JP2022103046A - Signal generating device - Google Patents

Abstract

To provide a signal generating device that can simplify software and hardware required for generating a bi-phase pulse signal having a predetermined phase difference and can be realized by an inexpensive and general-purpose microcomputer.SOLUTION: A signal generating device includes: a first timer for outputting an interruption signal at a predetermined period; a second timer for outputting a first pulse signal and a second pulse signal having a predetermined phase difference; a third timer for counting the number of edges of the first pulse signal and the second pulse signal; and a processing part for performing interruption processing for controlling an output mode of the second timer each time when the interruption signal is generated. The processing part performs short-period processing without performing long-period processing when the number of interruption times is not equal to an initial value while performing predetermined short-period processing after performing predetermined long-period processing when the number of interruption times is equal to the initial value when starting interruption processing.SELECTED DRAWING: Figure 1

Description

The present invention relates to a signal generator.

A general increment type rotary encoder outputs a two-phase pulse signal (A-phase pulse signal and B-phase pulse signal) having a phase difference of 90 ° according to the rotation angle of the rotating body. For example, in Patent Document 1 below, a rotation detection device that calculates the rotation angle of a rotating body based on the output signal of a magnetic sensor and generates an A-phase pulse signal and a B-phase pulse signal based on the calculation result of the rotation angle. Is disclosed.

Japanese Patent No. 4823021

In the rotation detection device of Patent Document 1, the software and hardware required to generate the A-phase pulse signal and the B-phase pulse signal are complicated. As a result, the processing load of the processor that executes the software increases, and the cost of the hardware also increases. Therefore, in order to realize the rotation detection device of Patent Document 1, an expensive and high-spec dedicated chip is required.

In view of the above circumstances, the present invention provides a signal generation device capable of simplifying software and hardware required for generating a two-phase pulse signal having a predetermined phase difference and being feasible by an inexpensive general-purpose microcomputer. One purpose is to do.

One aspect of the signal generator of the present invention is a first timer that outputs an interrupt signal at a predetermined cycle, a second timer that outputs a first pulse signal and a second pulse signal having a predetermined phase difference, and the like. A third timer that counts the number of edges of the first pulse signal and the second pulse signal, and a processing unit that performs interrupt processing that controls the output mode of the second timer each time the interrupt signal is generated. Be prepared.
When the number of interrupts, which is the number of times the interrupt signal is generated, is equal to the initial value at the start of the interrupt processing, the processing unit performs a predetermined long-period processing and then a predetermined short-period processing, while the number of interrupts is If it is not equal to the initial value, the short cycle process is performed without performing the long cycle process.
As the long cycle processing, the processing unit has an angle updating process for updating the absolute angle information indicating the absolute angle of the rotating body, and an absolute angle function for expressing the absolute angle indicated by the updated absolute angle information as a linear function of time. And the function calculation process to calculate.
As the short cycle processing, the processing unit has an angle estimation value calculation processing for calculating an estimated value of the absolute angle as an angle estimation value based on the absolute angle function calculated by the function calculation processing, and the third timer. The count value acquisition process for acquiring the count value of the number of edges from the above, the angle current value calculation process for calculating the current value of the absolute angle as the current angle value based on the count value, and the first The signal state value calculation process for calculating the signal state value representing the relationship between the level of the 1-pulse signal and the level of the second pulse signal, and the second timer based on the angle estimation value and the angle current value. The first pulse signal and the first pulse signal based on the signal state value when the timing determination process for determining whether or not the timing is to change the output mode and the timing for changing the output mode are determined in the timing determination process. The first output mode switching process for changing the output mode of one of the two pulse signals to one of the high level output mode and the low level output mode, the interrupt count update process for updating the interrupt count, and the interrupt count update process. When the number of interrupts after the update is equal to a predetermined first threshold value, the number of interrupts reset process for resetting the number of interrupts to the initial value is executed.

According to the above aspect of the present invention, there is provided a signal generation device that can simplify the software and hardware required to generate a two-phase pulse signal having a predetermined phase difference and can be realized by an inexpensive general-purpose microcomputer. can.

FIG. 1 is a block diagram schematically showing a configuration of a signal generation device according to the first embodiment of the present invention. FIG. 2 is a diagram showing an example of the relationship between the waveform of the A-phase pulse signal, the waveform of the B-phase pulse signal, the waveform of the Z-phase pulse signal, and the signal state value. FIG. 3 shows the temporal correspondence between the generation timing of the interrupt signal, the change timing of the number of interrupts, the execution timing of the interrupt processing, the operation timing of the A / D converter, and the execution timing of the angle calculation processing. It is a timing chart. FIG. 4 is a flowchart showing an interrupt process executed by the processing unit in the first embodiment. FIG. 5 is an explanatory diagram regarding delay compensation of an absolute angle. FIG. 6 is a diagram showing an example in which an A-phase pulse signal and a B-phase pulse signal are output from the second timer by the interrupt processing of the first embodiment. FIG. 7 is a diagram showing the operation of the third timer. FIG. 8 is a diagram showing the correspondence between the absolute angle and the edge count value. FIG. 9 is a flowchart showing an interrupt process executed by the processing unit in the second embodiment. FIG. 10 is a diagram showing an example in which an A-phase pulse signal and a B-phase pulse signal are output from the second timer by the high-speed mode processing included in the interrupt processing of the second embodiment. FIG. 11 is a block diagram schematically showing the configuration of the signal generation device according to the third embodiment of the present invention. FIG. 12 is a diagram showing the relationship between the count value of the fourth timer and the Z-phase pulse signal output from the fourth timer. FIG. 13 is a diagram used to explain processing when the rotor reverses during counting by the fourth timer.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.
[First Embodiment]
FIG. 1 is a block diagram schematically showing the configuration of the signal generation device 10 according to the first embodiment of the present invention. As shown in FIG. 1, the signal generation device 10 in the first embodiment processes the A / D converter 11, the first timer 12, the second timer 13, the third timer 14, and the storage unit 15. A unit 16 is provided. The signal generation device 10 is electrically connected to the sensor unit 20. The sensor unit 20 has three magnetic sensors 21, 22 and 23. Each of the magnetic sensors 21, 22 and 23 is a hall sensor that detects the magnetic flux strength that changes according to the rotation angle of the rotating body and outputs an analog signal indicating the detection result of the magnetic flux strength as the magnetic flux detection signal.

