JPH0157306B2 - - Google Patents

Description

[Detailed description of the invention] This invention relates to rotational accelerometers.

Conventional rotational accelerometers use the inertia of liquid to determine angular acceleration in an analog manner, and have limits in detection resolution and detection range. An object of the present invention is to provide a rotational accelerometer capable of detecting rotational acceleration with high resolution and over a wide area. This purpose consists of a stator that includes a plurality of excitation poles, a primary winding, and a secondary winding, and a rotor that has a shape that changes the reluctance of the magnetic path passing through each excitation pole of the stator in accordance with the rotation angle. By using a variable magnetic resistance type rotation detector with The rotational acceleration is detected by generating a frequency deviation according to the rotational speed of the motor, detecting the rotational speed moment by moment based on this frequency deviation, and calculating the change in the detected rotational speed per predetermined time. This is achieved using a rotational accelerometer.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

In the rotation detector 12 shown in Fig. 1, the stator (iron core) 1 is composed of four excitation poles A, B, C, and D arranged at 90 degree intervals in the circumferential direction, and are opposed in the radial direction. The two excitation poles A and C form one pair, and the excitation poles B and D form another pair. For the excitation pole pair A and C (or B and D),
Primary windings 2A and 2C (or 2B and 2D) are wound differentially. That is, each magnetic pole A,
Assuming that the direction of the magnetic flux toward the ends in B, C, and D is in positive phase, the windings A and C (or B and D) are wound so that the magnetic fluxes generated by each winding are in opposite phases to each other. Specifically, when the primary winding 2A generates a magnetic flux at the pole A in the direction of exiting from the extreme end as shown by the arrow 1 so that magnetic flux is generated in the direction of entering the extreme part as shown in
The next windings 2A and 2C are wound differentially. As a result, magnetic flux flows in the same direction in the excitation pole pair A and C via the rotor (iron core) 3 arranged in the central space of the stator 1. Similarly, for the other excitation pole pair B and D, the primary windings 2B and 2
D is wound. The reason why the primary winding is differentially wound around the excitation pole pair A and C (or B and D) is that each excitation pole pair A and C
Alternatively, this is because B and D are excited by different AC signals, and this is to ensure the flow of magnetic flux between poles (A and C or B and D) that are excited by the same AC signal.

A rotor 3, which faces the ends of each of the excitation poles A to D with a suitable gap interposed therebetween, rotates together with the rotating shaft 4. A rotation angle θ to be detected is given to this rotation axis 4. The rotor 3 has a shape that changes the reluctance of the magnetic path passing through each stator excitation pole A, B, C, and D according to the rotation angle θ. In the example shown in FIG. 1, the rotor 3 is connected to the rotating shaft 4.
It has a cylindrical shape and is mounted eccentrically with respect to the center. Due to this eccentric cylindrical shape, the gap distance between the cylindrical side surface of the rotor 3 and the end of each pole A, B, C, and D changes depending on the rotation angle θ. This gap change causes a reluctance change in each pole A, B, C, and D corresponding to one cycle of the trigonometric function per one revolution of the rotor 3.

The excitation pole pair consisting of A and C and the excitation pole pair consisting of B and D are separately excited by alternating current signals that are 90 degrees out of phase. In the figure, primary windings 2A and 2C of poles A and C are connected in series, and a sine wave signal i _a =Isinωt is applied from an oscillator 5. Also, poles B and D
The primary windings 2B and 2D of are connected in series, and a cosine wave signal i _b =Icosωt is applied from an oscillator 6. If you extract only the primary windings 2A and 2C, they appear to be connected in series in the same phase, but consider the direction of the magnetic poles A and C around which they are wound, that is, the direction of the magnetic flux generated by both. Then, both are substantially connected in reverse phase series (ie, differentially wound). The same applies to the primary windings 2B and 2D.

In the above configuration, each excitation pole A, B, C, D
2 to extract the voltages induced respectively by
The next winding 7 is wound around the stator 1. In the example shown in Figure 1, secondary windings 7A and 7C are wound around excitation poles A and C, respectively, in phase, and secondary windings 7B and 7D are wound around excitation poles B and D, respectively, in phase. ,7A
and 7C, 7B, and 7D are in opposite phases to each other. These secondary windings 7A to 7D are connected in series so that a composite signal E of voltages induced at each excitation pole A, B, C, and D is extracted. This output signal E is an excitation AC signal i _a =
This becomes an AC signal with a phase shift corresponding to the rotation angle θ of the rotor 3 with respect to Isinωt or i _b =Icosωt. This can be easily confirmed using a prototype, and the reason for this can be analyzed as follows.

An equivalent circuit of the magnetic circuit formed in the rotation detector 12 shown in FIG. 1 can be shown as shown in FIG. N is each primary winding 2A, 2B, 2C, 2
Indicates the number of turns of D. i _a and i _b are excitation AC signals
The instantaneous current values of Isinωt and Icosωt are shown. Therefore,
Ni _a , Ni _b , _-Nia , -Ni _b are each excitation pole A, B, C,
The magnetomotive force generated by the primary windings 2A to 2D of D is shown. P _A , P _B , P _C , P _D are each magnetic pole A, B,
The permeance caused by the gap between C and D and the rotor 3 is shown respectively. As mentioned above, since the rotor 3 is configured so that a reluctance change corresponding to one cycle of trigonometric functions is brought about per rotation, each permeance P _A to P _D can be expressed as the following equation. .

