IGBT Gate Waveform Monitoring for Failure Mode Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting failures in insulated gate type semiconductor devices, such as IGBTs and MOSFETs, require large circuit scales and have low accuracy due to the need for additional loads and amplifiers, and cannot accurately predict failure modes or classify failures.
Innovation Solution
A state determination device that includes a processor circuit to analyze time waveforms of gate voltage and element temperature to classify the state of the insulating film, using initial, current, and past waveforms to determine the state of the insulating film, reducing circuit scale and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional loads and amplifiers are used to detect failures in semiconductor devices, then detection capability is improved, but circuit scale increases
Solution Approach 1:
The patent extracts only the essential gate voltage waveform information needed for failure detection, eliminating the need for additional loads and amplifiers. By focusing on the gate voltage waveform characteristics that inherently contain failure information, the system achieves accurate failure detection without adding complex circuitry.
Solution Approach 2:
The patent uses waveform storage to create a historical copy of gate voltage waveforms for comparison with current waveforms. This copying mechanism enables failure detection through temporal comparison rather than requiring additional sensing circuitry, thus maintaining detection accuracy while minimizing circuit scale.
2Reliability
If conventional methods are used to detect semiconductor device failures, then detection is possible, but accuracy in predicting failure modes is low
Solution Approach 1:
The patent stores initial gate voltage waveforms as a reference before failures occur. By having this preliminary reference data available, the system can compare current waveforms against the initial state to detect subtle changes that indicate developing failures, enabling more accurate failure mode prediction and classification.
Solution Approach 2:
The patent implements a feedback mechanism by continuously comparing current gate voltage waveforms with stored historical waveforms. This feedback loop allows the system to track waveform evolution over time and identify patterns that indicate specific failure modes, thereby improving both reliability and measurement precision.
3Measurement precision
If more comprehensive waveform analysis is performed to improve state determination accuracy, then measurement precision improves, but processing time increases
Solution Approach 1:
The patent segments the waveform analysis into distinct phases: acquiring gate voltage waveforms, storing them for later comparison, and performing deterministic comparisons with stored waveforms. This segmentation allows for efficient processing by breaking down complex analysis into manageable steps, maintaining high measurement precision while controlling processing time.
Data Source
AI summary
A state determination device includes an initial waveform storage unit, a waveform acquisition unit, an element temperature acquisition unit, a past waveform storage unit, and a failure state determination unit. The failure state determination unit classifies the time waveform based on the initial time waveform of the gate voltage, the current time waveform of the gate voltage, the element temperature, and the past time waveform of the gate voltage, and determines a fault condition of the IGBT.