In the present embodiment, the rotating body is, for example, a rotor of a three-phase brushless DC motor. A control board that supplies a drive current to a three-phase coil is mounted on a three-phase brushless DC motor. The magnetic sensors 21, 22 and 23 are arranged on the control board in a state of facing the rotor magnet in the axial direction of the rotor. Further, the magnetic sensors 21, 22 and 23 are arranged on the control board at regular intervals along the rotation direction of the rotor when viewed from the axial direction of the rotor. The constant interval is, for example, a 120 ° interval.

The magnetic sensor 21 outputs a magnetic flux detection signal Hu indicating a detection result of the magnetic flux intensity in the U phase. The magnetic sensor 22 outputs a magnetic flux detection signal Hv indicating a detection result of the magnetic flux intensity in the V phase. The magnetic sensor 23 outputs a magnetic flux detection signal Hw indicating a detection result of the magnetic flux intensity in the W phase. The three magnetic flux detection signals Hu, Hv and Hw have a phase difference of 120 ° in electrical angle from each other.

The signal generation device 10 is a microprocessor such as an MCU (Microcontroller Unit) arranged on a control board, for example. The signal generation device 10 calculates the absolute angle θ of the rotor based on the magnetic flux detection signals Hu, Hv, and Hw output from the sensor unit 20, and has a predetermined phase difference based on the calculation result of the absolute angle θ. A pulse signal and a second pulse signal are generated. In this embodiment, the predetermined phase difference is, for example, 90 °. Hereinafter, the first pulse signal is referred to as an A-phase pulse signal PA, and the second pulse signal is referred to as a B-phase pulse signal PB. Further, the signal generation device 10 generates an origin pulse signal indicating the origin of the absolute angle θ. Hereinafter, the origin pulse signal is referred to as a Z-phase pulse signal PZ.

The magnetic flux detection signals Hu, Hv and Hw output from the sensor unit 20 are input to the A / D converter 11 of the signal generation device 10. The A / D converter 11 converts each of the magnetic flux detection signals Hu, Hv, and Hw into digital data by sampling at a predetermined sampling frequency, and converts the digital data of the magnetic flux detection signals Hu, Hv, and Hw into the processing unit 16. Output.

The first timer 12 outputs an interrupt signal INT to the processing unit 16 at a predetermined cycle. Specifically, the first timer 12 increments the timer count value in synchronization with a clock signal (not shown), and when the timer count value reaches the timer reset value TRES1, outputs an interrupt signal INT and outputs the timer count value. Reset. In this way, the cycle in which the interrupt signal INT is output from the first timer 12 is determined by the timer reset value TRES1. The timer reset value TRES1 is set in the first timer 12 by the processing unit 16.

The second timer 13 outputs the A-phase pulse signal PA and the B-phase pulse signal PB having a phase difference of 90 °. The second timer 13 has a high-level output mode, a low-level output mode, and a comparative output mode as output modes of the A-phase pulse signal PA and the B-phase pulse signal PB. When the output mode of the A-phase pulse signal PA is the high-level output mode, the second timer 13 sets the level of the A-phase pulse signal PA to a high level. When the output mode of the A-phase pulse signal PA is the low level output mode, the second timer 13 sets the level of the A-phase pulse signal PA to the low level. The same applies to the B-phase pulse signal PB.

On the other hand, when the output mode is the comparative output mode, the second timer 13 increments the timer count value in synchronization with a clock signal (not shown), and each time the timer count value reaches the timer reset value TRES2, the timer count value is reached. To reset. Further, when the output mode is the comparative output mode, the second timer 13 inverts the level of the A-phase pulse signal PA every time the timer count value reaches the first level inversion threshold value Acom, and the timer count value becomes the second. Every time the level inversion threshold value Bcom is reached, the level of the B-phase pulse signal PB is inverted.

The output mode of the second timer 13 is switched by the output mode setting signal MSET output from the processing unit 16 to the second timer 13. The timer reset value TRES2, the first level inversion threshold value Acom, and the second level inversion threshold value Bcom are set in the second timer 13 by the processing unit 16. The A-phase pulse signal PA and the B-phase pulse signal PB output from the second timer 13 are input to the third timer 14.

