JPH0132449B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in an angle measuring device in which the signal generating means of a resolver is photoelectric.

Resolvers are widely used as angle or position detectors in robots, NC machine tools, etc. The reason is that speed information and angle information can be obtained from one unit.
In addition, the structure is relatively simple, there are few failures, and high-resolution angular pulses can be easily obtained by combining with an integrated circuit signal converter (for example, Analog Devices' IRDC1730).

The resolver will be briefly explained below.

In FIG. 1, which shows the principle, two coils 2 and 3 are wound around a stator 1 in a perpendicular state, and inside the stator 1 there is a freely rotatable rotor 4, and a coil 5 is wound around it. has been done. In use, the coil 5 of the rotor 4 is excited and the voltage generated in the coils 2 and 3 of the stator 1 is used, and vice versa, the coil 2 and 3 are excited and the voltage generated in the coil 5 is used. There are cases where

That is, in the former case, as shown in FIG. 2, a voltage Vc of the following formula is applied to the coil ₅ , where _V1 : amplitude, ω: excitation angular velocity. As a result, the magnetic flux interlinking with the coils 2 and 3 changes in accordance with the rotation angle θ of the rotor 4, and as a result, the amplitudes of voltages Va and Vb generated in the coils 2 and 3 correspond to the rotation angle θ. and change. As a result, the rotation angle θ is converted into an amplitude signal as shown in the following equation.

Va=K ₁ V ₁ sinθsinωt Vb=K ₁ V ₁ cosθsinωt Here, K ₁ : Proportionality coefficient (constant) Next, the latter is determined by the coil 2,
When voltages Va, Vb Va=V ₂ sinωt Vb=V ₂ cosωt with a 90 degree phase difference are applied to 3, the magnetic flux interlinking with the coil 5 among the respective magnetic fluxes will be at the rotation angle θ of the rotor 4. As a result, the phase of the voltage Vc generated by the coil 5 changes in accordance with the rotation angle θ as follows. As a result, the rotation angle θ is converted into a phase signal as shown in the following equation.

Vc=K ₂ V ₂ cosωtsinθ+K ₂ V ₂ sinωtcosθ =K ₃ V ₂ sin(ωt+θ) From the above, the resolver has the rotation angle θ of the rotor 4.
It is converted into an angle signal whose amplitude or phase changes depending on the angle and is extracted.

Below, to simply find the angle θ, either measure Va and Vb in the above equation with a voltmeter, or measure the phase difference θ between the voltage Vc in the above equation and the excitation signal Va shown in the equation with a phase difference meter. .

Also, for example, the pulse train signal corresponding to the rotation angle θ used for the feedback signal of the servo system is the angular velocity signal.
In order to obtain dθ/dt, it is sufficient to combine the amplitude signal of the above formula with a converter such as the IRDC1730, which has been developed for the conversion.

An example of this type of converter will be briefly described below.

In FIG. 4, 6 and 7 are the output terminals of the output voltages Va and Vb of the coils 2 and 3 of the resolver, respectively, and are connected to one input terminal of the corresponding cosine multiplier 11 and sine multiplier 12, respectively. The other input terminals for angle data of the cosine multiplier 11 and sine multiplier 12 are connected to the counted value output terminal of a reversible counter 13, which will be described later, and the output terminals of the cosine multiplier 11 and sine multiplier 12 are connected to a deviation amplifier. 1
The output terminal of the deviation amplifier 14 is connected to the input terminal of a phase detection circuit 15, and the detection command input terminal is connected to a rectangular wave shaping circuit (not shown) for the excitation signal of the resolver. (not)
The detected signal output terminal of the phase detection circuit 15 is connected to the input terminal of a low-pass filter 21 of the addition/subtraction pulse generator 20, and the output terminal of the low-pass filter 21 is connected to the input terminal of the absolute value circuit 22. The output terminal of the absolute value circuit 22 is connected to the input terminal of the voltage controlled oscillator 23.
The output terminal of No. 3 is connected to the first and second gate circuits 24 and 25.
is connected to one input terminal of the first gate circuit 2.
The other input terminal of the second gate circuit 25 is connected to the polar signal output terminal of the phase detection circuit 15 via the inverter 26, and the other input terminal of the second gate circuit 25 is directly connected to the polar signal output terminal of the second gate circuit 25. 1. The output terminals of the second gate circuits 24 and 25 are external terminals 27 and 28
and the addition and reduction input terminals of the reversible counter 13, respectively.

FIG. 5 is a waveform diagram of the signal corresponding to the number indicated in a circle on the signal line in FIG. 4. Hereinafter, with reference to FIGS. The operation of the above converter will be explained using an example where the angle is changed from 30 degrees to 45 degrees.

