JPH06288787A - Method for correcting zero point error of torque sensor - Google Patents

Abstract

PURPOSE:To execute the correction of the zero point signal value by correctly measuring the zero point signal value of a torque sensor which is continuously used at a place where the torque signal is operated in the impulse shape. CONSTITUTION:The measurement of the torque for every constant period T is started, and the torque signal value Vtorque (t) is measured for every constant time T0 after starting. When the absolute value of the measured torque signal values of the specified number is under the constant value (a), the mean value of the torque signal values of these specified numbers is set to be a new zero point.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for correcting a zero point error of a torque sensor such as a magnetostrictive torque sensor or a strain gauge type torque sensor.

[0002]

2. Description of the Related Art For example, as a magnetostrictive torque sensor,
A magnetic anisotropy portion is formed on the outer circumference of the torque detection shaft, and a change in the magnetic permeability of the magnetic anisotropy portion when torque is applied to this shaft is measured by a coil provided near the magnetic anisotropy portion. It is known that the detection is performed in. In this magnetostrictive torque sensor, changes in the ambient temperature are detected in the sensor unit in which the magnetic anisotropy portion is formed on the shaft, the excitation circuit for exciting the coil, and the detection circuit for processing the signal from the coil. If a secular change occurs, it causes a zero point error in the torque signal value.

Therefore, in the prior art, in order to correct this zero point error, the torque signal value when the torque is not loaded on the torque detection shaft, that is, the zero point signal value is obtained, and this zero point signal value becomes a certain value or less. There have been proposed many methods for performing such correction.

On the other hand, in an impulse tool for screw tightening, a screw tightening electric tool, a nut runner, a screw tightener, and a torque tester for inspecting the performance of these, in order to measure the operating torque in real time, It has been proposed to incorporate the above-mentioned magnetostrictive torque sensor and the like.

[0005]

However, in an impulse tool for screw tightening or the like, the torque load acts in an impulse manner, so that torque is applied to the torque detection shaft when the tool is continuously operating. There is a problem in that it is a difficult task to select only when there is no torque and obtain the torque signal value, that is, the zero-point signal value at that time.

Therefore, the present invention solves such a problem, and corrects the zero-point signal value of the torque sensor continuously used at the place where the torque load acts in an impulse manner so that the correction can be performed. The purpose is to do.

[0007]

In order to achieve this object, the present invention starts a torque measurement at a constant period, measures a torque signal at a constant time after the start, and measures a predetermined number of torques. When the absolute values of the signal values are below a certain value, the average value of these predetermined numbers of torque signal values is set as a new zero point.

[0008]

By doing so, the average value is set as a new zero point only when the absolute values of the measured torque signal values of a predetermined number are each less than or equal to a fixed value.
Even if the start of torque measurement at fixed intervals and the torque acting on the shaft are not synchronized, the torque signal value when torque is acting is not used as the reference for zero point setting. A new zero is set exactly based only on the torque signal value when not loaded. Since the zero point is obtained by averaging a plurality of torque signal values, it is less likely to be affected by ripples superimposed on the torque signal, and an accurate zero point is set also from this point. Since a new zero point is obtained and a zero point correction is performed at regular intervals, zero point shift due to hysteresis and zero point drift due to changes in ambient temperature are canceled.

[0009]

EXAMPLE FIG. 5 exemplifies a waveform of a torque signal value of a torque sensor incorporating an impulse tool. The vertical axis is the signal output voltage, and the horizontal axis is the time axis, which is 100 msec / div. Figure 6
Shows one of the waveforms in FIG. 5 enlarged in time.
The scale of the vertical axis is the same as in Fig. 5, but the horizontal axis is 1 msec / di
enlarged to v. FIG. 7 shows an enlarged view of the portion A in FIG. 6, that is, the portion where no torque acts, in both the vertical and horizontal axes. It can be seen that ripple is added as noise.

FIG. 1 shows that a deviation due to hysteresis H occurs in the torque signal value at the time when the impulse torque is not loaded. However, the hysteresis H is exaggerated for easy understanding. In this case, if the zero point correction is not performed, an error will occur in the torque measurement value.

Next, a method of correcting the zero point will be described. When the torque measurement is activated every fixed period T, and when this torque measurement is activated, the torque signal value Vtorque (t) is sampled and measured continuously n times at every constant time T0. The torque signal value at this time is Vtorque (t ₁ ), Vtorque
(T ₂ ), ..., Vtorque (t _n ).

Then, each torque signal value Vtorque (t ₁ ),
When the absolute values of Vtorque (t ₂ ), ..., Vtorque (t _n ) are equal to or less than a constant value a, the torque signal value Vtorq
ue (t _i ), Vtorque (t _{i + 1} ), ..., Vtorque (t
The average value of _{n-j + 1} ) [i ≧ 1, j ≧ 1] is set as a new zero point. The positive or negative of the torque signal value changes depending on the direction of torque application.

