JPH0130083B2 - - Google Patents

Description

[Detailed Description of the Invention] This invention is applicable to the case where a fixed periodic vibration wave such as power supply noise or mechanical vibration is superimposed on the output of a detector such as a load cell or a differential transformer. This invention relates to a method for removing periodic vibration waves from output.

First, two of the inventors of this invention collaborated with others to sample the output of a detector on which a periodic oscillation wave was superimposed, at a sampling frequency that was approximately an integral multiple of the fundamental frequency of the periodic oscillation wave. The process of
A method for removing periodic oscillating waves has been proposed, which includes the step of calculating the average value of each of the sampling values corresponding to approximately an integer period of the periodic oscillating waves, which are shifted by one period of the sampling frequency. This method takes less time to measure than conventional removal methods, but since it calculates the average value of sampling values that correspond to approximately an integer period of the periodic oscillation wave, it There was a delay in measurement until a sampling value corresponding to approximately an integer period of the vibration wave was obtained. Especially recently,
A method of estimating the weight of an article to be weighed when it is placed on a scale by observing the movement of the pointer while it continues to vibrate has been considered. In this case, a small change in the situation can cause a large error in the results, so it is necessary to faithfully reproduce the measured waveform without time delay.

SUMMARY OF THE INVENTION An object of the present invention is to provide a periodic oscillation wave removal method for further shortening measurement time and faithfully reproducing measured waveforms.

This invention first samples the approximately integer period of the periodic oscillation wave superimposed on the output of the detector at a sampling frequency that is approximately an integer multiple of the frequency of the periodic oscillation wave, and then samples the periodic oscillation wave corresponding to the same phase. The average value is taken for each value, the average sampling value for one cycle is taken, and the output of the detector on which the constant frequency oscillation wave is superimposed is sampled at the sampling frequency, and from these sampling values, the corresponding constant period oscillation wave is obtained. This is a method of removing periodic oscillation waves by sequentially subtracting the average sampling value of .

The present invention will be explained below based on two illustrated embodiments. The first embodiment is applied when only power supply noise is superimposed on the output of the detector. First, as a preparatory work, the detector 1 shown in FIG.
For example, by setting the excitation voltage of the load cell to zero, only power supply noise 2 is generated on the output side of the detector 1 as shown in FIG. Then, the frequency multiplier 5 multiplies the synchronization signal generated by the disturbance synchronization signal generator 4 every cycle T of the power supply noise 2 in response to the signal from the power supply 3 by an integral multiple of the power supply noise 2. A/D
Generates a conversion command signal. Every time this A/D conversion command signal is generated, the power supply noise 2 amplified by the amplifier 6 is sequentially converted into digital data by the A/D conversion section 7. For example, ten cycles of these digital converted values are sequentially stored in the disturbance storage section 9 via the arithmetic unit 8. Then, the calculation unit 8 calculates the average value of the digital conversion values having the same phase in each cycle.
These average values are stored in the disturbance storage section 9 as noise sampling values. These noise sampling values are 2a, 2b, 2 shown in Figure 2.
This corresponds to digitized versions of c, 2d, 2e, and 2f. These noise sampling values are stored in the disturbance storage section 9 as 2a, 2b, 2c, 2d, 2
They are stored in the order of e and 2f. As the disturbance synchronization signal generator 4, one that generates, for example, an H level output when the input is greater than the ground potential after aligning the phase of the signal from the power supply 3 with the phase of the power supply noise 2 can be used. . According to this, the comparator generates an H level signal during a period corresponding to the positive half cycle of power supply noise 2. Each time this H level signal rises, it is determined that one period T has elapsed, and this rise is used as a synchronizing signal.

When the above preparations are completed, an excitation voltage is applied to the detector 1, and an output signal 10 on which the power supply noise 2 is superimposed is applied to the output side of the amplifier 6 as shown in FIG.
to occur. This output signal 10 is converted into a digital signal each time an A/D conversion command signal is supplied from the frequency multiplier 5 to the A/D converter, and these digital signals are shown in FIG.
The data shown as 0c, 10d, 10e, and 10f are digitized and are sequentially supplied to the arithmetic unit 8. The arithmetic unit 8 includes the disturbance synchronization signal generator 4
Using the synchronization signal from the A/D converter 7 as a start signal for calculation, each time a digital signal is supplied from the A/D converter 7, a noise sampling value corresponding to the input digital signal is read from the disturbance memory 9. , calculate the difference between this digital signal and the read noise sampling value.
For example, when the disturbance synchronization signal generator 4 generates a synchronization signal, the arithmetic unit 8 enters a calculation start state. First, when a digital signal corresponding to 10a is input, the noise sampling value corresponding to 2a is input from the disturbance storage section 9. The noise sampling value corresponding to 2a is subtracted from the digital signal corresponding to 10a. That is, the disturbance synchronization signal generator 4
The noise sampling value and the digital signal are synchronized by the synchronization signal from the A/D converter 7 and the digital signal from the A/D converter 7. The arithmetic unit 8 performs such subtraction every time a digital signal is supplied. As can be seen from FIG. 2, each of these subtracted values corresponds to the output signal 10 of the detector 1 with the power supply noise 2 removed.