Now, assuming that the rotation angle remains at 30 degrees and does not move, the voltages Va and Vb taken out from the output ends 6 and 7 of the coils 2 and 3 are respectively calculated from equation (2) as follows: Va = (K ₁ V ₁ /2 ) sinωt Vb=(√3K ₁ V ₁ /2) sinωt (5). Further, in this state, there is no input pulse to the reversible counter 13 (that is, no incremental signal is generated), and the count value of the reversible counter 13 corresponds to 30 degrees. That is, in the cosine multiplier 11 and the sine multiplier 12, Va・cos30,
Multiplication of Vb・sin30 is performed, and in the end, both multiplication results become the same (√3K ₁ V ₁ /4) sinωt, and as a result, the output of the deviation amplifier 14 becomes 0, and the output of the phase detection circuit 15 and the addition/subtraction pulse generator 20 Both input and output are 0. Next, when the rotor 4 of the resolver starts rotating from 30 degrees, the output voltage Va of the coil 2 increases and the output voltage Vb of the coil 3 decreases. As a result, the output of cosine multiplier 11 increases and the output of sine multiplier 12 decreases. As a result, the deviation amplifier 14 outputs a sine wave with an angular velocity ω having an amplitude corresponding to the difference between the two outputs, and the phase detection circuit 15 detects the output of the sine wave in the phase range of 0 to 180 degrees. is smoothed through a low-pass filter 21, and then through an absolute value circuit 22 (in this case, the detection output is positive, and the output of the low-pass filter 21 simply passes through the absolute value circuit 22). ) to a voltage controlled oscillator 23 which generates pulses with a frequency proportional to its input. At this time, the detection signal of the phase detection circuit 15 is positive, and as a result, no polarity signal is sent out, only the first gate circuit 24 is conductive, and the output pulse of the voltage controlled oscillator 23 is transmitted to the external terminal 2.
7 and is also introduced into the addition input terminal of the reversible counter 13. As a result, reversible counter 1
The count value of 3 is increased, and the count value is added to the cosine multiplier 11 and the sine multiplier 12. At this point, the rotor 4 of the resolver stops rotating, and Va,
If Vc remains at the above value, the multiplication causes the output of the cosine multiplier 11 to decrease from the above, and the output of the sine multiplier 12 to increase from the above, resulting in:
The output of the deviation amplifier 14 is reduced, and in response, a pulse is sent out from the voltage controlled oscillator 23 in the same manner as described above, and the count value of the reversible counter 13 increases, which is added to the cosine multiplier 11 and the sine multiplier 12. As a result, the deviation between both outputs is further reduced,
By repeating the above steps, the deviation is converged to 0. Also,
If the resolver rotor 4 rotates further,
At the same time as the count value of the reversible counter 13 increases, the output voltage Va of the coils 2 and 3 accompanying the rotation of the rotor 4,
Vb increases and decreases, but its moment-to-moment behavior is the continuous movement of the rotor 4 when it rotates by a minute angle, and in the end, the count value φ of the reversible counter 13 coincides with the rotation angle θ of the rotor 4, and the outputs of the cosine multiplier 11 and the sine multiplier (K ₁ V ₁ cosφsinθ) sinωt and (K ₁ V ₁ sinφcosθ)
Loop control is performed so that sinωt matches.
Therefore, the number of pulses sent out from the external output terminal 27 during this period corresponds to the increment of the rotation angle of the rotor 4 in the positive direction, and the number of pulses sent out from the output terminal 28 corresponds to the increment of the rotation angle of the rotor 4 in the negative direction. This is an incremental signal corresponding to the increment of the rotation angle. The count value of the reversible counter 13 that adds or subtracts this incremental signal directly corresponds to the rotational angle θ of the rotor 4, and the output voltage of the low-pass filter 21 corresponds to the rotational angular velocity of the rotor 4.
It corresponds to dθ/dt.

By the way, the above resolver uses electromagnetic induction to generate a signal, and therefore it is difficult to increase the excitation frequency. That is, the upper limit of the excitation frequency is about 1 kHz, and if it is higher than that, the secondary voltage waveform will be distorted, which will significantly reduce the accuracy. Also, rotor 4
Of course, it is inevitable that the penetrating moment of the entire rotor will be relatively large due to the coil 5 wound around it, and the rotating transformer for supplying excitation signals. This is a problem especially when the moment of inertia of the object to be detected is small, and causes a significant response delay. Further, although the structure is relatively simple, strict precision is required in the winding position of the coil in order to maintain uniformity of magnetic flux distribution during actual manufacture, which inevitably leads to a problem of high cost.

To solve this problem, photoelectric means are used as the signal generation means. In other words, the rotor is equipped with multiple light sources whose emission amount is controlled by sine wave drive signals that are in phase with each other or with a 90 degree phase, and multiple or shared light receiving elements. The photoelectric conversion means are sandwiched and faced each other, the light amount changing means is provided for the photoelectric conversion means and the rotor, and changes the amount of transmitted light between the light source and the light receiving element in a sine wave shape with a phase of 90 degrees from each other, and the photoelectric conversion means is provided to the rotor. It is conceivable to obtain an amplitude signal having an amplitude corresponding to the rotation angle of the rotor or a phase signal having a phase corresponding to the rotation angle of the rotor by the signal conversion means converting the light receiving element output of the conversion means into an angle signal. .
Hereinafter, the measurement principle of this type of well-known device will be explained with reference to mathematical formulas. Two or four light sources are arranged on one side of the rotor, and two or four light receiving elements facing the light sources, or a shared light receiving element facing all of the light sources, are arranged on the other side.
Then, each light source turns on its light emitting section in response to a sine wave-like drive signal, and changes the amount of light emitted in a sine wave form. Here, when extracting an amplitude signal corresponding to the rotor rotation angle, the phase of the drive signal of each light source is set to the same phase, and when extracting a phase signal corresponding to the rotor rotation angle, the phase of the drive signal of each light source is set to the same phase. mutually
It will be shifted by 90 degrees.

