JP2009272733A - Amplitude characteristic correction method for output signal of a/d converter, and seismometer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an amplitude characteristic correction method for output signal of an A/D converter, which realizes amplitude characteristics such that a relative ratio with the input signal becomes substantially "1", up to a predetermined frequency by means of a low-degree filter. <P>SOLUTION: The amplitude characteristic correction method of a digital signal of the A/D converter 20, which corrects a signal output from the A/D converter 20, which is constituted to include a ΔΣ modulator 21 and a Sinc3 filter 22, by using an IIR (infinite impulse response) filter 51. The transfer function of the IIR filter 51 consists of a formula, in which the transfer function of a low-pass filter, with a cut-off frequency as fh and a quality factor as Q is subjected to pre-warping processing and a bilinear transformation is carried out with the sampling frequency as fs. The transfer function sets the cut-off frequency fh and the quality factor Q to arbitrary values, and corrects the digital signal so that the relative ratio to an analog input signal becomes substantially 1, up until the predetermined frequency. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an amplitude characteristic correction method for an output signal of an A / D converter and a seismometer.

Conventionally, in seismometers and the like, it is known that a digital output signal desired by a user can be acquired from an analog input signal by an A / D converter including a ΔΣ modulator and a decimation filter (for example, Patent Documents). 1).
Japanese Patent Laid-Open No. 2003-258637

  However, in the digital signal output by the A / D converter as described above, the relative ratio of the amplitude to the input signal decreases with a high attenuation rate as the frequency increases, but it is sufficient in the passband up to a predetermined frequency. Therefore, there is a problem that a flat characteristic of the amplitude (a characteristic in which the relative ratio of the amplitude of the input signal to the output signal is approximately 1) cannot be obtained, and an accurate measurement result desired by the user cannot be obtained.

  Regarding the above problem, for example, when a sinc filter is used as a decimation filter, the digital signal output from the A / D converter is passed through an FIR (Finite Impulse Response) filter having reverse characteristics to the sinc filter. A correction method is known.

  However, when the FIR filter is used, there is a problem that the processing becomes high order and the amount of calculation is large, and a sufficient calculation processing speed cannot be obtained. However, there is a design problem that it must be long.

  An object of the present invention is to correct an amplitude characteristic of an output signal of an A / D converter that realizes an amplitude characteristic such that a relative ratio to an input signal is approximately 1 up to a predetermined frequency by a low-order filter, and an earthquake To provide a total.

In order to solve the above problems, the method for correcting the amplitude characteristic of the output signal of the A / D converter according to claim 1 comprises:
A ΔΣ modulator that converts an analog input signal into a digital signal;
A decimation filter that thins out a digital signal converted by the ΔΣ modulator at a predetermined decimation rate, and an A / D converter that corrects amplitude characteristics of the digital signal output from the A / D converter In the method for correcting the amplitude characteristic of the output signal,
Using an IIR filter,
The transfer function H (z) of the IIR filter is
The transfer function of the low-pass filter of Equation (1), where the cutoff frequency is fh and the quality factor is Q, is prewarped, and consists of Equation (2) obtained by bilinear transformation with the sampling frequency as fs.
The cutoff frequency fh and the quality factor Q are set to arbitrary values, and the digital signal output from the A / D converter is input to the IIR filter, so that the digital signal that has passed through the IIR filter is predetermined. The frequency characteristic is corrected so as to have an amplitude characteristic in which the relative ratio to the analog input signal is approximately 1.

The invention according to claim 2 is a seismometer,
A ΔΣ modulator that converts an analog input signal into a 1-bit digital signal;
A decimation filter that decimates the digital signal converted by the ΔΣ modulator at a predetermined decimation rate; and
Correction means for correcting the amplitude characteristics of the digital signal output from the A / D converter,
The correction means includes
With an IIR filter,
The transfer function H (z) of the IIR filter is
The transfer function of the low-pass filter of Equation (1), where the cutoff frequency is fh and the quality factor is Q, is prewarped, and consists of Equation (2) obtained by bilinear transformation with the sampling frequency as fs.
The IIR filter in which the cut-off frequency fh and the quality factor Q are set to arbitrary values is such that the digital signal output from the A / D converter that has passed through the IIR filter is 30 Hz or higher. The correction is performed so as to have an amplitude characteristic in which the relative ratio to the signal is approximately 1.

