is the primary vehicle control unit to perform fundamental hybrid powertrain operations such as energy optimization, battery charge control, engine start-stop, drivability control, electric traction control, and regenerative braking. In the student-designed vehicle, the MicroAutoBox communicates with several control modules through dual CAN buses. The versatile I/O interface simplifi ed the integration of several analog and digital I/Os into the controller for the added hybrid components. The fast numerical processor featured by the MicroAutoBox made it possible to implement computationally burdensome algorithms onboard the vehicle. Belted starter alternator Engine & transmission High-voltage battery pack Driver CAN A Auxiliary controller Exhaust system CAN B Rear electric motor GM LAN Data acquisition The MicroAutoBox interfaces with the powertrain control modules via dual CAN buses and several I/Os. 14

15 Rapid Control Prototyping / Success Stories Deutz: Developing Hybrid Drives for Mobile Machines Wheel Loader with Hybrid Drive In a joint project with wheel loader specialist Atlas Weyhausen, Deutz used dspace tools to develop what is called a mild hybrid system for their AR-65 Super wheel loader. Mild means that the electric motor is rigidly coupled to the diesel engine and supports frequent braking and acceleration. The following dspace tools were used to develop the software functions for the hybrid system s ECU: MicroAutoBox (as the hybrid system ECU) Real-Time Interface (for setting up the I/O interfaces for the MicroAutoBox) RTI CAN MultiMessage Blockset (for setting up CAN communication) ControlDesk (for calibrating the hybrid functions) By using RTI and the RTI CAN MultiMessage Blockset, Deutz was able to implement fully functioning system software on the MicroAutoBox in only 3 months. The RTI CAN Multi- Message Blockset proved to be a very easy-to-use tool, and its support for linking CAN configuration files (DBC fi les) enabled us to set up the CAN commu nication very quickly. Three CAN channels were set up in the wheel loader: engine CAN, hybrid CAN, and vehicle CAN. Because the system software was programmed directly in Simulink, it was possible to try out the software functions immediately on a plant model (MIL) containing the engine, electric machine, inverter, battery, work hydraulics and traction hydraulics. Deutz was therefore able to test the software functions long before the fi rst prototype components became available. This was absolutely essential in view of the very short development time assigned to this project. Using the pretested software functions and the inputs and outputs configured with RTI (digital, analog, PWM, CAN), Deutz produced a software version that would run on the MicroAutoBox and tested it on the test bench. Functions such as start/stop were tested and calibrated with ControlDesk. Finally, Deutz put the wheel loader into operation with the MicroAutoBox as a superordinate hybrid system ECU and implemented the functions for boosting power and raising/ shifting the load point. Schematic of the mild hybrid system in the wheel loader. The MicroAutoBox is used as a superordinate hybrid system ECU. 15

16 Rapid Control Prototyping / Success Stories Additional Information You can download success stories, articles and product information on drive applications at under "Downloads". Available Publications (Partial List) Title Author Published at Implementing Electromobile Ideas Holger Ross (dspace GmbH) Elektronik Automotive, Apr 2011 When Processor and FPGA Work Together Frank Mertens (dspace GmbH), Thomas Sander (dspace GmbH) Elektronik Automotive, May 2011 Get Your Ideas on Track Frank Mertens (dspace GmbH) Automobil Elektronik, Oct 2010 Intelligent I/O up Close Jürgen Klahold (dspace GmbH) Offprint translation from "Hanser Automotiv", Nov 2009 All Inclusive/Off-the-Shelf Flexibility Can Be So Compact Frank Mertens (dspace GmbH), Holger Ross (dspace GmbH) Frank Mertens (dspace GmbH), Holger Ross (dspace GmbH) Offprint translation from "Elektronik automotive", Oct 2009 Offprint translation from "AutomobilElektronik", Oct

17 Rapid Control Prototyping / Products dspace Products MicroAutoBox II Compact prototyping unit for electric motor controls Comprehensive I/O including CAN, LIN, K/L line, FlexRay, Ethernet, and LVDS/bypass interfaces Robust and compact design ideal for in-vehicle use IBM PowerPC running at 900 MHz Variant with Simulink -programmable FPGA AC Motor Control Solution (p. 19) NEW: Multistage watchdog mechanism Application Areas MicroAutoBox is a real-time system for performing fast function prototyping in fullpass and bypass scenarios. It operates without user intervention, just like an ECU. Key Benefits The special strength of the MicroAutoBox hardware is its unique combination of high performance, comprehensive auto motive I/O, and an extremely compact and robust design all for a favorable price. This lets you equip several vehicles or a whole test fl eet to check the reliability of your control functions. In addition to the standard I/O, MicroAutoBox offers variants with FPGA functionality for application-specific I/O extensions and for user-programmable FPGA applications. Moreover, there are MicroAutoBox variants with inter faces for all major auto motive bus systems: CAN, LIN, K/L line, FlexRay, and Ethernet. IBM PPC 750 GL 16 MB local RAM Performance timer 64-Bit Global Bus Signal Generation/ Measurement CAN/LIN/serial module Ethernet host interface 6 MB communic. memory 16 MB flash (non-volatile) Clock/ calendar Watchdog USB ECU interface ECU interface Ethernet I/O interface Signal Conditioning Physical CAN/serial Connector (LEMO) Connector (LEMO) Connector (LEMO) Connector (LEMO) Connector (LEMO) CAN/LIN/serial module Physical CAN/serial Local Bus/Intermodule Bus Digital I/O (FPGA-based) 16-channel 16-bit ADC 4-channel 12-bit DAC Signal Generation/ Measurement Signal conditioning & protection Signal conditioning & protection Signal driver & protection Signal Conditioning I/O Connector IP module slot (e.g., for FlexRay) MicroAutoBox II 1401/1511/1512 Programmable FPGA IP module slot (e.g., for FlexRay) FPGA extension slot Optional signal conditioning on DS1552 or ACMC Solution Add-On Modules 1) I/O Connector 17

18 Rapid Control Prototyping / Products Technical Details Parameter Specification MicroAutoBox II 1401/ /1511/1512 Processor Memory Boot time Interfaces Host interface Real-time I/O interface USB Interface CAN interface Serial interface (based on CAN processor) Dual-port memory inter face (ECU interface) IBM PPC 750GL, 900 MHz (incl. 1 MB level 2 cache) 16 MB main memory 6 MB memory exclusively for communication between MicroAutoBox and PC/notebook 16 MB nonvolatile flash memory containing code section and flight recorder data Clock/calendar function for time-stamping flight recorder data Depending on flash application size. Measurement examples: 1 MB application: 160 ms; 3 MB application: 340 ms 100/1000 Mbit/s Ethernet connection (TCP/IP). Fully compatible with standard network infrastructure. LEMO connector. Optional XCP on Ethernet interface to support third-party calibration and measurement tools 100/1000 Mbit/s Ethernet connection (UDP/IP). RTI Ethernet (UDP) Blockset (optional) for read/write access. LEMO connector. USB 2.0 interface for long-term data acquisition with USB mass storage devices. LEMO connector. 2 dual CAN interfaces; 4 CAN channels in total 2 x RS232 interface 2 x serial interface usable as K/L line or LIN interface 2 x dual-port memory interface, 16 K x 16-bit DPRAM FlexRay interface 2 slots 1) for FlexRay modules (i.e. 4 FlexRay channels) Programmable FPGA Xilinx Spartan -6 LX150 2) Analog input Analog output Digital I/O Resolution bit channels bit channels (additional channels with DS1552) Sampling Input voltage range 16 parallel channels with 1 MSPS conversion rate V Resolution 4 12-bit channels 4 12-bit channels (additional channels with DS1552) Output voltage range V V Output current General Bit I/O PWM generation/ measurement Signal conditioning Physical connections Physical characteristics Enclosure material 5 ma max. sink/source current FPGA-based digital I/O RTI software support for bit I/O, frequency, and PWM generation/measurements 40 inputs 40 outputs, 5 ma output current Input / output logic levels: 5 V or levels up to 40 V (depending on V Drive ), selectable All channels fully configurable as frequency or PWM inputs/outputs PWM frequency Hz KHz, duty cycle %, up to 21-bit resolution Signal conditioning for automotive signal levels, no power driver included Overvoltage protection Overcurrent and short circuit protection ZIF connector for I/O signals, mechanically secured, Sub-D connector for power supply 40 inputs (additional channels with DS1552) 40 outputs, 5 ma output current (additional channels with DS1552) LEMO connectors for 2 ECU interfaces, Ethernet I/O interface, USB interface, and Ethernet host interface Ethernet I/O interface for notebook/pc for program load, experiment configuration, signal monitoring and flight recorder read-out Integrated Ethernet switch Cast aluminum box Enclosure size Approx. 200 x 225 x 50 mm (7.9 x 8.9 x 2.0 in) Approx. 200 x 225 x 95 mm (7.9 x 8.9 x 3.8 in) Temperature Power supply Power consumption Operating (case) temperature: C ( F) Storage temperature: C ( F) V input power supply, protected against overvoltage, and reverse polarity Max. 25 W Max. 50 W 18 1) IP module slot. Can also be used for other IP modules such as an ARINC interface module (via dspace Engineering Services). 2) User-programmable via RTI FPGA Programming Blockset. Using the RTI FPGA Programming Blockset requires additional software.