P _A =P ₀ −P ₁ sinθ P _B =P ₀ −P ₁ cosθ P _C =P ₀ −P ₁ sinθ ………(1) P _D =P ₀ +P ₁ cosθ P ₀ and P ₁ are stator 1 and This is a constant determined depending on the size, magnetic permeability, etc. of the rotor 3. Above number
In equation (1), as shown in FIG. 1b, the rotation angle θ is set to 0 degrees when the gap distance between the magnetic pole D and the rotor 3 is the minimum. φ _A , φ _B , φ _C and φ _D represent the magnetic flux passing through the gap between each magnetic pole A, B, C, D and the rotor 3, respectively. As is clear from the equivalent circuit, the following relationship exists: φ _A + φ _B + φ _C + φ _D = 0 (2). U indicates the magnetic field of the entire equivalent circuit, U=Ni _a +φ _A /P _A = Ni _b +φ _B /P _B = −Ni _a
+φ _C /P _C =−Ni _b +φ _D /P _D ......(3) There is a relationship. Therefore, each magnetic flux φ _A to φ _D can be expressed as follows.

φ _A = (U−Ni _a )P _{A φB} = (U−Ni _b )P _{B φC} = (U+Ni _a )P _C ………(4) φ _D = (U+Ni _b )P _DHere , each When the number of turns of the secondary windings 7A, 7B, 7C, and 7D is respectively _N2 , the voltages e _A , e _B , and e induced in each of the secondary windings 7 A to 7 D according to the gap of each magnetic pole A to D are _C and e _D are expressed as follows.

e _A = N ₂ d/dtφ _A e _B = −N ₂ d/dtφ _B e _C = N ₂ d/dtφ _C e _D = −N ₂ d/dtφ _D ……(5) Secondary winding 7 (7A ~7D) composite output signal E
is the above equation (5), equation (4), equation (3), equation (1) and i _a =
From Isinωt, i _b =Icosωt, it can be shown as follows.

E=e _A +e _B +e _C +e _D =N ₂ d/dt (φ _A −φ _B +φ _C −φ _D )
=N ₂ d/dt {(U−Ni _a )P _A −(U−Ni _b )P _B +(U+Ni _a )P _C −(U+Ni _b )P _D }=N ₂ d/dt{U
(P _A − P _B + P _C − P _D ) − Ni _a (P _A − P _C ) + Ni _b (P _B − P _D )}
=N ₂ d/dt (−2Ni _a P ₁ sinθ−2Ni _b P ₁ cosθ) = 2N ₂ NP ₁
d/dt (−Isinωtsinθ−Icosωtcosθ) = 2N ₂ NP ₁ I (−cosωtsinθ + sinωtcosθ) = 2N ₂ NP ₁ I
sin(ωt−θ)……(6) However, from the above equation (1), P _A −P _B +P _C −P _D =0
It is. In the above equation (6), the coefficient (2N ₂ NP ₁ I)
Since is a constant, by replacing it with K, it can be expressed as E=Ksin(ωt−θ) (7). As is clear from this equation (7), the output signal E is an AC signal whose phase is shifted from the excitation AC signal Isinωt by a phase angle corresponding to the rotation angle θ.

Here, when the rotating shaft 4 rotates at an appropriate angular velocity or angular acceleration, the phase shift (i.e. rotation angle) θ in the above equation (7) is expressed as a function of time θ as shown in the following equation.
It is given by (t).

E=Ksin {ωt±θ(t)} ......(8) The sign (±) of the phase shift function θ(t) indicates the direction of the phase shift (advanced or slow), which is the axis 4 corresponds to the direction of rotation. For convenience of explanation, in the following it is assumed that a phase shift occurs in the phase advancing direction, and +θ
(t). This phase shift function θ(t)
includes a component of the angular velocity or angular acceleration of the rotating shaft 4.

When the shaft 4 is rotating at a constant angular velocity ω _M , d/dtθ(t) = ω _M (9), and the integral of this angular velocity ω _M is the phase shift amount θ(t). Therefore, the above equation (8) can be rewritten as follows. However, θ ₀ is the initial phase.

E=Ksin {(ω+ω _M )t+θ ₀ } ………(10) Also, if the axis 4 is rotating with a constant angular acceleration α _M , d/dtθ(t)=α _M t ………( 11) Therefore, θ(t)=∫α _M tdt=α _M /2t ² +θ ₀ (12), and the above equation (8) can be rewritten as follows.

E=Ksin {(ω+α _M /2t)t+θ ₀ } ......(13) As in the above equation (10) or (13), the phase shift of the rotation detection signal E obtained from the rotation detector 12 of the present invention Since the component of the rotational angular velocity ω _M or the angular acceleration α _M is included in the rotational velocity ω M , the rotational velocity or acceleration can be determined by analyzing the phase shift amount θ (generally θ(t)).