According to the present invention, in the transfer function of equation (2), by setting the cutoff frequency fh and the quality factor Q to arbitrary values, an AR that has passed through an IIR (Infinite Impulse Response) filter composed of the transfer function is obtained. The digital signal output from the / D converter can realize an amplitude characteristic in which the relative ratio to the input signal is approximately 1 up to a predetermined frequency.
That is, since the digital signal output from the A / D converter is corrected by the IIR filter composed of the transfer function of Equation (2), the user adjusts the cutoff frequency fh and the quality factor Q value. A flat characteristic of the amplitude of the output signal in a desired frequency band can be easily realized.
Furthermore, as shown in the equation (2), since it is a low-order filter, the calculation amount is small, and the calculation process can be speeded up.
Therefore, the present invention can be said to be an amplitude characteristic correction method for an output signal of an A / D converter that realizes an amplitude characteristic such that a relative ratio to an input signal is approximately 1 up to a predetermined frequency by a low-order filter. .

  Hereinafter, a method for correcting an amplitude characteristic of an output signal of an A / D converter and a seismometer according to the present invention will be described with reference to FIG. However, the scope of the invention is not limited to the illustrated examples.

  FIG. 1 is a block diagram showing the internal configuration of the seismometer 1. The seismometer 1 receives, for example, a vibration detector 5 that detects vibration caused by an earthquake as an analog signal, and an analog signal detected by the vibration detector 5 as an analog input signal 10 and converts it into a digital signal 40. FPGA (Field Programmable Gate Array) as a PLD (Programmable Logic Device) that corrects the digital signal 40 output from the A / D converter 20 by the D converter 20 and filtering processing and generates the corrected digital signal 60. ) 50, a control unit 70 that includes a predetermined control program and the like to control the processing of the seismometer 1, an input unit 80 for executing an input operation of the seismometer 1 by a user, a measurement result of the seismometer 1, and the like And a display unit 90 for displaying.

  The vibration detector 5 is composed of, for example, a servo-type vibration detector, and the deviation from the neutral position of the pendulum provided therein is returned to the neutral position by current amount control using the drive coil, and at that time the drive coil The acceleration in each direction can be detected by measuring the current flowing through

  The A / D converter 20 includes a ΔΣ modulator 21, a Sinc3 filter 22 as a decimation filter, and the like, and inputs an analog signal detected from the vibration detector 5 as an analog input signal 10. Processing to convert the signal 40 is performed.

The ΔΣ modulator 21 includes a quantizer and an integrator (not shown) and converts the analog input signal 10 into a digital signal.
The ΔΣ modulator 21 reduces the quantization error by performing oversampling, and suppresses the generation of noise by performing noise shaving.

The Sinc3 filter 22 functions as a decimation filter that thins out the digital signal oversampled by the ΔΣ modulator 21 based on a predetermined decimation rate N, and is a third-order filter represented by a transfer function of the following equation (3). It is a Sinc filter.

The FPGA 50 is composed of a plurality of logic circuits arranged in a lattice pattern, and the user can write an original logic circuit, so that the IIR filter 51 can be mounted. Since the IIR filter 51 can execute the filtering process, the digital signal 40 output from the A / D converter 20 can be corrected and the corrected digital signal 60 can be output. The FPGA 50 functions as a correction unit.
As long as the IIR filter 51 can be mounted, the FPGA 50 is not limited, and a one-chip microprocessor such as a DSP (Digital Signal Processor) may be used.

そして、上記伝達関数にプリワーピング処理を施し、サンプリング周波数をｆｓとして次式（５）に基づく双一次近似により双一次変換すると、次式（６）の伝達関数が導出される。

この式（６）の伝達関数と、式（４）の伝達関数の積により得られる関数において、遮断周波数ｆｈとクオリティファクタＱに任意の値を入力することにより、Ａ／Ｄ変換器２０の出力信号の振幅特性が決定される。 The IIR filter 51 is one of signal processing filters having infinite impulse response characteristics, and is implemented as an analog filter or a digital filter. In the present invention, a digital IIR filter that changes the amplitude characteristics of a digital signal is used. Is functioning as
The transfer function of the IIR filter 51 is designed based on the transfer function of an analog low-pass filter of the following equation (4) where the cutoff frequency is fh and the quality factor is Q.

