JPS6125101B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a zero-crossing point detection device for detecting a target zero-crossing point of one frequency from a superimposed signal of two or more frequencies. Furthermore, the object is to provide a device that can be easily detected.

First of all, apply this device and see the effect 1
As an example of use, there are the following methods for detecting the position of a train running between points A and B on a track. In other words, a guide line with an appropriate delay time for signal transmission is laid along the track on which the train runs, and from point A the frequency Fa (HZ),
Send a signal with frequency Fb (HZ) from point B. Assuming that the train is moving from point A to point B, a signal is received on the train in which the phase of signal Fa is delayed and the phase of signal Fb is advanced. By comparing the phases of both signals thus received, the position where the train is running can be determined. In this case, two signals Fa and Fb are transmitted in a superimposed manner on the guide wire, so in order to compare the phases of both signals, it is necessary to know the 0° phase (zero crossing point) of each signal. An easy way to do this is to use a normal LC filter to separate the signals at frequencies Fa and Fb and shape the waveform, but the phase between the input and output of the LC filter must be corrected to be exactly the same for frequencies Fa and Fb. That's almost impossible. In such cases, if the apparatus of the present invention is used, the Xerox point can be detected easily and with high precision.

Next, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a circuit constituting the present device, in which a sampling circuit 1 receives an input signal a(t).
This is a circuit that takes in (in the case of the above usage example, a signal in which frequencies Fa and Fb are superimposed) for a sufficiently short time compared to the sampling period. A smoothing circuit 2 connected next smoothes the output wave from the sampling circuit 1. Reference numeral 3 denotes a pulse oscillation circuit whose oscillation frequency can be controlled by the voltage applied to its input, and generates a clock pulse for controlling the variable phase shift circuit 4 at the next stage. The variable transfer circuit 4 can control the transmission time of the analog signal between its input and output depending on the frequency of the clock pulse.
There are some integrated circuits known as Bucket Brigade Devices (BBD). A bandpass filter 6 connected between the input circuit of the sampling circuit 1 and the input of the variable transfer circuit 4 selects a target frequency (for example, frequency Fa) at which a zero-crossing point is to be detected from the input signal a(t). There is no problem as long as the filter is for extracting only a signal and the phase delay between its input and output is within the range that can be controlled by the variable phase shift circuit 4. A delay equalizer may be used if necessary.

The pulse generating circuit 5 is a circuit that generates sampling pulses of the same frequency selected by the band filter 6 and synchronized with one target frequency (eg, frequency Fa) sent through the variable phase shift circuit 4.

Figure 2 shows an operational amplifier consisting of an amplifier A, resistors R ₁ and R ₂ , and a capacitor C. The circuit shown in Figure 1 is when the aforementioned operational amplifier is used in the smoothing circuit 2 of Figure 1. The transfer function of is calculated as follows.

That is, Φr/Φi=Ka/Y+Ka... However, Φr...response function for the phase of the feedback circuit Φi...input function for the phase of the feedback circuit a=1/R1C, Y...complex frequency K=K ₁・K ₂・K ₃ or K≫1/R ₂ holds K ₁ ... Gain of sampling circuit 1 K ₂ ... Gain of smoothing circuit 2 K ₃ ... Gain of pulse oscillation circuit 3 and variable phase shift circuit 4 where K ₂ < Since K ₃ <O, K ₁ >O, the transfer function expressed by the formula has a pole on the negative real axis of the complex plane. It can be seen that the operation of transfer circuit 4→pulse generation circuit 5→sampling circuit 1) is stable.

Therefore, according to the time chart in Figure 3,
To explain the operation of the circuit shown in the figure, chart a in FIG. 3 shows the waveform of input signal a(t), and to simplify the explanation, only the signal whose zero crossing point is to be detected is shown. The reason why only the signals whose zero-crossing points are to be detected is shown and explained is as follows.

That is, the input signal a(t) actually has a frequency Fa
and Fb, but assuming that the signal whose zero crossing point should be detected is the frequency Fa, the band filter 6 is set to extract only this frequency Fa, so the pulse generating circuit 5
outputs a sampling pulse of the same frequency synchronized with this frequency Fa. Therefore, when the input signal a(t) is sampled in the sampling circuit 1 using the sampling pulse of the frequency Fa, the signal of the frequency Fa of the input signal a(t) is always sampled at the same time position every cycle. do. On the other hand, the signal of frequency Fb is sampled at different time positions every cycle. This frequency is applied to smoothing circuit 2.
A composite wave of the sampling output for the two signals Fa and Fb will be input, but since the sampling output component of the signal at frequency Fb has a completely random sampling position with respect to frequency Fb, it is smoothed by smoothing circuit 2. When converted to direct current, the potential becomes zero and is erased. However, since the sampling output of frequency Fa that is always sampled at the same time position is a pulse train consisting of the same potential and polarity, the smoothed DC output becomes a certain DC potential with a specific polarity, and the smoothed output of the smoothing circuit 2 becomes only the signal component of the target frequency Fa whose zero crossing point is to be detected.

