Sensor Amplifier Offset Gain Calibration Using Product Reticle Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The miniaturization of patterns on reticles in semiconductor manufacturing makes it challenging to calibrate the offset and gain of sensor amplifiers using traditional methods, as black and white regions with sufficiently wide areas may not exist, leading to reduced throughput and potential errors in defect inspection due to differences in light transmittance across reticles with varying glass substrate thicknesses.
Innovation Solution
A reticle defect inspection method and apparatus that calibrate the offset and gain of sensor amplifiers using the product reticle itself, by imaging patterns, storing bottom and peak values of light signals, and setting offsets and gains for each pixel, allowing for normalization of signal amplitudes without relying on separate calibration reticles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a separate calibration reticle is used to calibrate offset and gain, then calibration accuracy is improved, but throughput is reduced and errors may occur due to differences in glass substrate thickness between calibration and product reticles
Solution Approach 1:
The product reticle inspects itself by using its own pattern regions (black and white regions within the pattern areas) to perform offset and gain calibration. This eliminates the need for separate calibration reticles and allows calibration to be performed on each product reticle, ensuring consistency in glass substrate thickness and eliminating reticle replacement steps.
Solution Approach 2:
The pattern regions on the product reticle serve dual purposes: they function as both the inspection target and the calibration reference. The black and white regions within the product pattern itself are utilized for calibration, making the product reticle multi-functional and eliminating the need for dedicated calibration reticles.
2Measurement precision
If black and white regions of sufficient area are added to the product reticle for calibration, then calibration accuracy is improved, but the occupied area of the product pattern is reduced
Solution Approach 1:
Instead of adding large separate calibration regions, the invention utilizes existing pattern regions with appropriate black and white areas distributed throughout the product reticle. These local pattern regions provide sufficient contrast for calibration while maintaining the overall product pattern area and design integrity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate calibration of sensor amplifiers using the product reticle, improving signal-to-noise ratio and defect inspection accuracy, even when wide black and white regions are not present, thereby enhancing throughput and reducing errors.
Implementation Method 1
an image sensor moves relative to a reticle and detects amount-of-light signal of light transmitted through or reflected from the reticle at plural pixels
Implementation Method 2
a sensor amplifier amplifies an output of each pixel of the image sensor to generate an optical image
Implementation Method 3
amount-of-light signal of light transmitted through or reflected from the reticle
Implementation Method 4
amount-of-light signal of light transmitted through or reflected from the reticle
Data Source
AI summary
The present invention provides a reticle defect inspection method and a reticle defect inspection apparatus capable of calibrating the offset and gain of a sensor amplifier using a product reticle even though black and white regions each sufficiently wider than a TDI sensor imaging area do not exist in the product reticle. An output of each pixel of the TDI sensor is amplified by the sensor amplifier. A bottom value of the amplified amount-of-light signal of each pixel is stored by bottom value storing means of offset/gain calibrating means, and a peak value thereof is stored by peak value storing means. The offset of each pixel is calculated by offset calculating means based on the bottom value of each pixel. The gain of each pixel is calculated by gain calculating means based on the offset of each pixel and the peak value of each pixel. The calculated offset and gain of each pixel are stored in a register and thereby the offset and gain of the sensor amplifier are calibrated.