The third timer 14 counts the number of edges of the A-phase pulse signal PA and the B-phase pulse signal PB output from the second timer 13. When the count value of the number of edges reaches the predetermined timer reset value TRES3, the third timer 14 resets the count value of the number of edges. Hereinafter, the count value of the number of edges is referred to as an edge count value EC. The third timer 14 compares the edge count value EC with a predetermined Z-phase output threshold value Zcom (third threshold value), and outputs a signal that becomes a high level when the edge count value EC is smaller than the Z-phase output threshold value Zcom. It is output as a Z-phase pulse signal PZ indicating the origin of the angle θ. The timer reset value TRES3 and the Z-phase output threshold value Zcom are set in the third timer 14 by the processing unit 16. The third timer 14 outputs the edge count value EC to the processing unit 16.

The storage unit 15 is used as a non-volatile memory for preliminarily storing programs and setting values necessary for the processing unit 16 to execute various processes, and as a temporary storage destination of data when the processing unit 16 executes various processes. Includes volatile memory. Examples of the non-volatile memory include EEPROM (Electrically Erasable Programmable Read-Only Memory) and flash memory. The volatile memory is, for example, a RAM (Random Access Memory) and a register. In the non-volatile memory, the timer reset value TRES1 of the first timer 12, the timer reset value TRES3 of the third timer 14, the Z-phase output threshold Zth, and the like are stored as set values. Although the details will be described later, the timer reset value TRES2 of the second timer 13, the first level inversion threshold value Acom, and the second level inversion threshold value Bcom are calculated by the processing unit 16.

The processing unit 16 is a processor core that executes various processes according to a program stored in the storage unit 15. The processing unit 16 executes an angle calculation process for calculating the absolute angle θ of the rotor based on the digital data of the magnetic flux detection signals Hu, Hv, and Hw input from the A / D converter 11. Further, the processing unit 16 performs interrupt processing for controlling the output mode of the second timer 13 each time an interrupt signal INT is generated. By controlling the output mode of the second timer 13 based on the absolute angle θ, the processing unit 16 outputs the A-phase pulse signal PA and the B-phase pulse signal PB having a phase difference of 90 ° from the second timer 13. To. When the A-phase pulse signal PA and the B-phase pulse signal PB are output from the second timer 13, the Z-phase pulse signal PZ is automatically output from the third timer 14.

FIG. 2 is a diagram showing an example of the relationship between the waveform of the A-phase pulse signal PA, the waveform of the B-phase pulse signal PB, the waveform of the Z-phase pulse signal PZ, and the signal state value State. In FIG. 2, assuming that the period of the A-phase pulse signal PA is “T”, the lengths of “a”, “b”, “c” and “d” are T / 4, respectively. As shown in FIG. 2, the B-phase pulse signal PB has a delay time of T / 4 with respect to the A-phase pulse signal PA. As described above, there is a phase difference of 90 ° between the A-phase pulse signal PA and the B-phase pulse signal PB. The rising edge of the Z-phase pulse signal PZ is generated in synchronization with the rising edge of the A-phase pulse signal PA. The pulse width of the Z-phase pulse signal PZ can be adjusted to an arbitrary value by the Z-phase output threshold value Zcom.

Although details will be described later, the signal state value State shown in FIG. 2 is a value calculated by the processing unit 16 based on the edge count value EC in the above interrupt processing. The signal state value State represents the relationship between the level of the A-phase pulse signal PA and the level of the B-phase pulse signal PB. For example, when the signal state value State is "0", the signal state value State represents a state in which both the A-phase pulse signal PA and the B-phase pulse signal PB are at a high level. When the signal state value State is "1", the signal state value State represents a state in which the A-phase pulse signal PA is at a high level and the B-phase pulse signal PB is at a low level. When the signal state value State is "2", the signal state value State represents a state in which both the A-phase pulse signal PA and the B-phase pulse signal PB are at a low level. When the signal state value State is "3", the signal state value State represents a state in which the A-phase pulse signal PA is at a low level and the B-phase pulse signal PB is at a high level.

Hereinafter, the operation of the signal generation device 10 configured as described above will be described with reference to FIGS. 3 to 8.

When the signal generation device 10 is switched from the power-off state to the power-on state, the processing unit 16 executes a predetermined initialization process. As one of the initialization processes, the processing unit 16 reads the timer reset value TRES1 of the first timer 12 from the storage unit 15, and sets the read timer reset value TRES1 in the first timer 12. Further, the processing unit 16 reads the timer reset value TRES3 and the Z-phase output threshold value Zcom of the third timer 14 from the storage unit 15 as one of the initialization processes, and reads the read timer reset value TRES3 and the Z-phase output threshold value Zcom. Set to the third timer 14. Further, the processing unit 16 sets the edge count value of the third timer 14 to the initial value corresponding to the absolute angle θ as one of the initialization processes, and the details of this setting process will be described later.

FIG. 3 shows the temporal correspondence between the generation timing of the interrupt signal INT, the change timing of the interrupt count, the execution timing of the interrupt processing, the operation timing of the A / D converter 11, and the execution timing of the angle calculation processing. It is a timing chart showing the relationship.
When the timer reset value TRES1 is set in the first timer 12, the first timer 12 increments the timer count value in synchronization with a clock signal (not shown), and when the timer count value reaches the timer reset value TRES1, an interrupt occurs. The signal INT is output and the timer count value is reset. As a result, as shown in FIG. 3, the interrupt signal INT is output from the first timer 12 in a predetermined period T _INT .

