Research On Theoretical Models And Test Methods In The Seismometer Design

The performance of the seismometer is an important factor affecting the quality of the seismic observation data.However,during the seismometer design(or development)process,improving the performance of the seismometer is difficult and time-consuming.This is because there are two problems in the seismometer design work:(1)In some seismometer design work,there is a lack of theoretical models;in other seismometer design work,the theoretical models are not complete.This leads to a lack of theoretical guidances for the seismometer designers in the high-performance design.Repeated attempts to "adjust the electromechanical parameters of the seismometer-test the performance of the seismometer" become a helpless choice.However,this "trial and error method" is very time-consuming.(2)There is a lack of methods for testing the subsystem noise of the feedback system of the seismometer.If the seismometer designer only knows the noise level of the seismometer,but does not know the noise level of the subsystem,it is difficult to locate the abnormal noise source.The efficiency of the seismometer frequency response test needs to be improved.The tedious frequency response testing process also reduces the efficiency of the seismometer design work.In order to solve the above problems,we studied some theoretical models and test methods which can be used when we design seismometers,including the following contents:(1)Establishing 3 theoretical models:a theoretical model for the self-noise of a velocity broadband seismometer,a theoretical model for the transfer function of a velocity broadband seismometer,and a new theoretical model for the Brownian noise of a mechanical pendulum of the seismometer.(2)Proposing 2 methods for testing the noise of the subsystems of the feedback system of the seismometer:a method for testing the self-noise of a seismometer mechanical pendulum(or a method for testing the self-noise of a low sensitivity seismic sensor),a method for testing the electronic noise in the feedback system of a seismometer.(3)Proposing 2 methods for testing the frequency response of a seismometer:an improved pseudo random binary calibration method for a seismometer,a method for measuring the phase response of a seismometer using superimposed sinusoidal signals.In Chapter 2,we established a theoretical model for the self-noise of a velocity-broadband seismometer.The model’s inputs are mechanical and electrical parameters of the seismometer and the output is the predicted self-noise of the seismometer.The noise sources involved in the model include not only noise sources in the forward path,but also noise sources in the feedback path and external to the feedback loop.Therefore,our model is more complete than the previous model.We compared the CS60 seismometer’s predicted self-noise calculated using this model with the measured self-noise,and found good agreement between them.Therefore,this model allows us to predict the self-noise level and it can help to reduce the time cost by avoiding unnecessary field testing.It is valuable in designing a seismometer.By studying the theoretical model for the seismometer self-noise,we found that the dominant noise source in CS60 is different over different frequency bands.In addition,some viewpoints about the low noise design of seismometers were proposed.In Chapter 3,we studied the theoretical model for the transfer function of a velocity-broadband seismometer.From this model,we can know how the electromechanical parameters of the seismometer feedback system determine the characteristics of the seismometer transfer function.Then,the efficiency of adjusting the characteristics of the seismometer transfer function can be greatly improved.We took CS60 seismometer as an example to introduce three applications of the model,including:obtaining the transfer function of a seismometer,flattening the amplitude response in the passband of a seismometer,and predictting the frequency characteristics of a seismometer.In addition,we also studied the LP(long period)sensitivity of the seismometer and obtained the following conclusions:When designing a seismometer,we can reduce the LP sensitivity of the seismometer by reducing the resistance of the resistor which is connected to the output terminal of the integrator in the feedback path,so that the possibility of seismometer output being clipped can be reduced,in environments with large temperature variation.In Chapter 4,we established a new theoretical model for the Brownian noise of a mechanical pendulum of the seismometer.The previous theoretical model was developed on the basis of a translational model of the mechanical pendulum.The new theoretical model was developed on the basis of a pendulum-type model of the mechanical pendulum.From the new theoretical model,the seismometer designer can know the relationship between the parameters related to the rotation of the rigid body(e.g.,the distance from the center of the mass of the pendulum to the rotating axis)and the Brownian noise level of the mechanical pendulum.However,from the previous theoretical model,this relationship can not be known.The basis for the development of the new theoretical model(the pendulum-type model of the mechanical pendulum)is closer to the actual structure of the mechanical pendulum.Therefore,the new theoretical model has more guiding significance in the low noise design of the mechanical pendulum.In Chapter 5,we introduced a new method for self-noise measurement of seismic sensors.Previously,we mainly used two methods to test the self-noise of seismic sensors:two-sensor method and three-sensor method.When we test with these two methods,if the sensor under test is a low-noise and low-sensitivity sensor,the self-noise of the digitizer will significantly affect the accuracy of the test results.In the new method,new experimental methods and data processing methods are adoptted to reduce the influence of the digitizer self-noise on the test results.We performed a series of computer simulations to verify the theoretical correctness of the new method.The experimental results showed that even when the level of the equivalent acceleration of the digitizer self-noise was 6dB higher than the level of the equivalent acceleration of the sensor self-noise,the new method could still accurately calculate the level of the sensor self-noise,while the Holcomb coherence function method would overestimate the level of the sensor self-noise.In Chapter 6,we introduced a method for testing the electronic noise in the feedback system of a seismometer:the two-channel method.Testing the electronic noise in the feedback system can help us efficiently locate the unexpected electronic noise sources in the feedback system.There are many articles in which the seismometer self-noise test is discussed.However,as far as we know,there is no article discussing the test of the electronic noise in the feedback system of the seismometer.We introduced the connections of the equipments in the two-channel method.We also introduced the principle and the data processing method of the two-channel method.When we tested with the two-channel method,the uncorrelated interference noise in the two channels was eliminated.Then,the accuracy of the test results could meet the needs of the practical application.The experimental results showed that,When we tested the resistance thermal noise whose level was approximately 0.7nV/(?)in the experimental environment that we had built,the measured results were very close to the theoretical results.This met the requirements for the accuracy of the measurement of the electronic noise in the feedback system of the seismometer.In Chapter 7,we proposed an improved pseudo random binary calibration method for a seismometer.When testing with this method,we used a pseudo-random binary signal as the calibration input signal in the low frequency band,and used a signal which was redesigned based on the pseudo-random binary signal in the high frequency band.This method not only inherits the advantages of the traditional pseudo-random binary calibration method(the calibration signal is simple and has only two levels),but also greatly improves the accuracy of the test results in the high frequency band.In Chapter 8,we proposed a method for measuring the phase response of a seismometer using superimposed sinusoidal signals.The traditional sine calibration can be used for measuring the phase response of a seismometer.However,it has two shortcomings:(1)Transient responses that can not be used for data processing appear multiple times.This increases the time consumption of the calibration test.(2)Testers have to manually select the steady-state response waveform from the sine response waveform of each frequency.This makes it difficult to realize automatic data processing.When we test with the new method,the transient response appears only once at the beginning of the response waveform.Therefore,the time span of the response waveform that can not be used for data processing is greatly reduced.(For example,when we test the phase response of the same seismometer on the same group of frequencies,the measurement time of the new method and the traditional sine calibration are 1560 seconds and 5089 seconds,respectively.)Moreover,when we select the steady-state response waveform,it is only necessary to avoid the transient response at the beginning of the response waveform.Therefore,it is easier to realize automatic data processing with this new method.In a word,all the above models and methods are proposed to meet the actual needs of the seismometer design work in the context of a rapid development of the research field.These models and methods are of great significance to the high-performance design of the seismometer.