JPH0374326B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an absolute linear position detection device that expands the absolute detectable range.

A linear position detector (or transducer) that converts the mechanical linear position of the object to be detected into an electrical signal.
There is a known actuating transformer, and there is a practical application currently pending by the applicant.
There is one disclosed in No. 56-22075. The disadvantage of these methods is that the measurable range is limited to a relatively narrow range. Therefore, in the specification of Utility Model Application No. 57-32127, a plurality of cores are provided at predetermined intervals in a core part that moves relative to a pick-up part consisting of a primary coil and a secondary coil to expand the measurable range. It is proposed that However, in the case disclosed in the same prior application, although the measurable range is expanded,
Only the linear position within one cycle of each core could be detected, and the range that could be detected absolutely was not expanded.

Furthermore, Japanese Patent Publication No. 50-23618 discloses a linear position detection device using a magnetic scale consisting of a magnetic track having a large number of magnetic gratings with a predetermined magnetization, and two magnetic gratings with different wavelengths. It has been shown that by providing the above magnetic tracks and comparing the phases of the phase modulated output signals of each magnetic track, an output signal with an expanded absolute position detection range can be obtained. However, this conventional device does not have sufficient countermeasures against output errors of the detector and data errors during calculation. When using two or more detectors or detection scales, if the origin of each detector is not aligned accurately, the data will be inaccurate near the cycle switching point of the detector output signal, causing errors. arise. Furthermore, when digital calculation is performed, a quantized data error similarly occurs near the cycle switching point of the detector output signal.

This invention was made in view of the above points,
Positioning that expands the absolute linear position detection range by using multiple linear sensors that each generate periodic position detection signals with different periods depending on the mechanical displacement of the detection target, and calculating the output signals. It is an object of the present invention to provide an absolute linear position detection device capable of obtaining a detection signal, in which an error that may occur near a cycle switching point can be appropriately corrected.

The absolute linear position detection device according to the first invention comprises (a) a linear scale member in which two substances exhibiting different magnetic properties are alternately arranged at a predetermined pitch along the direction of linear displacement; A periodic position in which one pitch of displacement of the linear scale member is arranged so as to be displaceable relative to the scale member, and one period corresponds to one pitch of displacement of the linear scale member, depending on the relative positional relationship of the two substances according to this displacement. Two linear sensors each having an electromagnetic detection unit that generates a detection signal as digital data are provided, and one pitch of the linear scale member in each linear sensor is different, and the pitch of the linear scale member in each linear sensor is different. One of the detection parts is integrally coupled so that it is integrally displaced linearly in accordance with the mechanical displacement of the detection target, and thereby each linear sensor is connected at a different period with respect to the mechanical linear displacement of the detection target. (b) the above-mentioned 2.
Regarding the two linear sensors, the difference between them is digitally calculated using the first position detection signal generated from the first linear sensor and the second position detection signal generated from the second linear sensor as calculation variables, and Calculating means is provided that uses this difference as a calculation variable to determine the number of periods of the detection target from the origin regarding the first linear sensor, and outputs a period number signal indicating the determined number of periods; (c) Determine whether the value of the first position detection signal belongs to a predetermined first range before the cycle switch or a predetermined second range after the cycle switch, and When it is determined that it is within the second range, at least one value of the calculation variable in the calculation means is added by a predetermined correction value depending on whether it belongs to the first or second range. Alternatively, it consists of performing a correction operation to change the value of the difference by subtraction.

As mentioned above, each linear sensor includes a linear scale member and an electromagnetic detection head, and the pitch of the linear scale member in each linear sensor is different. Then, either the linear scale member or the detection section in each linear sensor is integrally coupled so that they are integrally displaced linearly in accordance with the mechanical displacement of the detection target. The linear scale member is made by repeatedly disposing two materials exhibiting different magnetic properties at a predetermined pitch along the direction of linear displacement. The electromagnetic detection section is movable relative to the linear scale member, and depending on the relative positional relationship of the two substances according to this displacement, one pitch of displacement of the linear scale member is converted into one cycle. A periodic position detection signal is generated. As a result, the linear scale member or the detection part of each linear sensor can be connected to each other at different periods with respect to the mechanical linear displacement of the detection target, using an extremely simple configuration that requires only one of the linear scale member or the detection part of each linear sensor to be integrally connected without providing a special movement speed change mechanism. A position detection signal can be generated from each linear sensor.

For example, for the displacement of the detection target from the origin,
When the position detection signal of the first linear sensor changes by exactly one cycle, the position detection signal of another linear sensor does not show a change of exactly one cycle, and there is a difference between the position detection signals. There is. This difference gradually increases as the displacement of the detection target from the origin increases, and increases in proportion to the number of cycles of the detection target from the origin with respect to the first linear sensor. Therefore, depending on how large this difference is,
It becomes clear how many cycles from the origin the position detection signal of the first linear sensor is.