That is, the angular velocity of the drive signal is ω, and each light source 1
Let the light emission amounts of a, 1b, 1c, and 1d (if there are two light sources, only 1a and 1b) be e _a ~e _d , and let the minimum value of the light emission amount be 0 and the maximum value be A. The amount of light emitted e _a to e _d is determined by e = e _a = e _b = e _c = e _d = (A/2) (sinωt + 1) (1) when driven by in-phase drive signals, and the phase is changed by 90 degrees. When driven by a shifted drive signal, e _a = (A/2) (sinωt+1) e _b = (A/2) (cosωt+1) e _c = (A/2) (-sinωt+1) e _d = (A/2 )(−cosωt+1) (2).

Next, this amount of light is emitted toward each of the opposing light receiving elements 2a, 2b, 2c, and 2d or the shared light receiving elements 2a+2b+2c+2d (if there are two light sources, only 2a, 2b, or 2a+2b). However, in this optical path, there is interposed a light amount changing means provided for the rotor and the photoelectric conversion means, so that the phases of each light emission amount are shifted by 90 degrees from each other, and the rotation of the rotor is The light is changed into a sine wave shape corresponding to the moving angle and then introduced into the light receiving elements 2a to 2d.

Note that when a shared light receiving element is used instead of the light receiving elements 2a to 2d, the sum of the amounts of light introduced to each light receiving element 2a to 2d is introduced, and when two light receiving elements are used, Since two of the above-mentioned light receiving elements will be used, in the following explanation, a case will be explained in which there are four light receiving elements 2a to 2d.

Now, as the above-mentioned light amount changing means, there are those that use a pair of polarizer plates, those that use a transparent circular rotor formed eccentrically with respect to the rotor, etc. The rate of change in light amount due to each of these methods will be explained below. That is, the ratio between the input light amount and the output light amount of the light amount changing means will be explained with reference to mathematical expressions.

First, the light amount changing means using a pair of polarizer plates is
If the intersection angle of the transmission axes of a pair of polarizer plates is α, then
This takes advantage of the fact that the amount of light I〓 transmitted through it changes as shown in the following equation in response to α (Malus's law).

I = I ₀ {H ₀ − H ₉₀ ) cos ² α + H ₉₀ } (3) = I ₀ (K ₄ cos2α + K ₅ ) Here, I = amount of transmitted light when the intersection angle of the transmission axes is α I ₀ : Amount of transmitted light before inserting the polarizer plate H ₀ : Parallel position (the transmission axes of the two polarizer plates are parallel) Transmittance H ₉₀ : Orthogonal position (the transmission axes of the two polarizer plates are orthogonal) Transmittance α : Angle of intersection of transmission axes K ₄ : (1/2) (H ₀ - H ₉₀ ) K ₅ : (1/2) (H ₀ + H ₉₀ ) In other words, the polarizer plate is integrally fixed to the rotor. , a polarizer plate is also placed in the path of each light source and photodetector, and the transmission axes of the polarizer plates are mutually aligned.
It is shifted by 45 degrees. In addition, contrary to the above, donut-shaped polarizer disks with their transmission axes shifted by 45 degrees are integrally fixed to the inner and outer circumferences of the rotor.
Even if the transmission axes of the polarizer plates interposed between the light source and the light receiving element are the same, they are exactly the same. As a result, the ratio of the input light amount to the output light amount of a pair of polarizer plates whose transmission axes are mutually shifted by 45 degrees, that is, the transmittance of the pair of polarizer plates β _a , β _b , β _c , β _d (when the light source is If there are two pieces, β _a and β _b only) are: β _a = K ₄ cos2θ + K ₅ β _b = K ₄ cos2 (θ + 45°) + K ₅ = −K ₄ sin2θ + K ₅ β _c = K ₄ cos2 (θ + 90°) +K ₅ = -K ₄ cos2θ + K ₅ β _d = K ₄ cos2 (θ + 135°) + K ₅ = K ₄ sin2θ + K ₅ (4), and in the end, the transmittance β _a to β _d is 90 degrees corresponding to the rotation angle θ. The signal changes in the form of a sine wave with a phase shift. However, this transmittance is proportional to the sine function value of twice the rotation angle θ, and therefore, a sine wave-like transmittance change pattern occurs twice during one rotation of the rotor.

Next, the light amount changing means using the eccentric light transmission circle is as follows.
By rotating the rotor forming the eccentric light-transmitting circle, the overlapping state of the light-transmitting circle and the X and Y axes drawn on the plane facing the rotor can be changed.
That is, on the above plane, draw an X axis and a Y axis with the origin as the same as the center of the rotor, and draw the X axis on the negative side, the Y axis on the negative side, the X axis on the positive side, and the positive side The overlapping line segment between the Y axis of and the eccentric transparent circle l _a , l _b , l _c , l _d
Now, if we take the rotational position of the rotor where the center of the eccentric transparent circle is lined up to the left of the rotor center as a reference, and let the rotational angle of the rotor in the counterclockwise direction be θ, these line segments are , which utilizes the fact that it changes in a substantially sinusoidal waveform with a phase shift of 90 degrees as shown below.

l _a = Rcosθ+√L ² (Rsinθ) ² l _b = Rsinθ+√L ² (Rcosθ) ² l _c = −Rcosθ+√L ² (Rsinθ) ² l _d = Rsinθ+√L ² (Rcosθ) ² (5) Here , L: Radius of the eccentric light-transmitting circle R: Amount of eccentricity Therefore, the photoelectric conversion means consisting of a light source and a light receiving element placed between the rotor also becomes a part of the light amount changing means, and the light emitting surface and light receiving surface thereof are One or both of them are formed into a rectangular shape and are arranged on the positive and negative X and Y axes.

Note that instead of forming the projection and reception surfaces into rectangular shapes,
The same effect can be obtained even if a mask having a rectangular slit is placed in front of the mask.