Therefore, even if the signal with the non-target frequency Fb is omitted, there will be no problem in explaining the operation, so the chart a in FIG. Only the above is shown. Chart b is the waveform of the sampling pulse when the phase is advanced with respect to the input signal of chart a, and chart c is the waveform of the input signal sampled by the sampling pulse of chart b, which is smoothed by the smoothing circuit 2. In this case, the output of the smoothing circuit 2 becomes a negative voltage v. Similarly, chart d shows a case where the sampling pulse has a delayed phase.
In this case, the output of the smoothing circuit 2 becomes a positive voltage v as shown in the chart e.

The pulse oscillation circuit 3 is configured such that the oscillation frequency becomes higher as the output voltage v of the smoothing circuit 2 that controls the oscillation frequency becomes higher. Therefore, when the phase of the sampling pulse is delayed, the frequency of the clock pulse input to the variable phase shift circuit 4 increases, and the signal transmission time between the input and output of the variable phase shift circuit 4 becomes shorter. This acts to advance the phase of the sampling pulse generated by the pulse generator 5, thereby correcting the phase shift between the signal and the sampling pulse.

Next, frequency selection characteristics will be explained. Since noise can be synthesized using a sine wave, a sine wave with one frequency will be explained. Now input an(t)=sin
(2πfnt+) and the frequency of the sampling pulse is fs, the output of sampling circuit 1 is bn(t)=sin(2πfnN/fs+)...where N=0, 1, 2,..., an integer. expressed. When fn=Mfs (M is an integer), the formula is bn(t)=sin(2πNM+)=sin... When the noise represented by the equation is smoothed by the smoothing circuit 2, it becomes a direct current, which causes an error in zero-crossing point detection. However, in the above usage example, if the frequencies Fa and Fb are determined to be integral multiples of frequencies, position detection becomes impossible. Therefore the frequencies Fa and Fb
This device can be used by making sure that and is not an integral multiple.

Furthermore, mfn=Mfs (m≠1, M≠1,
When m≠M (an integer), it is expressed as bq(t)=sin(2π·MN/m+)... This is a pulse amplitude modulation (PAM) wave with one period of m pulses, and the average value of the amplitude is 0. That is, the smoothing circuit 2 can sufficiently attenuate it to a required value. if
If mfn=Mfs does not hold, an appropriate integer m′,
Since it is possible to approximate with the required accuracy using M′, it is possible to sufficiently attenuate as in the case of Eq.

Also, if there is no correlation between the sampling pulse and the noise, the variance of the average sample value is equal to the variance of the noise function divided by the sample size, and the degree of improvement in the signal-to-noise ratio (S/N) G is G=10logS, where S is expressed by the sample size.

For the reasons stated above, it is sufficiently practical if the conditions of use are limited as in the case of the above-mentioned usage example.

Note that the pulse that coincides with the zerox point of the input signal as an output signal is generated by the sampling pulse generation circuit 5.
A sine wave whose phase matches that of the input signal is obtained from the output of the variable phase shift circuit 4. In the case of the above usage example, two circuits shown in FIG. 1 are used to obtain pulses at the zerox point of signals of frequencies Fa (HZ) and Fb (HZ), and the phase between these pulses is examined.

Since the present invention has the configuration and operation as described above, it is possible to easily and highly accurately detect the zero-crossing point of a target signal of one frequency from a mixed signal of two or more frequencies superimposed. This has an extremely excellent effect of being able to obtain a zero-crossing output that is completely in phase with the detected frequency.

[Brief explanation of the drawing]

The drawings relate to an embodiment of the signal wave zero-crossing point detecting device of the present invention; FIG. 1 is a block diagram of a circuit constituting the device; FIG. 2 is an example diagram of an operational amplifier used in a smoothing circuit; FIG. 3 is a time chart showing the operation of the circuit shown in FIG. 1: Sampling circuit, 2: Smoothing circuit, 3: Clock pulse oscillation circuit, 4: Variable transfer circuit, 5:
Sampling pulse generation circuit.

Claims (1)

[Claims] 1. A sampling circuit that samples a superimposed input signal of two or more frequencies using a sampling pulse sent from a pulse generation circuit, a smoothing circuit that smoothes the sampling output of the sampling circuit to convert it into a direct current, and the smoothing circuit. a pulse oscillation circuit that generates a clock pulse whose oscillation frequency increases as the DC output voltage of the input signal increases; a bandpass filter that extracts only the target frequency signal whose zero-crossing point is to be detected from the input signal; a variable phase shift circuit that controls the signal transmission time between the input and output of the frequency signal extracted by the band filter so that it becomes shorter as the oscillation frequency of the clock pulse of the pulse oscillation circuit becomes higher; A signal wave zero-crossing point detection device comprising a pulse generation circuit that generates synchronized sampling pulses of the same frequency.