Therefore, if the calculation means calculates the difference between the first position detection signal of the first linear sensor and the second position detection signal of the second linear sensor, the first position detection signal is calculated based on this difference. The number of cycles from the origin of the detection target for the linear sensor can be determined.

In this way, the obtained first period number signal indicates the period number of the first position detection signal based on the output signal of the first linear sensor. The combination with the period number signal specifies the absolute position of the detection target from the origin over a wide range.

As described above, in the present invention, the absolute linear position detection range can be expanded, and one of the linear scale member or the detection head part of each linear sensor can be integrally connected without providing a special motion transmission mechanism. The configuration is simple, and the overall detection resolution depends on the resolution of the output signal for one period of one linear sensor, so the resolution is extremely high, and the accuracy is high. It will be good.

At the cycle switching point of the first position detection signal, the data changes discontinuously from the maximum value to the minimum value, so before and after this cycle switching point, the first position detection signal and the second position detection signal Due to quantization errors and detector accuracy errors, the difference between the two values does not follow the logical value, and increases or decreases somewhat. Therefore, when it is determined that the value of the first position detection signal is within a predetermined first or second range before and after the cycle change, the value of the difference is calculated by adding or subtracting a predetermined correction value. By performing a correction calculation to change it, it is possible to obtain a correct period number signal with the error corrected.

In the examples described below, those corresponding to linear sensors are indicated by symbols S1 and S2, and those corresponding to linear scale members are as follows.
The rod-shaped core parts 2-1 and 2-2, which correspond to the electromagnetic detection part, are the primary coil and secondary coil parts housed in the casing 4-1, and the digital phase difference associated therewith. detection circuit 37-1,
This is part 37-2. In addition, the arithmetic unit COM corresponds to the arithmetic means and the means for performing correction arithmetic.
It is.

In the second invention, a third linear sensor is further provided. The 1-pitch difference between the first linear sensor and the third linear sensor is smaller than the 1-pitch difference between the first and second linear sensors, and the first position detection generated from the first linear sensor The second difference, which is the difference between the signal and the third position detection signal generated from the third linear sensor, is the difference between the first position detection signal generated from the first linear sensor and the second position detection signal generated from the first linear sensor.
The period is longer than the first difference between the second position detection signal and the second position detection signal generated from the linear sensor.
Therefore, using this second difference as a calculation variable, a calculation is performed to determine the number of periods of the first period number signal corresponding to the first difference, and a second period number signal indicating the determined number of periods is obtained. I can do it. This makes it possible to detect the absolute linear position in a further expanded range. To show the correspondence with the example, the sensor S3 corresponds to the third linear sensor,
Similar to the above, the arithmetic unit COM corresponds to the means for performing the arithmetic operation to obtain the second period number signal.

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

As the simplest example, an example using two linear sensors will be explained in principle with reference to FIG.
Linear sensors S1 and S2 that generate electrical output signals at a predetermined period in response to mechanical linear displacement are described, for example, in the specification of Utility Model Application No. 57-32127 (Utility Model Application No. 58).
136718), in which a plurality of cores are provided at predetermined intervals. The amount of mechanical displacement of the detection target corresponding to one cycle of the output signal of one sensor S1 is P ₁ , and the amount of mechanical displacement of the detection target corresponding to one cycle of the output signal of the other sensor _S2 is P1.
Let it be P ₂ . This invention is characterized in that the amount of mechanical displacement of the detection target corresponding to one cycle of the output signal of each sensor is made different, so that P ₁ ≠ P ₂
It is. Each sensor S1, S2 is at an absolute position (relative absolute position within one cycle) within the range of mechanical linear displacement P ₁ , P ₂ for each cycle.
Detection is possible. It is assumed that each sensor generates a value of N ₁ for S1 and N ₂ for S2 as an output signal for one cycle. In other words, in the case of a digital type, sensor S1 generates an absolute output signal with an accuracy of one period divided by _N1 , and sensor _S2 generates an absolute output signal with an accuracy of one period divided by N1.
N Generates an absolute output signal with an accuracy divided into _two .

In other words, when the amount of mechanical displacement of the object to be detected is P ₁ , the sensor S1 outputs an output signal for one period, that is, the maximum output N ₁ , and when the amount of mechanical displacement of the detection target is P ₂ , the sensor S2 outputs the output signal for that one period.
Outputs the output signal for the period, that is, the maximum output N ₂ .
When the amount of mechanical displacement of the detection target from the origin is greater than P ₁ or P ₂ , it is not possible to determine how many cycles from the origin these output signals are from sensors S1 and S2 alone, and the absolute position cannot be detected. Can not. However, if the output signals of sensors S1 and S2 are used in combination as described below, the absolute position can be determined.

