JP3747153B2 - Square circuit and signal processing circuit using square circuit - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
In the present invention, when a vector quantity (acceleration, etc.) is input to a physical quantity sensor (acceleration sensor, etc.) having a plurality of reference axes, the sum of squares of the components of the vector quantity on the plurality of reference axes, and thus the magnitude of the vector quantity. The present invention relates to a signal processing circuit that calculates
[0002]
[Prior art]
In a physical sensor having two orthogonal axes such as a force sensor, pressure sensor, acceleration sensor, angular acceleration sensor, magnetic sensor, optical sensor, pyroelectric sensor, etc., when an absolute quantity independent of the direction of the physical quantity detected by the sensor is required, Conventionally, the square root of the sum of squares of two orthogonal axes has been obtained, but the circuit to be realized is expensive and complicated. In general, a circuit for calculating the square root of the sum of squares is mainly composed of a square circuit and a circuit for obtaining a root. A square sum square root circuit composed of a square circuit and a circuit for obtaining a root requires a well-matched transistor pair in order to make it less susceptible to environmental influences. Therefore, the square sum square root circuit has a complicated circuit configuration and is expensive.
[0003]
For example, as a square sum square root circuit composed of integrated circuits, a circuit as shown in FIG. 4 has been published. FIG. 4 is a circuit diagram of an operational amplifier described in P327-P332 of “Operational Amplifier” published by McGraw-Hill in 1983. In FIG. 4, a square circuit can be configured by setting V1 = V2. In the circuit of FIG. 4, the field effect transistors Q1 and Q2 need to have well-matched characteristics. In the example of FIG. 4, the maximum error is estimated to be 4.4% of full scale. The root generation circuit for finding the root can be realized by a circuit that feeds back the output of the divider to the input. The total error in the root generation circuit having such a circuit configuration is considered to be about 10%.
[0004]
As described above, the conventional square sum square root circuit is a complicated circuit, and requires further adjustment of the circuit for precise physical quantity measurement, and cannot be a simple and inexpensive configuration. As a method for calculating the square sum of squares with a simple and inexpensive configuration, a method described in Japanese Patent Publication No. 58-34865 has been proposed. This square sum of squares calculation method pays attention to the first quadrant (0 degree to 90 degree) of two orthogonal axes, and divides the angle into three equal parts (0 degree, 30 degree, 60 degree), and outputs the sensor at each angle. This is a method of performing correction processing. Specifically, in the range from 0 degree to 30 degrees, the sensor output directed to 0 degree is applied. In the range from 60 degrees to 90 degrees, the sensor output for 90 degrees is applied. In the range from 30 degrees to 60 degrees, a value obtained by multiplying the sum of the sensor output directed to 0 degrees and the sensor output directed to 90 degrees by COS45 ° = 0.707 as a coefficient is applied. According to this method, the error is estimated to be 7.6%. In such a disclosed invention, it was divided into three equal parts, but the error can be further reduced by increasing the angle of equally dividing. Further, this square sum square root calculation method can be realized by a circuit that has a relatively simple configuration and can be operated without adjustment.
[0005]
[Problems to be solved by the invention]
However, the method of Japanese Patent Publication No. 58-34865 focuses only on the first quadrant of two orthogonal axes. In order to realize a circuit that applies the method to the sensor output in the range of 0 to 360 degrees, it is necessary to provide an absolute value circuit for obtaining an absolute value of the orthogonal two-axis sensor output as a pre-circuit. Since an absolute value circuit is necessary as a pre-circuit, the sum-of-squares root calculation circuit to which the method of Japanese Patent Publication No. 58-34865 is applied in the range of 0 to 360 degrees cannot be realized with a simple configuration, and the cost is reduced. It was difficult. The present invention has been made in view of the above problems, and is applied to a physical quantity sensor having sensitivity to a range of 0 degrees to 360 degrees, and a vector quantity component detected with respect to a plurality of axes of the physical quantity sensor. It is an object of the present invention to provide a signal processing circuit that can calculate the sum of squares of the sq.
[0006]
[Means for Solving the Problems]
The gist of the present invention for achieving the object lies in the inventions of the following items.