Next, how to obtain the angular velocity ω _M and the angular acceleration α _M will be explained in principle with reference to FIG.

An example of a rotation detection signal E (indicated by E _S ) obtained when the rotating shaft 4 rotates at an angular velocity ω _M is shown by a dashed line in FIG. 3 . The waveform indicated by the solid line indicates the excitation AC signal Isinωt, and the waveform indicated by the broken line indicates the rotation detection signal E when the stationary state is at a constant rotation angle θ ₀ .
(denoted as E ₀ ). The signal E _S during rotation is θ ₀
The figure shows the state when rotated with , as the initial phase. t ₀ is one cycle of the rotation detection signal E ₀ at rest, which is the same as the cycle of the excitation AC signal Isinωt. t _S is one cycle of the rotation detection signal E _S during rotation. It can be seen from FIG. 3 that during rotation, the frequency of the rotation detection signal E (ie, E _S ) deviates from the reference frequency (ω). This is also clear from equation (10) above, and the amount of frequency deviation corresponds to the angular velocity ω _M . If the angular frequency of the rotation detection signal E _S during rotation is ω _S , E _S can be expressed as follows from equation (10).

E _S = K'sin (ω _S t+θ ₀ ) = K'sin {(ω+ω _M )t+θ ₀ } ......(14) In Figure 3, △θ is the reference signal Isinωt and the rotation detection signal at a certain point in time. E (i.e. E _S )
It is the difference between the phase shift θ ₀ with respect to θ 0 and the phase shift θ _S after t _S seconds. If it is stationary, θ ₀ = θ _S and △θ
is 0, but when rotating, this Δθ takes a value corresponding to the angular velocity ω _M of the rotating shaft 4. In other words, as is clear from Fig. 3, the reference signal
If one period t ₀ of Isinωt is 2π (radians), the phase value corresponding to time t _S is ∫ ^tS ₀ ωdt, and △θ is expressed as △θ=2π−∫ ^tS ₀ ωdt (15) Ru. Here, from the above equation (14), ω _S = ω + ω _M ω = ω _S − ω _M ………(16), and substituting this into equation (15), △θ=2π−∫ ^tS ₀ ω _S dt +∫ ^tS ₀ ω _M dt＝∫ ^tS ₀ ω _M dt (17). Furthermore, since ω _S =2π1/t _S , ∫ ^tS ₀ ω _S dt=
2π. As is clear from equation (17) above, △θ
is a function of the angular velocity ω _M. Solving this, we get △θ=ω _M・t _S ω _M = △θ/t _S (18). Therefore, based on △θ and t _S, the angular velocity ω _M
can be found.

Specifically, t _S is the rotation detection signal E (i.e. E _S )
This can be determined by counting one period of the clock pulse CP. The count value corresponding to t _S is n _S , and 1 of the clock pulse CP is
If the period is φ (seconds), then t _S =n _S・φ (19). Δθ can also be determined based on this count value n _S . If the count number of the clock pulse CP corresponding to one period t ₀ of the excitation signal Isinωt is n ₀ , its angular frequency ω can be expressed as ω=2π1/t ₀ =2π1/n ₀・φ (20) Ru. Substituting this equation (20) into the above equation (15) and solving the integral term, △θ=2π−2π・n _S /n ₀ =2π・n ₀ /n ₀ −2π・n _S /n ₀ = 2π/n ₀ (n ₀ −n _S ) (21). Substituting equations (21) and (19) into equation (18) yields ω _M =2π/n ₀ ·φ (n ₀ −n _S /n _S ) (22). Here, 2π, n ₀ and φ are constants, so
The angular velocity ω _M can be obtained by simply counting one cycle of the rotation detection signal E and calculating the equation (22) based on the count value n _S .

In addition, by replacing the above equation (16), we set ω _M = ω _S −ω = 2π1/t _S −2π1/t ₀ = 2π (1/n _S・φ−1/n ₀・φ) and solve this. However, the same solution as the above equation (22) can be obtained.

The following relationship holds true between angular acceleration α _M and angular velocity ω _M.

α _M =d/dtω _M △ω _M /△t (23) Here, △ω _M is the change in the angular velocity ω _M during the time change △t. Let ω _M1 be the angular velocity determined at time t ₁ , and then at time t ₂ t _S (seconds) later.
If the angular velocity found in is ω _M2 , then △t=t _S , △ω _M = ω _M2 −ω _M1 , and from the above equation (19), t _S = n _S・φ, so the above (23) The formula is α _M = ω _M2 −ω _M1 /t _S = ω _M2 −ω _M1 /n _S・φ……(2
4) It can be rewritten as

Therefore, the angular velocity ω _M is detected every period t _S of the rotation angle detection signal E (that is, E _S ), the difference between the currently detected angular velocity ω _M2 and the previously detected angular velocity ω _M1 is determined, and the difference is calculated as follows: The angular acceleration α _M can be determined by dividing by the product of the count value n _S and the clock pulse period φ.

Next, an example of a circuit that executes the above operation is shown in the fourth section.
As shown in the figure.