The detected value of sensor S1 per unit displacement can be expressed by a constant N ₁ /P ₁ , and similarly the detected value of sensor S2 can be expressed by a constant N ₂ /P ₂ . Therefore, the detection target is at the origin (at the origin, sensors S1 and S
Considering the case where the sensor S ₁ moves by P 1 from 0 (both outputs of 2 are 0), the output of sensor S ₁ is N ₁ , which corresponds to exactly one period. Also, the output of sensor S ₂ is
_N2・_P1 / _P2 . Therefore, if the values of the output signals obtained from sensors S1 and S2 are expressed as D ₁ and D ₂ for the current position of the detection target, then D ₁ and D ₂ are as follows when the absolute position P ₁ .

D ₁ = N ₁ D ₂ = N ₂・P ₁ /P ₂ (1) From this, for every linear displacement P ₁ of sensor S1 for one period, the value of the output signal between both sensors S1 and S2 is as follows. It is clear that the values shift sequentially according to the amount of change (constant). Here, it is assumed that D ₁₂ =D ₁ −D ₂ .

Change in D ₁₂ per P ₁ = N ₁ · P ₂ - N ₂ · P ₁ /P ₂ (2) Therefore, in the calculation device COM, each sensor S1, S2 corresponding to the current position of the detection target is If we calculate the output difference "D ₁₂ = D ₁ - D ₂ " and divide it by the constant in equation (2) above as shown below, the current output signal D ₁ of sensor S1 will be calculated from the origin. What cycle (C _X )
It becomes clear that it belongs to

_{C _ _ _ _ _ _ _ _} (i.e., round down the decimal part or remainder of C _X obtained by equation (3) and use D ₁ as the decimal part),
The linear position of the detection target can be determined absolutely. Note that D ₁ may become smaller than D ₂ due to a difference in the periods of S1 and S2, but in that case, N ₁ is added to the value of D ₁ to obtain D ₁₂ .

Sensor S by phase shift type linear position detector
1 and S2 are shown in FIG. 2. first,
To explain the sensor S1, the sensor S1 is as follows.
1 stored in a predetermined arrangement within the casing 4-1
It includes a secondary coil, a secondary coil, and an elongated core portion 2-1 inserted so as to be linearly movable within these coils. The core portion 2-1 includes a plurality of cores 3-1 arranged at predetermined intervals in the axial direction, and each core 3-1.
1, and a sleeve 6-1 that covers the core 3-1 and the spacer 5-1. The core 3-1 is made of a magnetic material, and the spacer 5-1 is made of air or other non-magnetic material. This core portion 2-1 is linearly displaced in response to a linear motion applied from the outside as a detection target. As an example, each core 3-1 has a cylindrical shape with a predetermined length "P ₁ /2" (P ₁ is an arbitrary number), and the spacer 5-1
The length of the core 3-1 is approximately equal to the length of the core 3-1. Therefore, the distance of one pitch in the array of cores 3-1 is " _P1 ". In this embodiment, the coils are arranged to operate in four phases. For convenience, these phases are distinguished using symbols A, B, C, and D. Each phase A~ depending on the position of core 3-1
The reluctance generated in D is shifted by 90 degrees, for example, if phase A is a cosine phase,
B phase is sine phase, C phase is negative cosine phase, D
The phase is becoming a minus sign phase.

In FIG. 2, the primary coils 7, 8, 9, 10 and the secondary coils 11, 1 are individually shown for each phase A to D.
2, 13, and 14 are provided. The secondary coils 11-14 of each phase A-D are wound outside the corresponding primary coils 7-10, respectively. The length of each coil is approximately equal to the length of the core 3-1, which is "P ₁ /2". In the example of FIG. 2, the A-phase coils 7 and 11
and C-phase coils 9, 13 are provided adjacent to each other, B-phase coils 8, 12 and D-phase coil 1.
0 and 14 are also provided adjacently. In addition, the interval between the A-phase and B-phase or C-phase and D-phase coils is "P ₁
(n±1/4" (n is any natural number). This allows the phase of the reluctance change of the magnetic circuit in the B phase and D phase to be 90 degrees ( 1/4 of 1 pitch P ₁ )
It can be shifted.

Further, the arrangement of the coils of each phase A to D in the sensor 1 is not limited to that shown in FIG. 1. That is, each phase A to D corresponds to the linear displacement of the core part 2-1.
The reluctance of the magnetic circuit changes, and the phase of the reluctance change shifts by 90 degrees for each phase (therefore, the A phase and C phase shift by 180 degrees,
Since the B phase and the D phase are also shifted by 180 degrees, it is only necessary to bring about such a change in reluctance.

