Hybrid Defect Detection for Magnetic Disk Drives
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional defect detection mechanisms in magnetic disk drives face challenges in accurately identifying long shallow and short deep defects, leading to a trade-off between the probability of miss and false alarm, which reduces drive yield and capacity, and are ineffective for dynamic defect detection.
Innovation Solution
A hybrid defect detection mechanism using multiple sub-detectors with different sliding window lengths and thresholds is employed to detect various classes of defects, maximizing detection accuracy while minimizing the probability of miss or false alarm.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional defect detection mechanisms are used, then the detection process is simple, but the accuracy of identifying long shallow and short deep defects is poor, leading to high probability of miss or false alarm
Solution Approach 1:
The defect detection mechanism is segmented into multiple sub-detectors, each specialized for detecting specific defect types (long shallow defects, short deep defects, etc.). Each sub-detector applies different parameters such as unique thresholds and sliding window lengths optimized for its target defect class, enabling accurate detection without requiring a single complex universal detector
Solution Approach 2:
The patent introduces additional detection dimensions by applying multiple sets of parameters (different thresholds, window lengths, and detection criteria) to the same signal samples. This multi-dimensional approach allows the system to detect defects that would be invisible or ambiguous under a single parameter set, thereby improving detection accuracy without increasing physical hardware complexity
2Adaptability or versatility
If conventional defect detection with single threshold is used, then the device complexity is low, but it cannot effectively detect various classes of defects simultaneously
Solution Approach 1:
The defect detection mechanism achieves universality by designing multiple sub-detectors that can collectively identify various classes of defects including long shallow defects, short deep defects, and other defect types. Each sub-detector is configured with specific parameters to handle different defect characteristics, making the overall system versatile across multiple defect scenarios
Solution Approach 2:
The patent systematically changes detection parameters across different sub-detectors, including thresholds, sliding window lengths, and detection criteria. These parameter variations allow each sub-detector to be optimized for specific defect types, enabling the system to adapt to and detect a wide range of defect classes without requiring physical reconfiguration
3Reliability
If conventional defect detection is used, then the processing speed is fast, but the drive yield and capacity are reduced due to high probability of miss or false alarm
Solution Approach 1:
By segmenting the detection task into multiple specialized sub-detectors, the system improves reliability through targeted detection of specific defect types. Each sub-detector focuses on particular defect characteristics, reducing false alarms and misses compared to a single general-purpose detector, thereby improving drive yield without significantly impacting processing efficiency
Data Source
AI summary
Methods, systems and computer program products for performing hybrid defect detection are disclosed. In some implementations, an apparatus includes a signal module to process data signals corresponding to data on a storage medium to generate signal samples. The apparatus includes a first defect detector to identify a first portion of the signal samples, determine a number of the signal samples in the first portion that are associated with abnormal signal characteristics, and generate a first output based on the number of the signal samples in the first portion that are associated with abnormal signal characteristics. The apparatus includes a second defect detector to identify a second portion of the signal samples different from the first portion, and generate a second output based on a correlation between data bits and signal samples in the second portion.