[0007]
[1] A change amount of the pulse width of the rectangular wave in the first rectangular wave signal having the repetition frequency f represents a certain amount A, and the first rectangular wave signal is processed to represent a square value A ² of the amount A. A square circuit that generates a square value signal;
First averaging means for generating a first average signal representing an average amplitude of the first rectangular wave signal;
The first average signal and a constant value such as a ground voltage are alternately selected in synchronization with the first rectangular wave signal, and the amplitude is a difference value between the first average value and the constant value. Selecting means for generating a second rectangular wave signal whose pulse width is the pulse width of the first rectangular wave signal;
And a second averaging means for generating a second average signal representing an average amplitude of the second rectangular wave signal and outputting a change amount of the second average signal as the square value signal. A square circuit.
[0008]
[2] The first averaging means includes a first filter that inputs the first rectangular wave signal, and amplifies or attenuates the output of the first filter to obtain the first average signal. With the amplitude adjustment circuit to generate,
The square circuit according to [1], wherein the second averaging means is a second filter that receives the second rectangular wave signal.
[0009]
[3] When acceleration and other vector quantities are input, the components of the vector quantities relating to the n-axis (n is 2 or 3) orthogonal to each other are sensed, and the n-axis sensor that outputs each component of the n-axis In a signal processing circuit that receives each n-axis component and generates a sum of squares of each n-axis component,
The n square circuits according to the above [1] or [2], and an adder that adds the square signals that are outputs of the n square circuits and outputs the sum as a square sum.
The n square circuits are provided corresponding to the n-axis,
Each square circuit includes a pulse width modulator that generates the first rectangular wave signal;
Each of the pulse width modulators modulates the pulse width of the first rectangular wave signal according to the corresponding component of the axis.
[0010]
[4] a clock circuit that outputs a clock signal that is a rectangular wave signal;
The pulse width modulator generates the first rectangular wave signal by receiving the clock signal and modulating the pulse width of the clock signal with the corresponding component of the axis. ]. The signal processing circuit of description.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a configuration diagram showing a signal processing circuit according to an embodiment of the present invention. However, FIG. 1 includes a physical quantity sensor 10 to which this signal processing circuit is applied. The physical quantity sensor 10 is a capacitance detection type sensor such as an acceleration sensor, and uses the orthogonal X axis and Y axis as sensing axes. Therefore, when a vector quantity such as acceleration is input to the physical quantity sensor 10, the physical quantity sensor 10 outputs an X-axis component x and a Y-axis component y of the vector quantity. The capacitance detection type sensor converts the X-axis component x and the Y-axis component y of the input vector quantity into a change amount of the capacitance and outputs it as an X-axis sensor signal x and a Y-axis sensor signal y, respectively.
[0012]
The X-axis sensor signal x is transmitted to the first pulse width modulator 20, and the Y-axis sensor signal is transmitted to the second pulse width modulator 30. First, a circuit for processing the X-axis sensor signal x will be described. The clock circuit 11 outputs a clock signal that is a rectangular wave to the first pulse width modulator 20 and the second pulse width modulator 30. The first pulse width modulator 20 synchronizes with the clock signal output from the clock circuit 11 and outputs a pulse width modulation signal obtained by modulating the pulse width in accordance with the level of the X-axis sensor signal x output from the physical quantity sensor 10. Since the X-axis sensor signal x is a change in capacitance, the first pulse width modulator 20 generates a pulse width modulation signal by changing the pulse width of the input clock signal in accordance with the change in capacitance. The pulse width modulation signal is a rectangular wave and corresponds to the first rectangular wave signal described above. The pulse width modulation signal is a signal having a constant amplitude and a pulse width change amount corresponding to the X-axis sensor signal x. Therefore, the average value of the pulse width modulation signal is proportional to the pulse width of the pulse width modulation signal. Therefore, the amount of change in the average value of the pulse width modulation signal corresponds to the X-axis sensor signal x (change in capacitance in the X-axis capacitance element of the physical quantity sensor 10).
