SNUG Saber 2008 Session Information
June 3 - 4 , 2008 | University of Michigan-Dearborn Fairlane Center

Tuesday, June 3, 2008

Synopsys Solutions

Design and Verification of the In-Vehicle Network Physical Layer (FlexRay, CAN, LIN)
As the complexity of In-Vehicle Networks continues to grow, it is becoming an increasingly difficult task to manually predict the behavior of the network at the physical layer. Simulation-based methods provide a powerful means of designing and optimizing IVN physical layer implementations of FlexRay, CAN, and LIN networks much earlier in the development phase.  Using simulation, it becomes possible to analyze topology tradeoffs and network variants in order to maximize the signal integrity characteristics of the network.  Further, capabilities to account for component and environmental variability in a simulation environment allows the IVN physical layer designer to optimize the system to be robust in the presence of variability.  This session describes an simulation-based approach for analyzing the IVN physical layer, including simulation models, simulation scenarios, and post-processing mechanisms needed to sufficiently evaluate system behavior.

Lee Johnson, Synopsys, Inc.

Integrating Electrical Simulation & Analysis into the Wire Harness Design Process
Demands for improved robustness of vehicle electrical systems are driving the need for tighter integration of electrical simulation & analysis into the wire harness design process. In addition to their application for the functional design of electrical systems, simulation-based methods are also important in the design of the physical interconnections between functions: the wire harness . This session discusses simulation tool capabilities and analysis techniques that enable robust design of wire harnesses in a vehicle electrical system – managing system complexity and variants, modeling approaches, and more.

Lee Johnson, Synopsys, Inc.

Grid Computing for Robust Design
The need for and cost benefits of Robust Design processes like Design-for-Six-Sigma (DFSS) are well established. However, doing so with physical prototypes can be expensive. Moreover, the ability to achieve statistical significance with physical prototypes is often cost prohibitive. However, these two problems are easily overcome using simulation. This session demonstrates how simulation can be used to develop a virtual manufacturing process that allows designers to perform Robust Design reducing or eliminating the need for physical prototypes. This virtual assembly line allows designers to focus on the design parameters that contribute most and least to design variation thereby improving quality and eliminating costly overdesign. The demonstration focuses on the use of a mini High Performance Compute cluster to show how parallelization can provide the performance to assemble the high numbers of virtual builds required for statistical significance.

Jim Patton, Synopsys, Inc.

Platforms for Automotive Applications
Automotive systems are currently experiencing an explosive growth in software complexity, which can be found in various in-car systems, including electronic control systems (ECUs) for engine, braking & stability control, in-car telematics and entertainment, and hybrid and electric vehicle control systems. At the same time, the industry is challenged by the need to reduce design cycle times, higher safety and reliability and emission standards, and cost reduction. Virtual Platforms are accurate software models of automotive hardware, like ECUs, and include processor, bus, on-chip and off-chip component models, complemented with environment models of sensors and systems being controlled. This model based approach allows automotive semiconductor and subsystem suppliers to quantitively evaluate and select architecture candidates and to use this as a specification and collaboration tool with supply chain customers. Both automotive suppliers and integrators leverage the virtual platform for ECU software development, and use their higher visibility, early availability and testing capabilities to respectively lower engineering and bench cost, reduce cycle time and improve quality and reliability.

Filip Thoen, Synopsys, Inc.

VHDL-AMS Trends & Insights / Saber Product Update
As a leading provider of VHDL-AMS modeling and simulation tools, Synopsys will provide an update of trends and insights adoption of VHDL-AMS for mechatronic design and simulation, followed by highlights of new features and capabilities in recent releases of the Saber products.

Andre Jennert, Synopsys, Inc.

 

Saber User Track
User Paper - Model-Based Voltage Compensation of an Engine’s Solenoid Valve
GM Powertrain has various engine applications that contain a solenoid-controlled electrohydraulic system to disable engine valves for Active Fuel Management (Cylinder Deactivation). Timing of the solenoid-controlled system is critical to guarantee the engine valves get enabled/disabled within a required time window.

Response time of the solenoid valve is heavily dependent upon the supplied Voltage. The solenoid response time gets characterized by the solenoid supplier under laboratory conditions, but the electrical architecture of the real-world vehicle differs from the laboratory. This presentation focuses on the modeling/simulation techniques used to create a compensation method for variation caused by the real-world wiring harness and electrical loads across engine operating conditions and part/part variations.

