IC Trimming Circuit Using Statistical Parameters to Cut Fuse Usage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional trimming techniques for integrated circuits (ICs) require excessive fuse usage for recording measurements, leading to increased manufacturing costs and memory usage, especially as the precision of IC characteristics increases.
Innovation Solution
A trimming method and system that calculates statistical parameters based on output values from a parametric model function and stores only the optimal trim settings in memory, reducing the need for extensive data storage by using a characteristic outputting circuit, data memory, and trim memory to determine and write optimal trim settings across various operating conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional trimming techniques are used to record precise characteristics of each IC, then measurement precision is improved, but fuse usage increases leading to higher manufacturing cost
Solution Approach 1:
The patent extracts only the essential statistical parameters (mean and standard deviation) from the complete set of measurement data under multiple operating conditions. Instead of recording all raw measurement values which would require extensive fuse usage, only the critical statistical characteristics are stored in the data memory, significantly reducing fuse consumption while preserving the ability to determine optimal trim settings.
Solution Approach 2:
The patent transforms the raw measurement data into statistical parameters (mean and standard deviation) that capture the essential characteristics of IC performance across operating conditions. This parameter transformation reduces the data volume stored in fuse while maintaining measurement precision for trim optimization.
2Measurement precision
If more fuse usage is allocated for recording measurements, then measurement precision is improved, but manufacturing cost increases
Solution Approach 1:
The patent extracts only the essential statistical parameters (mean and standard deviation) from the complete set of measurement data under multiple operating conditions. Instead of recording all raw measurement values which would require extensive fuse usage, only the critical statistical characteristics are stored in the data memory, significantly reducing fuse consumption while preserving the ability to determine optimal trim settings.
Solution Approach 2:
The patent transforms the raw measurement data into statistical parameters (mean and standard deviation) that capture the essential characteristics of IC performance across operating conditions. This parameter transformation reduces the data volume stored in fuse while maintaining measurement precision for trim optimization.
3Reliability
If complete measurement data is stored for all operating conditions, then reliability is improved, but memory usage increases
Solution Approach 1:
The patent extracts only the essential statistical parameters (mean and standard deviation) from the complete set of measurement data under multiple operating conditions. Instead of recording all raw measurement values which would require extensive fuse usage, only the critical statistical characteristics are stored in the data memory, significantly reducing fuse consumption while preserving the ability to determine optimal trim settings.
Solution Approach 2:
The patent introduces statistical parameters (mean and standard deviation) as intermediary representations of the complete measurement data. These statistical parameters serve as mediators that capture the essential information needed for reliable trim setting determination while occupying minimal memory space, thus bridging the gap between data completeness and memory efficiency.
Data Source
AI summary
The disclosure provides a trimming method, a trimming circuitry, and a trimming system for an IC with memory usage reduction. The method is applicable to a system including a tester, a characteristic adjustable circuit, and a trimming circuitry having a characteristic outputting circuit, a data memory, and a trim memory. The method includes the following steps. Under each condition, output signals respectively corresponding to trim settings are received from the characteristic adjustable circuit by the characteristic outputting circuit to obtain output values of the condition, a statistical parameter associated with the output values of the condition is calculated by the tester. The statistical parameter of at least one of the conditions is written into the data memory by the tester. An optimal trim setting of the characteristic adjustable circuit is determined according to the statistical parameters under all the conditions and written into the trim memory by the tester.