The connection format of the primary coils 7 to 10 and the secondary coils 11 to 14 in FIG. 2 is, for example, as shown in FIG. 3. Figure 3 shows a wiring configuration in which the A-phase and C-phase primary coils 7 and 9 are excited in opposite phases by a sine signal sinωt, and the outputs of the secondary coils 11 and 13 are added in phase. It shows. Similarly to the above, for the B phase and D phase, the primary coils 8 and 10 are excited in opposite phases with the cosine signal cosωt, and the secondary coils 12 and 1
Add the outputs of 4 in phase.

In short, the wiring format shown in FIG. 3 can be expressed as follows. In other words, two phases (A and C or B and D) whose reluctance changes are shifted by 180 degrees are operated in opposite phases, and two pairs whose reluctance changes are shifted by 90 degrees (a pair of A and C and a pair of B and One of the pair D) is excited by a sine signal sinωt, and the other is excited by a cosine signal cosωt. In other words, the two pairs (the pair A and C and the pair B and D) are the same as two differential transformers whose reluctance changes are out of phase by 90 degrees. Each is individually excited by two types of alternating current signals (sinωt, cosωt) having an electrical phase shift corresponding to the phase shift of the reluctance change. Sensor S1 is the sum of the secondary coil outputs of the A, C phase pair and the B, D phase pair.
The secondary side output signal _Y1 is obtained. This output signal Y ₁ is
The reference AC signal (sinωt or cosωt) is phase-shifted by a phase angle _φ1 corresponding to the linear position of the core portion 2-1. The reason for this is that the reluctance of each phase A to D is shifted by 90 degrees, and the electrical phases of the excitation signals of one pair (A, C) and the other pair (B, D) are shifted by 90 degrees. It's for a reason. This point can be expressed briefly as follows.

That is, if the phase corresponding to the linear position of the core part 2-1 is φ ₁ , the function of reluctance change according to the linear position is cosφ ₁ for the A phase, sinφ ₁ for the B phase,
It can be expressed simply as _-cosφ1 for the C phase and _-sinφ1 for the D phase. The A and C phases are operated in opposite phases to each other by a sine signal sinωt, and the B and D phases are operated in opposite phases to each other by a cosine signal cosωt, and the resulting secondary coil output is additively Therefore, the output signal Y ₁ can be substantially expressed in the following informal formula.

Y ₁ = sinωtcosφ ₁ −(−sinωtcosφ ₁ ) +cosωtsinφ ₁ −(−cosωtsinφ ₁ ) =2sinωtcosφ ₁ +2cosωtsinφ ₁ =2sin(ωt+φ ₁ ) The coefficient indicated as “2” in the above formula for convenience is determined according to various conditions. By replacing it with a constant K determined by , it can be expressed as Y ₁ =Ksin(ωt+φ ₁ ). Here, since the phase φ ₁ of the reluctance change is proportional to the linear position x of the core part 2-1 according to a predetermined proportionality coefficient (or function), the reference signal sinωt (or cosωt) in the output signal Y ₁
The linear position x can be detected by measuring the phase shift φ ₁ from . However, when the phase shift amount φ ₁ is a full width 2π, the linear position x corresponds to the aforementioned distance P ₁ (one pitch length of the core 3-1). That is, according to the electrical phase shift amount φ ₁ in the signal Y ₁ , a relative linear position can be detected within the range of the distance P ₁ .

Sensor S2 has the same configuration as S1, but the distance P ₂ between cores 3-2 and 4-2 in core part 2-2 is
is different from P ₁ . The core parts 2-1 and 2-2 of the sensors S1 and S2 are connected by a connecting member 50,
They move linearly together in accordance with the linear displacement x of the detection target. Then, an AC signal Y ₂ =Ksin including an electrical phase shift φ ₂ corresponding to that linear position x of the core part 2-2
(ωt+φ ₂ ) is obtained from the secondary side of sensor S2.
However, when this phase shift amount φ ₂ is 2π, the linear displacement x corresponds to one pitch length P ₂ of the core 3-2.

Secondary output signals Y ₁ , Y 2 of both sensors S1, _S2
A phase difference detection circuit 37-1 as shown in FIG.
37-2, each pitch length P ₁ ,
It is possible to obtain periodic electrical output signals D ₁ and D ₂ in which one period is the mechanical linear displacement corresponding to P ₂ .

In FIG. 4, the oscillation unit 32 is a reference sine signal.
A circuit that generates sinωt and a cosine signal cosωt, a phase difference detection circuit 37-1, detects the phase shift φ ₁ as shown in FIG.
2 is a circuit for measuring φ ₂ respectively. Clock pulse oscillated from clock oscillator 33
CP is counted by a counter 30. The counter 30 is, for example, a modulo M, and its counter value is given to the register 31. From the 4/M frequency-divided output of the counter 30, a pulse Pc obtained by dividing the clock pulse CP by 4/M is taken out and applied to the C input of a flip-flop 34 for 1/2 frequency division. Pulse Pb from the Q output of this flip-flop 34
is applied to the flip-flop 35, the pulse Pa from the output is applied to the flip-flop 36,
The outputs of these 35 and 36 are sent to the low pass filter 2.
1 and 22 and amplifiers 23 and 24, a cosine signal cosωt and a sine signal sinωt are obtained, and the sensor S
1, input to S2. M at counter 30
The counter corresponds to the phase angle of 2π radians of these reference signals cosωt and sinωt. That is, one count value of the counter 30 indicates a phase angle of 2π/M radians.