As shown in FIG. 3, the processing unit 16 performs interrupt processing every time an interrupt signal INT is generated, and the interrupt count count, which is the number of times the interrupt signal INT is generated at the start of the interrupt processing, is equal to the initial value "0". In this case, the predetermined short cycle processing is performed after the predetermined long cycle processing is performed, while the short cycle processing is performed without performing the long cycle processing when the interrupt count count is not equal to the initial value “0”.

As shown in FIG. 3, the interrupt count count is reset to the initial value “0” when the maximum value “N” is exceeded. In this way, the period T _period in which the interrupt count count is reset is referred to as a "control period". The control period T _period is expressed by the following equation (1). The long-period processing included in the interrupt processing is processing that is repeatedly executed in a period equal to the control period T _period . The short-cycle processing included in the interrupt processing is a processing that is repeatedly executed in a cycle equal to the generation cycle T _INT of the interrupt signal INT.
T _period = (N + 1) × T _INT … (1)

When the rotor rotates, the sensor unit 20 outputs magnetic flux detection signals Hu, Hv, and Hw having a phase difference of 120 ° in electrical angle from each other. As shown in FIG. 3, when the interrupt count count is the initial value “0”, the A / D converter 11 starts digital conversion of the magnetic flux detection signals Hu, Hv, and Hw, and the interrupt count count is, for example, “2”. At, the digital conversion ends. That is, digital data of the magnetic flux detection signals Hu, Hv, and Hw in one control cycle can be obtained during the period when the interrupt count count changes from the initial value “0” to “2”. As shown in FIG. 3, when the processing unit 16 acquires the digital data of the magnetic flux detection signals Hu, Hv and Hw, the processing unit 16 executes the angle calculation process during the period when the interrupt process is not executed.

The processing unit 16 calculates the absolute angle θ of the rotor based on the digital data of the magnetic flux detection signals Hu, Hv, and Hw in the angle calculation process. As the calculation algorithm of the absolute angle θ, for example, the algorithm described in Japanese Patent No. 6233532 can be used. Therefore, in the present specification, the description of the calculation algorithm of the absolute angle θ will be omitted. However, the algorithm for calculating the absolute angle θ is not limited to the algorithm described in Japanese Patent No. 6233532. Any algorithm that can calculate the absolute angle of the rotating body may be used.

As shown in FIG. 3, when the interrupt count count is the maximum value “N”, the angle calculation process executed within one control cycle ends. When the angle calculation process is completed, the processing unit 16 substitutes the calculation result of the absolute angle θ into the global variable gwTheta. The period in which the value of the global variable gwTheta indicating the absolute angle θ is rewritten to a new value is equal to the control period T _period . In other words, the control period T _period is the period during which the absolute angle θ of the rotor is updated.

FIG. 4 is a flowchart showing interrupt processing executed by the processing unit 16. As described above, the processing unit 16 executes the interrupt processing shown in FIG. 4 when the interrupt signal INT is input from the first timer 12. As shown in FIG. 4, when the interrupt processing is started, the processing unit 16 first determines whether or not the interrupt count count is equal to the initial value "0" (step S1). The processing unit 16 shifts to the processing of step S2 in the case of “Yes” in step S1, that is, when the interrupt count count is equal to the initial value “0”. On the other hand, when the processing unit 16 is "No" in step S1, that is, when the interrupt count count is not equal to the initial value "0", the processing unit 16 shifts to the processing of step S5.

In the interrupt processing shown in FIG. 4, the processing from step S2 to step S4 is a long-period processing. Further, in the interrupt processing shown in FIG. 4, the processing from step S5 to step S13 is a short cycle processing. That is, when the interrupt count count is equal to the initial value "0", the processing unit 16 executes the long-period processing including the processing from step S2 to step S4, and then the short period including the processing from step S5 to step S13. Execute periodic processing. On the other hand, when the interrupt count count is not equal to the initial value "0", the processing unit 16 executes the short-period processing without performing the long-period processing.

The processing unit 16 executes an angle acquisition process for acquiring the current value of the absolute angle θ of the rotor as one of the long-period processes (step S2). Specifically, in step S2, the processing unit 16 acquires the value of the global variable gwTheta as the current value Theta of the absolute angle θ. As shown in FIG. 3, the current value Theta of the absolute angle θ is the value of the absolute angle θ calculated in the control cycle immediately before the current control cycle.

Next, as one of the long-period processing, the processing unit 16 creates an absolute angle function that expresses the absolute angle θ as a linear function of time based on the current value Theta of the absolute angle θ and the previous value Theta_prev of the absolute angle θ. The function calculation process to be calculated is executed (steps S3 and S4). The previous value Theta_prev of the absolute angle θ is the value of the absolute angle θ calculated in the control cycle two before the current control cycle.

Specifically, as one of the function calculation processes, the processing unit 16 executes an intercept calculation process for calculating the intercept of the absolute angle function by applying delay compensation to the current value Theta of the absolute angle θ (step S3). ). The current value Theta of the absolute angle θ includes the following time delay components. As described above, the current value Theta of the absolute angle θ is the value of the absolute angle θ calculated in the control cycle immediately before the current control cycle. Therefore, the current value Theta of the absolute angle θ has a time delay corresponding to one control period. Further, the current value Theta of the absolute angle θ has a time delay due to the response delay of the magnetic sensors 21, 22 and 23. Further, when a low-pass filter is provided for the magnetic flux detection signals Hu, Hv and Hw, the current value Theta of the absolute angle θ has a time delay due to the frequency characteristic of the low-pass filter.