The modeling/simulation work included electrical models of various engine control module (ECM) circuits, dynamic solenoid valve models, wiring harnesses, and sensors/actuators that will be drawing various amounts of current at various engine operating conditions. Monte Carlo analysis was performed at different system Voltage levels. Monte Carlo data was then used to adjust calibrations and ‘mean-shift’ the system response to create an optimal compensation for the differences between the laboratory characterization data and the real-world application of the solenoid valves.

Mike McDonald, GM Powertrain
User Paper - Saber Worst Case Modeling Techniques for Semiconductor Devices
Since semiconductor model parameters are usually defined based on semiconductor device physics and fabrication process, these model parameters are not expressed in the same terms as device ratings found in data sheet specifications. In addition, most semiconductor device model libraries included with the simulation tool or found in the component vendor websites contain only nominal model parameters. Although all the model parameters can be adjusted during a simulation analysis, changing model parameter from a nominal value to an extreme value and matching the model performance with device worst case ratings require a fully understanding of every model parameters. All these factors inhibit engineers to utilize simulation tool with satisfied effectiveness. The modeling technique presented in this presentation takes advantage of some advanced Saber modeling features such as a parameter group feature. With this technique, worst case semiconductor device model parameters are pre-grouped and mapped to common engineering terms. It virtually combines multiple models into a single model and eliminates the needs of swapping symbols and/or retyping model names in the circuit schematic. Design engineers can implement various worst case circuit/module analyses more effectively and efficiently during design validation process.

Guangyu Liu, Visteon Corp
User Paper - Using Saber to Simulate Automotive Conducted Emissions Testing
The following documents a method of using the Saber Circuit Simulator as an EMI receiver. This enables the design engineer to use simulation to predict the LW (Long Wave) and AM Band (150 kHz – 2 MHz) conducted emissions (Automotive EMC Requirement: CISPR 25) of circuits which draw switching currents from the vehicle power supply and/or return such currents to ground. A custom windowing function and Saber macros have been developed that allow simulation results equivalent to those obtained using a CISPR 16 compliant EMI receiver with Peak (PK), Average (AVG), and Quasi-Peak (QP) detector response.

Jeff Gloudemans,Delphi
Michael McDermott, Delphi
User Paper - Using Saber Harness / Bundle in a Global Product Development
This presentation will discuss the different solutions the Volvo Group has chosen to be able to meet the challenges of working in a global PD environment The Volvo Group has Saber Harness users on almost all continents and in a lot of cases design data and workload is shared between the sites The major challenges are to make sure acceptable performance even for remote sites with low bandwidth and/or high latency is being met. To meet the business requirements regarding performance and access to common data several solutions have been successfully implemented.

Jesper Grasjo, Volvo
User Paper - Saber Analysis and Control of a PWM Inverter Supplied by Unbalanced Bus in an Aerospace Application
Saber-based analysis and control method aimed at eliminating dc content in the output currents of a 3-phase pulse-width modulated (PWM) inverter supplied by an unbalanced split-source dc voltage bus and used in an aerospace application is presented. After a thorough examination of the problem using the simulator, from which a particular case associated with the non-zero dc current content has been identified, a control scheme is proposed and demonstrated for both sinusoidal and space vector PWM inverter whereby dc components in the inverter load currents are removed.

Novica A. Losic, Honeywell



Wednesday, June 4, 2008

Saber Robust Design Workshop
Learn how to apply robust design methodology using a virtual prototype Workshop Overview This hands-on Robust Design workshop teaches a practical methodology for implementing a robust design flow using simulation. The class concentrates on methodology and walks you through the steps that lead to quality, cost-effective design. During the course of this workshop, you will learn and apply the following skills:

At the end of this workshop you will be able to:

Audience Profile: This workshop is intended for design engineers or engineering managers who are interested in improving the quality or cost-effectiveness of their products. This class applies to both those with and without Saber experience. The lab work is setup to accommodate new users. Current Saber users will also find the class helpful for taking full advantage of the advanced features of the simulator such as Sensitivity and Monte Carlo analyses.