In circuit 37-1, the output signal of sensor S1
Y ₁ is applied to a comparator 26 via an amplifier 25, and a square wave signal corresponding to the positive or negative polarity of the signal Y is output from the comparator 26. In response to the rise of the output signal of the comparator 26, a pulse T _S is output from the rise detection circuit 28, and the count value of the counter 30 is loaded into the register 31 in response to this pulse T _S. As a result, a digital value D ₁ corresponding to the phase shift φ ₁ is taken into the register 31.

The other circuit 37-2 has the same configuration as the circuit 37-1, and inputs the secondary output signal Y ₂ of the sensor S2 to the comparator 39 via the amplifier 38, and controls the register 41 via the rising edge detection circuit 40. Then, the digital value D ₁ corresponding to the phase shift φ ₂ is taken into the register 41 from the counter 30 . Note that only one phase difference detection circuit 37-1 is provided, and the register 3
Only sensors 1 and 41 may be provided corresponding to the sensors S1 and S2, and the phase difference detection circuits may be used in a time-sharing manner.

The output signals D ₁ and D ₂ of both sensors S1 and S2 obtained as above are supplied to the calculation COM (FIG. 1) as described above, and calculation is performed according to the above-mentioned equation (3). However, as shown in FIG.
7-2, N ₁ =N ₂ , and if this is represented by N, the equation (3) becomes Cx = D ₁₂ ÷N (P ₂ −P ₁ /P ₂ (4)). Of course, N ₁ ÷ N ₂ is also acceptable.

In addition, in the example of Fig. 2, if φ ₁ = x・2π/P ₁ , φ ₂ = x・2π/P ₂ and P ₂ −P ₁ =a, then φ ₁ −φ ₂ = 2πx1/P ₁ (a/P ₁ +a), which corresponds to D ₁₂ =D ₁ -D ₂ . Here, φ ₁ −φ ₂
The maximum value 2π (that is, one period of D ₁ - D ₂ ) of (that is, the phase value corresponding to D ₁₂ ) is 2π = 2πx1/P ₁ (a/P ₁ + a), and the value of x that satisfies this x _MAX x _MAX = P ₁ (P ₁ +a/a) (5) This equation (5) indicates the maximum absolute detectable range when two sensors S1 and S2 are used. For example, when P ₁ = 10 mm and a = 1 mm, x _MAX is "110", and the absolute detection range is expanded to 11 times (110 mm) the absolute detection range P ₁ (10 mm) of sensor S1 alone. I understand.

In order to further expand this absolute detectable range, the number of sensors may be further increased. For example, a third sensor S3 may be further provided to detect the mechanical displacement amount.
Output signal whose period is P ₃ (P ₃ ≠P ₁ ≠P ₂ )
Output D ₃ according to the linear displacement x.
Preferably, P ₁ <P ₃ <P ₂ , and the period of "D ₁ - D ₃ " is longer than the period of "D ₁ - D ₂ ", and the period of "D ₁
−D ₃ ”, the periodicity of “D ₁ −D ₂ ” may be determined.

In other words, the third linear sensor S3 is provided so as to be integrally displaced linearly like the first and second linear sensors S1 and S2. Since the pitch P ₃ of this third linear sensor S3 has the above relationship, the difference of 1 pitch between the first linear sensor S1 and the third linear sensor S3 is the difference between the pitch P 3 of the first linear sensor S1 and the second linear sensor S3. This is smaller than the 1 pitch difference of the linear sensor S2. Therefore, the period of the difference D1- _D3 between the position detection signal _D1 of the first linear sensor S1 and the position detection signal _D3 of the third linear sensor _S3 is:
As mentioned above, it is longer than the period of D ₁ −D ₂ . D ₁ −
The value of D ₁ - _{D 3} per period of D ₂ is constant, and D ₁
The value of −D ₃ is correlated to the number of periods of D ₁ −D ₂ . Therefore, if we calculate the difference D ₁ − D ₃ in the same way as above, we get
Based on this difference D ₁ -D ₃ , the number of cycles of D ₁ -D ₂ can be determined. Therefore, the first linear sensor S
In addition to the combination of the position detection signal D ₁ of 1 and the number of cycles Cx, the number of cycles of D ₁ - _{D 2} found here (second
(referred to as the period number signal), the absolute linear position can be specified in a further expanded range.