In step S3, the processing unit 16 applies delay compensation to the current value Theta of the absolute angle θ having the time delay component as described above. In FIG. 5, θ is the absolute angle θ calculated by the angle calculation process, θ _true is the true value of the absolute angle θ, and θ _new is the delay-compensated absolute angle θ. As shown in FIG. 5, when a time delay T _delay exists at the absolute angle θ, an angle error θ _delay occurs at the absolute angle θ with respect to the true value θ _new . In this case, the delay-compensated absolute angle θ _new is expressed by the following equation (2). In the following equation (2), θ (k) is equal to the current value Theta of the absolute angle θ, and θ (k-1) is equal to the previous value Theta_prev of the absolute angle θ. In the following equation (2), the second term on the right side is equal to the angle error θ _delay .

In step S3, the processing unit 16 calculates the delay-compensated absolute angle θ _new based on the above equation (2) as the intercept of the absolute angle function. The processing unit 16 acquires the calculation result of the delay-compensated absolute angle θ _new as the intercept Theta_new_low. As the time delay T _delay included in the above equation (2), the calculated value obtained by performing the simulation considering the above-mentioned factor of the time delay may be used, or it may be obtained by conducting an experiment. The measured value may be used.

Next, as one of the function calculation processes, the processing unit 16 executes a slope calculation process for calculating the slope of the absolute angle function by subtracting the previous value Theta_prev from the current value Theta of the absolute angle θ (step S4). ). Specifically, in step S4, the processing unit 16 calculates the slope Theta_extr of the absolute angle function based on the following equation (3).
Theta_extr = Theta --Theta_prev… (3)

When the processing unit 16 executes the function calculation process including the above steps S3 and S4, the absolute angle function θ (count) represented by the following equation (4) is finally obtained.
θ (count) = Theta_new_low + Theta_extr / (N + 1) × count… (4)

The above processes from step S2 to step S4 are long-period processes that are repeatedly executed in a cycle equal to the control cycle T _period . Subsequently, the short cycle processing will be described.

As one of the short-period processes, the processing unit 16 executes an angle estimation value calculation process for calculating an estimated value of the absolute angle θ as an angle estimated value based on the absolute angle function calculated by the function calculation process (step S5). ). Specifically, in step S5, the processing unit 16 acquires the value of θ (count) calculated by substituting the value of the interrupt count into the above equation (4) as the angle estimation value Theta_est.

Next, the processing unit 16 executes a count value acquisition process for acquiring the edge count value EC from the third timer 14 (step S6), and sets the current value of the absolute angle θ to the current angle based on the acquired edge count value EC (step S6). The angle current value calculation process to be calculated as a value is executed (step S7). Specifically, in step S7, the processing unit 16 calculates the current angle Theta_edge_now based on the following equation (5).
Theta_edge_now = edge_now × EDGE2THETA… (5)

In the above equation (5), edge_now is a variable to which the edge count value EC is assigned. EDGE2THETA is a coefficient for converting an edge count value EC into, for example, a 32-bit angle. Assuming that the resolution of the signal generator 10 is N_EDGE, EDGE2THETA is expressed by the following equation (6). The resolution is the number of edges generated in the A-phase pulse signal PA and the B-phase pulse signal PB during one rotation of the rotor.
EDGE2THETA = 2 ^ 32 / N_EDGE… (6)

Next, the processing unit 16 executes a signal state value calculation process for calculating the signal state value State based on the edge count value EC (step S8). As described above, the signal state value State is a value representing the relationship between the level of the A-phase pulse signal PA and the level of the B-phase pulse signal PB (see FIG. 2). Specifically, in step S8, the processing unit 16 calculates the signal state value State based on the following equation (7). The following equation (7) means that the logical product of the value of the variable edge_now and 3 (11 in binary notation) is calculated in order to acquire the lower 2 bits of the variable edge_now indicating the edge count value EC.
State = edge_now ・ (11)… (7)

Next, the processing unit 16 determines whether or not it is the timing to change the output mode of the second timer 13 based on the angle estimated value Theta_est calculated in step S5 and the current angle value Theta_edge_now calculated in step S7. The determination timing determination process is executed (step S9). Specifically, in step S9, whether the processing unit 16 has an absolute value of the difference value Theta_diff_low obtained by subtracting the current angle value Theta_edge_now from the angle estimated value Theta_est is larger than the predetermined threshold value Dth (second threshold value). Judge whether or not. The processing unit 16 calculates the difference value Theta_diff_low based on the following equation (8).
Theta_diff _low = Theta_est --Theta_edge_now… (8)

In the case of "Yes" in step S9, that is, when the absolute value of the difference value Theta_diff_low is larger than the predetermined threshold value Dth, the processing unit 16 determines that it is the timing to change the output mode of the second timer 13, and determines in step S10. Move to processing. On the other hand, when the processing unit 16 is "No" in step S9, that is, when the absolute value of the difference value Theta_diff_low is equal to or less than the predetermined threshold value Dth, the processing unit 16 shifts to the processing of step S11 without performing the processing of step S10.

When the processing unit 16 determines that the timing is to change the output mode in the timing determination process, the processing unit 16 sets the output mode of one of the pulse signals of the A-phase pulse signal PA and the B-phase pulse signal PB based on the signal state value State. , The first output mode switching process for changing to one of the high level output mode and the low level output mode is executed (step S10).