19 Rapid Control Prototyping / Products AC Motor Control Solution Control of diverse AC motors Fast current/voltage measurements Control of AC motors, e.g., asynchronous motors, brushless DC motors (BLDCs) and permanent magnet synchronous motors (PMSMs) Suitable PWM generation for electric drives Connection of diverse position encoders RTI Blockset for MATLAB /Simulink Purpose The AC Motor Control Solution is based on the DS5202 FPGA Base Board and on the MicroAutoBox with DS1512 I/O Board. These are specially designed for fast current / voltage measurements, connecting diverse position encoders and control ling AC motors such as ASMs, BLDCs and PMSMs. Rapid prototyping of AC motors requires fast I/O, which is fullfilled by using an FPGA Base Board and the appropriate piggyback module, with the control algorithms running on the main processor. The interface between the user's control model and the AC Motor Control hardware is realized by the AC Motor Control Solution RTI Blockset. Use Cases Typical use cases for the AC Motor Control Solution are highly dynamic control systems for different types of AC motors. Some scenarios are: Field-oriented control of PMSMs or ASMs Controlling BLDCs Prototyping new methods for sensorless control Using RapidPro Power Unit with suitable modules or customer-specific power converters To use the AC Motor Control Solution in a vehicle, you can install it in the dspace AutoBox or use the MicroAutoBox variant. For stationary use, it can be installed in an Expansion Box. Applications Measurements (of phase currents and/or voltages, and DC link currents or voltages) and ADC starts can be synchronized to PWM signals Position and speed measurements using typical sensors such as Hall sensors or incremental encoders, resolvers, or typical single/multiturn encoders with serial interfaces (EnDat or SSI) Generation of gate driver signals (center-aligned 3/6 PWMs for sinusoidal commutation or PWM signals for block commutation or advanced customer-specific PWM patterns) Synchronization of ADC measurement with centeraligned PWM signals (regular sampling) Model synchronization by interrupt generation at the center position of the PWM or at user-defined motor angles Trigger signal to external devices at the center position of the PWM 19

20 Rapid Control Prototyping / Products PHS Bus DS1005/ DS1006 Control signals or Current signals DS5202 EV1048 Piggyback module used inside MicroAutoBox II Resolver, SSI, EnDat Hall / Encoder RapidPro Power Unit RTI Blockset Motor 20

21 Rapid Control Prototyping / Products AC Motor Control Solution for MicroAutoBox II Electric Drives Control The AC Motor Control (ACMC) Solution upgrades the MicroAutoBox II to a compact, flexible development system for electric motor control applications. The ACMC solution consists of an add-on hardware module which provides the I/O interfaces for diverse AC motors and utilizes the new FPGA functionality of the MABX II 1401/1511/1512. It also comes with a dedicated RTI Blockset as the interface to the user s control model. This includes: Control of AC motors, e.g., asynchronous motors (ASMs), brushless DC motors (BLDCs) and permanent magnet synchronous motors (PMSMs) Flexible I/O interfaces for diverse position sensors such as hall sensors, encoders and resolvers Suitable PWM generation for electric drives Dedicated RTI Blockset for MATLAB /Simulink Available with MicroAutoBox variant 1401/1511/1512 Technical Data (AC Motor Control Solution for MicroAutoBox II) Parameter Digital input Digital output ADC DAC Resolver interface RS422/RS485 Power Supply for sensors Specification 8 channels, V, differential or single-ended, configurable by software. Example: 3 x single-ended for Hall sensor, 3 x differential for incremental encoder, 2 x single-ended for bit in, frequency and duty cycle measurement 24 channels, V, single-ended to generate gate driver signals, PWM synchronization signals, bit out Gate driver frequency 10 Hz... 1 MHz 8 channels, software-configurable inpult voltage range (±5 V, ±15 V, ±30 V), differential, 10 MSPS 2 channels, V (single-ended) or V (differential, reference to GND) Max. position resolution 16 bit (depending on motor velocity). Generation of excitation signal (3,7,10 Vrms; excitation frequency from khz within 250 Hz steps (software-configurable) SSI oder EnDat for connection of single/multi-turn encoder 4 RS485 transceivers (Endat or SSI Interface) 12 V: max. 50 ma 5 V: max. 50 ma (use the MicroAutoBox II VSENS-PIN for currents up to 250 ma) Scope of Delivery (AC Motor Control Solution for MicroAutoBox II) I/O piggyback module for AC motor control applications FPGA firmware for AC motor control applications Simulink interface for AC motor control applications 21

22 Rapid Control Prototyping / Products RapidPro Hardware Electric drive power stages Scalable, modular, and configurable system architecture Compact and robust enclosure For in-vehicle, laboratory, and test bench use Comprehensive software support Application-specific confi gurations for common application areas Power Stages for Electric Drive Applications Especially in the rapid prototyping phase, versatile power stages are required for driving different electric motors. Ideally, only a minimum of hardware development, or none at all, should be necessary to connect electric motors to the prototyping system. In reality, the design and implementation of such circuits can be an expensive and time-consuming task. Configuring the power stage hardware later on during the course of a project also usually involves a lot of work. The flexibility and intelligent assistance provided by the RapidPro hardware and the corresponding software from dspace will help you achieve challenging tasks for power stages with high efficiency. RapidPro Modul Examples 1) 22 Power Stage Module (PS Module) PS-HSD 6/1 PS-HCFBD 1/2 PS-HCHBD 2/2 Description 6-channel high-side driver module Requires 1 slot in a RapidPro unit 4 output channels with up to 5 A, clamping voltage 48 V 2 output channels with up to 1 A, clamping voltage 63 V Max. supply voltage: 36 V continuous, 40 V peak Switching time: <30 µs Load failure diagnosis Current measurement with hardware-adjustable low-pass fi lter (1st order) on channel 1 and 2 Overload protection, overtemperature protection, short-circuit protection to ground, V BAT and across the load, active output clamping Integrated on carrier board: load dump protection up to +100 V (only valid for internal voltage supply of the module), reverse voltage protection up to -100 V 1-channel high-current full bridge driver module Requires 3 slots in a RapidPro unit Up to 60 A DC peak current (1 s), 42 A rms continuous (T ambient = 25 C, fi lter frequency 1 khz, corresponding duty cycle) Up to 60 A DC peak current (1 s), 29 A rms continuous (depends on ambient temperature) Max. supply voltage: <20 V continuous Current measurement with hardware-adjustable low-pass fi lter (1st order) Internal free-wheeling diodes Protection against: short circuit, overtemperature, and overvoltage Load failure diagnostics 2-channel, high-current, half-bridge driver module Requires 3 slots in a RapidPro unit Each channel up to 30 A peak current (1 s), 25 A rms continuous (depends on ambient temperature) Parallel mode possible (30 A DC peak per channel, 19 A rms continuous per channel) Usable as half-bridge or low-side or high-side driver output Max. supply voltage: <20 V continuous Current measurement with hardware-adjustable low-pass fi lter (1st order) for each channel Internal free-wheeling diodes Protection against short circuit, overtemperature, and overvoltage Load failure diagnostics 1) Further signal conditioning and power stage modules available.

23 Rapid Control Prototyping / Products DS1103 PPC Controller Board Powerful controller board for rapid control prototyping Single-board system with real-time processor and comprehensive I/O CAN interface and serial interfaces ideally suited to automotive applications High I/O speed and accuracy PLL-driven UART for accurate baud rate selection Application Areas The DS1103 controller board is designed to meet the requirements of modern rapid control proto typing and is highly suitable for applications such as: Automotive controllers Electric motor control Robotics Positioning systems and stepper motors Active vibration control An integrated Infi neon CAN microcontroller makes the board an attractive tool for automotive and automation applications. Key Benefits The DS1103 is an all-rounder in rapid control prototyping. You can mount the board in a dspace Expansion Box or dspace AutoBox to test your control functions in a laboratory or directly in the vehicle. Its processing power and fast I/O are vital for applications that involve numerous actuators and sensors. Used with Real-Time Interface (RTI), the controller board is fully programmable from the Simulink block diagram environment. You can confi gure all I/O graphically by using RTI. This is a quick and easy way to implement your control functions on the board. Comprehensive Interfaces The unparalleled number of I/O interfaces makes the DS1103 a versatile controller board for numerous applications. It provides a great selection of interfaces, including 50 bit- I/O channels, 36 A/D channels, and 8 D/A channels. For additional I/O tasks, a DSP controller unit built around Texas Instruments TM320F240 DSP is used as a subsystem. Recording and Output of I/O Values The control of electrical drives requires accurate recording and output of I/O values. It is possible to synchronize the A/D channels and D/A channels, and the position of the incremental encoder interface, with an internal PWM signal or an external trigger signal. Also, the serial interface (UART) is driven by a phase-locked loop to achieve absolutely accurate baud rate selection. 23

24 Rapid Control Prototyping / Products Technical Details Parameter Specification Processor PowerPC Type PPC 750GX CPU clock 1 GHz Memory Local memory 32 MB application SDRAM as program memory, cached Global memory 96 MB communication SDRAM for data storage and data exchange with host A/D converter Channels 16 multiplexed channels equipped with 4 sample & hold A/D converters (4 channels belong to one A/D converter. 4 consecutive samplings are necessary to sample all channels belonging to one A/D converter.) 4 parallel channels each equipped with one sample & hold A/D converter Note: 8 A/D converter channels (4 multiplexed and 4 parallel) can be sampled simultaneously. Resolution Input voltage range Overvoltage protection 16-bit ±10 V ±15 V Conversion time Multiplexed channels: 1 µs 1) D/A converter Channels 8 channels Resolution Output range Parallel channels: 800 ns 1) 16-bit ±10 V Digital I/O Channels 32-bit parallel I/O Organized in four 8-bit groups Each 8-bit group can be set to input or output (programmable by software) Digital incremental encoder interface Analog incremental encoder interface Channels Position counters Encoder supply voltage Channels Position counters 6 independent channels Single-ended (TTL) or differential (RS422) input (software programmable for each channel) 24-bit resolution Max MHz input frequency, i.e., fourfold pulse count up to 6.6 MHz Counter reset or reload via software 5 V/1.5 A Shared with analog incremental encoder interface 1 channel Sinusoidal signals: 1 Vpp differential or 11 µapp differential (software programmable) < 5 resolution 32-bit loadable position counter Max. 0.6 MHz input frequency, i.e., fourfold pulse count up to 2.4 MHz CAN interface Configuration 1 channel based on SAB 80C164 microcontroller ISO DIS CAN high-speed standard Serial interface Configuration TL6C550C single UART with FIFO PLL-driven UART for accurate baud rate selection RS232/RS422 compatibility Baud rate Up to kbd (RS232) Up to 1 MBd (RS422) Slave DSP Type Texas Instruments TMS320F240 DSP Host interface Physical characteristics I/O channels 2) Physical size Ambient temperature Cooling 16 A/D converter inputs 10 PWM outputs 4 capture inputs 2 serial ports Plug & Play support Requires a full-size 16-bit ISA slot 340 x 125 x 45 mm (13.4 x 4.9 x 1.77 in) 0 50 ºC ( ºF) Passive cooling 24 1) Speed and timing specifications describe the capabilities of the hardware components and circuits of our products. Depending on the software complexity, the attainable overall performance figures can deviate significantly from the hardware specifications. 2) The exact number of I/O channels depends on your configuration and is described in the user documentation.