FIG. 5 shows an example in which three linear sensors S1 to S3 are provided. In this case, it goes without saying that the specific configuration of the third sensor S3 may be the same as that shown in FIG.
It will be easily understood from the explanation of FIG. 2 that the coil portions are integrally connected by the coil portions so as to be integrally displaced relative to the corresponding coil portions.

Assuming that the output signal for one cycle of the third sensor S3, that is, the maximum output, is N ₃ , and considering D ₃ corresponding to the absolute position P ₁ as in equation (1) above, D ₃ =
(N ₃ · P ₁ ) / P ₃ ...(6), and if D ₁ − D ₃ = D ₁₃ and the change in D ₁₃ per P ₁ is expressed in the same way as in equation (2) above, (N ₁・P ₃ −N ₃・P ₁ )/P ₃ …(7). Here, to simplify the explanation, if we set N ₁ = N ₂ = N ₃ = N as before, the above formula (7), that is, the change in D ₁₃ per P ₁ , is expressed as {N( P ₃ −P ₁ )}/P ₃ …(8).

The first periodicity signal Cx obtained from the above equation (4) based on the above difference D ₁₂ has a maximum value of (P ₁ +a)/a=P ₂ /(P ₂ −P ₁ ) from the above equation (5). ...(9), which is the number of cycles of the output signal of the first sensor S1 corresponding to one cycle of the first cycle number signal Cx. The value of D ₁₃ at this time is the value of equation (8) multiplied by the number of periods in equation (9), so {N(P ₃ −P ₁ )P ₂ }/{P ₃ (P ₂ −P ₁ )} …(10). Therefore, by dividing the value of D ₁₃ by the constant of equation (10) as shown in the following equation, it is possible to obtain the second period number signal Cy indicating the period number of the first period number signal Cx.

Cy＝D ₁₃ ÷ {N (P ₃ − P ₁ ) P ₂ } / {P ₃ (P ₂ − P ₁ )}
...(11) In the arithmetic unit COM, three sensors S1, S2,
Based on the output of S3, the first difference D ₁₂ =D ₁ −D ₂
and the first difference D ₁₃ =D ₁ −D ₃ , and the above-mentioned
Calculations for obtaining the first period number signal Cx and the second period number signal Cy may be performed according to equations (4) and (11). As a result, the output signal of the first sensor S1
D ₁ , the first period number signal Cx, and the second period number signal
The combination of Cy provides expanded absolute linear position detection. That is, D ₁ is the distance
Indicates the absolute linear position in the range of P ₁ , Cx
represents the repetition period number of D ₁ as an absolute value from the origin (however, P ₂ / (P ₂
-P ₁ ) is the maximum value), Cy indicates the number of repetition cycles of Cx as an absolute value from the origin, and the combination of these three specifies the absolute linear position. Furthermore, it is possible to use a plurality of sensors with the relationship P ₁ <...<P ₄ <P ₃ <P ₂ .

In other words, if a fourth linear sensor S4 (this one pitch is indicated by _P4 ), which has an even smaller one-pitch difference from the first linear sensor S1, is similarly provided, for the same reason, D ₁ - D ₃ It is clear that the number of periods of can be determined further. It is clear that it is possible to push this idea further, as shown in the above inequality.