[0013]
The first filter 21 filters a component near the clock frequency included in the input pulse width modulation signal, and generates a DC signal obtained by rectifying the pulse width modulation signal. The amplitude of the DC signal represents the average value of the amplitude of the pulse width modulation signal. The first amplifier 22 receives the DC signal output from the first filter 21, amplifies the level of the DC signal, and supplies the DC signal whose level is adjusted to the first switch 23. The amplification of the level of the DC signal by the first amplifier 22 causes the level of the input DC signal to be equal to the average value of the pulse width modulation signal output from the first pulse width modulator 20 and guided to the first switch 23. This is performed in order to adjust the level and supply the DC signal whose level has been adjusted to the first switch 23. Therefore, the level of the DC signal output from the first amplifier 22 represents the X-axis sensor signal x. The circuit composed of the first filter 21 and the first amplifier 22 corresponds to the first averaging means described above.
[0014]
The first switch 23 alternately switches the DC signal output from the first amplifier 22 and the ground voltage every time the level of the pulse width modulation signal transmitted from the first pulse width modulator 20 increases or decreases. Select and output. Therefore, the output of the first switch 23 is a rectangular wave signal. The amplitude of this rectangular wave signal is a DC level output from the first amplifier 22. The pulse width of the rectangular wave signal is the pulse width of the pulse width modulation signal transmitted from the first pulse width modulator 20. Therefore, in the rectangular wave signal output from the first switch 23, both the amplitude and the pulse width represent the X-axis sensor signal x. The first switch 23 corresponds to the first selection means described above. The rectangular wave signal that is the output of the first switch 23 corresponds to the second rectangular wave signal described above. The third filter 24 filters components around the clock frequency of the input rectangular wave signal, and generates a DC signal obtained by rectifying the rectangular wave signal. The amount of change in the amplitude of the DC signal represents the average value of the rectangular wave signal input to the third filter 24. Then, the amplitude of variation of the direct current signal is representative of the square x ² of the X-axis sensor signal x.
[0015]
Next, the Y-axis sensor signal processing circuit will be described. The Y-axis sensor signal processing circuit has basically the same configuration as the X-axis sensor signal processing circuit. The difference between the two processing circuits is that the signal to be processed is the X-axis sensor signal x in the X-axis sensor signal processing circuit, but the Y-axis sensor signal y in the Y-axis sensor signal processing circuit. Therefore, in the rectangular wave signal output from the second switch 33, both the amplitude and the pulse width represent the Y-axis sensor signal y. The direct current signal output from the fourth filter 34 is input to the adder 12. The output of the fourth filter 34 is a DC signal, and the amount of change in the amplitude of this DC signal represents the average value of the rectangular wave signal input to the fourth filter 34. The amount of change in the amplitude of the DC signal represents the square y ² of the Y-axis sensor signal y.
[0016]
The adder 12 is a DC signal representative of a change amount of the square x ² of the X-axis sensor signal x (output of the third filter 24), a DC signal representative of a change amount of the square y ² of Y-axis sensor signal y ( And the output of the fourth filter 34) is added to generate a square sum x ² + y ² of the X-axis sensor signal x and the Y-axis sensor signal y. Details of the operation of the circuit of FIG. 1 will now be described in detail with reference to FIG.
[0017]
As described with reference to FIG. 1, the switching process is performed by the first switch 23 and the second switch 33, and the outputs of these switches are output by the third filter 24 and the fourth filter 34. A rectangular wave signal was averaged. The switching process and the averaging process have the effect of obtaining the square of the amount of change in the pulse width of the pulse modulated wave signal generated by the first pulse width modulator 20 and the second pulse width modulator 30, respectively. Based on
[0018]
FIG. 2 shows a timing diagram of each signal in the circuit of FIG. 1 that processes the sensor signals x and y. FIG. 2A is a timing diagram of the clock signal output from the clock circuit 11. The clock signal is a rectangular wave signal with a duty ratio of 50%, and its clock cycle is 1 μS. The amplitude of the clock signal is 5 Vpp, and its low potential is the ground potential. FIG. 2B illustrates a sensor signal output from the physical quantity sensor 10 as a change in capacity. The sensor signals x and y are changes in capacitance. Here, instead of the sensor signal x, the sensor signal x is represented by a change amount a of the capacitance. Similarly, instead of the sensor signal y, the sensor signal y is represented by a capacitance change amount b.
[0019]
In the X-axis sensor signal x and the Y-axis sensor signal y output from the physical quantity sensor 10, it is assumed that the capacitance change of the X-axis sensor is a [pF] and the capacitance change of the Y-axis sensor is b [pF]. FIG. 2B shows that the X-axis sensor signal x has been reduced by a [pF] from the reference capacitance of 50 [pF].