25 Rapid Control Prototyping / Products Technical Data (AC Motor Control Solution for PHS-bus-based systems) Parameter Digital input Digital output ADC DAC Resolver interface RS422/RS485 Power supply for sensors Specification 8 channels, V, differential or single-ended. Default: 3 x single-ended for Hall sensor, 3 x differential for incremental encoder, 2 x single-ended for bit in, frequency and duty cycle measurement 10 channels, V, single-ended; 6 gate driver signals, 4 generic digital outputs (e.g. PWM synchronization signals or bit out), optional: 12 additional gate driver signals with ACMC PWM Extension Board Gate driver frequency 10 Hz... 1 MHz 8 channels, software-configurable inpult voltage range (± 5 V, ± 15 V, ± 30 V), differential, 10 MSPS 2 channels, V (single-ended) or V (differential, reference to GND) Max. position resolution 16 bit (depending on motor velocity). Generation of excitation signal (3,7,10 Vrms); excitation frequency from khz within 250 Hz steps (software-configurable) SSI oder EnDat for connection of single/multi turn encoder 4 RS485 transceivers (Endat or SSI Interface) 5 V and 12 V, 140 ma Scope of Delivery (AC Motor Control Solution for PHS-bus-based systems) FPGA Base Board I/O piggyback module for AC motor control applications FPGA firmware for AC motor control applications Simulink interface for AC motor control applications Bracket for connecting I/O and mating connectors 25

26 Rapid Control Prototyping / Products DS5203 FPGA Board FPGA programmable per application Completely user-programmable via RTI FPGA Programming Blockset Utilizes the Xilinx System Generator (XSG) Simulink Blockset Offl ine Simulation in Simulink Basic set of I/O drivers on board, I/O extendable by piggyback modules Purpose The DS5203 FPGA Board can be adapted to various tasks, so you can react flexibly to tougher requirements like signal conditioning, using new interfaces, or speeding up model parts. FPGAs are especially useful for relieving the processor board of tasks such as signal preprocessing during ECU development. Application Areas Running at 100 MHz, the DS5203 board is ideal for application fields like engine knock, cylinder pressure analyses and electric drive projects. The DS5203 works together closely with application-specific XSG model libraries. Programming via the RTI FPGA Programming Blockset The DS5203 FPGA Board is programmed via the RTI FPGA Programming Blockset from dspace and the Xilinx System Generator. These let you develop applications for the processor board and the DS5203 together. You can test the interaction between the processor application and the FPGA application in offline simulation before implementing them on the real-time hardware. This enables you to react flexibly to new requirements such as new interfaces or having to accelerate the execution of submodels. You can also use the RTI FPGA Programming Blockset Handcode Interface to program the DS5203. Two Variants The DS5203 is available with two different FPGAs: DS5203 LX 50 includes a Xilinx Virtex -5 LX50T-1C FPGA consisting of 46,080 logic cells and 48 special DSP blocks. This board offers a cost-effective solution for smaller applications and starter systems. DS5203 SX95 includes a Xilinx Virtex -5 SX95T-2C FPGA consisting of 94,298 logic cells and 640 special DSP blocks. The large amount of DSP blocks help by performing tasks such as fast, resource-saving multiplication. 26

27 Rapid Control Prototyping / Products Technical Details Parameter General FPGA Device timing Digital I/O Input Output Specification DS5203 LX50 User-programmable FPGA Xilinx Virtex -5 LX50T-1C Logic cells: (Virtex-5 slices: 7200; DSP slices: 48) Distributed RAM: 480 kbits Block RAM: 2160 kbits 100 MHz 16 channels, usable as input or output DS5203 SX95 Maximum input voltage 15 V Digital input: Threshold adjustable for each channel from 1 V to 7.5 V Xilinx Virtex -5 SX95T-2C Logic cells: (Virtex-5 slices: 14720; DSP slices: 640) Distributed RAM: 1520 kbits Block RAM: 8784 kbits Digital output: Push-pull drivers; one output voltage can be selected for all channels: 3.3 V or 5 V Analog I/O Input 6 channels Resolution 14-bit pipelined Sampling rate 10 MSPS Input voltage range selectable for each channel: ±5 V or ±30 V Further interfaces Output 6 channels Resolution 14-bit Update rate 10 MSPS Output voltage range: ±10 V Slot for one I/O module for extending the analog and digital I/O Connection for the APU (angular processing unit) bus Physical characteristics Physical size 340 x 125 x 15 mm (13.4 x 4.9 x 0.6 in) Ambient temperature Power supply ºC ( ºF) +5 V ±5%, 2.5 A +12 V ±5%, 0.7 A -12 V ±5%, 0.1 A Relevant Software and Hardware Software Required Optional Real-Time Interface (RTI) RTI FPGA Programming Blockset FPGA Interface RTI FPGA Programming Blockset Handcode Interface Xilinx ISE Foundation and System Generator for DSP XSG Electric Components Library XSG Utils Library XSG ACMC Library Hardware Optional DS5203M1 Multi-I/O Module EV1099: Resolver SC Module for DS

28 Rapid Control Prototyping / Products DS5203M1 Multi-I/O Module The DS5203M1 Multi-I/O Module is a piggyback module for the DS5203 FPGA Board. It extends the available digital and analog I/O to give you more flexibility. Technical Details Parameter Digital I/O Input Output Specification 16 channels, usable as input or output Maximum input voltage: 15 V Threshold for each channel adjustable from 1 V to 7.5 V Push-pull drivers One output voltage can be selected for all channels: 3.3 V or 5 V Analog I/O Input 6 channels Resolution 14-bit pipelined Sampling rate 10 MSPS Input voltage range selectable for each channel: ±5 V or ±30 V Sensor supply Output 6 channels Resolution 14-bit Update rate 10 MSPS Output voltage range: ±10 V Adjustable Output voltage range: 2 V to 20 V EV1099: Resolver SC Module for DS5203 The EV1099 Resolver SC Module is a transfer element for the DS5203 FPGA Board or the DS5203M1 Multi-I/O Module. It offers special signal conditioning for electric drive applications, such as transformers for resolver simulation. Additional features: Configurable audio transformer for each of the 6 DAC channels Switchable 220 Ω resistor for each of the 6 ADC channels The module can be installed in a dspace Simulator Full-Size or Mid-Size. It is connected via ribbon cable to the DS5203 or DS5203M1, which are pin-compatible. 28

29 Rapid Control Prototyping / Products Battery Cell Voltage Measurement and Balancing Application Areas The Battery Cell Voltage Measurement and Balancing system enables highly precise measurement and control of cell voltages in lithium-ion batteries and allows the development of algorithms for battery management. The system can be installed directly in a vehicle and features cell-balancing functions that maintain the charge states of individual cells at the same level to ensure safe operation. This prevents thermal instability and extends battery life. Manual and Automatic Balancing Two operation modes are available: The manual balancing mode gives users complete freedom to balance cells individually or collectively, and at any desired intervals. The automatic balancing mode is a comfort function that specifies target voltages and switch-off times, leaving users free to focus on the more important algorithms. Reliable Safety Features Because of the high voltages of Li-ion batteries, the system provides various safety features. These include warnings about hardware, communication and synchronization errors, and also about overheating, isolation faults, and cell undervoltages and overvoltages. Technical Details Modular system supporting 6 to approx. 200 cells, installable in a vehicle Intersil ISL78600 BMS IC Cell voltage measured with ±3 mv accuracy User-defined sampling rate (max. 1 ksps) Plug-on modules for quick replacement of balancing resistances. Resistance values up to 10 Ohm S-function-based Simulink blockset (RTI Ethernet (UDP) blockset additionally required) Two balancing modes: Manual mode with full user control Configurable automatic mode Synchronized battery cell measurement Comprehensive error detection features Isolation monitoring device connectable to each EV1093 For the emulation of high-voltage batteries, please refer to the EV1077 Battery Cell Voltage Emulation Board (p. 52). 29

30 Testing with Hardware-in-the-Loop Simulation Advantages of HIL Simulation After the ECU functions have been developed and implemented on the production ECU, they have to be tested thoroughly. With hardware-in-the-loop (HIL) simulation, you can easily cover all the different motor varieties and their ECUs. The ECU s environment (interacting components or even a whole system) is simulated. This has several advantages: Function tests are possible at an early development stage, even before all parts are available in reality Laboratory tests reduce time and costs and take place under controlled conditions Failures, and the ECU s behavior in what are normally dangerous situations, can be tested with no risk for the driver or the controlled machine. The tests are reproducible and can be automated Challenges of Testing ECUs for Electric Motors Electric motors have been becoming more and more powerful in a wide range of applications. The conventional brushed direct current (BDC) motors were replaced by brushless direct current (BLDC) motors. The ECUs controlling the electric motors provide the actuation power directly. This is unlike other applications, where thermodynamic or hydraulic power is controlled by means of low auxiliary power coming from the ECU. dspace offers products and solutions for PHS-bus-based HIL simulation as well as for SCALEXIO HIL systems. ECUs for controlling electric motors are often incorporated into complex and distributed vehicle functions, so it is essential to test their interaction with other ECUs. Special solutions are needed for interfacing the ECU: High power level High dynamics Special I/O, e.g., for encoders and resolvers HIL Interfaces An ECU or other system for controlling electric motors can be accessed by the HIL simulator at different levels. Which interface to use depends on the testing purpose and project conditions: Signal level: Simulation of the power electronics, the electric motor, and the mechanical environment Very scalable, as parameters can be set fl exibly regardless of power level Full access to the model ECU must be opened Electric power level: Simulation of the electric motor and the mechanical environment Production ECU can be used Full access to the model Motor parameters can be set fl exibly within a certain power range Mechanical level: Simulation of the mechanical environment Testing of mechanical parts 30