Now, the ratio of the input light amount to the output light amount in this light amount changing means, that is, the transmittance, is determined by the rotation angle with respect to the maximum overlap line segment (R+L) _.
.about.ld , and now let the transmittances be _{.gamma..sub.a} to _{.gamma..sub.d} , they are expressed by the following equation.

γ _a = (Rcosθ+√L ² − (Rsinθ) ² )/(R+L) γ _b = (−Rsinθ+√L ² − (Rcosθ) ² )/(R+L) γ _c = (−Rcosθ+√L ² −(Rsinθ) ² /(R+L) γ _d = (Rsinθ+√L ² −(Rcosθ) ² /(R+L) (6) The above is the light amount changing means, and the amount of light emitted from the light source to the light receiving element reaches the light receiving element after being multiplied by the transmittance shown in equation (4) or equation (6) above.

The above explanation is based on the case where the light amount changing means is a polarizer plate or an eccentric transparent circular rotor. As disclosed in Japanese Patent Publication No. 58-85115, etc., the light amount changes in a substantially sinusoidal waveform due to the polymerization of optical scales (geometrically, the transmission surface corresponds to the polymerization area of the rectangular slits, so the change is in the form of a triangular wave). However, in reality, it approximates a sine wave due to the diffraction of light, the shape of the receiving surface of the receiver, etc., and therefore,
Its transmittance is the same as (4) and (6) above, and the amount of transmitted light also has a positive magnitude corresponding to the sum of the constant amount and the amount of sine wave change. .

Below, we will explain the case where the light emission amount is in-phase for the purpose of generating an amplitude signal, that is, the case where the light emission amount is given by e in equation (1) above, and the case where the light emission amount is set at 90 degrees phase for the purpose of generating a phase signal. e _a to e _d in the above formula (2)
The amount of light received by each of the light receiving elements 2a to 2d and the voltage generated thereby will be explained using mathematical formulas.

First, in a case where the light amount changing means is a pair of polarizer plates, let the amounts of light received by each light receiving element be f _a to f _d when the emitted light amounts are in the same phase, and the amounts are as follows.

f _a = β _a e f _b = β _b e f _c = β _c e f _d = β _d e (7) As a result, the voltages g _a to g _d generated by each light receiving element correspond to the amount of light. , now, let M be the light amount-voltage conversion coefficient, K ₆ = AK ₄ M/2, K ₇ = AK ₅ M/2
Then, the generated voltage is as follows.

g _a =f _a・M=K ₆ cos2θsinωt＋K ₇ sinωt＋K ₇
cos2θ＋K ₇ g _a =f _a・M=K ₆ cos2θsinωt＋K ₇ sinωt＋K ₇
cos2θ＋K ₇ g _b =f _b・M=−K ₆ sin2θsinωt＋K ₇ sinωt−K ₇ cos2θ＋
K ₇ g _a =f _a・M=K ₆ cos2θsinωt＋K ₇ sinωt＋K ₇
cos2θ＋K ₇ g _b =f _b・M=−K ₆ sin2θsinωt＋K ₇ sinωt−K ₇ cos2θ＋
K ₇ g _c =f _c・M=−K ₆ cos2θsinωt＋K ₇ sinωt−K ₆ cos2θ+
K ₇ g _a =f _a・M=K ₆ cos2θsinωt＋K ₇ sinωt＋K ₇
cos2θ＋K ₇ g _b =f _b・M=−K ₆ sin2θsinωt＋K ₇ sinωt−K ₇ cos2θ＋
K ₇ g _c =f _c・M=−K ₆ cos2θsinωt＋K ₇ sinωt−K ₆ cos2θ+
K ₇ g _d =f _d・M=K ₆ sin2θsinωt＋K ₇ sinωt＋K ₆ sin2θ＋K ₇
(8) Similarly, if the amount of light received by each light-receiving element is set to f' _a to f' _d when the amount of light emitted is set to a 90-degree phase, it becomes as follows.

f′ _a = β _a e _a f′ _b = β _b e _b f′ _c = β _c e _c f′ _d = β _d e _d (9) Therefore, the generated voltages g′ _a ~ g′ _d are as follows. become that way.

g′ _a =f′ _a・M=K ₆ cos2θsinωt＋K ₇ sinωt
=+K ₆ cos2θ+K ₇ g′ _a =f′ _a・M=K ₆ cos2θsinωt+K ₇ sinωt
=+K ₆ cos2θ+K ₇ g′ _b =f′ _b・M=−K ₆ sin2θcosωt+K ₇ cosωt−K ₆ sin2
θ＋K ₇ g′ _a =f′ _a・M=K ₆ cos2θsinωt＋K ₇ sinωt
=+K ₆ cos2θ+K ₇ g′ _b =f′ _b・M=−K ₆ sin2θcosωt+K ₇ cosωt−K ₆ sin2
θ＋K ₇ g′ _c ＝f′ _c M＝K ₆ cos2θsinωt＋K ₇ sinωt＋K ₆ cos2θ＋
K ₇ g′ _a =f′ _a・M=K ₆ cos2θsinωt＋K ₇ sinωt
=+K ₆ cos2θ+K ₇ g′ _b =f′ _b・M=−K ₆ sin2θcosωt+K ₇ cosωt−K ₆ sin2
θ＋K ₇ g′ _c ＝f′ _c M＝K ₆ cos2θsinωt＋K ₇ sinωt＋K ₆ cos2θ＋
K ₇ g′ _d =f′ _d M=K ₆ sin2θcosωt−K ₇ cosωt−K ₆ sin2θ+
K ₇ (10) Next, in a case where the light amount changing means is an eccentric transparent circular rotor, if the amount of light received by each light-receiving element is set as h _a to h _d when the emitted light amount is in the same phase, it is as follows. become.