Specifically, in step S10, when the signal state value State is "0", the processing unit 16 sets both the output mode of the A-phase pulse signal PA and the output mode of the B-phase pulse signal PB to the high-level output mode. set. In step S10, when the signal state value State is "1", the processing unit 16 sets the output mode of the A-phase pulse signal PA to the high-level output mode and outputs the output mode of the B-phase pulse signal PB to the low level. Set to mode. In step S10, when the signal state value State is "2", the processing unit 16 sets both the output mode of the A-phase pulse signal PA and the output mode of the B-phase pulse signal PB to the low-level output mode. In step S10, when the signal state value State is "3", the processing unit 16 sets the output mode of the A-phase pulse signal PA to the low-level output mode and outputs the output mode of the B-phase pulse signal PB to the high-level output. Set to mode.

The processing unit 16 executes an interrupt count update process for updating the interrupt count count after performing the process in step S10 or when determining "No" in step S9 (step S11). Specifically, in step S11, the processing unit 16 increments the value of the interrupt count count.

Next, the processing unit 16 executes an interrupt count reset process for resetting the interrupt count to the initial value "0" when the updated interrupt count count is equal to a predetermined threshold value M (first threshold value) (step). S12 and S13). Specifically, the processing unit 16 determines whether or not the interrupt count count is equal to the threshold value M (step S12). The threshold value M is a value obtained by adding "1" to the maximum value "N" of the interrupt count. In the case of "Yes" in step S12, that is, when the interrupt count count is equal to the threshold value M (= N + 1), the processing unit 16 resets the interrupt count count to the initial value "0" and ends the interrupt processing (step). S13). On the other hand, when "No" in step S12, that is, when the interrupt count count is not equal to the threshold value M, the processing unit 16 ends the interrupt processing without performing the processing in step S13.

Each time the interrupt signal INT is generated, the processing unit 16 performs the interrupt processing as described above, so that the A-phase pulse signal PA and the B-phase pulse signal PB are output from the second timer 13. Hereinafter, an example in which the A-phase pulse signal and the B-phase pulse signal are output from the second timer 13 by the interrupt processing of the first embodiment will be described with reference to FIG.

In FIG. 6, the straight line L indicates an absolute angle function θ (count) calculated when the interrupt count count is “0”. The slope of the straight line L is the slope Theta_extr of the absolute angle function θ (count). The value of the point P0 on the straight line L is the angle estimation value Theta_est calculated when the interrupt count count is “0”. The value of point P0 is equal to the intercept Theta_new_low of the absolute angle function θ (count). The value of the point P2 on the straight line L is the angle estimation value Theta_est calculated when the interrupt count count is “2”. The value of the point P3 on the straight line L is the angle estimation value Theta_est calculated when the interrupt count count is “3”. The value of the point P5 on the straight line L is the angle estimation value Theta_est calculated when the interrupt count count is “5”. The value of the point P7 on the straight line L is the angle estimation value Theta_est calculated when the interrupt count count is “7”.

As shown in FIG. 6, the signal state value State is “0” before the interrupt count count changes to “2”, and both the output mode of the A-phase pulse signal PA and the output mode of the B-phase pulse signal PB. Is set to high level output mode. In this case, both the A-phase pulse signal PA and the B-phase pulse signal PB output from the second timer 13 become high levels before the interrupt count count changes to “2”.

As shown in FIG. 6, when the interrupt count count changes to "2", the signal state value State becomes "3", and the difference value between the angle estimated value Theta_est (value of point P2) and the current angle value Theta_edge_now is Theta_diff_low. It is assumed that the absolute value of is larger than the threshold value Dth. In this case, the processing unit 16 changes only the output mode of the A-phase pulse signal PA to the low-level output mode based on the signal state value State. As a result, as shown in FIG. 6, only the A-phase pulse signal PA output from the second timer 13 changes to the low level at the timing when the interrupt count count changes to “2”.

As shown in FIG. 6, when the interrupt count count changes to "3", the signal state value State becomes "2", and the difference value between the angle estimated value Theta_est (value of point P3) and the current angle value Theta_edge_now is Theta_diff_low. It is assumed that the absolute value of is larger than the threshold value Dth. In this case, the processing unit 16 changes only the output mode of the B-phase pulse signal PB to the low-level output mode based on the signal state value State. As a result, as shown in FIG. 6, only the B-phase pulse signal PB output from the second timer 13 changes to the low level at the timing when the interrupt count count changes to “3”.

As shown in FIG. 6, when the interrupt count count changes to "5", the signal state value State becomes "1", and the difference value between the angle estimation value Theta_est (value of point P5) and the angle current value Theta_edge_now is Theta_diff_low. It is assumed that the absolute value of is larger than the threshold value Dth. In this case, the processing unit 16 changes only the output mode of the A-phase pulse signal PA to the high-level output mode based on the signal state value State. As a result, as shown in FIG. 6, only the A-phase pulse signal PA output from the second timer 13 changes to a high level at the timing when the interrupt count count changes to “5”.

As shown in FIG. 6, when the interrupt count count changes to "7", the signal state value State becomes "0", and the difference value between the angle estimated value Theta_est (value of point P7) and the current angle value Theta_edge_now is Theta_diff_low. It is assumed that the absolute value of is larger than the threshold value Dth. In this case, the processing unit 16 changes only the output mode of the B-phase pulse signal PB to the high-level output mode based on the signal state value State. As a result, as shown in FIG. 6, only the B-phase pulse signal PB output from the second timer 13 changes to a high level at the timing when the interrupt count count changes to “7”.