31 Hardware-in-the-Loop Simulation ECU 3 Phase Voltages Electric Motor Controller Current Signal Power Converter Transmission Vehicle Application Application Controller Position Signal Sensor Signals Controller Power Stage Electric Motor Mechanics Signal Level Electric Power Level Mechanical Level Battery Simulation dspace offers special hardware and software for battery simulation: Real-time hardware for HIL tests with high voltage accuracy and galvanic separation Simulation models for lithium-ion batteries and nickle-metal hydride batteries for realistic battery management tests Simulation Models For real-time simulation of an electrical system, dspace provides the ASM Electric Components Library for processorbased simulation and the XSG Electric Components Library for FPGA-based Simulation. Applications can range from electric drives and inverters for closed-loop simulation with an electric drive controller, to a complete automotive electrical system including a battery, starter, alternator, and loads. Typical use cases are the simulation of realistic battery behavior during starter activation, electric drives that are integrated into a hybrid electrical vehicle powertrain, etc. For further product information, please see: DS5203 FPGA Board (PHS-bus only), page 26 Electric Motor HIL Solution (PHS-bus only), page 40 Programmable Generic Interface, page 47 Electrical Load Modules, page 48 Battery Cell Voltage Emulation, page 52 ASM Electric Components Model, page 54 XSG Electric Component Library, page 56 XSG Utils Library, page 57 JMAG-RT Parameterization Support, page 58 DS2655 FPGA Base Module (SCALEXIO only), page 45 31

32 Hardware-in-the-Loop Simulation / Use Cases Use Cases Simulating Brushless DC Motors at Electric Power Level Task Drives with brushless direct current (BLDC) motors are popular because they are simple and robust. Often they are operated without any position sensors like Hall sensors and are used in continuously running pumps, electric fuel pumps, selective catalytic reduction (SCR) systems, and so on. Stator windings Stator Permanent magnets Rotor Hall sensors C h B B A h C A B ha C i A i B i C h A h B h C z A+ z B+ z C+ z A- z B- z C- commutation logic U D Challenge During the operation of BLDC motors only two of the three phases are triggered at a time. In the third, untriggered phase the electromotive voltage is induced, which affects the ECU terminals. In sensorless control, the ECU measures this voltage for position detection. Solution A dspace Simulator equipped with a DS5203 board and XSG Electric Component Models enables the complete simulation of BLDC motors at electric power level. No real parts are required. Due to the characteristics of BLDC operation described above, the simulation comprises a current emulation and also a voltage emulation. Above: Sensorless control of a simulated BLDC motor. The triggered phase and freewheeling phase are simulated in current mode, the floating phase is simulated in voltage mode. Below: Detail measurement of the floating phase: Accurate reproduction of the pulsing feedback voltage 32

33 Hardware-in-the-Loop Simulation / Use Cases Testing a Servocontroller at Signal Level Task In this application, the software for a servocontroller can be used for almost all electric motors in a power range from a few watts to several hundred kw. The ECU for the servocontroller contains a wide range of functions that have to work with various configurations of motors, power stages, sensors and bus systems. A wide range of configurations have to be tested. Challenge The test system has to cover a wide range of electric motor power variations. As real parts would necessitate time-consuming modifications to the test system, simulation models are used. They have to be as precise as possible. Solution A DS1005 Processor Board and the DS5202 Electric Motor HIL Solution are used for the real-time simulation. The simulation models use components from the ASM Electric Components Library. The FPGA of the DS5202 handles the time-critical I/O parts for the simulation model, enabling moderate sample rates. The dspace EMH Solution (p. 40) is ideal for this application, as it offers emulation for almost all industrial position sensor systems such as resolvers, TTL encoders, sine encoders and Hall sensors, and also protocol-based sensors such as SSI, EnDat 2.a and Hiperface. L3 L2 L1 Power Supply Network Interface Hiperface EnDat 2.1 Encoder Encoder SSI Encoder ANALOGIN1 ANALOGIN2 START RESET TTL Encoder Resolver U V W Encoder Resolver QUICKSTOP ENABLE L- L+ RB D.C. link Braking resistor Example application: either the servo controller or the electric motor can be simulated with dspace Simulator. 33

34 Hardware-in-the-Loop Simulation / Use Cases Simulating Electric Power Steering Systems at Electric Power Level Task Electric power steering (EPS) systems support the driver during steering. A torque sensor measures the steering movement and sends this data to the EPS ECU, which causes the EPS electric motor to support and enforce the movement. As the EPS electric motor acts directly on the steering rod, the vehicle can be steered even without the driver's interaction. This enables fully automated parking as well as interaction with the electronic stability control (ESC) to support the driver. Challenge The signals of ECUs for EPS often cannot be accessed at signal level. HIL simulation is therefore performed at power level. The ECUs have to be connected to the real motor either at mechanical level or by simulation at electric power level. The electric power level requires real currents and a simulation for the motor. This solution is quite flexible and can be adapted quickly: for example, to simulate different motor types. The simulation can also be combined with HIL simulation for an ESP. Solution In both cases dspace Simulator is equipped with a DS5203 FPGA Board running the XSG electric component s e-motor models. Electronic load modules (p. 48) provide the real current for simulation at electric power level. dspace Automotive Simulation Models for vehicle dynamics are used for simulating the actual physical vehicle characteristics, including the steering system for the EPS and the brake hydraulics for the ESP. Real Sensor Cluster ESP ECU Node #1: ESP HIL ASM Driver, Maneuver Scheduler ASM Traffic CAN Gateway for Sensor Cluster I/O for ESP ASM (e.g. Gasoline, Transmission) ASM Vehicle Steering System MDoF Vehicle Dynamics Node #2: EPS HIL FEPS xrod xrod Fast Tasks i EPS i EPS i EPS TEPS weps Electronic Loads DS5380 Electric Motor Simulation DS5203 with XSG EC models Position Sensor Simulation APU DS5203 EPS ECU eeps VSD (Valve Current Detection) I/O for VSD Brake Hydraulics T Sensor Torque Sensor Simulation Overall integration of EPS electric motor simulation at electric power level, together with an HIL simulation for an ESP. 34

35 Hardware-in-the-Loop Simulation / Use Cases Developing Mechatronic Steering Systems Task The steering system and its characteristics decisively affect a vehicle's driving behavior and feeling. Haptic feedback plays a vital role here, as it gives the driver vital information about the road and the vehicle. It is only when the overall steering system is integrated into the vehicle that developers can actually experience it. As this subjective impression is very important and the integration process is expensive, another approach is needed. Challenge The test system needs to provide haptic feedback and offer a close-to-reality environment simulation for a realistic steering feeling and driving behavior. Its purpose is to give an initial impression at an early stage of product development. The simulation model has to take extra features into account, such as automatic parking and lane-keeping assistants for enhancing comfort and active safety. Driving simulator Optical feedback Steering wheel angle Steering torque Solution dspace combines a HIL steering test bench and a static driving simulator that enables pre-calibration on a virtual experimental vehicle. The test apparatus consists of the real steering system with its actuators, and the driving simulator with a load machine for the steering wheel, the accelerator pedal, and the brake pedal. Both are coupled to the dspace Automotive Simulation Models (ASMs) running on a quad-core DS1006 Processor Board. The ASMs are open Simulink models for the real-time simulation of passenger cars, trucks and trailers that simulate the vehicle's vertical, longitudinal and lateral dynamics as a multibody system with 24 degrees of freedom. EPS test system with HIL Rack force Torque interface Rack position Visualization is done with dspace MotionDesk using small LCD monitors representing the outside and inside rear-view mirrors. The sensation of driving is further intensifi ed by road and engine noises. The system is used not only to investigate and adjust steering systems, but also to run driver assistance systems such as lane departure warning systems. Visualization Vehicle response Vehicle simulation Accelerator and brake pedal positions Functional diagram of the HIL steering test bench driving simulator and the camera HIL. 35

36 Hardware-in-the-Loop Simulation / Use Cases Simulating Automated Manual Transmissions Task Automated manual transmission (AMT) operates similarly to manual transmission, except that it does not require clutch actuation or shifting by the driver. Automatic shifting is controlled electronically (shift-by-wire) and performed by electric motors or hydraulically. In this application, an ECU for an AMT controls three electric direct current (DC) motors, one for the clutch and one each for the longitudinal and the lateral movement of the gear selector level. The ECU chooses the gear according to the motor rotation speed and accelerator pedal position. It activates the clutch and engages the appropriate gear via the shift and the selector motor. It is also possible to shift up and down manually without engaging the clutch. Challenge To test the ECU with real motors as well, the test system allows switching between electric motors as real parts and simulated electric motors by using simulation models. High currents up to 60 A are needed to simulate the inrush current of the DC motors. Solution A dspace hardware-in-the-loop simulator is equipped with a DS2211 HIL I/O Board, which provides various interfaces for connecting the ECU. New ECU variants can easily be adapted to the simulation just by changing the cable harness. Electronic load modules and a DS5203 running XSG Electric Components Library models emulate the electric motors. To test the ECU's behavior during electric failures, high-current failure simulation can be performed. dspace Automotive Simulation Models (ASM) such as ASM Drivetrain are used for simulating the actual physical gearbox characteristics. Hardware-in-the-loop simulator for AMT simulation with 12 electronic loads (page 48). 36

