Plasma Frequency Tuning Algorithm Global Optimum Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Frequency tuning algorithms for plasma processing power supplies face challenges in finding the global optimum frequency without extinguishing the plasma load, particularly due to the nonlinear nature of plasma loads and impedance changes over time, which can lead to local minima traps and power delivery issues.
Innovation Solution
A method involving a flowchart-based algorithm that probes frequencies, sets initial and probe frequencies, and adjusts the generator's operation based on reflection coefficient measurements, ensuring minimal time at each frequency to prevent plasma extinction, and uses a matched source impedance to simplify the frequency tuning process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If frequency tuning algorithms are used to find optimal operating frequency, then reflected power is reduced and power delivery is improved, but the algorithms may become trapped in local minima and fail to find the global optimum
Solution Approach 1:
The system performs a preliminary frequency sweep to identify multiple local minima before applying the hill-climbing algorithm. This preliminary action provides multiple starting points for optimization, reducing the risk of getting trapped in a single local minimum and improving the likelihood of finding the global optimum frequency
Solution Approach 2:
The patent introduces an intermediary frequency sweep step that acts as a bridge between initial frequency estimation and final optimization. This intermediate phase maps the reflection coefficient landscape and identifies promising regions, enabling the main algorithm to start from informed positions rather than random guesses
2Measurement precision
If the frequency tuning algorithm dwells at frequencies with high reflection coefficients to find the global optimum, then frequency optimization is improved, but the plasma may extinguish due to insufficient power delivery
Solution Approach 1:
The system applies partial action by limiting the dwell time at each frequency to only what is necessary for measurement, rather than prolonged dwelling that would cause plasma extinction. The algorithm takes just enough time to sample the reflection coefficient but quickly moves on, accumulating information without compromising plasma stability
Solution Approach 2:
The frequency tuning employs periodic sweeping through the frequency range, repeatedly sampling different frequencies in a systematic sequence. This periodic action ensures comprehensive coverage of the frequency space while maintaining brief exposure at each point, preventing plasma extinction while still gathering sufficient data to identify the global optimum
3Measurement precision
If the generator spends more time probing frequencies to accurately identify the global optimum, then frequency tuning precision is improved, but plasma extinction risk increases due to prolonged exposure at high reflection frequencies
Solution Approach 1:
The algorithm performs partial measurements at each frequency point, collecting just enough data to determine whether that frequency is promising. Rather than exhaustive measurement at each point, the system takes minimal necessary samples and quickly transitions to more promising frequencies, achieving adequate precision without excessive time exposure
Data Source
Figure 1
Figure 2A~2B
Figure 3A~3B
AI summary
A generator and method for tuning the generator are disclosed. The method includes setting the frequency of power applied by the generator to a current best frequency and sensing a characteristic of the power applied by the generator. A current best error based upon the characteristic of the power is determined, and the frequency of the power at the current best frequency is maintained for a main-time-period. The frequency of the power is then changed to a probe frequency and maintained at the probe frequency for a probe-time-period, which is less than the main-time-period. The current best frequency is set to the probe frequency if the error at the probe frequency is less than the error at the current best frequency.