In FIG. 4, the rotation detector 12 is connected to the primary winding 2.
Only A, 2C, 2B, 2D and the secondary winding 7 are schematically shown, and the others are not shown, but it is the same as in FIG. 1. First, the excitation AC signal Isinωt,
The circuit that forms Icosωt will be explained. Oscillator 15 generates a high speed clock pulse CP.
The frequency dividing circuit 16 divides the frequency of this clock pulse CP by 1/M and outputs a pulse Pb with a duty of 50% and an inverted signal Pa of this pulse Pb. (However, M is any integer). Specifically, it includes a 2/M frequency divider 17 and a flip-flop 18 for 1/2 frequency division, and a pulse Pc obtained by dividing the clock pulse CP by 2/M is obtained from the frequency divider 17, and this pulse Pc is sent to the flip-flop 18. Divide the frequency by 1/2. As a result, flip-flop 18
The output (Q) of is a square wave pulse Pb with a duty of 50% and a frequency of 1/M of the clock pulse CP.
is output, and its inverted output () outputs the pulse
A square wave pulse Pa, which is the inversion of Pb, is output. 180
The pulses P _b and P _a with a degree phase shift are input to the flip-flops 19 and 20 for 1/2 frequency division, respectively, and the pulses 1/2 P _b and P _a are obtained by dividing these pulses P b and P _a by 1/2, respectively. 1/2P _a is obtained. Also, from the inverted output () of flip-flop 20, a pulse 1/2P _a
A pulse 1/2P _a ′ is obtained by inverting . Therefore,
Each pulse CP, P _c , P _b , P _a , 1/2P _b , 1/2P _a , 1/2
The relationship between P _a ' is shown in Figure 5. That is, the pulses 1/2P _b and 1/2P _a output from each flip-flop 19 and 20 are clock pulses.
It has a frequency of 1/2M of CP and is 90 degrees out of phase. Furthermore, 1/2P _a and 1/2P _a ′ are out of phase by 180 degrees.

Pulse 1/2P _b is input to the low pass filter 21 . One of the pulses 1/2P _a and 1/2P _a ' is selected by the switching circuit 40 and input to the low-pass filter 22 . For now, the explanation will proceed assuming that pulse 1/2P _a has been selected. In the low-pass filters 21 and 22, each pulse 1/2P _b , 1/2
The fundamental wave component of P _a is extracted. Therefore, if the signal output from the low-pass filter 21 is a cosine wave signal cosωt, the low-pass filter 22
The signal output from is a sine wave signal sinωt.
The output cosωt of the low-pass filter 21 is
The output Icosωt is amplified by the primary winding 2B of one excitation pole pair B and D (or 9B and 9D).
and applied to 2D. The output sinωt of the low-pass filter 22 is amplified by the amplifier 24, and its output
Isinωt is applied to the primary windings 2A and 2C of the other excitation pole pair A and C (or 9A and 9C).

As mentioned above, the output winding 7 generates an alternating current signal E=E _S = with a frequency deviation ω _M corresponding to the rotational speed.
K′sin(ω _S t+θ ₀ ) is obtained. The direction of the frequency deviation ω _M with respect to the reference frequency ω differs depending on the direction of rotation of the rotating shaft 4, but for now it is assumed that the rotation is in the positive direction, that is, in the direction in which the frequency ω _S is higher than the reference frequency ω. , this direction is defined as clockwise.

The speed detection circuit 51 receives the rotation detection signal E (i.e.
This is a circuit for determining the rotational speed based on E _S ), and includes a period calculation circuit 41, an arithmetic circuit 48, and a latch circuit 49.

The period calculating circuit 41 is for calculating the period t _S of the output signal E (ie, E _S ) of the rotation detector 12.
By counting one period of the signal E _S using the clock pulse CP, a count value n _S corresponding to t _S is output. Rotation detection signal E (i.e. E _S )
is input to the comparator 42, and as shown in FIG. 6, "1" or "0" is output from the comparator 42 depending on its polarity. The monostable multivibrator 43 receives the output signal F of the comparator 42.
A single pulse G (with a time width of about 100+1 seconds, for example) is output at the rising edge of . The monostable multivibrator 44 outputs one pulse H at the falling edge of the pulse G (see FIG. 6). Therefore, the pulse G is generated corresponding to one period t _S of the rotation detection signal E (E _S ), and the pulse H is generated slightly later than the pulse G. The frequency dividing circuit 45 divides the clock pulse CP into 1/2.
The frequency is divided and the output is counted by a counter 46. A pulse H is supplied to the reset input of this counter 46. Further, the count output of the counter 46 is input to the register 47. A pulse G is supplied to the load control input of register 47. Therefore, the counter 46 is reset immediately after the count value is taken into the register 47 by the pulse G. Since the counter 46 is reset by the pulse H every period t _S of the rotation detection signal E (that is, E _S ), the counter 46 holds the count value n _S corresponding to t _S at the timing of the immediately preceding pulse G. Ori,
This n _S is stored in register 47.