h _a = γ _a e h _b = γ _b e h _c = γ _c e h _d = γ _d e (11) As a result, the voltages i _a to i _d generated by each light receiving element correspond to the amount of light. , now let the light amount-voltage coefficient be M, K ₈ = ARM/2 (R + L), K ₈ =
If we set AM/2(R+L), the generated voltage will be as follows.

i _a = h _a M=K ₈ cosθsinωt＋K ₈ √L ² −(Rsinθ) ²
sinωt＋K ₈ cosθ＋K ₈ √L ² −(Rsinθ) ² i _a = h _a M=K ₈ cosθsinωt＋K ₈ √L ² −(Rsinθ) ²
sinωt＋K ₈ cosθ＋K ₈ √L ² −(Rsinθ) ² i _b =h _b M=−K ₈ sinθsinωt＋K ₈ √L ² −(Rcosθ) ² sin
ωt−K ₈ sinθ＋K ₈ √L ² −(Rcosθ) ² i _a = h _a M=K ₈ cosθsinωt＋K ₈ √L ² −(Rsinθ) ²
sinωt＋K ₈ cosθ＋K ₈ √L ² −(Rsinθ) ² i _b =h _b M=−K ₈ sinθsinωt＋K ₈ √L ² −(Rcosθ) ² sin
ωt−K ₈ sinθ＋K ₈ √L ² −(Rcosθ) ² i _c =h _c M=−K ₈ cosθsinωt＋K ₈ √L ² −(Rsinθ) ² sin
ωt−K ₈ cosθ＋K ₈ √L ² −(Rsinθ) ² i _a = h _a M=K ₈ cosθsinωt＋K ₈ √L ² −(Rsinθ) ²
sinωt＋K ₈ cosθ＋K ₈ √L ² −(Rsinθ) ² i _b =h _b M=−K ₈ sinθsinωt＋K ₈ √L ² −(Rcosθ) ² sin
ωt−K ₈ sinθ＋K ₈ √L ² −(Rcosθ) ² i _c =h _c M=−K ₈ cosθsinωt＋K ₈ √L ² −(Rsinθ) ² sin
ωt−K ₈ cosθ＋K ₈ √L ² −(Rsinθ) ² i _d = h _d M=K ₈ sinθsinωt＋K ₈ √L ² −(Rcosθ) ² sinωt
+K ₈ sinθ+K ₈ √L ² − (Rcosθ) ² (12) Similarly, if the amount of light received by each light receiving element is set to _{90 degrees} phase, then it is as follows. become.

h _a ′＝γ _a e _a h _b ′＝γ _b e _b h _c ′＝γ _c e _c h _d ′＝γ _d e _d (13) Therefore, the generated voltage i _a ′ ~ i _d ′ is become that way.

i _a ′=h _a ′M=K ₈ cosθsinωt＋K ₉ √L ² −(Rsin
θ） ² sinωt＋K ₈ cosθ＋K ₉ √L ² −(Rsinθ) ² i _a ′＝h _a ′M＝K ₈ cosθsinωt＋K ₉ √L ² −(Rsin
θ) ² sinωt+K ₈ cosθ+K ₉ √L ² −(Rsinθ) ² i _b ′=h _b ′M=−K ₈ sinθcosωt+K ₉ √L ² −(Rcosθ) ²
cosωt−K ₈ sinθ＋K ₉ √L ² −(Rcosθ) ² i _a ′=h _a ′M=K ₈ cosθsinωt＋K ₉ √L ² −(Rsin
θ) ² sinωt+K ₈ cosθ+K ₉ √L ² −(Rsinθ) ² i _b ′=h _b ′M=−K ₈ sinθcosωt+K ₉ √L ² −(Rcosθ) ²
cosωt−K ₈ sinθ＋K ₉ √L ² −(Rcosθ) ² i _c ′=h _c ′M=K ₈ cosθsinωt−K ₉ √L ² −(Rsinθ) ² si
nωt−K ₈ cosθ+K ₉ √L ² −(Rsinθ) ² i _a ′=h _a ′M=K ₈ cosθsinωt+K ₉ √L ² −(Rsin
θ) ² sinωt+K ₈ cosθ+K ₉ √L ² −(Rsinθ) ² i _b ′=h _b ′M=−K ₈ sinθcosωt+K ₉ √L ² −(Rcosθ) ²
cosωt−K ₈ sinθ＋K ₉ √L ² −(Rcosθ) ² i _c ′=h _c ′M=K ₈ cosθsinωt−K ₉ √L ² −(Rsinθ) ² si
nωt−K ₈ cosθ＋K ₉ √L ² −(Rsinθ) ² i _d ′=h _d ′M=−K ₈ sinθcosωt−K ₉ √L ² −(Rcosθ) ²
cosωt+K ₈ sinθ+K ₉ √L ² −(Rcosθ) ² (14) As described above, due to the coupling between the photoelectric conversion means and the light amount changing means, each of the light receiving elements 2a to 2d has information about the rotation angle θ of the rotor ( The voltage signals shown in 8), (10), (12), and (14) are obtained.

Next, this voltage signal is sent to a signal converter, where it is converted into an angle signal having an amplitude or phase corresponding to the rotation angle θ.

