Predictive Plasma Power Coupling Control for Impedance Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In plasma processing tools, the impedance matching network struggles to quickly and accurately adjust to changes in plasma density, leading to suboptimal power delivery due to lagging control responses.
Innovation Solution
An adaptive control architecture with a feedforward control system between control loops, utilizing a process power controller with a source predictor and uniformity controller to anticipate and adjust impedance changes proactively, ensuring cohesive control of plasma parameters and minimizing disturbances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the impedance matching network adjusts impedance after plasma density changes, then the system maintains stable operation, but the control response is slow and power delivery efficiency deteriorates
Solution Approach 1:
The patent implements a feedforward control mechanism that predicts plasma density changes and proactively adjusts the impedance matching network before the actual changes occur. This preliminary action eliminates the lag inherent in traditional feedback-only systems, maintaining both stability and efficiency by anticipating and preparing for impedance variations in advance.
Solution Approach 2:
The system combines feedback control with feedforward prediction, where real-time plasma density measurements are fed back to validate predictions and continuously refine the control algorithm. This hybrid approach ensures stable operation through feedback while achieving rapid response through the predictive feedforward component.
2Manufacturing precision
If multiple antennas are used to control plasma density at different locations, then plasma uniformity is improved, but system impedance becomes complex and difficult to match
Solution Approach 1:
The patent employs parameter sweeping and optimization techniques to characterize the complex impedance landscape created by multiple antennas. By systematically varying and measuring impedance parameters across different operating conditions, the system builds a predictive model that simplifies real-time control, transforming the complex multi-antenna impedance problem into a manageable set of pre-characterized parameters.
3Device complexity
If the impedance matching network uses traditional feedback control, then the system remains simple, but the control accuracy and response speed deteriorate
Solution Approach 1:
The patent introduces a computational prediction layer that performs preliminary impedance calculation based on anticipated plasma conditions. This adds intelligence to the control system without requiring complex hardware modifications, achieving high control accuracy through software-based prediction while maintaining relative simplicity in the physical impedance matching network.
Data Source
AI summary
Embodiments disclosed herein include a process power controller for a plasma processing tool. In an embodiment, the process power controller includes a process power source optimizer, a source predictor, and a process uniformity controller. In an embodiment, the source predictor is communicatively coupled to the process power source optimizer and the process uniformity controller.