When the absolute value of the torque signal value exceeds the constant value a, no new zero point is set. Further, in the above, when i = 1 and j = 1 are set, all the sampled torque signal values are used for the calculation of the average value. However, when i is set to 2 or more, the average value is calculated from a large number of sampled and measured torque signal values except for the first part. When j is set to 2 or more, the average value is calculated except for the last part of many torque signal values. After all, by setting both i and j to 2 or more, only a plurality of torque signal values in the remaining intermediate part except for the first part and the last part of a large number of torque signal values are used. The zero point is calculated.
By doing so, a new zero point can be accurately obtained by using only data having a small absolute value.

Next, a specific example will be described with reference to FIG.
Here, period T = 20 msec, constant time for sampling T0 = 0.5 msec, sample number n = 9, i = 3, j =
3. When the voltage of the torque signal value when the torque is applied is at a level of several V, the constant value a = 100 mV.

At this time, the torque measurement is started every 20 msec of the constant period, and after the start, the torque signal value Vtorque (t) is sampled and measured 9 times continuously every 0.5 msec of the constant time. Then, the result is stored in a memory or the like by A / D conversion or the like. The torque signal value at this time is V
torque (t ₁ ), Vtorque (t ₂ ), ..., Vtorque (t ₉ ).

If the absolute value of these torque signal values is 100 mV or less, the torque signal values Vtorque (t ₃ ) and Vtorque
The average value of five data of (t ₄ ), ..., Vtorque (t ₇ ) is calculated, and the value is applied as a new zero point, and the magnitude of the applied torque is obtained from the subsequent torque signal value.

By doing so, the deviation of the zero point due to the presence of the hysteresis H shown in FIG. 1 is corrected, and the next applied torque value is accurately measured. In this way, since a new zero point is calculated and updated by using only data whose absolute value is equal to or less than the constant value a, the torque signal value at the time of torque application is not used as data for calculating the zero point. Even if the pulse waveform of the torque signal at the time of applying the torque and the start of the torque measurement are not synchronized, the new zero point value can be accurately calculated.

Further, as described above, a new zero point can be calculated by using only a plurality of torque signal values in the remaining intermediate portion excluding the first portion and the last portion of a large number of torque signal values. , More accurately, a new zero point is sought.

Moreover, since the zero point is obtained by averaging a plurality of torque signal values, it becomes difficult to be influenced by the ripple superimposed on the torque signal, and the zero point can be obtained accurately.

FIG. 2 shows a state in which the horizontal axis, that is, the time axis is extremely compressed as compared with FIG. Here, in addition to the hysteresis H, a zero-point drift D that occurs due to a change in ambient temperature is also shown. However, again,
The hysteresis H and the zero point drift D are exaggerated for easy understanding. As shown in FIG. 1 and FIG. 2, according to the present invention, a new zero point is obtained and a zero point correction is performed every fixed period T. Therefore, not only the shift of the zero point based on the hysteresis H can be corrected but also the zero point drift D is caused. The underlying zero shift can also be corrected.

FIG. 3 shows an example of the circuit configuration of the magnetostrictive torque sensor. Here, 10 is a torque detection shaft, and the magnetic anisotropy portion 12 inclined with respect to the axis is formed at only one location on the outer peripheral surface thereof. A coil is placed around the magnetically anisotropic portion 12.
14 are arranged.

The coil 14 supplies an alternating current to the coil 14 via a resistor 16 and a temperature-sensitive resistor 18 shown by a solid line in the figure for coarsely adjusting a zero-point variation due to a temperature change. Is connected to the oscillator circuit 20 for. The temperature-sensitive resistor 18 can be installed between the resistor 16 and the coil 14 as shown by a broken line in the figure, or can be installed in another place. The resistor 16 has a resistance value substantially equal to the impedance of the coil 14. Oscillator circuit
An exciting voltage detecting circuit 22 having a rectifying circuit and a low-pass filter is connected in parallel between the temperature sensitive resistor 20 and the temperature sensitive resistor 18.

The output line from the coil 14 constitutes a torque detecting circuit 24 having a rectifying circuit and a low pass filter. The output side of the torque detection circuit 24 and the output side of the excitation voltage detection circuit 22 are both connected to the input side of the differential amplifier 26. The output side of the differential amplifier 26 is connected to a differential amplifier 28 for fine adjustment of zero point fluctuation. In the differential amplifier 28, zero point data for making the sensor output zero when it is determined that the torque signal value is equal to or less than the constant value a and the torque is zero is stored in the constant voltage generating circuit.
It is input through the D / A converter 32 in which 30 is arranged in parallel.

The output side of the differential amplifier 28 is connected via an amplifier 34 to a polarity judgment circuit 36 for detecting the direction of the torque applied to the torque detection shaft 10 by a known method, and a gain switching circuit 38. ing. The gain switching circuit 38 is a polarity determination circuit.
Based on the output signal of 36, the gain of the torque detection signal can be switched according to the direction of the torque applied to the torque detection shaft 10. 40 is a torque signal line, and the above-mentioned torque signal value appears.