Then, pre-warping processing is performed on the above transfer function, and bilinear transformation is performed by bilinear approximation based on the following equation (5) with the sampling frequency set to fs, the transfer function of the following equation (6) is derived.

In the function obtained by the product of the transfer function of the equation (6) and the transfer function of the equation (4), an arbitrary value is input to the cutoff frequency fh and the quality factor Q, whereby the output of the A / D converter 20 is obtained. The amplitude characteristic of the signal is determined.

  The control unit 70 includes, for example, a CPU (Central Processing Unit) that executes various programs, a ROM (Read Only Memory) that stores various programs, and a RAM (Random Access Memory) that temporarily stores various data. The configured and corrected digital signal 60 is converted into seismic motion data (acceleration data) and processed, and output display is controlled via the display unit 90.

  The input unit 80 includes, for example, an input key, an input button, and the like, and outputs a predetermined input operation signal to the control unit 70 in accordance with a user's input operation of the seismometer 1.

  The display unit 90 includes, for example, a liquid crystal display, and displays the vibration measurement result of the seismometer 1 based on the seismic motion data processed by the control unit 70.

(Amplitude characteristics correction processing)
Hereinafter, as an example, the amplitude characteristic correction processing when the method for correcting the amplitude characteristic of the output signal of the A / D converter according to the present invention is used in the seismometer 1 will be described with reference to FIGS. .

  First, it is assumed that the analog input signal 10 detected by the vibration detector 5 of the seismometer 1 is oversampled by the ΔΣ modulator 21 of the A / D converter 20 and the sampling frequency is 125000 Hz. The sampling frequency is thinned by the Sinc3 filter 22 with a decimation rate N of 625.

In the above case, in FIG. 2, the amplitude characteristic of the digital signal when it is output from the A / D converter 20 without using the IIR filter 51, that is, for each frequency (Hz), using the equation (3). The relative ratio of the amplitude of the digital signal 40 to the analog input signal 10 is shown.
In general, in seismometers, it is desirable to form a flat region with a relative ratio of 1 up to a frequency of at least 30 Hz. However, as shown in FIG. 2, without using a correction filter such as the IIR filter 51. When a digital signal is output, the relative ratio starts to drop from around 10 Hz, and it can be seen that the relative ratio drops by nearly 10% at the time when the frequency is 30 Hz.
Therefore, depending on the digital signal 40 as it is, a value lower than the actual vibration measurement value is output, and thus the user cannot grasp an accurate vibration measurement value.

Next, FIG. 3 shows the amplitude characteristics of the IIR filter 51 according to the present invention when the cutoff frequency fh is 52 Hz and the quality factor Q is 1.055, using the transfer function of Equation (6).
According to FIG. 3, since the relative ratio exceeds 1 in the frequency region lower than the cut-off frequency fh, it is possible to correct the relative ratio shown in FIG. 2 by passing the IIR filter 51. Recognize.

Next, FIG. 4 shows the amplitude characteristic (curve A) of the digital signal 40 output without passing through the IIR filter 51 of FIG. 2, the amplitude characteristic (curve B) of the IIR filter 51 of FIG. ) And the equation (3), and the amplitude characteristic (curve C) of the corrected digital signal 60 obtained by passing the digital signal 40 through the IIR filter 51 is shown.
From the curve C, it can be seen that the relative ratio is substantially 1 until the frequency is about 40 Hz, and the desired flat region is formed in the seismometer 1 by the correction function of the IIR filter 51. Furthermore, the curve C has a relative ratio that is drastically reduced in the region where the frequency is 40 Hz or higher compared to the curve A, and therefore, it is possible to suitably cut off the output of excess high-frequency noise.
Therefore, the user can accurately grasp vibration measurement values up to a predetermined frequency.