By the way, if the period number Cx is simply calculated using the above equation (3) or (4), an error may occur due to the detection resolution of the sensors S1 and S2. Theoretically, "D ₁ - D ₂ " is a number that changes continuously as D ₁ and D ₂ change, but in reality, "D ₁ - D 2" is a number that changes continuously as D 1 and D 2 change.
This is because the change in D ₂ is step-like, and the steps in which D ₁ and D ₂ change are also asynchronous. This can be explained using a diagram as shown in Figure 6. For example, if N ₁ = N ₂ = N = 1024, the change in D ₁₂ per P ₁ {N ( ₂ - P ₁ )} / P ₂ = 32 When the first
Each sensor S1, S when the cycle of the output signal _D1 of sensor S1 switches from the first cycle to the second cycle
The state of outputs D ₁ and D ₂ of 2 is illustrated. Figure a
Figure b shows the case where there is no error in the sensor outputs D ₁ and D ₂ , Figure b shows the case where the sensor output D ₂ has an error in the forward direction, Figure c shows the case where the sensor output D ₂ lags behind. Indicates a case where an error occurs in the direction.
As shown in a, even under normal conditions, in a certain range just before the rotation speed changes, there is a part where "D ₁ - D ₂ " becomes "32", and in this part, the equation (4) above is not satisfied. As a result, the number of cycles Cx obtained becomes "1".
This is because, although theoretically "D ₁ - _{D 2} " is a number that changes continuously as D ₁ and D ₂ change, "D ₁ - _{D 2} "
The change step is 1 step for every 32 steps of D ₁ , and the change steps of D ₁ and D ₂ are not the same but gradually deviate, so
The theoretical one-change step of “D ₁ − D ₂ ” (that is, D ₁
32 steps), the actual "D ₁ - D ₂ " does not maintain a constant value, but the theoretical value and that value differ by 1.
The value added by 1 is repeated alternately, and the ratio of the value added by 1 to 1 becomes higher, and eventually when the theoretical change step switches, the actual ``D ₁
This is due to the fact that "-D ₂ " also switches to the theoretical value (the value obtained by adding 1 to the theoretical value of the previous step).
Therefore, in the range where the theoretical value of D ₁₂ switches from 31 to 32, that is, in the range "992≦D ₁ ≦1023", Fig. 6a
In some cases, D ₁₂ =32. Therefore,
For example, the position D ₁ = 1023, D ₂ = 991 is actually the 1023rd address of the first cycle (Cx = 0), but if you simply apply the above equation (4), _Rx =1
Therefore, it becomes address 1023 of the second rotation.
In addition, if an error as shown in Figure b occurs, simply applying equation (4) above at the positions D ₁ = 0 and D ₂ = 992 will result in Cx = 1, which means that Cx = 1 in the second period. The correct absolute position of address 0 can be found, but D ₁ = 0,
At the position D ₂ = 993, simply applying equation (4) yields Cx =
0, and the incorrect position of the 0th address of the first cycle, that is, the origin, is determined. In addition, if there is an error as shown in c in the same figure, D ₁ = 1023,
Even though the position D ₂ = 990 is the 1023rd address of the first rotation, simply applying equation (4) yields Cx =
1, and becomes address 1023 in the second cycle. In order to eliminate the error caused by this, the sensor S
An appropriate correction value for error correction may be added to or subtracted from the value of "D ₁ - D ₂ " in a predetermined range before and after the cycle of the output D1 changes.

That is, in a predetermined range before the period of the output signal D ₁ of the first sensor S1 is switched, D ₁ −D ₂
may be larger than the theoretical value, so it is sufficient to compensate for the error within that range. Therefore, it is determined whether the value of the output signal D ₁ of the first sensor S1 belongs to a predetermined first range immediately before the cycle changes, and in this range, a predetermined correction is made to D ₁ −D ₂ . The above compensation can be achieved by subtracting the value. In addition, in a predetermined range after the period of the output signal D ₁ of the first sensor S1 is switched, D ₁ - _{D 2} may become smaller than the theoretical value, so compensation for this error is performed. That's fine. Therefore, it is determined whether the value of the output signal D ₁ of the first sensor S1 belongs to a predetermined first range immediately after the cycle switching, and in this range, a predetermined correction value is set for D ₁ - _{D 2} . The above compensation can be achieved by adding the values. D ₁ is the predetermined first
Alternatively, the determination as to whether or not it is within the second range and addition/subtraction for correction using correction values may be performed by the arithmetic unit COM. This correction value can be set to an appropriate value. For example, if it is set to "1", D ₁ - D ₂ is corrected by ±1, so it is effective when the change step deviation of D ₁ and D ₂ is within 1 step. If it is "2",
Since D ₁ -D ₂ is corrected by ±2, it is effective when the change step deviation of D ₁ and D ₂ is within 2 steps. In this way, this compensation value may be set as appropriate in the design depending on how many steps the deviation between the change steps of D ₁ and D ₂ is allowed (how many steps are compensated for).

As is clear from the arithmetic, the actually calculated difference D ₁ −
D ₁ is not limited to adding or subtracting from D ₂ .
It may be added to or subtracted from D ₂ , or an appropriate value may be added to or subtracted from both D ₁ and D ₂ . Note that an error of the same nature as the error regarding D ₁ −D ₂ may also occur regarding D ₁ −D ₃ .
It is also possible to compensate for this by dealing with it in the same way as above.

In addition, in the above, core parts 2-1 and 2-2 are moved,
Although I explained that the coil should be kept stationary, you can do the opposite. In addition, linear type sensors are not limited to the phase shift type as shown in Figure 2, but also have multiple cores as shown in Figure 2 to obtain periodic analog DC voltage according to the linear position. Any other linear sensor may be used, such as a dynamic transformer type sensor.