In the signal converter, first, the output of the light-receiving element is introduced into a high-pass filter, and low-frequency components, which vary depending on the rotational angular velocity of the rotor, and DC components are blocked. As a result, when the above-described light amount changing means is a pair of polarizer plates, first, when the amount of light emitted is in the same phase, only the following components j _a to j _d are extracted from the output of equation (8). .

j _a =K ₆ cos2θsinωt+K ₇ sinωt j _b =−K ₆ sin2θsinωt+K ₇ sinωt j _c =−K ₆ cos2θsinωt+K ₇ sinωt j _d =K ₆ sin2θsinωt+K ₇ sinωt (15) Similarly, when the light emission amount is at 90 degrees phase, Only the following components j _a ′ to j _d ′ are extracted from the output of equation (10).

j _a ′=K ₆ cos2θsinωt+K ₇ sinωt j _b ′=−K ₆ sin2θcosωt+K ₇ cosωt j _c ′=K ₆ cos2θsinωt−K ₇ sinωt j _d ′=−K ₆ sin2θcosωt−K ₇ cosωt (16) In addition, the light amount changing means When is an eccentric transparent circular rotor, first, when the amount of light emission is in the same phase, only the following components m _a to m _d are extracted from the output of equation (12).

m _a =K ₈ cosθsinωt+K ₉ √L ² −(Rsinθ) ² sinωt m _b =−K ₈ sinθsinωt+K ₉ √L ² −(Rsinθ) ² sinωt m _c =−K ₈ cosθsinωt+K ₉ √L ² −(Rsinθ) ² sinωt m _d = K ₈ sin θ sin ωt + K ₉ √L ² − (R cossin θ) ² sin ωt (17) Similarly, the following components m _a ′ to m _d ′ from the output of equation (14) when the luminescence amount is 90 degrees phase only are extracted.

m _a ′=K ₈ cosθsinωt+K ₉ √L ² ‐(Rsinθ) ² sinωt m _b ′=‐K ₈ sinθcosωt+K ₉ √L ² ‐(Rcosθ) ² cosωt m _c ′=‐K ₈ cosθsinωt−K ₉ √L ² ‐ (Rsinθ) ² sinωt m _d ′=-K ₈ sinθcosωt−K ₉ √L ² -(Rcosθ) ² cosωt (18) Next, in the signal converter, the output of this high-pass filter is introduced into the addition/subtraction operator, Perform predetermined addition and subtraction.

That is, in order to form an amplitude signal corresponding to the rotation angle θ, the extracted components shown by equation (15) or (17) above are subtracted from each other as shown below, so that the amplitude becomes a sine wave or a cosine wave, respectively. It is converted into two outputs n ₁ , n ₂ or n ₁ ′, n ₂ ′ that change in a waveform.

That is, for the extracted components j _a to j _d of the above equation (15), n ₁ = j _a − j _c = 2K ₆ cos2θsinωt n ₂ = j _d −j _b = 2K ₆ sin2θsinωt (19) Similarly, , for the extracted components m _a to m _d in equation (17) above, the subtraction is n ₁ ′=m _a −m _c =2K ₈ cosθsinωt n ₂ ′=m _d −m _b =2K ₈ sinθsinωt is carried out, thereby forming an amplitude signal. Next, the phase signal corresponding to the rotation angle θ of the rotor is obtained by adding the extracted components shown by the above equation (16) or (18) as shown below, so that the phase corresponds to the rotation angle θ. It is converted into a signal P or P'.

That is, for the extracted components j _a ′ to j _d ′ in the above equation (16), P=j _a ′+j _b ′+j _c ′+j _d ′ =2K ₆ (cos2θsinωt−sin2θcosωt) =2K ₆ sin(ωt− 2θ) …(21) Similarly, for the extracted components m _a ′ to m _d ′ in equation (18), P′=m _a ′+m _b ′+m _c ′+m _d ′=2K ₈ (cosθsinωt− sinθcosωt) = 2K ₈ sin(ωt−θ) (22) is performed, thereby forming a phase signal.

The above explanation is based on the case where the number of light receiving elements is four, and the amplitude signal and phase signal are obtained by adding and subtracting the outputs of the four light receiving elements. This is similar to the prior art disclosed in the above-mentioned Japanese Patent Publication No. 48-37107, and consists of four light sources and four light receiving elements. Note that the above-mentioned prior arts all use an optical scale as the light amount changing means. For this reason, the sinusoidal change in transmittance occurs for two periods per rotation when using the polarizer plate described above, and one period per rotation when using the eccentric transparent circular rotor, but due to this optical scale. However, except for this point, the output of the light-receiving element is substantially the same.

By the way, the above method requires four pairs of light sources and light receiving elements, making the configuration complicated. In addition, the optical scale as a light amount changing means needs to have four index scales with a 90 degree phase difference facing the pulse scale.
It is unavoidable that the assembly and adjustment is complicated and takes a lot of time.

This configuration was simplified in JP-A-56-94216.
This was disclosed in No. In this method, the number of light sources is reduced from four to two, two of the four light receiving elements are arranged in pairs to face each light source, and the difference in output between the pairs is calculated and added. However, even in this case, the problem of assembly and adjustment of the light amount changing means remains unsolved.

All of the above technologies change the amount of light emitted from the light source in a sine wave pattern,
As disclosed in No. 85115, the amount of light is a constant amount,
The same result can be obtained even if the output of the light receiving element is electrically multiplied by a sine wave.