The arithmetic circuit 48 is a circuit that executes an arithmetic operation corresponding to equation (22) above. That is, since 2π, n ₀ and φ are constants known in advance, register 47
Calculations can be performed in the following order using the count value n _S stored in .

n ₀ −n _S →R ₁ R ₁ /n _S →R ₂ 2π/n ₀ ·φ×R ₂ →R ₃ Note that R ₁ , R ₂ , and R ₃ indicate the calculation results at each calculation step. In this example, the clock pulse
Since the frequency of CP is divided by 1/2 and counted by the counter 46, φ in the above equation indicates a period twice the period of the clock pulse CP.
That is, in this example, it is assumed that the period of the clock pulse CP is φ/2. By the way, concrete examples of each constant are as follows.

The reference AC signal Isinωt is the clock pulse CP divided by a ratio of 1/2M, so its period t ₀ is the period of the clock pulse CP multiplied by 2M, t ₀ =
It becomes φM. The count value n ₀ for this one cycle is n ₀ =
t ₀ /φ=M. Here, if the frequency division ratio M of the frequency dividing circuit 16 is 9766, n ₀ =9766. Also, if the frequency of clock pulse CP is 3.2MHz, φ=
It is 1/1600000.

Furthermore, what is found by calculating equation (22) is the angular velocity ω _M , so the number of revolutions per second (rps)
To find , just divide the angular velocity ω _M by 2π. Generally, the rotational speed is expressed not by the angular velocity ω _M but by the number of rotations, so in the calculation step described above, the arithmetic circuit 48 multiplies the coefficient 1/n ₀ ·φ, which is obtained by dividing 2π in advance. good. That is, the above is changed as follows.

'1/n ₀ ·φ×R ₂ →R ₃ The calculation result R ₃ (ie, ω _M /2π) thus obtained represents the rotation speed of the rotating shaft 4. This calculation result _R3 is latched in the latch circuit 49. This latch circuit 49 takes in data at the timing of pulse G. Therefore, the data X indicating the rotation speed held in the latch circuit 49 is rewritten every cycle _tS of the rotation detection signal E. Note that the arithmetic circuit 48 receives a clock pulse to control the arithmetic timing.
CP and pulse G are input. Of course, the angular velocity ω _M may be obtained by executing the calculation steps as they are.

The digital comparator 50 is for discriminating the rotational direction of the rotary shaft 4, and compares whether the count value n _S output from the register 47 is larger than the reference count value n ₀ . The comparator 50 is n _S > n ₀
When , outputs “1” at the timing of pulse G,
In other cases, "0" is output. Comparator 50
The output of is applied to the switching circuit 40. Switching circuit 4
0 switches the pulse to be selected from 1/2P _a to 1/2P _a ' or vice versa every time "1" is applied from the comparator 50.

Now, pulse 1/2P _a is selected in the switching circuit 40,
It is assumed that the rotation direction is clockwise and that the frequency ω _S of the rotation detection signal E (ie, E _S ) is higher than the reference frequency (ω _S >ω). At this time, n _S <n ₀ and the output of the comparator 50 is "0". Therefore, pulse 1/2P _a is still selected. If the rotation direction changes counterclockwise in this state, the angular velocity ω _M becomes a negative value - ω _M , so ω _S < ω, and the count value becomes
n _S > n ₀ . Therefore, the output of the comparator 50 becomes "1", and the pulse selected by the switching circuit 40 is switched from 1/2P _a to 1/2P _a '. The output of the filter 22 corresponding to 1/2P _a ' is -sinωt, which means that i _a in the above equation (6) changes to -Isinωt. This(6)
As can be seen by replacing _ia in the equation with -Isinωt, only the polarity of the rotation detection signal E is reversed, and the direction of the phase shift θ does not change. In other words, if the direction of the phase shift of the primary winding excitation signals i _a and i _b does not change, when the rotation direction changes, the direction of the phase shift θ of the rotation detection signal E will also be reversed to positive or negative; When the direction changes, the excitation signal i _a of the primary winding,
By reversing the direction of the phase shift of i _b , the direction of the phase shift θ of the rotation detection signal E can always be kept constant. In this way, by switching the excitation signal, n _S <n ₀ is established for counterclockwise rotation. Therefore, the output of the comparator 50 becomes "0" immediately.

As mentioned above, when the pulse 1/2P _a ' is selected in the switching circuit 40, when the rotation direction changes from counterclockwise to clockwise, n _S > n ₀ holds, and the comparator 50 outputs "1". ” is output. As a result, the switching circuit 40 is switched to a state in which pulse 1/2P _a is selected. Therefore, n _S <n ₀ soon comes to hold for clockwise rotation.

As described above, by the functions of the comparator 50 and the switching circuit 40, n _S <n ₀ can always be satisfied regardless of the rotation direction, and the configuration of the arithmetic circuit 48 can be simplified (n ₀ - n Since _S is always positive), the processing of the rotational speed data X output from the latch circuit 49 is also simplified. That is, if the comparator 50 and the switching circuit 40 are not provided, the rotation speed data X output from the latch circuit 49 will be a positive or negative value depending on the rotation direction, so a circuit for calculating this absolute value is required. , such a circuit becomes unnecessary. Of course, the present invention can be practiced without providing the comparator 50 and the switching circuit 40.

