Bond Wire Parasitic Extraction via 2D EM Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for modeling bond wire parasitics in advanced wire bond packages are either time-consuming, inaccurate, or require extensive computational resources and expertise, making it challenging to predict electrical noise and ensure signal integrity in high-speed, high-density designs.

Innovation Solution

A 2D EM simulation-based method that breaks bond wires into segments, generates LC values for each segment, and constructs multi-segment equivalent circuit models, considering current return paths and curvature, to quickly and accurately predict noise effects, reducing design turnaround time and costs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If full wave 3D EM simulation tools are used to model bond wires, then the accuracy of parasitic inductance and capacitance extraction is improved, but the simulation time increases significantly to 3-4 hours per model

Engineering Contradiction:
Improveaccuracy of parasitic extractionVSAvoidsimulation time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent divides the bond wire structure into multiple 2D cross-sectional segments along its length. Each segment is modeled independently using 2D EM simulation, which is computationally much faster than full 3D simulation. The individual segment results are then combined to obtain the overall parasitic parameters of the complete bond wire, achieving both accuracy and speed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the 3D EM simulation problem into a series of 2D simulation problems by taking cross-sectional slices of the bond wire at different positions. This dimensional reduction allows use of faster 2D solvers while capturing the essential 3D effects through proper stacking and combination of segment results.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If traditional empirical formulas are used to calculate bond wire parasitics, then the computation time is reduced, but the accuracy deteriorates for complex bond wire geometries with curvature and multiple current paths

Engineering Contradiction:
Improvecomputation speedVSAvoidaccuracy of parasitic extraction
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The bond wire is divided into multiple 2D segments, allowing the model to capture geometric variations and curvature effects at each segment level. This segmentation enables accurate representation of complex geometries while maintaining computational efficiency through 2D analysis.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different modeling approaches to different segments based on their local characteristics. Each segment's parasitic parameters are calculated considering its specific geometric features and local current distribution, providing locally accurate results that combine to give overall accuracy for the complete structure.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If full wave 3D EM simulation is used, then the most accurate results are obtained, but the method requires extensive EM knowledge and expertise that package designers typically lack

Engineering Contradiction:
Improveaccuracy of EM modelingVSAvoidease of use
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

By segmenting the 3D problem into 2D slices, the patent reduces the complexity and knowledge barrier. 2D EM simulation is more intuitive and easier to set up than full 3D simulation, making the tool accessible to package designers without requiring deep EM expertise, while still achieving accurate results.

Inventive Principle:
Principle #1Segmentation

4Reliability

If full wave 3D EM simulation is used for analyzing entire wire bond packages, then the most accurate noise prediction is achieved, but the computational resources and time required become impractical

Engineering Contradiction:
Improvenoise prediction accuracyVSAvoiddesign turnaround time
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments both the bond wire geometry and the simulation process into manageable 2D sections. This enables rapid extraction of parasitic parameters for multiple bond wires in a package, making it practical to analyze entire wire bond packages with high I/O density, thereby achieving reliable noise prediction within practical design timelines.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS8312404B2Multi-segments modeling bond wire interconnects with 2D simulations in high speed, high density wire bond packages
Publication Date: 2012.11.13 MARVELL ASIA PTE LTD
  • US8312404B2 patent drawing
  • US8312404B2 patent drawing
  • US8312404B2 patent drawing

AI summary

A method for modeling bond wires in an IC package for predicting noise effects generated by electromagnetic coupling in complex bond wire configurations. A look-up table of equivalent LC circuit models for the bond wires is generated that accurately predicts the effects of the bond wire circuitry of a signal transmission system. Switch and mirror techniques are applied to reduce the bond wire configurations necessary to simulate. The method includes: setting parameters related to the IC package layout of groups of bond wires; sub-dividing each group of bond wires into regions, each including a portion of the bond wire or its corresponding pad, and generating dissection planes for the regions consisting of the bond wires; performing a 3D simulation on the regions consisting of the corresponding pads, and a 2D simulation for each dissection plane; constructing equivalent circuit models for groups of bond wires and corresponding pads based on the 3D and 2D simulations results; inputting the equivalent circuit models into a circuit simulator to measure the noise effects; and modifying the layout geometry to meet noise targets.