An output line from the arithmetic control circuit 42 is connected to the D / A converter 32 and the gain switching circuit 38. A memory 44 is provided in parallel with the arithmetic control circuit 42, and a data input / output line 46 is connected to the arithmetic control circuit 42. Reference numeral 48 is a temperature sensor, the output line of which is connected to the arithmetic and control circuit 42 through the A / D converter 50. 52 is a feedback line from the torque signal line 40 to the A / D converter 54
It is connected to the arithmetic and control circuit 42 via.

According to this structure, the temperature-sensitive resistor 18 acts to roughly adjust the zero-point fluctuation associated with the temperature fluctuation. The torque signal value appearing on the torque signal line 40 is
It is input to the arithmetic control circuit 42 via the feedback line 52 and subjected to arithmetic processing. Then, in order to correct the shift of the zero point due to the presence of the hysteresis H and the zero point drift D that occurs due to the change of the ambient temperature, the zero point data is obtained as described above, and the zero point data is D / It is input to the differential amplifier 28 via the A converter 32.

In the gain switching circuit 38, sensitivity correction and correction of sensitivity fluctuation due to temperature change are performed based on the signal from the arithmetic control circuit 42. As shown in FIG. 1, although the sampling is performed only about 10 times at every constant time T0 = 0.5 msec, the start of the torque measurement for the correction of the zero point fluctuation is performed at the constant cycle T.
= It takes place every relatively long cycle of every 20 msec. The reason is that the arithmetic processing circuit 42 has to perform many arithmetic processings such as correction of the sensitivity variation due to the temperature change described above in addition to the zero point correction, and secures the time.

FIG. 4 shows an example of a digital correction type magnetostrictive torque sensor. Here, the output line from the differential amplifier 26 is connected to the arithmetic control circuit 42 via the A / D converter 56. The torque signal line 40 is derived from the arithmetic control circuit 42 via the D / A converter 58.

In the digital correction type torque sensor 1, the operation control circuit 42 performs the correction operation of the zero point according to the present invention and various corrections similar to those of other analog type.
In digital operation.

In the case where the correction is performed by the digital calculation as described above, in order to accurately measure the pulse torque signal as shown in FIG. 1, this torque signal is converted into an A / D converter.
56 must be converted at high speed. Specifically, for example, the sampling interval is 0.02 msec, the number of samples is 1
00 to 200 is required.

The case where the method of the present invention is applied to a magnetostrictive torque sensor has been described above as a specific example.
Of course, it can be applied to other types of torque sensors such as a strain gauge type torque sensor. Further, it is obvious that the technical idea of the present invention can be applied to various sensors such as a load cell and a pressure sensor other than the torque sensor described above.

[0032]

As described above, according to the present invention, the average value is set as a new zero point only when the absolute values of the measured predetermined number of torque signal values are equal to or less than the fixed values. Even if the start of torque measurement for each cycle and the torque acting on the shaft are not synchronized, the torque signal value when torque is acting is not used as the reference for zero point setting, and therefore torque is not applied to the load. A new zero point can be accurately set based only on the torque signal value when not being performed. Moreover, since the zero point is obtained by averaging a plurality of torque signal values, the influence of ripple superimposed on the torque signal is less likely to occur, and the zero point can be set accurately from this point as well. Furthermore, since a new zero point is obtained and a zero point correction is performed at regular intervals, zero point shift due to hysteresis and zero point drift due to changes in ambient temperature can be canceled.

[Brief description of drawings]

FIG. 1 is a diagram showing a state in which a zero point is changed due to a deviation due to hysteresis in a torque signal value when an impulse-shaped torque is not applied.

FIG. 2 is also a diagram showing a zero-point drift caused by a change in ambient temperature by compressing the time axis as compared with FIG.

FIG. 3 is a diagram showing an example of a circuit configuration of a magnetostrictive torque sensor to which the present invention is applied.

FIG. 4 is a diagram showing another example of the circuit configuration of the magnetostrictive torque sensor to which the present invention is applied.

FIG. 5 is a diagram illustrating a waveform of a torque signal value of a torque sensor incorporating an impulse tool.

6 is a diagram showing one of the waveforms in FIG. 5 in an enlarged scale with respect to time.

7 is an enlarged view showing a portion A in FIG.

[Explanation of symbols]

 H Hysteresis D Zero point drift T Constant period T0 Constant time a Constant value

Claims (2)

[Claims] 1. A torque measurement is activated at regular intervals,
The torque signal is measured at fixed time intervals after startup, and when the absolute values of the measured torque signal values are below a certain value, the average value of these torque signal values is set as a new zero point. A method for correcting a zero point error of a torque sensor, comprising: 2. A plurality of remaining torque signal values excluding a first partial torque signal value and a final partial torque signal value out of a predetermined number of torque signal values measured at fixed time intervals. 2. The average value is set as a new zero point.
A method for correcting a zero point error of the torque sensor described.