On the other hand, the device disclosed in Japanese Patent Publication No. 49-221 has two light sources, a light amount changing means consisting of a scale scale and two groups of slits with a 90 degree phase difference opposed to it, and two light sources. This is an optical reading device having a light receiving element. However, this is because the light amount changing means changes the amount of light from the light source in a sine wave shape.
In other words, the device is based on the premise that it has the function of periodically changing the amount of light between positive and negative around zero (the minimum actual change in light amount is zero even assuming an ideal state, and within the positive range). ), and it is necessary to resolve this difficult issue in implementation.

The present invention obtains a desired angle signal from the output of the light receiving elements using a simple signal generating means consisting of two light sources, a light amount converting means consisting of a pair of polarizer plates, and two light receiving elements. Therefore, unnecessary portions generated by the signal generating means are removed by the signal converter.

That is, the present invention provides a detection shaft that is rotatably supported within a sealed housing and has at least one end protruding to the outside, and a layer/transmission shaft that is fixed to the detection shaft within the housing, and the inside and outside are mutually shifted by 45 degrees. A two-layer polarizer rotor, an inner and outer two-layer/or one-layer polarizer plate whose transmission axes are mutually shifted by 45 degrees, which are arranged opposite to the polarizer rotor, and the polarizer rotor and the polarizer plate. Two light sources are arranged inside and outside on one side directly or via a fiber, and are controlled by a sine wave with a phase difference of 90 degrees, and on the other side, they are arranged directly or via a fiber, facing the above light source. an addition/subtraction calculator that calculates the difference between the sum of the outputs of the two light receiving elements and the sum of a signal obtained by multiplying the amplitude of the lighting control sine wave of the two light sources by a predetermined value; ,
It consists of a signal path between the addition/subtraction calculator and the light receiving element or a high-pass filter interposed at the output end of the addition/subtraction calculator.

In this case, when the detection axis rotates, depending on the rotation angle θ, the light receiving element receives the signal as shown in equation (10) above.
g _a ′ and g _b ′ are output. Then, signals K ₇ sin ωt and K ₇ cos ωt, which are obtained by multiplying the amplitude of the sine wave for lighting control of the light source by a predetermined value, are subtracted from the sum as shown in equation (23) below, thereby removing unnecessary components,
Furthermore, a low frequency component corresponding to the rotation angle θ is removed by a high-pass filter, and an output identical to the output shown in equation (21) above is formed.

Next, preferred embodiments of the present invention will be described in detail.

In FIG. 6 and FIG. 7 showing a cross section in the AA direction, an angle detection shaft 103 is rotatably supported on the front cover 101 of the cylindrical casing 100 via a bearing 102. A polarizer rotor 108 in which a polarizer disk 105 is sandwiched between transparent thin glass disks 106 and 107 and bonded together is integrally fixed to the rear end 104 of the detection shaft 103 that penetrates into the detection shaft 103 . On the front cover 101 of the housing 100 located on the left side of the rotor 108, there are two light emitting diodes 110, 111 (however, 111 in the figure is not shown because it is located on the rear side of the detection axis 103).
is fixed to the rear bottom plate 109 of the casing 100, which is located on the right side with the polarizer rotor 108 in between, phototransistors 112 and 113 are fixed at positions facing the light emitting diodes 110 and 111,
The front surfaces of the holding cases 114 and 115 of the phototransistors 112 and 113 have their transmission axes aligned at 45
Polarizer plates 116 and 117 are fixed with a difference of one degree.

FIG. 8 shows a light source 120 including light emitting diodes 110 and 111 and phototransistors 112 and 1.
2 is a block diagram of an addition/subtraction calculator 130 and a filter 140 forming a signal converter for output voltages of No. 13. FIG. In the figure, the configuration of the light source 120 includes a clock pulse oscillator 121, a counter 122 into which the output of the clock pulse oscillator 121 is introduced, a count value output of the counter is introduced as an address designation signal, and a predetermined A read-only memory (ROM) 123 that outputs a "" or "L" signal, and a drive circuit 12 into which 90-degree phase outputs (to be described later) of the read-only memory are respectively introduced to light up the corresponding light emitting diodes 110 and 111.
Consisting of 4,125. Then, as shown in FIG. 9, the read-only memory 123 has the following information:
(A/2) (sinωt+1) waveform [However, A/2
A triangular wave with a minute period is superimposed on the triangular wave (amplitude), and a pulse width modulation signal is written with the period of crossing as "H" and the rest as "L", and that time is written as an address. A pulse width modulation signal with the phase shifted by 90 degrees is also written. Therefore, from the read-only memory 123,
Pulse width modulation signals corresponding to a sine wave and a cosine wave of angular velocity ω are sent to drive circuits 124 and 125, respectively.
The light emitting diodes 110 and 111 emit light with the amounts of light shown as e _a and e _b in equation (2) above, respectively. Next, the signal converter includes an addition/subtraction calculator 130 and a high-pass filter 140.
Note that a signal converter for extracting a phase signal is illustrated here. Now, the operational amplifier 135 of the addition/subtraction calculator 130 receives the pulse-width modulated sine wave and cosine wave output from the read-only memory 123 to each inverter 13.
1,132, input resistor 133,134
is multiplied by a predetermined ratio and added, and the output voltages of the phototransistors 112 and 113 are also added. The output of the operational amplifier 135 is sent to the output terminal 141 via a high pass filter 140.