[0020]
As described with reference to FIG. 1, the processing of the X-axis and Y-axis sensor signals has the same contents up to the adder 12, and therefore in the following diagrams (FIGS. 2-3 to 2-7). Processing of the X-axis sensor signal x will be described. As shown in FIG. 2-3, the first pulse width modulator 20 corresponds to the capacitance change amount a [pF], which is the output of the physical quantity sensor 10, in accordance with the pulse width modulation characteristics of the first pulse width modulator 20. The pulse width modulation signal in which the pulse width of the clock signal is changed by the time width α to be generated is generated. In FIG. 2-3, a change is given by α / 2 hours at the rising and falling times of the rectangular wave that is the clock signal, and a pulse width modulation signal having a pulse width narrower by α time than the clock signal is generated. Since the first filter 21 outputs the average of the amplitude of the pulse width modulation signal as a DC signal, the level of the DC signal decreases by a value proportional to α. The level of the DC signal that is the output of the first filter 21 is slightly lower than 1/2 of 5 Vdc when the duty ratio of the input pulse width modulation signal is 50%.
[0021]
FIG. 2-4 shows the degree of amplification by the first amplifier 22. The operation of the first amplifier 22 has been described with reference to FIG. The first amplifier 22 has a voltage equal to the average voltage (1/2 of 5 Vdc) of the pulse width modulation signal from the first pulse width modulator 20 branched and input to the first switch 23. The DC signal output from the first filter 21 is amplified. In the pulse width modulation signal from the first pulse width modulator 20, the pulse widths of continuous rectangular waves are all decreased by α. Therefore, considering the area of the ON region of the pulse width modulation signal which is a rectangular wave, only α is obtained from 1/2 (reference voltage) of the average voltage 5 Vdc of the pulse width modulation signal output from the first pulse width modulator 20. It can be seen that the value decreases. Therefore, FIG. 2-4 shows that the amplification effect in the first amplifier 22 is adjusted to a value that is reduced by α from ½ of 5 Vdc.
[0022]
FIG. 2-5 shows a rectangular wave signal output by the switching process in the first switch 23. That is, a pulse width modulation signal having a pulse width narrower by α than the clock signal is input from the first pulse width modulator 20, and the timing of the switching process in the first switch 23 is given. Also, a DC signal that is reduced by α from the reference voltage of 5 Vdc is input to the first switch 23 from the first amplifier 22. When switching processing is performed at the above-described timing according to these signal inputs, a rectangular wave signal in which the pulse width and height shown in FIG. 2-5 are reduced by α is generated and output. Considering the amount of change in pulse width and amplitude in this rectangular wave signal, the average level of the rectangular wave signal output from the first switch 23 is the same as that in the pulse modulated signal transmitted from the first pulse width modulator 20. This corresponds to the square of the change amount α of the pulse width.
[0023]
2-6 shows the level of the DC signal obtained by filtering the rectangular wave signal transmitted from the first switch 23 by the third filter 24 (the average level of the rectangular wave signal output from the first switch 23). ) Represents a voltage value of 5 Vdc-α ² .
[0024]
As for the processing of the Y-axis sensor signal, if the voltage change amount is β with respect to the capacitance change amount b [pF] output from the physical quantity sensor 10 in the second pulse width modulator 30, the above-described series of processing is performed. In the same manner, the fourth filter 34 outputs the DC signal level 5Vdc-β ² to the adder 12. FIG. 2-7 shows that 5Vdc− (α ² + β ² ), which is a value obtained by adding the DC signals related to the X axis and the Y axis by the adder 12, is output. However, since the adder 12 is a logic circuit made of a semiconductor integrated circuit, the addition in the adder 12 is performed for the changes α ² and β ² from the fixed value 5 Vdc component at the input. Further, due to the characteristics of this logic circuit, the output of the adder 12 is also at a level obtained by subtracting the change (α ² + β ² ) from the fixed value 5 Vdc. An added value can be obtained as a change in the output of the adder 12.