37 Hardware-in-the-Loop Simulation / Use Cases Simulation for Battery Management Systems Task The battery management system (BMS) monitors the electric and thermal state of the batteries used in hybrid or electric vehicles. It takes the drive's requirements and environment impacts into account, and influences each battery and its cells to provide the energy needed and to maintain optimal operation conditions for good performance and long battery life. Challenge Batteries for electric vehicles have extremely high voltages and currents, so the BMS is safety-critical. To ensure safety during HIL simulation, the overall voltage might have to be scaled down. Electrical failure simulation is also needed to make sure the BMS reacts correctly in all circumstances. These are typical failure tests: Broken wires Short circuits Loose contacts Solution The typical HIL simulation setup for battery management system tests comprises a processor board, HIL I/O boards for I/O interfaces, a board for CAN interfaces, and a failure insertion unit for testing electric failures. Restbus simulation is used for simulating unavailable cell stacks. dspace offers specialized hardware and software for testing a BMS, for example, the EV1077 Battery Cell Voltage Emulation Board (p. 52) for simulating high-voltage batteries at cell level and the ASM Multicell Models (p. 55). The test system can virtually represent the electrical and thermal properties of a battery down to cell level. Other components are high-precision voltage sources from 0 to 6 V, which can take the load of the current flowing in cell balancing. Typical requirements for cell voltage simulation are a precision of about 2 mv and a current up to a few hundred ma. The voltage sources are galvanically isolated and can be switched in sequence to form cell modules. The voltage of the entire battery can be simulated this way. Failure simulations such as a break in the measurement cable or the cell connectors (galvanic disconnection of the cell stack) can be run. The voltage sources are connected to the processor board via an LVDS or Ethernet interface, with connection distances of up to 5 m with copper cabling and up to 100 m with optic cabling. All the cell voltages in a battery can be adjusted in less than 1 ms. Relay control Switch (relay) DS1006 High voltage measurement High voltage simulation Isolation monitoring Isolation fault simulation LVDS DS4121 CAN restbus simulation Electric I/O BMS Temperature sensor simulation CAN CE 1... n Cell voltage/temperature simulation Failure simulation Failure simulation I/0 CAN DS2211 DS4302 PHS bus Failure simulation BMS: Battery Management System CE: Cell ECU LVDS: Low Voltage Differential Signaling Example of a HIL simulation setup for a battery management system. 37

38 Hardware-in-the-Loop Simulation / Use Cases Testing ECU Networks of Hybrid Electric Powertrains Task A hybrid electric powertrain typically contains several networked ECUs which the functions are distributed to. The functions, such as overlaid hybrid control functions, can be implemented on separate ECUs or combined with other functions. As these functions need to be extremely reliable, the development and test requirements are high. Challenge To set up a typical realistic hybrid powertrain, two parallel CAN structures have to be built: A powertrain or vehicle CAN and a private hybrid CAN. The powertrain CAN connects the standard ECUs such as the engine ECU and transmission ECU, and others such as the ESP ECU which can also be simulated by their CAN messages. The hybrid-specific ECUs are usually connected to the hybrid CAN. Solution For hardware-in-the-loop simulation, you can connect all existing powertrain ECUs with the HIL simulator. The simulator is equipped with at least one processor board and various interface boards. Powertrain components that are not yet available are emulated via restbus simulation. Testing usually also covers several CAN networks. The modularity of the dspace hardware means that the simulators can be configured for various applications. All known hybrid vehicle versions and ECU or CAN confi gu rations are possible. For integration testing, a hybrid electric powertrain simulator can be extended to simulate a full virtual hybrid electric vehicle by adding further racks to cover all the other ECUs in the vehicle, such as ESP. If the HIL tests cover the simu lation of the electric motor and the battery, safety aspects make it necessary to separate the HIL simulations. The simulator racks are then connected via Gigalink. This high-speed serial data transmission via fiber-optic cable and 1.25G bit/s technology provides very fast information exchange. Transmission Powertrain CAN Engine Hybrid CAN Electric Motor Battery Transmission ECU Engine ECU E-Motor ECU Valve Drawer Load FIU Gigalink Hybrid EPS Powertrain ECU ECU Gigalink Battery Management System Extension by further racks and ECUs to full Virtual Vehicle Node #1 Node #2 Node #3 Node #X 38

39 Hardware-in-the-Loop Simulation / Publications Additional Information You can download success stories, articles and product information on drive applications at under "Downloads". Available Publications (Partial List) Title Author Published at Hardware-in-the-Loop Test of Battery Management Systems Hardware-in-the-Loop Simulation of Electrified Powertrains Leistungselektronikmodelle für Hardware-in-the-Loop-Simulation Hardware-in-the-Loop-Testing of Battery Management Systems Concept of a New Hardware-in-the-Loop Driving Simulator for the Model-Based Design of Mechatronic Steering Systems HIL Simulation of Power Electronics and Electric Drives for Automotive Applications, HIL-Test von Batteriemanagementsystemen HIL-Prüfstand zum Test von Batterie-Management-Systemen Hardware-in-the-Loop-Simulation for electric Drives Real-Time Simulation of Electric Drives by Electronic Load- Emulation Electric Motors: Hardware-in-the- Loop Testing at Full Power Hardware-in-the-Loop: The Technology for Testing Electronic Controls in Automotive Engineering Hardware-in-the-Loop Simulation for Hybrid Electric Vehicles HIL Simulation for Mechatronic Automotive Electronic Control Units: Current applications in vehicle dynamics and electric power steering HIL Simulation for Mechatronic Automotive Electronic Control Units Hardware-in-the-Loop Test Systems for Electric Motors in Advanced Powertrain Applications Test elektrischer Antriebe für Hybridfahrzeuge mittels Hardwarein-the-Loop Simulation Markus Ploeger (dspace GmbH), Joerg Bracker (dspace GmbH), Dr. Hagen Haupt (dspace GmbH) Tino Schulze (dspace GmbH), Matthias Deter (dspace GmbH), Markus Ploeger (dspace GmbH) Dr. Thomas Schulte (HS OWL), Axel Kiffe (HS OWL), Frank Puschmann (dspace GmbH) Dr. Claus Abicht (dspace GmbH, ), Markus Ploeger (dspace GmbH), Joerg Bracker (dspace GmbH), Dr. Hagen Haupt(dSPACE GmbH) Steffen Stauder (University Kaiserslautern), Prof. Dr. Steffen Müller (University Kaiserslautern), Markus Ploeger(dSPACE GmbH), Andre Lehnsmeier (dspace GmbH) Dr. Thomas Schulte (HS OWL), Axel Kiffe(HS OWL), Frank Puschmann (dspace GmbH) Markus Ploeger (dspace GmbH), Joerg Bracker (dspace GmbH), Dr. Hagen Haupt (dspace GmbH) Markus Ploeger (dspace GmbH), Joerg Bracker (dspace GmbH), Dr. Hagen Haupt (dspace GmbH) Dr. Thomas Schulte (dspace GmbH), Frank Puschmann (dspace GmbH), Dr. Harald Wertz (LTi DRiVES GmbH) Dr. Thomas Schulte (dspace GmbH), Jörg Bracker (dspace GmbH) Nils Holthaus (dspace GmbH), Markus Plöger (dspace GmbH), Dr. Thomas Schulte (dspace GmbH) Dr. Peter Wältermann (dspace GmbH) Syed Ali (dspace Inc), Amanjot Dhaliwal (dspace Inc), Shreyas C. Nagaraj (dspace Inc) Andreas Filgerdamm (dspace GmbH), Markus Plöger (dspace GmbH), Dr. Thomas Schulte (dspace GmbH) Andreas Filgerdamm (dspace GmbH), Markus Plöger (dspace GmbH), Dr. Thomas Schulte (dspace GmbH Dr. Thomas Schulte (dspace GmbH), Dr. Herbert Schütte (dspace GmbH), Dr. Andreas Wagener (dspace GmbH), Dr. Peter Wältermann (dspace GmbH) Jürgen Klahold (dspace GmbH), Dr. Thomas Schulte (dspace GmbH), Dr. Andreas Wagener (dspace GmbH) SAE World congress, Apr MTZ - Motortechnische Zeitschrift, Dec 2012 SPS IPC Drives 2012, Nürnberg, Sep 2012 HDT, München, Apr 2012 FKFS Symposium, Stuttgart, Mar th International Symposium on Power Electronics, EE2011, Novi Sad, Oct 2011 Automobil Elektronik, Oct 2011 Electronic Automotive, Jul 2011 SPS/IPC/DRIVES10, Nürnberg, Nov 2010 IFAC Symposium, Munich, Jul 2010 Automobil-Elektronik, Feb th Paderborn Workshop "Designing Mechatronic Systems", Paderborn, Apr 2009 SAE World Congress, Apr 2009 Elektronik automotive, Mar 2009 Elektronik automotive, Mar 2009 SAE World Congress, Apr 2007 HDT, March