When the A-phase pulse signal PA and the B-phase pulse signal PB are output from the second timer 13 by the interrupt processing as described above, as shown in FIG. 7, the third timer 14 has the A-phase pulse signal PA and the B-phase. The number of edges of the pulse signal PB is counted, and when the edge count value EC reaches the timer reset value TRES3, the edge count value EC is reset. Further, as shown in FIG. 7, the third timer 14 compares the edge count value EC and the Z-phase output threshold value Zcom, and outputs a signal that becomes a high level when the edge count value EC is smaller than the Z-phase output threshold value Zcom. It is output as a Z-phase pulse signal PZ. In FIG. 7, ENC _res indicates resolution. The timer reset value TRES3 is set to the value of "ENC _res -1".

When the signal generator 10 is switched to the power-on state, the edge count value of the third timer 14 is "0". Therefore, as shown in FIG. 8, the initial edge count value is set according to the absolute angle. You need to set the value. Therefore, in the present embodiment, when the signal generation device 10 is switched to the power-on state, the processing unit 16 counts the edges based on the following equations (9) and (10) as one of the initialization processes. Performs the process of setting the initial value of the value EC. In the following equations (9) and (10), TIM_CNT represents the initial value of the edge count value. Further, in the present specification, (uint32) (X) means that X is cast to unsigned 32 bits, and Y >> Z means that Y is shifted to the right by Z bits.

As described above, the signal generation device 10 of the first embodiment has a first timer 12 that outputs an interrupt signal INT in a predetermined cycle, and an A-phase pulse signal PA and a B-phase pulse signal PB having a predetermined phase difference. A second timer 13 that outputs the above, a third timer 14 that counts the number of edges of the A-phase pulse signal PA and the B-phase pulse signal PB, and an output mode of the second timer 13 each time an interrupt signal INT is generated. A processing unit 16 that performs interrupt processing to be controlled is provided.
When the number of interrupts is equal to the initial value at the start of interrupt processing, the processing unit 16 performs predetermined short-period processing after performing predetermined long-period processing, while long-period processing is performed when the number of interrupts is not equal to the initial value. Perform short-period processing without performing.
The processing unit 16 executes an angle acquisition process and a function calculation process as long-period processes. The processing unit 16 performs angle estimation value calculation processing, count value acquisition processing, angle current value calculation processing, signal state value calculation processing, timing determination processing, first output mode switching processing, interrupt count update processing, and as short-cycle processing. Executes interrupt count reset processing.
According to the signal generation device 10 of the first embodiment as described above, A from the absolute angle of the rotating body by software-like processing by the processing unit 16 which is the processor core without using hardware other than the three timers. The phase pulse signal PA and the B phase pulse signal PB can be generated. The timer is a circuit generally mounted on a general-purpose microcomputer such as an MCU. Therefore, according to the present embodiment, there is a signal generation device 10 that can simplify the software and hardware required to generate a two-phase pulse signal having a predetermined phase difference, and can be realized by an inexpensive general-purpose microcomputer. Can be provided.

In the first embodiment, in the function calculation process, the processing unit 16 performs the section calculation process of calculating the section of the absolute angle function by applying delay compensation to the current value of the absolute angle, and the previous value from the current value of the absolute angle. By subtracting, the slope calculation process for calculating the slope of the absolute angle function is executed.
In this way, the intercept of the absolute angle function required for calculating the estimated value of the absolute angle (angle estimated value) is calculated in consideration of the delay time existing in the current value of the absolute angle, so that the angle estimated value can be obtained. It can be obtained with high accuracy.

In the first embodiment, the processing unit 16 determines that the absolute value of the difference value obtained by subtracting the current angle value from the estimated angle value is larger than a predetermined second threshold value (Dth) in the timing determination process. It is determined that the timing is to change the output mode of the second timer 13.
As a result, chattering generated in the A-phase pulse signal PA and the B-phase pulse signal PB can be suppressed. Further, the balance between the degree of chattering suppression and the responsiveness can be adjusted by the setting value of the second threshold value (Dth).

In the first embodiment, the third timer 14 resets the edge count value when the edge count value reaches a predetermined timer reset value.
According to this, the resolution can be arbitrarily adjusted by setting the timer reset value of the third timer 14.

In the first embodiment, the third timer 14 compares the edge count value with a predetermined third threshold value (Zcom), and outputs a signal having a high level when the edge count value is smaller than the third threshold value at an absolute angle. It is output as a Z-phase pulse signal PZ indicating the origin.
According to this, it is possible to generate a Z-phase pulse signal PZ indicating the origin of an absolute angle based on the number of edges of the A-phase pulse signal PA and the B-phase pulse signal PB.

[Second Embodiment]
Next, a second embodiment of the present invention will be described.
The configuration of the signal generation device 10 in the second embodiment is the same as that in the first embodiment. However, the interrupt processing executed by the processing unit 16 in the second embodiment is different from the interrupt processing in the first embodiment. In the interrupt processing of the second embodiment, the operation mode is switched between the low speed mode and the high speed mode according to the rotation speed of the rotor, so that the A-phase pulse signal PA is maintained in an appropriate accuracy according to the rotation speed. And the B-phase pulse signal PB is generated.

FIG. 9 is a flowchart showing an interrupt process executed by the processing unit 16 in the second embodiment. When the interrupt signal INT is input from the first timer 12, the processing unit 16 executes the interrupt processing shown in FIG.

