Controlled Toggle Rate for Modular Scan Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modular testing of semiconductor chips faces challenges in managing switching activity levels, which can lead to false rejects or damage due to excessive power consumption during testing, and affects neighboring modules even when they are not being tested, resulting in inaccurate test results.
Innovation Solution
A method and circuitry that control the toggle rate of neighboring modules during testing to mimic normal operational conditions, using toggle generators to manage switching activity and ensure accurate testing of modules-under-test by adjusting the toggle rate to match or differ from normal operational effects on power delivery, thermal coupling, and noise interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If high switching activity is used during testing to achieve high fault coverage, then test effectiveness is improved, but power consumption increases causing false rejects or brown-out
Solution Approach 1:
The patent applies local quality by differentiating the switching activity levels for different modules. The module under test (MUT) operates at high switching activity to achieve high fault coverage, while neighboring modules operate at lower switching activity levels. This localized control allows the MUT to maintain high test effectiveness while neighboring modules consume less power, preventing false rejects and brown-out conditions.
Solution Approach 2:
The patent implements dynamics by making the switching activity level adjustable and controllable for each module. Test controllers can dynamically set different toggle rates for MUTs and neighboring modules based on test requirements. This dynamic control allows optimization of both test effectiveness and power consumption during different phases of the testing process.
2Use of energy by moving object
If neighboring modules are quieted by switching off their scan chains to reduce power consumption, then false rejects are reduced, but test escapes may occur due to overly optimistic test environment
Solution Approach 1:
The patent applies partial action by partially activating neighboring modules during testing. Instead of completely switching off scan chains (which would create an overly optimistic environment), the patent activates neighboring modules at reduced switching activity levels. This partial activation maintains enough activity to prevent test escapes while consuming significantly less power than full activation, thus avoiding false rejects.
3Use of energy by moving object
If test frequency is lowered to reduce switching activity, then power consumption is reduced, but timing defect detection capability is inhibited
Solution Approach 1:
The patent applies local quality by allowing different modules to operate at different test frequencies. The MUT can operate at high test frequencies to maintain timing defect detection capability, while neighboring modules operate at lower frequencies to reduce power consumption. This localized frequency control resolves the contradiction between power consumption and timing defect detection.
4Productivity
If modular testing is implemented to reduce test application time, then productivity is improved, but managing switching activity in neighboring modules becomes complex
Solution Approach 1:
The patent applies segmentation by dividing the test control into independent module-level controllers. Each module has its own test controller that can independently manage switching activity, scan chain operations, and test pattern application. This segmentation enables parallel testing of multiple modules (improving productivity) while simplifying the overall control architecture, as each controller manages only its local module rather than the entire system.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method is provided to test a modular integrated circuit (IC) 100 comprising: testing a module-under-test (MUT) 101B within the IC 100 while causing a controlled toggle rate within a first neighbor module 101A of the MUT 101B; wherein the controlled toggle rate within the first neighbor module 101A is selected so that toggling within the first neighbor module 101A has substantially the same effect upon operation of the MUT 101B as operation of the first neighbor module 101A would have during actual normal functional operation of the first neighbor module 101A.