40 Hardware-in-the-Loop Simulation / Products dspace Products EMH Solution Electric motor hardware-in-the-loop simulation High-precision digital capturing of 3-phase PWM signals Fast signal measurement and analysis Simulation of various position sensors High timing resolution Various additional multipurpose signals Purpose The EMH (electric motor HIL) solution is based on the DS5202 FPGA Base Board. It gives you all the I/O channels that are needed for HIL simulation of electric motors, such as highprecision digital capturing of 3-phase PWM signals and position sensor simulation, plus several digital and analog I/O channels. It combines the features of the DS5202 PWM and PSS Solutions with many additional multipurpose channels, and enables efficient testing of electronic control units (ECUs) for electric motors on a single I/O board. Use Case A typical use case is a hardware-in-the-loop (HIL) simulation where the electric motor including the electronic power stage is simulated with the dspace modular real-time processing hardware (DS1005/1006). Simulation models like the ASM Electric Components models are used for simulating the electric components. The gate driver signals (typically PWM signals) coming from a controller are measured by the DS5202 EMH Solution, and calculated motor current signals are sent back to the controller by means of analog voltage signals, which can also be provided by the DS5202 EMH Solution. In addition, the DS5202 EMH Solution provides the necessary position sensor signals for the ECU. Applications The DS5202 EMH Solution combined with a simulation model allows you to measure the signals of up to 2 electric machines with up to 8 power switch control signals each, such as IGBTs (insulated-gate bipolar transistors). There is a choice of three operating modes: software polling, external interrupt source, and internal pulse center interrupt source, meaning clock generation based on the measured PWM signals, which is the most suitable way of avoiding beat effects. The current feedback signals for the ECU can be simulated by using the board s fast analog output channels. For position sensor simulation, the board is equipped with four independent angular processing units (APUs) that receive the angular velocity from the model and calculate the position signal. Since each APU s sample time is 12.5 ns, it provides a position signal with a high timing and angular resolution. 40

41 Hardware-in-the-Loop Simulation / Products PHS bus DS1005/DS1006 DS5202 Gate control signals I/O module Current signals ECU Hall/encoder RTI Blockset Resolver PWM Measurement Features Measuring duty cycles and periods of up to 16 PWM signals Separate access by groups of 8 channels via Simulink interface; capture mode adjustable independently for each group Evaluating dead time between adjacent channels Generating interrupts and triggers to external devices at the center position of the PWM period Oversampling and downsampling for interrupt generation Using an external trigger as a latch source for the time measurement and interrupt for the real-time model 8 channels can alternatively be used as general purpose inputs (digital or PWM inputs) High timing resolution Adjustable debounce time Position Sensor Simulation Key Features 1 digital and 1 analog sensor (three signals each, one protocol-based sensor) and 3 independent angle-based digital signals can be simulated in parallel Different sensor types are possible: Analog type: resolver, sinus encoder, user-defined waveform Digital type: TTL encoder, Hall position sensors, user-defi ned waveform Protocol type: SSI, Hiperface or EnDat sensor simulation All sensor simulation groups can be arbitrarily allocated to up to four independent simulated shafts (APUs) Resolver failure simulation (please refer to PSS solution) APU-angle-dependent arbitrary waveform generation High precision and high timing resolution Onboard signal conditioning Full differential I/O for all analog signals 41

42 Hardware-in-the-Loop Simulation / Products Analog Sensors Resolver: Number of pole pairs Output mode Direct (single-ended) Direct (differential) Transformer (standard) Amplitude of excitation signal Transformation ratio Sinus encoder: Lines per revolution Output mode Direct (single-ended) Direct (differential) Amplitude of output voltage signal DC offset of output signal Default or user-defined index signal Protocol Sensors SSI Encoder: Data frame length Monoflop time Bit code Type of parity bit Number of multiturn bits Definition of data frame Hiperface Encoder: Baud rate Type of parity bit Number of multiturn bits Driver active time Pause time Bus dead time Enable parameter channel Definition of data frame User-defined waveform: Separate waveform definition for each of the three analog channels Waveform repetition per revolution DC offset Digital Sensors TTL encoder: Lines per revolution Default or user-defined index signal EnDat Encoder: Data frame length Recovery time Calculation time Bit code Number of multiturn bits Enable memory access Memory access time Definition of data frame Hall encoder: Number of pole pairs Start and end positions of pulse for all three digital channels separately User-defined waveform: Waveforms per revolution Waveform definition for each of the three digital channels separately 42

43 Hardware-in-the-Loop Simulation / Products Multipurpose Channels Features 7 analog output channels 6 analog input channels (2 channels shared with resolver simulation) 10 digital outputs 3 digital and PWM outputs (shared with position sensor simulation) 1 digital input; 16 digital/pwm input (freely accessible to each of the 16 center aligned input channels) 1 RS485 interface for LTi ServoOne TWINsync interface In addition to the EMH Solution, a PWM Measurement and a Position Sensor Simulation Solution are also available. Other dspace products for hardware-in-the-loop simulation: DS5203 FPGA Board (p. 26) FPGA programmable per application Completely user-programmable via RTI FPGA Programming Blockset Utilizes the Xilinx System Generator (XSG) Simulink Blockset Offline simulation in Simulink Basic set of I/O drivers on board, I/O extendable by piggyback modules 43

44 Hardware-in-the-Loop Simulation / Products Technical Data for PWM Measurement Parameter Timing resolution Frequency range Digital inputs Specification 12,5 ns Max. frequency range 39 Hz... 2 MHz 16 channels: maximum input voltage +5 V, overvoltage protection ±50 V Technical Data for Position Sensor Simulation Parameter Timing resolution Angular precision (APU) Angular precision (resolver) Delay for resolver feedback signals Resolution of analog input signals Resolution of analog output signals Analog input range Analog output range Specification 12,5 ns digital / 100 ns analog 32 bits 0.1 (depending on the settings) Min. 1.6 µs (adjustable up to 400 µs) 14 bits (10 MSPS) 12 bits (10 MSPS), based on the user-specified voltage range (10 MSPS) ± 30 V differential, ± 15 V single-ended, overvoltage protection ±50 V ± 20 V differential, ± 10 V single-ended, overvoltage protection ±50 V Digital output Encoder 5 V, TTL, max. 40 ma, overvoltage protection ±50 V Independent anglebased signals 5 V, TTL, max. 40 ma, overvoltage protection ±50 V Technical Data for Multipurpose Channels Parameter Analog outputs Analog inputs Digital outputs Digital inputs RS485 Specification 6 channels: ± 10 V, 12-bit (10 MSPS), overvoltage protection ±50 V 1 channel: V, 14-bit (1 MSPS), overvoltage protection ±50 V 3 channels: inpult voltage range (±30 V ), differential, 14-bit (10 MSPS), overvoltage protection ±50 V 1 channel: inpult voltage range (±30 V ), differential, 16-bit (1 MSPS), overvoltage protection ±50 V 13 channels: 5 V, TTL, max. 40 ma, overvoltage protection ±50 V, (3 channels shared with position sensor simulation) 1 channel: inpult voltage range, threshold adjustable from 1V to 8.5V, overvoltage protection ±50 V 16 channels: inpult voltage range, up to 80MHz PWM measurement, threshold adjustable from 1V to 8.5V, overvoltage protection ±50 V SSI sensor simulation Optionally: Hiperface or EnDat sensor simulation LTi ServoOne TWINsync Interface (to control a ServoOne inverter by dspace real-time hardware) PWM control Torque control Speed control Position control Scope of Delivery FPGA Base Board I/O piggyback module for DS5202 EMH applications Slot module for DS5202 EMH applications FPGA firmware for DS5202 EMH applications Simulink interface for DS5202 EMH applications 44

45 Hardware-in-the-Loop Simulation / Products DS2655 FPGA Base Module I/O board with user-programmable FPGA for use in SCALEXIO Systems Highlights User-progammable FPGA Flexible board for special I/O solutions Up to 5 piggyback modules for I/O can be added Application Area The DS2655 FPGA Base Module has been designed for SCALEXIO HIL applications that require very fast, highresolution signal processing, for example: Hybrid vehicle applications Electric drive applications Wind energy converters Processor-based electric motor simulation FPGA-based electric motor simulation Key Benefits The DS2655 includes a powerful, freely programmable fieldprogrammable gate array (FPGA), the Xilinx Kintex T. To include I/O channels, you connect up to five I/O modules to the board. Failure simulation can be added for each I/O module by using an additional FIU module. Programming the FPGA Programs for the DS2655 FPGA Base Module's FPGA are programmed with the RTI FPGA Programming Blockset. These programs are downloaded to the FPGA via dspace ConfigurationDesk. You can test the program in offline simulation before implementing it on the real-time hardware. This enables you to react flexibly to new requirements, such as new interfaces or having to accelerate the execution of submodels. The DS2655 works together closely with application-specifi c XSG model libraries (p. 56/57). DS2655M1 I/O Module The DS2655M1 is a piggyback module for the DS2655 FPGA Base Module. It contains the digital and analog I/O channels needed for electric drives applications. 45

46 Hardware-in-the-Loop Simulation / Products Technical Details DS2655 Parameter General FPGA Connector for I/O modules 5 Device timing Internal communication interface Specification User-programmable FPGA Xilinx Kintex T Logic cells: (DSP slices: 600) Distributed RAM: 480 kbits Block RAM: kbits 125 MHz IOCNET Physical characteristics Physical size 205 x 100 x 20 mm (8.1 x 3.9 x 0.8 in) Power supply 24 V Technical Details DS2655 M1 Parameter Digital I/O Input Output Specification 10 channels, usable as input or output Maximum input voltage: 15 V Threshold for each channel adjustable from 1 V to 7.5 V Push-pull drivers One output voltage can be selected for all channels: 3.3 V or 5 V Analog I/O Input 5 channels Resolution 14 bit Sampling rate 4 MSPS SAR Input voltage range selectable for each channel: ±5 V or ±30 V Output 5 channels Resolution 14 bit Update rate 7,8125 MSPS Output voltage range: ±10 V Sensor supply Adjustable Output voltage range: 2 V to 20 V Physical characteristics Physical size 208 x 100 x 18 mm (8.2 x 3.9 x 0.7 in) Power supply 24 V Relevant Software Software Required Optional RTI FPGA Programming Blockset FPGA Interface ConfigurationDesk Implementation Version (SCALEXIO) Xilinx ISE Design Suite WebPACK and System Generator for DSP (only for blockset) or Xilinx ISE Design Suite DSP / System Edition RTI FPGA Programming Blockset Handcode Interface XSG Electric Components Library XSG Utils Library 46