The acceleration calculation circuit 52 executes the calculation of equation (24) based on the speed data obtained by the speed detection circuit 51, and calculates the acceleration α _M. The acceleration calculation circuit 52 receives the data latched in the latch circuit 49 as data _X1 indicating the previous rotational speed, and outputs the data from the calculation circuit 48 as data _X2 indicating the current rotational speed. The calculation result R ₃ at ' is input, and the count value n _S indicating one period t _S of the rotation detection signal E stored in the register 47 as data indicating the time difference between the previous time and the current time is input. Based on these data, the following calculation is executed: X ₂ −X ₁ /n _S ·φ (25) to obtain the rotational acceleration. here
Assuming that X ₂ = ω _M2 /2π and X ₁ = ω _M1 /2π, the above equation can be expressed as ω _M2 −ω _M1 /2π・n _S・φ=α _M /2π, which is similar to equation (24) above. It can be seen that equivalent operations are being performed. That is, (24)
In the equation, the angular acceleration α _M is obtained, but in the equation (25), the change in the number of revolutions per second, that is, the acceleration α _M /2π is directly obtained. Of course, X ₁ and X ₂ are angular velocities ω _M1 ,
If it represents ω _M2 , as in equation (24),
The arithmetic circuit 52 calculates the angular acceleration α _M. Data indicating the acceleration α _M /2π determined by the arithmetic circuit 52 is input to the latch circuit 53 and taken into the latch circuit 53 at the timing of the pulse G. In this way, the latch circuit 53 outputs data indicating rotational acceleration. This data is 1 of rotation detection signal E.
It is rewritten every cycle.

Next, other embodiments of the rotation detector 12 will be listed.

FIG. 7 shows an embodiment in which the structure of the stator 1 is the same as that in FIG. 1, but the shape of the rotor 8 is different from that of the rotor 3 in FIG. This rotor 8 has the shape of a cylinder cut diagonally, and the center of this cylinder and the center of the rotating shaft 4 coincide. Although the ends of each of the magnetic poles A to D face the cylindrical side surface of the rotor 8, the gap distance does not change, and the opposing area changes according to the rotation angle θ of the rotor 8. Therefore, with the rotor 8 of FIG. 7, the reluctance of the gap of each magnetic pole A to D can be changed in accordance with the rotation angle .theta., similarly to the rotor 3 of FIG. 1. That is, it is possible to obtain the same permeance change as in equation (1) above.

In the embodiments shown in FIGS. 8 and 9, the structure of the stators 9 and 9' is slightly different from that of the stator 1 shown in FIGS. 1 and 7. In FIG. 8, the stator 9 includes excitation poles 9A, 9B, 9C, and 9D arranged at intervals of 90 degrees in the circumferential direction, and an output magnetic pole 9E located on an extension of the rotating shaft 4. Similar to FIG. 1, primary windings 2A and 2C are differentially wound around one excitation pole pair 9A and 9C, and are excited by a sinusoidal signal i _a =Isinωt. Further, primary windings 2B and 2D are differentially wound around the other excitation pole pair 9B and 9D, and are excited by a cosine wave signal i _b =Icosωt. A secondary winding 7 is wound around the output magnetic pole 9E. In this case, one secondary winding 7
As a result, a composite signal E of induced voltages caused by all the excitation poles 9A to 9D can be extracted. In FIGS. 1 and 7, the ends of each magnetic pole A to D face in the radial direction, but in FIGS. 8 and 9, each magnetic pole 9A
The ends of ~9E point in the axial direction. In FIG. 8, the rotor 10 has a disc shape eccentrically attached to the shaft 4, and although the gap distance from the end of each excitation pole 9A to 9D does not change, the opposing area is the angle of rotation. It is designed to change according to θ. Therefore, even in the configuration shown in FIG.
Equation (1) can be modified to obtain similar permeance changes.

The stator 9' in FIG. 9 has almost the same configuration as the stator 9 in FIG. 8, except that the output magnetic pole 9E' is slightly longer than the other magnetic poles 9A to 9D. The rotor 11 is a swash plate, and each excitation pole 9A to 9
The distance between the gap and D changes according to the rotation angle θ. Therefore, even in the configuration shown in FIG. 9, it is possible to obtain the same permeance change as in equation (1).

Note that the AC signals that excite each excitation pole pair A and C (9A and 9C) and B and D (9B and 9D) are not limited to sine waves and cosine waves, but also inverted signals (-cosωt) of sine waves and cosine waves. Alternatively, it may be an inverted signal (-sinωt) of a cosine wave and a sine wave, as long as the phases are shifted by 90 degrees.

Note that the excitation pole pairs may be arranged not on the same circumference but on different circumferences having the same central axis. An example is shown in FIG. In FIG. 10, the stator section includes two axially arranged
It consists of two stators 1A and 1B. The stator 1A has two excitation poles A and C facing each other in the radial direction.
primary windings 2A and 2C wound around each
are connected in series so as to generate magnetic flux in opposite directions, and are excited by a sinusoidal signal (Isinωt).
The stator 1B similarly has two excitation poles B and D facing each other in the radial direction, and the primary windings 2B and 2D wound around each are connected in series so as to generate magnetic flux in opposite directions. It is excited by a wave signal (Icosωt). Both stators 1A and 1B are arranged such that the positions of the respective excitation pole pairs A and C and B and D are shifted by 90 degrees. Each excitation pole A,
The outputs of the secondary windings 7A to 7D wound around B, C, and D are combined and extracted in the same manner as in FIG. The rotor 32 is similar to the rotor 3 in FIG.
It consists of a cylindrical iron core mounted eccentrically with respect to the shaft 4.

