Font Size: a A A

Research On The Dynamic Behavior Characteristics And Control Technology Of High-frequency Ultrasonic Piezoelectric Transducer System

Posted on:2020-03-10Degree:MasterType:Thesis
Country:ChinaCandidate:J W HuangFull Text:PDF
GTID:2432330575469081Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Piezoelectric ultrasonic transducer is one of the most important components of high-frequency(HF)thermosonic wire bonding system,and its dynamic behavior characteristics have significant impacts on the reliability and quality of bonding interconnections.Therefore,the investigations about the dynamic behavior features and control technologies are valuable.The aims of this study are to build up an exciting and multi-sensor monitoring system for HF transducers,and to find out the influencing factors and the relevant rules on the dynamics characteristics of the HF ultrasonic transducer,and to achieve the optimal assembly preload force and driving frequency,and to realize the real-time measuring of the dynamic impedance phase of the transducer system.The following results have been achieved:(1)Based on direct digital frequency synthesis(DDS)technology,an exciting and multi-sensor monitoring system for HF transducers was built up,in which a high-performance STM32 single-chip microcomputer was used.The system mainly consists of a dynamic impedance phase detecting circuit,a true RMS detecting circuit for current and voltage signals,and a power amplifier.A data acquisition system depending on multi-sensor and virtual instrument technique was specially developed and it realized the real-time measurement and storage of several vital processing signals,such as current flowing through the piezoelectric driver,tip vibration amplitude output of HF transducer,and dynamic impedance phase of the transducer,and so forth.All experiments and tests in this research are carried out based on this self-developed exciting and monitoring system.(2)The equivalent circuit admittance model near the resonant frequency of the piezoelectric ultrasonic transducer was established,and then the resonant vibration control strategies for HF ultrasonic transducer based on phase locked loop were summarized.Furthermore,the correlations between the assembly preload force and the electrical parameters of the transducer were investigated.(3)Relying on the equivalent circuit model of piezoelectric transducer,the mathematical model of the dynamic response of the transducer was established.The model accuracy was verified through the comparisons between the experiments and simulations,and three key factors affecting the dynamic response characteristics of the transducer were achieved,involving assembly preload force,voltage amplitude and frequency of the excitation signal.Indexes of overshoot and adjustment time were innovatively applied to study and analyze the influences of these three important factors on the dynamic response properties of the transducers.After a series of experimental tests,the relevant affecting regulations of three factors on the dynamic response,especially for the initial transient response,were made clear.(4)To ensure the proper exciting frequency for the HF transducer and achieve a desired vibration response,the RBF neural network was introduced to predict the resonant frequency of the transducer.In addition,a new method to get the optimal assembly preload force for the ultrasonic transducer had been established and it is based on the random forest algorithm.The test results of these two methods show that they are feasible and effective,in the meanwhile,they also can be improve the ultrasonic energy utilization efficiency of the transducer system.To realize the resonance frequency tracking of HF transducer,and obtain the good tip vibration amplitude output,a new method to acquire the dynamic impedance phase was proposed,in which the Hilbert transform was used.A critical comparison between the measured DIP results respectively from sensors and the newly proposed method was performed,confirming the accuracy and effectiveness of this method.
Keywords/Search Tags:high frequency ultrasonic transducer, dynamic behavior characteristics, direct digital frequency synthesis, preload force, RBF neural network, random forest, Hilbert transform
PDF Full Text Request
Related items