In the above, the light emitting diode 110,
111 is controlled by a path width modulated sine wave and a cosine wave, respectively, and as a result, the amount of light emitted from the light emitting diodes 110 and 111 is as described above.
Emit light with the amount of light e _a and e _b shown in equation (2). The amount of light transmitted to the phototransistors 112 and 113 is determined by the polarizer disk 105 of the polarizer rotor 108.
and the respective polarizer plates 116 and 117 depending on the intersection angle of the transmission axes thereof. Therefore,
When the rotor 108 rotates integrally with the rotation of the detection shaft 103, each phototransistor 112,
113, voltage signals g _a ′ and g _b ′ shown in the above equation (10) are generated which change in a sine shape and a cosine shape in proportion to the double angle of the rotation angle θ of the detection shaft 103, and the voltage signals are as follows. The signal is added to the addition/subtraction calculator 130 as an addition signal. At the same time, the read-only memory 12
The sine wave and cosine wave drive signals generated by 3 are inverted via inverters 131 and 132, and then their amplitudes are matched with K ₇ shown in equation (10) above by resistors 133 and 134, and addition/subtraction operations are performed. Vessel 1
30 as a subtraction signal. Therefore,
The output q of the addition/subtraction operator 130 is q=g _a ′+g _b ′−K ₇ sinωt−K ₇ cosωt =K ₆ cos2θsinωt−K ₆ sin2θcosωt +K ₆ cos2θ−K ₆ sin2θ (23). Subsequently, this output q is filtered by a high-pass filter 140 to remove the low frequency component therein, that is, the third
The components of the term and the fourth term are removed, and eventually the output terminal 14
1, the phase is relative to the detection axis 1 as shown in equation (21) above.
A phase signal proportional to the rotation angle θ of 03 is transmitted. In the above embodiment, filtering is performed after addition and subtraction operations, but the same effect can be obtained even if the order is reversed. Further, in the above embodiment, the light emitting diode 1 is directly inside the housing 100.
10, 111 and phototransistors 112, 11
3 are placed facing each other, but in order to further reduce the size of the mechanical parts, these may be provided externally, and the transmission and reception of light between them may be performed using optical fibers. Further, in the above embodiment, the case where two sets of light sources and light receiving elements are used is illustrated, but the inverter 13
1,132 outputs are added to the separately provided drive circuits of the third and fourth light emitting diodes, and the transmission axes are set at 45 degrees to each other on the front surfaces of the third and fourth phototransistors facing the third and fourth light emitting diodes. Even if shifted third and fourth polarizer plates are provided and added to the outputs of the first to fourth phototransistors, the above (21)
Similarly, as shown in Eq.

Further, in the above embodiment, the transmission axis of the polarizer plate 105 on the rotor 108 side is single, and the light receiving element 112,
Although the case where the transmission axes of the polarizer plates 116 and 117 on the front side of the 113 are shifted by 45 degrees from each other is illustrated, the 10th
As shown in FIG. 11, which shows the cross section in the C-C direction, the transmission axes are mutually arranged around the inside and outside of the rotor 108'.
The donut-shaped first and second polarizer discs 105'' and 105' shifted by 45 degrees are fixed, and the two light receiving elements 1
The same effect can be obtained even if a polarizer plate 116' having a single transmission axis is disposed in front of the polarizers 12 and 113.

As described above, in the present invention, the signal generating means has a simple configuration of only two light sources, two pairs of polarizers, and two light receiving elements, and unnecessary components can be removed when converting the signal by using the light source. This is done using a lighting control signal, which simplifies assembly and adjustment, and also makes it easier to match the characteristics of the elements, improving reliability.

[Brief explanation of drawings]

Figure 1 is a model diagram showing the principle of a known resolver, Figures 2 and 3 are waveform diagrams showing the excitation signal waveform and output signal waveform of the resolver, and Figure 4 is a model diagram showing the resolver's amplitude signal in a predetermined form. A block diagram showing an example of a signal converter that converts into an angle signal, FIG. 5 is a waveform diagram showing waveforms at each part of FIG. 4, and FIG. 6 is a partial cross section showing the first embodiment of the present invention. 7 is a cross-sectional view taken along the line AA in FIG. 6, FIG. 8 is a block diagram showing an embodiment of the light source of the present invention and a signal converter for forming a phase signal, and FIG. FIG. 10 is a partially sectional front view showing a second embodiment of the present invention, and FIG. 11 is a sectional view along the line CC in FIG. 10. 103: detection axis, 108, 108': rotor,
110, 111: Light emitting diode, 112, 11
3: Phototransistor, 105, 105': Polarizer disk, 116, 116', 117: Polarizer plate,
131, 132: Inverter, 133, 134:
Resistor, 135: Addition/subtraction calculator.

Claims (1)

[Claims] 1. A detection shaft that is rotatably supported within a sealed housing and has at least one end protruding to the outside, and a layer/or transmission shaft that is fixed to the detection shaft within the housing are mutually 45
Two layers of polarizer rotors, the inner and outer layers, are shifted by 45 degrees, and the transmission axis, which is placed opposite to the polarizer rotor, is
Two layers of polarizer plates (inside and outside) shifted by degrees, or one layer of polarizer plates, and the polarizer rotor and the polarizer plate are placed inside and outside of the first side directly or via a fiber, and the light is illuminated by a sine wave with a 90 degree phase difference. Two light sources to be controlled, two light receiving elements placed directly or via a fiber opposite the light sources on the other side, the sum of the outputs of the two light receiving elements, and the two light sources. an addition/subtraction calculator that calculates the difference between the amplitude of the lighting control sine wave and the sum of the signals multiplied by a predetermined value; A photoelectric angle measuring device consisting of a high-pass filter.