47 Hardware-in-the-Loop Simulation / Products Programmable Generic Interface (PGI1) Generic interface box to connect sensors and actuators to dspace rapid prototyping systems via diverse serial interfaces and protocols Emulation of sensor signals (e.g. of yaw rate or crash sensors) with hardware-in-the-loop (HIL) simulation Hardware Details Highly flexible adaptation to customer requirements via piggyback modules and programmable FPGA Decentralized connection of sensors and actuators to dspace systems via 250 Mbit/s LVDS interface Software Details Configurable via Simulink Blockset Available Solutions SPI Master Solution (supports up to 4 masters) and SPI Slave Solution (supports up to 16 slaves), including Simulink blocksets I²C Master/Slave Solution (supports up to 16 masters and 256 slaves), including Simulink blockset Interface for direct coupling of dspace real-time platforms with LTi Drives servo controllers (LTi TWINsync protocol) including Simulink blockset PSI5 Master/Slave Solution (supports up to 4 masters and 10 slaves), including Simulink blockset SPI and I 2 C Master Solution In battery management systems, the sensors for cell voltage, current and temperature are often connected to the microcontroller via SPI or I 2 C. These buses are typically used in the electronic control unit (ECU), or the sensors are installed close to the ECU. Emulating sensors in a hardware-in-theloop (HIL) system requires a decentralized, flexible solution that can be installed in a HIL system near the ECU for easy modeling of sensor-specific interfaces. The Programmable Generic Interface (PGI) from dspace is an ideal FPGA-based platform for emulating interfaces such as SPI or I 2 C Slave and substituting the real sensors. If a project requires signal conditioning, this can be implemented by an integrated plug-on module. The resulting decentralized I/O interface can be galvanically isolated and connected to a HIL simulator or MicroAutoBox up to 5 m away by LVDS, and can be addressed from Simulink. LTi TWINsync Solution The LTi ServoOne is a high-performance motor controller for precise, dynamic movement in a wide variety of linear and rotary motor systems. Two ServoOnes are synchronized via the LTi TWINsync protocol so that the TWINsync master can set the rotor position, speed, torque or duty cycles of the TWINsync slave. The PGI LTi TWINsync Solution is an interface for directly coupling dspace real-time platforms with LTi Drives servo controllers (LTi TWINsync protocol). A Simulink blockset is used to make the global interface settings. Typical use cases are: In-vehicle prototyping Electric motor HIL simulation on mechanical benches with linear and rotary motion load motors, e.g., for electric power steering 47

48 Hardware-in-the-Loop Simulation / Products Electronic Load Modules DS5380 Electronic Load Module (Voltage Range: 30 V) Electronic Load Module For the hardware-in-the-loop emulation of electrical machines, such as motors or generators, dspace offers the DS5380 Electronic Load Module. The module is optimized for high-speed operation as required for emulating electric motors such as those in electric steering systems. It can work as both a current sink and a current source to provide bidirectional current flow, i.e., it generates or consumes real current on ECU motor outputs. The DS5380 Electronic Load Module can be combined with the DS5203 FPGA Base Board (p. 26) the DS2655 FPGA Base Module (p.45) and the XSG Electric Component Library (p. 56) to provide the fast reaction times required for controlling electrical machines. The FPGA board computes parts of the simulation model for the electrical machine, e.g., from XSG Electric Components Library, and operates the Electronic Load Module. Key Features Current sink and source capability High-speed current regulation ideal for loading PWM power stages Simulation of current ripple Different types of electric motors Remote-controlled by standard 10 V analog signals Technical Details The DS5380 Electronic Load Module contains two independent load channels which can control a unipolar current through its output stage. It provides high-speed current regulation of less than 5 µs. Each module can provide continuous current of 30 A and 300 W, and a maximum voltage of 30 V can be applied. Several modules can be connected in parallel to increase the current. Cooling is performed by a temperature-controlled fan. The modules are protected against overload and overtemperature. The control options can be configured via jumper switches. The two channels can be used as a sink and source device (bidirectional current) or they can be switched in parallel to double the maximum current if a current flow in only one direction is required. This expands the range of applications beyond electric motor simulation. For further information, please refer to the DS5203 FPGA Board (p. 26), the DS2655 FPGA Base Module (p.45) and the XSG Electric Component Library (p. 56). 48

49 Hardware-in-the-Loop Simulation / Products DS5381 Electronic Load Module (Voltage Range: up to 60 V) The new dspace DS5381 Electronic Load Module emulates motor and generator currents at voltages of up to 60 V for the hardware-in-the-loop (HIL) simulation of electric motors. Highly dynamic switching between the current and voltage control modes enables emulation of floating brushless DC (BLDC) motor phases without additional booster components. With a voltage range up to 60 V, the module is also ideal for use with higher in-vehicle voltages of 42 and 48 volts or with numerous electric components running in parallel. The module is perfect for emulating three-phase electric motor units. Energy recuperation is also included to boost the energy efficiency of the overall system. Typical test application areas are electrically supported steering, starter and generator systems, and mild hybrid drives. Several loads can be operated in parallel to achieve higher electric currents. Power Recovery The DS5381 Electronic Load Module uses the same supply voltage as the device under test (the ECU). It is equipped with bidirectional working voltage regulators so that the current that is sunk on one pin can be sourced to another pin. This means that the effective power which is simulated on the load pins can be much higher than the power consumption of the electronic load. Operation Modes The DS5381 Electronic Load Module can be operated in three diffe rent modes: Current control mode (typically for motor appli cations, not BLDC motors) Voltage control mode (general-purpose applications) Mixed current and voltage control mode (BLDC motor applications) The mode parameter can be set from the host via protocol. For further information, please refer to the DS2655 FPGA Base Module (p. 45). 49

50 Hardware-in-the-Loop Simulation / Products DC-link+ ECU + FPGA + MOSFET driver E-load type-b Microcontroller + MOSFET driver Battery DC-link- - dspace host system ECU diagnostics DS5831 Electronic Load Module Control supply voltage Control supply current consumption Load supply voltage (battery voltage) for operation Load supply current consumption Internal DC-link voltage Motor phase current Maximum power output (three-phase motor simulation) Dimensions Weight 12 V 3.8 A (includes fans at full speed) V Approx. 10 A (48 V battery voltage / 50 Arms motor phase current on 3 phases) V 0-50 Arms continuous, 100 A peak 3300 W 483 x 88 x 645 mm (19.0 x 3.5 x 25.4 in) 15.5 kg 50

51 Hardware-in-the-Loop Simulation / Products Electronic Load Module (Voltage Range: up to 800 V) Electronic Load Module for Simulating an Electric Motor at Power Level at High Voltages If the HIL tests for an electric drive system have to include the power stages, testing at signal level is not enough. Testing at electrical power level is required. One way is to operate a real drive motor on a test bench. Another is to simulate the electric motor at the electrical power level. This involves simulating the electrical behavior of a real motor by mapping the real terminal voltages and currents and feeding them to the ECU. Compared with a mechanical drive test bench, a purely electrical test bench of this kind is easier and safer to operate. Tests can be run at a very early stage, even if the real drive motor is not yet available. Moreover, it is also possible to simulate different motor types. Unlike mechanical test benches, these simulators have no restrictions on dynamic processes. The electronic load emulator covers voltages of more than 800 V and power outputs of up to 100 kw. Thus, it is suitable for the HIL simulation of current and future electrical drive systems. How the Electronic Load Emulator Works The electronic load emulator emulates the variable, active parts of the voltages u EMK induced in the motor coils, while the inductive behavior of the motor coils is represented by equivalent substitute inductivities L Motor. The induced voltages u EMK are calculated in real time by an electric motor model and implemented by the electronic load simulator. How the Electronic Load Emulator is Implemented The load emulator uses inverters from the ServoOne series by LTi. The electric motor model for computing the induced voltages is implemented on a dspace real-time system by means of Simulink. The model components that can be simulated include the drivetrain. Various sensor and actuator simu lations are added to the real-time system for this, according to project-specific requirements. A hybrid ECU requires at least one appropriate simulation of an engine speed sensor (such as a resolver). Applications The concept of the electronic load emulator can be used for simulating all types of motors. The physical properties of each motor, such as motor inductivity, torque generation and power consumption, are represented very realistically. For variable inductivities (such as in an interior permanent magnet or IPM motor, or with saturation effects), mean values have to be used in the load emulator due to the constant substitute inductivities. Nevertheless, correct representation of the torque and the power is possible. Any desired hybrid and electrical vehicle configuration can be simulated by using different electric motor models in conjunction with variable drivetrain models (for example, Automotive Simulation Models from dspace). The concept is also suitable for various industrial HIL applications. 51

52 Hardware-in-the-Loop Simulation / Products EV1077 Battery Cell Voltage Emulation Board For the HIL test for battery management systems (BMS), high-voltage batteries have to be simulated at cell level. To make this possible dspace provides the high-precision EV1077 Battery Cell Voltage Emulation Board. The EV1077 battery cell voltage emulation board emulates a controllable, highly precise terminal voltage for single battery cells. Emulation Electronics Setup Cell voltage emulation is performed with several EV1077s. The number of these controllable buffer amplifi er boards is configured to match the battery type. The boards supply an adjustable voltage in the range 0 to 6 volt. This relatively wide range means that damaged cells can be emulated. For example, a short-circuited cell can be emulated by outputting 0 V, and a voltage higher than the nominal voltage simulates a cell's increased internal resistance during charging. The voltage is output with a precision of ±1.5 mv across the entire working temperature range. The voltage is galvanically isolated, allowing the modules to be connected in series up to a voltage of 800 V. A reference value step is corrected completely in less than 500 µs. Fast data transmission means that a change to all the cell's voltages takes less than 1 ms. The maximum current that can be supplied or sunk is 1 A, which is sufficient for the usual balancing currents. For special requirements, up to four modules can be connected in parallel to quadruple the maximum current. dspace's Automotive Simulation Models are ideal battery simulation models. Technical Data EV1077 Battery Cell Voltage Emulation Board 1) Hardware structure Output voltage 32 cells per 19" 3-HE module V Resolution 120 µv Precision (across working temperature range) ±1.5 mv Working temperature (environment) C Maximum current (sink/source) Isolation Connection Maximum update rate for all cells Fault simulation 1) Technical modifications possible. 1 A, switchable in parallel 60 V between the cells of a module 1000 V between cell and environment Ethernet, e.g., as interface to SCALEXIO 1 khz Broken wire between ECU and battery Broken wire between cells (cell connectors) 52