[0025]
By the signal processing described above, the sensor output of the orthogonal two axes is converted into the square value of the absolute amount that does not depend on the direction of the physical quantity α ² + β ² . In the next processing, the square value can be used as it is, or a square root circuit can be added and converted to a root √ (α ² + β ² ). Although the configuration in which the square value of the sensor output of the orthogonal two axes is processed by the adder through the filter is adopted in FIG. 1, the processing of FIG. As in the case of, the square value of the absolute quantity independent of the direction of the physical quantity can be obtained. In addition, when the orthogonal 2-axis sensor output of the capacitance detection type physical quantity sensor is output as a pulse width modulation value, the X-axis and Y-axis pulse width modulation output of the physical quantity sensor and the output obtained by averaging these are output by a switch. By turning on and off, the square value of the output of each axis of the original physical quantity sensor can be obtained. This value is processed by the filter described with reference to FIG. 1 and added by an adder, thereby obtaining the sum of squares of the absolute quantity independent of the direction of the physical quantity. According to this configuration, since it is not configured to convert to a voltage signal as shown in FIG. 1, the number of elements that are the cause of error can be reduced, and the total error of the switcher and the adder is only about 2%. It has been found. The processing configuration for conversion to the square value of the sensor output described above is not limited to the orthogonal 2-axis physical quantity sensor, but can be applied to an orthogonal 3-axis physical quantity sensor. What is necessary is just to take the structure which carries out process conversion to the square value of a sensor output about each of 3 axis | shaft output, and calculates the square sum for 3 axes | shafts by an adder.
[0026]
FIG. 3 is a circuit configuration diagram showing the configuration of FIG. 1 with more specific elements. Assuming that the biaxial capacitance detection type accelerometer 100 is used as the physical quantity sensor 10 of FIG. The X-axis and Y-axis capacitances, which are two orthogonal axes of the biaxial capacitance detection type accelerometer 100, have two outputs as differential capacitances (unit pF) that increase and decrease according to changes in the respective accelerations. Further, a first exclusive OR element unit 200 and a second exclusive OR element unit 300 are provided corresponding to the first pulse width modulator 20 and the second pulse width modulator 30.
[0027]
A flow of processing conversion of the X-axis sensor output of the biaxial capacitance detection type accelerometer 100 is described. First, the internal configuration of the first exclusive OR element unit 200 will be described. The rectangular wave output from the clock circuit 11 is branched and input to the first exclusive OR element unit 200. The X-axis positive and negative two sensor outputs merge with each signal line extending from the parallel resistor 201 and are connected to the exclusive OR element 202. With this configuration, the rectangular wave output of the clock circuit 11 is modulated in pulse width by the sensor capacitance of the biaxial capacitance detection type accelerometer 100 and the parallel resistance 201, and further, the exclusive OR element 202 performs the first filtering as a charge / discharge waveform. Is branched and output to the device 210 and the first switch 230.
[0028]
The first filter 210 specifically represents the first filter 21 of FIG. 1 and includes a resistor 211 and a ground capacitor 212. The operation is the same as that of the first filter 21, and the rectangular wave is filtered and converted into a DC voltage signal and transmitted to the first amplifier 22. The first amplifier 22 adjusts and amplifies the DC signal from the first filter 210 so as to be equal to the average voltage of the voltage signal input to the first switch 230, and inputs it to the first switch 230. . The first switch 230 is a C-MOS transistor. A DC signal corresponding to the square of the voltage signal output from the first pulse width modulator 200 is generated by the switching operation described with reference to FIG. 1 and output to the second filter 240. The second filter 240 specifically represents the third filter 24 of FIG. 1, and includes a resistor 241 and a ground capacitor 242. The operation is the same as that of the third filter 24, and the rectangular wave is filtered, converted again into a DC signal, and transmitted to the adder 12. The flow of processing conversion of the Y-axis sensor output of the biaxial capacitance detection type accelerometer 100 is exactly the same as the above-described X-axis processing flow. The square-processed voltage signal transmitted from the second filter 240 and the square-processed voltage signal transmitted from the fourth filter 340 are both transmitted to the adder 12, and are added in the direction of acceleration. The sum of squares of the absolute value of acceleration that does not depend can be obtained.