The second embodiment is applied to the case where power supply noise 12 and mechanical vibration 14 of a lower frequency are removed from the output signal 16 of the detector 1 on which these are superimposed. First, the first preparatory work is performed. as the first
In the same manner as in the embodiment described above, the averaged noise sampling values for one cycle (digitized values of 12a, 12b, 12c, and 12d shown in FIG. 4) are stored in the disturbance storage section 9. The sampling frequency at this time is power supply noise 12 and mechanical vibration 14.
It is desirable that it is an integral multiple of both. Next, as a second preparatory work, mechanical vibration is added to the state of the first preparatory work. The output of the amplifying section 6 at this time is a superimposition of the power supply noise 12 and the mechanical vibration 14. This is sampled at the above-mentioned sampling frequency in the A/D conversion section 7 and sequentially supplied to the calculation section 8. The calculation unit 8 reads the noise sampling value corresponding to the supplied sampling value from the disturbance storage unit 9, subtracts the read noise sampling value from the supplied sampling value, and stores it in the storage unit 24. These correspond to how many cycles (for example, 5 cycles) of the sampled values of the mechanical vibration 14. Next, using the vibration detector 20 and the frequency multiplier 22, the arithmetic unit 26 calculates the average value of the same phase among the values stored in the storage unit 24, and uses this as the sampling value of the mechanical vibration 14 (the fourth 14a to 14j shown in the figure) are sequentially stored in the disturbance storage unit 18.

After the above preparations are completed, an excitation voltage is applied to the detector 1. Then, as in the first embodiment, an output signal 16 in which power supply noise 12 and mechanical vibration 14, which is a composite of two signals indicated by dotted lines in FIG. The D converter 7 is used to perform sampling, and the arithmetic unit 8 and disturbance storage section 9 are used to subtract the corresponding noise samplings from these sampling values. These subtraction values are 16a to 16 shown in FIG.
This is a digital version of m. These subtraction values are sequentially stored in the storage unit 24, and the frequency multiplier 2
16a-14a, 16 using the arithmetic unit 26 and the disturbance storage unit 18 every time a signal is supplied from 2.
b-14b, 16c-14c...16i-14i,
Subtraction is performed as 16j-14j, 16k-14a, and so on. Note that the sampling frequency oscillated by the frequency multiplier 22 is equal to the sampling frequency of the frequency multiplier 5 and the frequency of the power supply noise 12, as is clear from the relationship with the power supply noise waveform shown by the dotted line in FIG. (For example, if mechanical vibration is 6Hz and power supply noise is 50Hz, the sampling frequency is 150Hz.)
Hz). If it is not an integral multiple, this subtracted value (output of the arithmetic unit 26) cannot completely eliminate mechanical vibrations, but if a large number of mechanical vibrations are sampled, errors will not occur much.

In such a removal method, the noise sampling values obtained by sampling only power supply noise and mechanical vibration are stored in the disturbance storage sections 9 and 18 before the start of measurement, and after the start of measurement, each of the output signals of the detector 1 is Since the corresponding noise sampling values are sequentially subtracted from the sampling values, power supply noise and mechanical vibration can be removed immediately after measurement starts, and the measured waveform can be faithfully reproduced. In both embodiments, the disturbance storage units 9 and 18 store the average value of sampling values of the same phase, but it is also possible to simply store one period worth of samples.

In the above second embodiment, for convenience of explanation, both the power supply noise 12 and the mechanical vibration 14 do not contain harmonics, but even if they contain harmonics, they can be removed in the same way.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a device used in the first embodiment of the method for removing periodic vibration waves according to the present invention, FIG. 2 is an explanatory diagram of the method for removing periodic vibration waves in the first embodiment, and FIG. FIG. 4 is a block diagram of the apparatus used in the second embodiment, and FIG. 4 is an explanatory diagram of the removal method of the second embodiment. 1...Detector, 5, 22...Frequency multiplier,
7... A/D conversion section, 8, 26... Arithmetic device,
9, 18...Disturbance memory section.

Claims (1)

[Claims] 1. In a detector in which a periodic oscillation wave is superimposed on the output in an excited state, the detector is set in a non-excited state to generate the periodic oscillation wave, and approximately one period of the periodic oscillation wave is generated. A first process of obtaining a sampling value by sampling at a sampling frequency that is approximately an integral multiple of that frequency, and sampling the output of the detector in an excited state on which the periodic oscillation wave is superimposed at the sampling frequency. a second process to calculate the corresponding first values from each sampling value obtained in the second process.
and a third process of sequentially subtracting each sampling value obtained in the process. 2. In a detector in which two periodic oscillation waves are superimposed on the output in an excited state, one of the two periodic oscillation waves is generated when the detector is in a non-excited state, and approximately one of the generated periodic oscillation waves is superimposed on the output. a first process of obtaining a sampling value by sampling the period at a sampling frequency that is approximately an integer multiple of the frequency of the two periodic oscillation waves; and a step of setting the detector in a non-excited state and generating the two periodic oscillation waves. a second process of obtaining a sampling value by sampling approximately one period of the other one at the sampling frequency; and a second process of obtaining a sampling value by sampling approximately one period of the other one at the sampling frequency; A third process of sampling with
and a fourth process of sequentially subtracting each sampling value obtained in the first and second processes, which substantially correspond to each sampling value obtained in the third process, from each sampling value obtained in the third process. Vibration wave removal method.