As described above, according to the amplitude characteristic correcting method according to the present invention, by setting the cutoff frequency fh and the quality factor Q to arbitrary values in the transfer function of Equation (6), the IIR comprising the transfer function is set. The corrected digital signal 60 output from the A / D converter 20 that has passed through the filter 51 can realize an amplitude characteristic in which the relative ratio to the analog input signal is approximately 1 up to a predetermined frequency.
That is, since the digital signal 40 output from the A / D converter 20 is corrected by the IIR filter 51 having the transfer function of Expression (6), the user adjusts the cutoff frequency fh and the quality factor Q. Thus, the flatness characteristic of the amplitude of the corrected digital output signal 60 can be easily realized in a desired frequency band.
Further, since the filter is a low-order filter as shown in Expression (6), the amount of calculation is small, and the calculation process can be speeded up.
Therefore, the present invention can be said to be an amplitude characteristic correction method for an output signal of an A / D converter that realizes an amplitude characteristic such that a relative ratio to an input signal is approximately 1 up to a predetermined frequency by a low-order filter. .

In the seismometer 1 having the amplitude characteristic correction method using the Sinc3 filter 22 as a decimation filter, the cutoff frequency fh is set to 52 Hz, and the quality factor Q is set to 1.055. The corrected digital signal 60 output from the D converter 20 realizes an amplitude characteristic in which the relative ratio to the analog input signal is approximately 1 until the frequency is at least 30 Hz or higher.
That is, when the frequency is at least 30 Hz or higher, in the seismometer where it is desirable to form a flat region with an amplitude (relative ratio of 1), the frequency is set to 30 Hz or higher by performing the above device configuration and numerical setting. It can be said that a suitable output as a seismometer can be realized because a flat region with an amplitude up to can be formed.

The decimation filter according to the present invention is not limited to the Sinc3 filter, and may be another filter that can thin out an oversampled digital signal.
In the present embodiment, the user can arbitrarily input the cutoff frequency fh and the quality factor Q to obtain the flat characteristic of the amplitude of the output signal in the desired frequency band. Bands are set in advance via the input unit 80 and the like, and the fh and Q values are automatically changed by the processing of the control unit 70 until the desired frequency band is obtained. You may make it search or set the value of Q.

It is a block diagram which shows the principal part structure of the seismometer concerning this invention. It is an amplitude characteristic view of an A / D converter without passing through an IIR filter. It is an amplitude characteristic view of an IIR filter to which the present invention is applied. It is an amplitude characteristic figure of the A / D converter which passed the IIR filter to which the present invention is applied.

Explanation of symbols

1 Seismometer 10 Analog Input Signal 20 A / D Converter 21 ΔΣ Modulator 22 Sinc3 Filter (Decimation Filter)
50 FPGA (correction means)
51 IIR filter 60 Digital output signal corrected (digital signal)

Claims (2)

A ΔΣ modulator that converts an analog input signal into a digital signal;
A decimation filter that thins out a digital signal converted by the ΔΣ modulator at a predetermined decimation rate, and an A / D converter that corrects amplitude characteristics of the digital signal output from the A / D converter In the method for correcting the amplitude characteristic of the output signal,
Using an IIR filter,
The transfer function H (z) of the IIR filter is
The transfer function of the low-pass filter of the following equation (1), where the cutoff frequency is fh and the quality factor is Q, is prewarped, and consists of the following equation (2) obtained by bilinear transformation with the sampling frequency as fs:

The cutoff frequency fh and the quality factor Q are set to arbitrary values, and the digital signal output from the A / D converter is input to the IIR filter, so that the digital signal that has passed through the IIR filter is predetermined. A method for correcting an amplitude characteristic of an output signal of an A / D converter, wherein the amplitude characteristic is such that a relative ratio with respect to the analog input signal is substantially 1 up to a frequency equal to A ΔΣ modulator that converts an analog input signal into a 1-bit digital signal;
A decimation filter that decimates the digital signal converted by the ΔΣ modulator at a predetermined decimation rate; and
Correction means for correcting the amplitude characteristics of the digital signal output from the A / D converter,
The correction means includes
With an IIR filter,
The transfer function H (z) of the IIR filter is
The transfer function of the low-pass filter of the following equation (1), where the cutoff frequency is fh and the quality factor is Q, is prewarped, and consists of the following equation (2) obtained by bilinear transformation with the sampling frequency as fs:

The IIR filter in which the cut-off frequency fh and the quality factor Q are set to arbitrary values is such that the digital signal output from the A / D converter that has passed through the IIR filter is 30 Hz or higher. A seismometer which is corrected so as to have an amplitude characteristic in which a relative ratio to a signal is approximately 1.

Images