Large-scale MOSFET and interconnect circuit simulation using waveform relaxation and transmission line time step control

2012 ◽  
Author(s):  
Chun-Jung Chen ◽  
Allen Y. Chang
2011 ◽  
Vol 383-390 ◽  
pp. 2031-2037
Author(s):  
Chun Jung Chen ◽  
Chih Jen Lee ◽  
Chang Lung Tsai ◽  
Allen Y. Chang ◽  
Tien Hao Shih

In this paper, we propose a modified Waveform Relaxation algorithm to perform large-scale circuit simulation for MOSFET circuits containing lossy coupled transmission lines that have been encountered in modern circuit design community, in which a full time-domain transmission line calculation algorithm based on the Method of Characteristic is adopted. New software techniques are proposed to enhance the robustness as well as efficiency of the simulation process. All proposed methods have been implemented and executed to justify the claimed advantages.


2012 ◽  
Vol 433-440 ◽  
pp. 2762-2768
Author(s):  
Chun Jung Chen ◽  
Chih Jen Lee ◽  
Chang Lung Tsai ◽  
Tai Ning Yang ◽  
Allen Y. Chang

Relaxation-based algorithms are efficient in simulating large-scale circuits. One of the Relaxation-based algorithms, ITA (Iterated Timing Analysis), has even been widely used in industry. Hence we present accelerating techniques to fasten ITA’s speed. The basic idea is to let ITA utilize local time steps of subcircuits as much as possible. Our techniques include a heuristic approach, a method based on Strength of Signal Flow, and the method combining the previous two techniques. Proposed methods are quite advantageous, which have been justified by real experiments.


Transmission Line model are an important role in the electrical power supply. Modeling of such system remains a challenge for simulations are necessary for designing and controlling modern power systems.In order to analyze the numerical approach for a benchmark collection Comprehensive of some needful real-world examples, which can be utilized to evaluate and compare mathematical approaches for model reduction. The approach is based on retaining the dominant modes of the system and truncation comparatively the less significant once.as the reduced order model has been derived from retaining the dominate modes of the large-scale stable system, the reduction preserves the stability. The strong demerit of the many MOR methods is that, the steady state values of the reduced order model does not match with the higher order systems. This drawback has been try to eliminated through the Different MOR method using sssMOR tools. This makes it possible for a new assessment of the error system Offered that the Observability Gramian of the original system has as soon as been thought about, an H∞ and H2 error bound can be calculated with minimal numerical effort for any minimized model attributable to The reduced order model (ROM) of a large-scale dynamical system is essential to effortlessness the study of the system utilizing approximation Algorithms. The response evaluation is considered in terms of response constraints and graphical assessments. the application of Approximation methods is offered for arising ROM of the large-scale LTI systems which consist of benchmark problems. The time response of approximated system, assessed by the proposed method, is also shown which is excellent matching of the response of original system when compared to the response of other existing approaches .


Sign in / Sign up

Export Citation Format

Share Document