Browse Prior Art Database

Method and system for executing designed experiments in a circuit simulation environment

IP.com Disclosure Number: IPCOM000007898D
Publication Date: 2002-May-02
Document File: 5 page(s) / 146K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method and system for executing designed experiments in a circuit simulation environment. Benefits included improved development environment and improved product quality.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 50% of the total text.

Method and system for executing designed experiments in a circuit simulation environment

Disclosed is a method and system for executing designed experiments in a circuit simulation environment. Benefits included improved development environment and improved product quality.

Background

              During circuit design, especially large complex circuits, engineers tune many variables that effect performance. The difficulty of this process increases exponentially relative to the number of devices in the circuit due to the complex interactions between devices. Further, with complex circuits, simulation time can be much greater than for simpler circuits, often requiring many hours for a single simulation. Without an automated and scientific process for evaluating the simultaneous effects on many variables, the result is a tradeoff between a large expenditure of simulation time and the risk of a poorly tuned (or even faulty) circuit.

              This type of problem has been addressed in other industries by the use of standard mathematical techniques comprising the body of knowledge called Design of Experiments (DOE). This body of knowledge is a specialty area within the field of statistical methods, which itself is a formal academic area within the science of mathematics. Through these methods, experiments are defined for the specific goal of identifying optimal process settings. Applications range widely with examples in agriculture, manufacturing, and chemical industries.

General description

              Disclosed is the definition of a method and software system for executing designed experiments in a computer-aided design (CAD) tool circuit simulation environment. This system speeds up the evaluation of circuit performance by enabling the use of industry-standard DOE methodology for the efficient simultaneous investigation of multiple circuit variables. The system facilitates the design, generation, and execution of experiments using circuit-simulation software. The results of these experiments indicate whether the circuit is optimally tuned and, if not, how to tune the circuit to improve performance. In addition, the experiments can determine robustness to operating conditions such as temperature and supply voltage.

              The disclosed system incorporates DOE capability in an automated fashion into a production circuit-simulation environment. The disclosed method enables the rapid acquisition of experimental data, which is analyzed to understand the effect of design variables on circuit performance. The data enables creation of statistical response surface models that are used for circuit optimization and for sensitivity analysis.

              The components of the disclosed DOE system include:

§         The Main System Controller. It is used for managing the design, generation, and execution of experiments (see Figure 1). This platform is started by selecting a Start DOE menu item located in a User Function drop-down menu on the Analog Artist main screen. The key features on this screen are: Design Expe...