As explained above, according to the present invention, the absolute linear position detection range can be expanded, and one of the linear scale member or the detection head part of each linear sensor can be integrated into one body without providing a special motion transmission mechanism. The configuration is simple, as it only needs to be an extremely simple configuration that can be coupled to each other, and the overall detection resolution depends on the resolution of the output signal for one period of one linear sensor, so the resolution is extremely high. , the accuracy is good.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing one embodiment of the present invention, Fig. 2 is a sectional view specifically showing an example of the sensor portion used in this invention, and Fig. 3 is the primary and secondary coils of Fig. 2. 4 is a block diagram showing an example of a circuit for detecting a phase shift according to the amount of linear displacement included in the AC signal output from each sensor in FIG. 2. The figure is a block diagram showing an embodiment of the present invention when three sensors are used, and FIG. 6 is a diagram showing an example of the sensor output to explain the possibility of errors before and after the cycle switching point. S1, S2... Linear sensor, COM... Arithmetic unit, 2-1, 2-2... Core section, 3-1, 3-
2... Core, 5-1, 5-2... Spacer, 4-
1... Casing, 7, 8, 9, 10... Primary coil, 11, 12, 13, 14... Secondary coil,
37-1, 37-2...Phase difference detection circuit.

Claims (1)

[Claims] 1 (a) A linear scale member in which two substances exhibiting different magnetic properties are alternately arranged at a predetermined pitch along the direction of linear displacement, and a The linear scale member is disposed so as to be displaceable, and according to the relative positional relationship of the two substances according to the displacement, a periodic position detection signal whose cycle is one pitch of displacement of the linear scale member is converted into digital data. Two linear sensors are provided, each having a different pitch of the linear scale member in each linear sensor, and one of the linear scale member or the detection part in each linear sensor is integrated. The position detection signals are coupled together to cause an integral linear displacement in accordance with the mechanical displacement of the detection target, and thereby the position detection signals are transmitted from each linear sensor at different periods with respect to the mechanical linear displacement of the detection target. (b) Regarding the two linear sensors, a first position detection signal generated from the first linear sensor, a second position detection signal generated from the second linear sensor, and The difference between the two is digitally calculated using as a calculation variable, and this difference is used as a calculation variable to perform calculation to determine the number of cycles of the detection target from the origin regarding the first linear sensor, and the period indicating the determined number of cycles is calculated. and (c) whether the value of the first position detection signal falls within a predetermined first range before the cycle changes or within a predetermined second range after the cycle changes. and when it is determined that it is within the first or second range, at least one of the calculation variables in the calculation means is determined according to whether it belongs to the first or second range. An absolute linear position detection device that performs a correction operation to change the difference value by adding or subtracting one value by a predetermined correction value. 2. The detection unit includes a plurality of primary coils arranged at predetermined intervals in a linear displacement direction, a secondary coil provided corresponding to the primary coils, and
Each of the coils is individually excited by a plurality of phase-shifted reference AC signals, thereby outputting an output AC signal in which the reference AC signal is phase-shifted in accordance with the relative linear position of the linear scale member with respect to the coil. A circuit for generating a digital position detection signal in the secondary coil, and a circuit for obtaining a digital position detection signal by digitally measuring a phase shift between the output AC signal and the reference AC signal. Absolute linear position detection device as described in . 3 (a) A linear scale member in which two substances exhibiting different magnetic properties are arranged alternately at a predetermined pitch along the direction of linear displacement, and a linear scale member arranged so as to be displaceable relative to this linear scale member. and an electromagnetic device that generates a periodic position detection signal using digital data, with one pitch of displacement of the linear scale member as one period, in accordance with the relative positional relationship between the two substances according to this displacement. Three linear sensors each having a detecting section are provided, and one pitch of the linear scale member in each linear sensor is different,
1 of the first linear sensor and the third linear sensor
The difference in pitch is 1 between the first and second linear sensors.
The difference in pitch is smaller than the difference in pitch, and one of the linear scale member or the detection part in each linear sensor is integrally connected and linearly displaced in response to the mechanical displacement of the detection target. (b) the position detection signal is generated from each linear sensor at different periods with respect to the mechanical linear displacement of the detection target; (b) the first position detection signal generated from the first linear sensor;
Digitally calculates a first difference between the position detection signal and the second position detection signal generated from the second linear sensor as calculation variables,
Furthermore, using this first difference as a calculation variable,
A first circuit that performs an operation to determine the number of cycles from the origin of the detection target regarding the linear sensor, and outputs a first cycle number signal indicating the determined number of cycles.
(c) whether the value of the first position detection signal belongs to a predetermined first range before the cycle changes or to a predetermined second range after the cycle changes; and when it is determined that it is within the first or second range, at least one of the calculation variables in the first calculation means, depending on whether it belongs to the first or second range. (d) performing a correction operation of changing the difference value by adding or subtracting two values by a predetermined correction value; (d) a first linear sensor generated from the first linear sensor;
Digitally calculates a second difference between the position detection signal and the third position detection signal generated from the third linear sensor as calculation variables,
Furthermore, using this second difference as a calculation variable,
1. An absolute linear position detection device comprising: a second calculation means for performing a calculation to determine the period number of the period number signal and outputting a second period number signal indicating the determined period number.