Furthermore, an E-shaped stator 33 as shown in FIG. 11 may also be used. Primary windings 34A and 34B are wound around excitation poles 33A and 33B located at both ends of the stator 33, respectively. 34A is excited by a sine wave signal (Isinωt), and 34B is excited by a cosine wave signal (Icosωt). A secondary winding 7 is wound around a magnetic pole 33E located at the center of the stator 33. The rotor 35 has a shape in which both ends of a cylindrical core are cut diagonally, and the center of this cylinder coincides with the center of the shaft 4. Both ends of the rotor 35 are not cut in the same direction (parallel),
The slope at one end is twisted 90 degrees to the slope at the other end. Ends of each magnetic pole 33A and 33B and rotor 35
The cylindrical side surface of the rotor 3 faces the rotor 3, and its opposing area is
By changing the reluctance according to the rotation angle θ of No. 5, it is possible to obtain a reluctance change according to the rotation angle θ. Here, due to the 90 degree twist of the slopes at both ends of the rotor 35, a 90 degree difference occurs between the reluctance change at the pole 33A and the reluctance change at the pole 33B. As a result, the pole 33A is excited by the sine wave and the pole 3A is excited by the cosine wave.
3B and 3B by 90 degrees, and similarly to the embodiments shown in FIG. 1 or FIGS. 7 to 9, the phase shift according to the rotation angle θ of the rotor 35 The generated AC signal is connected to the 2nd pole of pole 33E.
It can be obtained from the next winding 7. As shown by the dashed line, a similar E-shaped stator 33' is provided on the opposite side of the stator 33, and a primary winding excited by a sine wave signal is differentially wound around the excitation poles 33A and 33C. In addition, primary windings excited by a cosine wave signal may be differentially wound around the excitation poles 33B and 33D. In that case, the output of the secondary windings provided at the central magnetic poles 33E and 33E' of both stators 33 and 33', respectively, is synthesized as an AC signal (Ksin(ωt −θ).

Further, as shown in FIG. 12, slot teeth may be provided on each of the magnetic poles A to D of the rotor 36 and stator 37. Primary windings 2A to 2D and 2 of each magnetic pole A to D
The next windings 7A to 7D are constructed in the same manner as in FIG. The correspondence between the slot teeth of the rotor 36 and the teeth of each magnetic pole A to D of the stator 37 is such that a mechanical phase shift equivalent to 1/2 pitch occurs between the paired magnetic poles (A and C or B and D). It's becoming like that. As a result, the permeance between one magnetic pole A (or B) and the rotor 36 and the permeance between the other magnetic pole C (or D) and the rotor 36 change differentially, with one tooth pitch being one cycle. do. Further, it is designed so that a mechanical phase shift of 1/4 pitch occurs between each pair of magnetic poles (A, C and B, D). Therefore, the same permeance changes P _A to P _D as in equation (1) above correspond to one cycle of the slot teeth.
Occurs as a cycle. As a result, when the number of pitches of the slot teeth for one rotation is N, the phase shift θ in equation (7) is expanded by N times and appears as Nθ. Therefore, the angular velocity ω _M obtained according to the above equation (19) can also be obtained by enlarging the actual angular velocity by N times, increasing the detection resolution.

The rotation detector 12 is not limited to the one shown in the above embodiment, but in short, it is a variable magnetic resistance type detector equipped with a primary winding and a secondary winding on the stator side, and is capable of detecting changes in permeance of each magnetic pole. Any structure may be used as long as each pole is individually excited by a plurality of alternating current signals having an electrical phase shift corresponding to a mechanical phase shift.

Furthermore, the output signal E of the rotation detector 12 (i.e.
It goes without saying that the circuit for determining the rotational speed and acceleration based on E _S ) is not limited to the one shown in FIG. 4, and the design can be modified as appropriate. In Figure 4, the period
The rotation speed is detected based on t ₀ , t _S (n ₀ , n _S ), but as shown in Fig. 13, the rotation detection signal E
The rotational angular velocity ω _M may be determined by directly measuring the frequency ω _S of (that is, E _S ).
In FIG. 13, symbols 18, 19, 20, 2
1, 22, 23, and 24 are circuits that perform the same functions as the circuits with the same symbols in FIG.
A sine wave signal is generated based on the oscillation output of the oscillation circuit 60.
Isinωt and a cosine wave signal Icosωt are generated. The comparator 61 to which the rotation detection signal E (i.e. E _S ) output from the rotation angle detector 12 is inputted is the fourth comparator 61 .
It performs the same function as the comparator 42 shown in the figure, and outputs a pulse signal F having the same frequency as the rotation detection signal E. The frequency measuring circuit 62 measures the frequency ω _S of the pulse signal F, and based on this, outputs data X ₂ indicating the current rotational angular velocity ω _M . That is,
Since ω _M =ω _S −ω from the above equation (13), ω _M can be obtained by measuring the frequency ω _S of the signal F (ie, the rotation detection signal E).