[0029]
【The invention's effect】
As described in the above embodiment, the present invention modulates the pulse width modulator with the detection value of the physical quantity sensor, generates a pulse width modulation signal of a rectangular wave, and detects the detection value by changing the pulse width of the rectangular wave. The average signal of the pulse width modulation signal is generated by the averaging means, and the product of the change in the pulse width and the change in the amplitude is generated by the selection means, thereby obtaining the square of the detection value of the physical quantity sensor. Yes. Further, a process of converting each of the detected values along the plurality of reference axes of the physical quantity sensor into the above-described product and finally adding all the products with an adder to calculate the sum of squares is adopted. With this configuration, in a signal processing circuit applied to a physical quantity sensor that can output a square value of an absolute quantity that does not depend on the direction of the physical quantity and has a sensitivity with respect to a range of 0 to 360 degrees, It is possible to provide a signal processing circuit that performs an accurate calculation with an inexpensive and simple configuration.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a signal processing circuit including a physical quantity sensor according to an embodiment of the invention.
FIG. 2 is a timing chart when sensor signals are processed in the configuration of FIG.
FIG. 3 is a circuit configuration diagram illustrating the configuration of FIG. 1 with more specific elements.
FIG. 4 is a circuit diagram of an operational amplifier described in P327-P332 of “Operational Amplifier” published in 1983 by McGraw-Hill, which is an example of a conventional square circuit.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 ... Physical quantity sensor 11 ... Clock circuit 12 ... Adder 20 ... 1st pulse width modulator 21 ... 1st filter 22 ... 1st amplifier 23 ... 1st switch 24 ... 3rd filter 30 ... 2nd pulse width modulator 30 ... 2nd pulse width modulator 31 ... 2nd filter 32 ... 2nd amplifier 33 ... 2nd switch 34 ... 4th filter 100 ... 2 axis capacity detection type Accelerometer 200 ... first exclusive OR element 201 ... parallel resistor 202 ... exclusive OR element 210 ... first filter 211 ... resistor 212 ... grounding capacitor 230 ... first switch 240 ... second Filter 241 ... Resistor 242 ... Ground capacitor 300 ... Second exclusive OR element 301 ... Parallel resistor 302 ... Exclusive OR element 310 ... Second filter 311 ... Resistor 312 ... Ground capacitor 330 ... Second Switcher 340 ... Fourth filter 41 ... resistance 342 ... ground capacitor

Claims (4)

A square value signal representing a certain amount A by a change amount of a pulse width of the rectangular wave in the first rectangular wave signal of the repetition frequency f, processing the first rectangular wave signal, and representing a square value A ² of the amount A Is a square circuit that generates
First averaging means for generating a first average signal representing an average amplitude of the first rectangular wave signal;
The first average signal and a constant value such as a ground voltage are alternately selected in synchronization with the first rectangular wave signal, and the amplitude is a difference value between the first average value and the constant value. Selecting means for generating a second rectangular wave signal whose pulse width is the pulse width of the first rectangular wave signal;
And a second averaging means for generating a second average signal representing an average amplitude of the second rectangular wave signal and outputting a change amount of the second average signal as the square value signal. A square circuit. The first averaging means includes a first filter for inputting the first rectangular wave signal, and an amplitude for generating the first average signal by amplifying or attenuating the output of the first filter. With the adjustment circuit,
2. The squaring circuit according to claim 1, wherein the second averaging means is a second filter having the second rectangular wave signal as an input. When acceleration and other vector quantities are input, the components of the vector quantities relating to the n-axis (n is 2 or 3) orthogonal to each other are detected, and the n-axis sensor outputs each component of the n-axis. In the signal processing circuit that generates the sum of squares of each component of the n-axis,
An n number of square circuits according to claim 1 and an adder that adds the square signals that are outputs of the n number of square circuits and outputs the sum of the squares.
The n square circuits are provided corresponding to the n-axis,
Each square circuit includes a pulse width modulator that generates the first rectangular wave signal;
Each of the pulse width modulators modulates the pulse width of the first rectangular wave signal according to the corresponding component of the axis. It has a clock circuit that outputs a clock signal that is a rectangular wave signal,
4. The pulse width modulator generates the first rectangular wave signal by receiving the clock signal and modulating the pulse width of the clock signal with a corresponding component of the axis. A signal processing circuit according to 1.

Images