53 Hardware-in-the-Loop Simulation / Products Electric Components Simulation Real-time models for simulating vehicle electronics and traction systems Simulation Packages and Models Electric components including drives and batteries FPGA-based plant models Parameterization of vehicle electric systems, drives, and further electric components. Use Cases: Battery Management Task Developing and testing battery management functions. Challenge To simulate multicell battery packs with serial cell connection for voltage increase and parallel cell connection for current increase. Solution The multicell battery model in the ASM Electric Components Library consists of a serial connection of up to 500 individual cells. Several instances of this model can be connected in parallel. For cell balancing purposes, each of the parallel circuits can be separated from the others. 53

54 Hardware-in-the-Loop Simulation / Products ASM Electric Components Model Simulating automotive electrical systems and electric drives Main Model Components Battery Multicell battery Starter Alternator Loads Electric motors (DC, BLDC, PMSM, asynchronous AC induction motor) Controllers Various auxiliary blocks Three-level inverter Features at a Glance Ready-to-use components with automotive features Prepared for testing battery management controllers Simulated battery voltage as set point for HIL power supply Simulation of electric drive components and power electronics in a closed loop with ECU Simulation of hybrid powertrain together with ASM engine simulation models Variable sample times for pulse width modulated (PWM)- synchronous calculation Simulation of a complete automotive electrical system Graphical parameterization in ModelDesk NEW: PMSM machines with current-dependent inductances NEW: Parallel connectivity of battery modules Simulation Model Characteristics ASM Electric Components provides models for the realtime simulation of a vehicle s electrical system. Applications can range from electric drives and inverters for closed-loop simulation with an electric drive controller to a complete automotive electrical system including the battery, starter, alternator, and loads. Typical use cases are the simulation of realistic battery behavior during starter activation, electric drives that are integrated into a hybrid electrical vehicle (HEV) power train, etc. The ASM Electric Components Model consists of automotive electrical system simulation components and closed-loop simulation components. The former can be used directly to create the electric circuits of an automotive system, since they already have all the necessary automotive features. These models are also optimized for real-time HIL simulation. The closed-loop components are ideal for the HIL simulation of electric devices such as drives or inverters in a closed control loop. The models offer variable sample times for pulse width modulated (PWM)-synchronous model calculation and optimized solvers for real-time simulation. ASM Electric Components can be combined with other ASM products such as the engine models and the vehicle dynamics model. More detailed information available Product Brochure: ASM Electric Components Model Electric Motor Transmission AC/DC 3-Phase Power Converter Battery Schematic of a basic electrical system. 54 Schematic of a hybrid powertrain system.

55 Hardware-in-the-Loop Simulation / Products ASM Multi Cell Model To simulate high voltage batteries like Li-ion batteries consisting of series of multiple battery cells, the ASM Electric Component Model features a cell simulation model. The ASM cell model consists of a cell voltage model and a charge state model. With the cell voltage model, individual physical effects such as internal resistance, diffusion and double-layer capacity can be parameterized. The charge state model deals with the cell s charge and discharge currents, and also with leakage currents such as those caused by gassing effects in the charging of NiMH cells. Reference and Delta Models The approach used in ASM is to connect single cells of identical design to create a series string of cells. This consists of a reference cell model that describes the basic behavior of the cell type used, and a delta model that computes the deviation of each individual cell s voltage from the reference voltage. The capacity, initial charge state and deviation from the reference value of the internal resistance can be specified for each cell. Components and Characteristics Real-time capable simulation of multiple battery cells Complexity of the model independent of the number of cells Parameterization for Li-Ion, NiMH, Pb, etc. Individual physical effects such as internal resistance, diffusion and double-layer capacity Supports charge, discharge, and leakage currents Online and offline simulation Supports dspace s cell voltage emulation hardware Graphical parameterization in ModelDesk Supports simulation of serial and parallel connected battery modules Terminal current of battery Reference cell model provides reference terminal voltage Terminal voltage of reference cell ECU battery management EV1077 V cell I bal C1 C2 Balancing current Delta model for calculating deviations in cell voltage based on individual parameters Reference resistance and charge state Voltage differences of cells Terminal voltages of cells CAN bus EV1077 C3 C4 C5 The ASM cell model consists of a reference cell model, and a delta model that computes the deviation of each individual cell s voltage from the reference voltage. Cell module C6 C7 C8 ASM Multi Cell Model Vehicle ECUs Simulator Cell voltage emulation with high-precision voltage amplifi ers (EV1077, p. 52) controlled by the ASM Multicell Model. 55

56 Hardware-in-the-Loop Simulation / Products XSG Electric Component Library Plant Models for FPGA-based Simulations Simulation of extremely time-critical applications Support of very high oversampling rates Direct I/O access Open models Application Examples Some electric motor control applications demand outstanding precision and correspondingly high sample rates that only very fast computing based on field-programmable gate array (FPGA) boards can provide. The XSG Electric Component Models are plant models that perform at very high speed on a dspace DS5203 or DS2655 FPGA Board to support such applications. Characteristics of the XSG Model The ASM Electric Component Models (closed-loop simulation components) are implemented as open Xilinx System Generator (XSG) models that run on a dspace DS5203 or DS2655 FPGA Board. Closed-loop simulations of electric devices and their controls are supported at very high sample rates in real time. In addition to the plant models, the XSG Electric Component Library is supplemented by enhanced I/O functions on the DS5203 or DS2655 FPGA Board and its I/O modules, e.g., for timing analysis and capturing digital input sources. The XSG Electric Components Library and the DS5203 or DS2655 FPGA Board can be used together for e-motor simulation both on signal and on power level. In comparison to processor-based models, the measurable latency between the hardware input and the hardware output usually decreases from 50 µs to approx. 1 µs. Components and Characteristics Permanent magnet synchronous motor (PMSM) Brushless DC motor (BLDC) Three-phase inverter Resolver and sine, TTL, and Hall encoders NEW: Highly Nonlinear Electric Motor Models Inductance and flux depending on stator current Spatial harmonics Continuous integrated parameterization workflow from FEA tool JMAG -RT to FPGA model Available on request Key Benefits High precision and stability Very high oversampling rate in relation to the PWM switching frequency No PWM synchronization necessary Current ripple (PWM effects) can be simulated Better precision in simulating higher fundamental frequencies 56 Open models can be modified or partly replaced Ideal for testing ECUs with variable PWM switching frequencies Run-time license available

57 Hardware-in-the-Loop Simulation / Products XSG Utils Library Ready-to-use function blocks for speeding up the implementation of FGPA models Completely open models for Simulink and XSG Real-time FPGA programming Wide range of function blocks Application Areas The XSG Utils Library offers users of real-time FPGAs a high number of enhanced function blocks to implement their own projects. It can be applied in rapid control prototyping projects using dspace MicroAutoBox II or hardware-in-theloop simulation with the DS5203 FPGA Board. Key Benefits The open XSG Utils Library contains essential, often needed function blocks, similar to standard Simulink functions. The high-quality function blocks are ready-to-use, easy to adapt to your project, and therefore greatly facilitate your FPGA programming. They range from enhanced I/O, scope, and look-up table functions to an average calculator, sine generator and wavetable encoder. The XSG Utils Library functions are a subset of the functions included in the XSG Electric Component Models, so users can pick just the function set needed for their application. Available Function Blocks The main functions of the XSG Utils Library are: Scope: Captures 8 (out of a selection of 16) high frequency signals within the FPGA clock rate and sends the captured data synchronously to the processor, where the data can be displayed and stored in implemented ControlDesk XY plotters, for example. PWM measurement: Measures the dead time (between HSD and LSD), high time and period time of a singleor three-phase signal. PWM Generator: Generates a pulse-center-aligned PWM signal (single-phase and three-phase). The dead time and the duty cycle can be set on the processor side (online tunable). Look-up table: Configures the accuracy of the normed table, the minimum and maximum data value which will be covered and lets the amount of bits be calculated automatically. Linear interpolation algorithms or the Use Input Below method can be configured online. 1-D, 2-D and 3-D look-up tables are available. Multiscale DAC: Enables flexible programming and run-time parameterization of the onboard FPGA I/O as well as the stimulus modus. Further function blocks include: Average Calculator Sine Generator Discrete PT1 Scaling Wavetable Encoder APU Small Apps Version Info 57

58 Hardware-in-the-Loop Simulation / Products JMAG-RT Parameterization Support for ASM and XSG Electric Components Library Application JMAG is a graphical development tool for electromechanical design that can be used to define the key characteristics of electric motors. JMAG-RT now supports both the ASM Electric Component model and the XSG EC FPGA-based models. With its new export feature, the detailed characteristics of an electric motor can be exported in ASM parameter files to parameterize the ASM electric motor models. Together with the XSG EC models the simulation of non-linear FPGA-based spatial harmonics FPGA-based motor models is possible. Features Graphical definition of motor characteristics ASM-compatible export of motor characteristics Easy and precise parameterization of the ASM electric motor models Parameterization and calculation of FPGA-based spatial harmonic motor models 58

59 Hardware-in-the-Loop Simulation / Products Workflow for ASM Electric Components Models 1. Definition of e-motor characteristics 2. Export to ASM Electric Components 3. Definition of e-motor characteristics The real-time-capable ASM Electric Components are parameterized via JMAG-RT data export to simulate the dynamic behavior represented in the detailed JMAG models. 59