FIG. 14 shows the frequency/frequency measurement circuit 62.
An example will be shown in which the voltage converter 62A is used to output an analog voltage V (ω _M ) indicating the rotational angular velocity ω _M . In this case, frequency/voltage converter 6
If the offset frequency of 2A is set to the reference frequency ω, the output voltage V corresponding to the input signal F
(ω _M ) is proportional to the difference ω _S −ω between the input frequency ω _S and the offset frequency ω, that is, the rotational speed ω _M. Therefore, the rotation speed can be determined extremely easily.

In FIG. 15, a frequency counter 62B is used as a frequency measuring circuit 62 to output digital data C(ω _M ) indicating the rotational speed. Data C (ω _S ) indicating the counted frequency ω _S
and data C(ω) indicating the reference frequency ω are subtracted by a subtracter 62C to obtain speed data C(ω _M ).

The method of directly measuring the frequency as shown in Fig. 13 uses a high-resolution rotation detector 1 as shown in Fig. 12.
It is particularly advantageous when 2 is used. because,
This is because a high-resolution rotation detector produces a large frequency deviation corresponding to the rotational angular velocity, which improves the accuracy of frequency measurement.

The data output from the frequency measurement circuit 62 is held in the storage circuit 63 at an appropriate timing. The output of the memory circuit 63 is the _previous speed data
The output of the frequency measurement circuit 62 is input to the acceleration calculation circuit 52 as the current speed data _X2 ,
Acceleration α _M is determined based on the amount of change (X ₂ −X ₁ ).

As explained above, according to the present invention, the rotational speed is detected based on the frequency deviation (period shift) of the rotation detection signal with respect to the reference AC signal, and the rotational acceleration is detected based on the change. Therefore, the range of rotational acceleration that can be detected becomes extremely wide, and the resolution can also be increased, which is an excellent effect.

[Brief explanation of drawings]

FIG. 1a is a schematic side cross-sectional view of an embodiment of the rotation detector of the rotational accelerometer according to the present invention, FIG. 1b is a schematic front view of a, and FIG. The equivalent circuit diagram of the magnetic circuit, Figure 3 is the second embodiment of the above example.
A waveform diagram illustrating that the frequency of the rotation detection signal obtained from the next winding shifts depending on the rotation speed, and FIG. 4 is a block diagram illustrating an embodiment of the electric circuit portion of the rotational accelerometer according to the present invention. , FIG. 5 is a timing chart showing the operation of the circuit section that forms the excitation AC signal in FIG. 4, and FIG. 6 is a timing chart showing the operation of the circuit section that detects the cycle of the rotation detection signal in FIG. 4. Figure 7a is the first
FIG. 8a is a schematic side cross-sectional view of another embodiment of the rotation detector portion shown in FIG. Figure b is a schematic front view of a, Figure 9 a is a schematic side sectional view showing another embodiment of the same part, Figure b is a schematic front view of a, and Figure 10 a is a schematic side view of another embodiment of the same part. Fig. 11a is a schematic side view showing another embodiment of the same part, Fig. 12b is a schematic front view of a, and Fig. 12a is a schematic front view of the same part. A radial cross-sectional view showing the embodiment, FIG. b is a side cross-sectional view of a, and the first
FIG. 3 is a block diagram showing another embodiment of the electric circuit portion of the rotational accelerometer according to the present invention, and FIGS. 14 and 15 are diagrams showing an example of the frequency measuring circuit shown in FIG. 13, respectively. 1, 9, 9', 37...Stator, 2A to 2D
...Primary winding, 3, 8, 10, 11, 36...Rotor, 4...Rotating shaft, A, B, C, D, 9A~9
D... Excitation pole, 5, 6... Oscillator, 7, 7A~7
D... Secondary winding, 12... Rotation detector, 15...
Clock oscillator, 16... Frequency dividing circuit, 41... Period calculation circuit, 48... Arithmetic circuit, 51... Speed detection circuit, 52... Acceleration computing circuit.

Claims (1)

[Claims] 1. A stator including a plurality of excitation poles, a primary winding, and a secondary winding, and a shape that changes the reluctance of a magnetic path passing through each excitation pole of the stator according to the rotation angle. a rotor, and a circuit that generates a plurality of phase-shifted alternating current signals of a predetermined frequency by frequency-dividing a predetermined clock pulse, and separately excites the excitation poles with the plurality of phase-shifted alternating current signals. a period calculation circuit that calculates digital data indicating the period of the output AC signal of the secondary winding by counting it using the clock pulse; A digital arithmetic circuit that digitally executes an arithmetic expression for calculating the rotational speed based on a prepared standard count value; A rotational accelerometer comprising an acceleration calculation circuit that digitally calculates rotational acceleration. 2. The acceleration calculation circuit calculates the speed based on the difference between the speed calculated one cycle ago by the digital calculation circuit and the speed calculated this time, and the count value for one cycle calculated by the period calculation circuit. The rotational accelerometer according to claim 1, which is a circuit for calculating rotational acceleration.