As shown in FIG. 9, when the interrupt processing is started, the processing unit 16 first determines whether or not the interrupt count count is equal to the initial value “0” (step S100). The processing unit 16 shifts to the processing of step S101 in the case of "Yes" in step S100, that is, when the interrupt count count is equal to the initial value "0". On the other hand, when the processing unit 16 is "No" in step S100, that is, when the interrupt count count is not equal to the initial value "0", the processing unit 16 shifts to the processing of step S108.

In the interrupt processing shown in FIG. 9, the processing from step S101 to step S107 and the processing from step S113 to step S117 are included in the long-period processing. Further, in the interrupt processing shown in FIG. 9, the processing from step S108 to step S112 is included in the short cycle processing. That is, when the interrupt count count is equal to the initial value "0", the processing unit 16 executes long-period processing including steps S101 to S107 and steps S113 to S117, and then steps S108 to S112. Perform short-period processing including. On the other hand, when the interrupt count count is not equal to the initial value "0", the processing unit 16 executes the short-period processing without performing the long-period processing.

The processing unit 16 executes an angle acquisition process for acquiring the current value Theta of the absolute angle θ of the rotor as one of the long-period processes (step S101). Since the process of step S101 is the same as the process of step S2 in the first embodiment, the description thereof will be omitted.

Next, the processing unit 16 executes an intercept calculation process for calculating the intercept Theta_new_low of the absolute angle function by applying delay compensation to the current value Theta of the absolute angle θ as one process of the long cycle process (step S102). ). Since the process of step S102 is the same as the process of step S3 in the first embodiment, the description thereof will be omitted.

Next, the processing unit 16 executes a speed calculation process for calculating the rotation speed of the rotor based on the current value Theta and the previous value Theta_prev of the absolute angle θ as one process of the long period process (step S103). Specifically, in step S103, the processing unit 16 calculates the rotation speed Speed of the rotor based on the following equation (11) and the following equation (12).

Next, the processing unit 16 executes a first operation mode determination process for determining whether or not the operation mode is a high-speed mode as one of the long-period processes (step S104). The processing unit 16 determines whether or not the rotation speed Speed is equal to or less than a predetermined lower limit threshold value when the operation mode is determined to be the high-speed mode in the first operation mode determination process, that is, when “Yes” in step S104. The lower limit threshold value determination process is executed (step S105).

For example, the reference value of the rotation speed for switching the operation mode is expressed by the following equation (13). In the following equation (13), ENC _res is the resolution, and f _{LOW_algo} is equal to the generation frequency of the interrupt signal INT (the reciprocal of the period T _INT ). For example, the lower limit threshold value in the following description is set by -20% of the reference value represented by the following equation (13), and the upper limit threshold value is set by + 20% of the reference value represented by the following equation (13). You may.

The processing unit 16 switches the operation mode from the high speed mode to the low speed mode when it is determined in the lower limit threshold value determination process that the rotation speed Speed is equal to or less than the lower limit threshold value, that is, when “Yes” in step S105. The process is executed (step S106). Then, the processing unit 16 executes the inclination calculation process, which is one of the function calculation processes, as the low-speed mode process A (first process in the low-speed mode) (step S107). Since the inclination calculation process is the same as the process of step S4 in the first embodiment, the description thereof will be omitted. The processing unit 16 acquires the absolute angle function θ (count) represented by the above equation (4) by performing the low-speed mode processing A.

The processing unit 16 operates in the low-speed mode after performing the above steps S101 to S107 as long-period processing, or when the interrupt count count is not equal to the initial value "0" at the start of interrupt processing. The second operation mode determination process for determining whether or not to perform is executed as a short-period process (step S108).

The processing unit 16 performs low-speed mode processing B (low-speed mode processing B) performed as the second processing of the low-speed mode when the operation mode is determined to be the low-speed mode in the second operation mode determination processing, that is, when “Yes” in step S108. Step S109), the interrupt count update process (step 110), and the interrupt count reset process (steps S111 and S112) are executed as short cycle processes.

The low-speed mode process B includes an angle estimation value calculation process (step S5), a count value acquisition process (step S6), an angle current value calculation process (step S7), and a signal state value calculation process (step S8) in the first embodiment. The timing determination process (step S9) and the first output mode switching process (step S10) are included. Therefore, the description of the low-speed mode process B executed as the process of step S109 will be omitted. Further, since the processing from step S110 to step S112 is the same as the processing from step S11 to step S13 in the first embodiment, the description thereof will be omitted.

Further, the processing unit 16 determines whether or not the rotation speed Speed is equal to or higher than a predetermined upper limit threshold value when the operation mode is determined not to be the high-speed mode in the first operation mode determination process, that is, when “No” in step S104. The upper limit threshold value determination process for determination is executed (step S113).

When the rotation speed Speed is determined not to be equal to or higher than the upper limit threshold value in the upper limit threshold value determination process, that is, when “No” in step S113, the processing unit 16 calculates the inclination, which is one of the function calculation processes, as the low speed mode process A. The process is executed (step S114). The processing unit 16 acquires the absolute angle function θ (count) represented by the above equation (4) by performing the low-speed mode processing A.

The processing unit 16 performs steps S101 to S104 and steps S113 to S114 as long-period processing, or when the interrupt count count is not equal to the initial value "0" at the start of interrupt processing. The second operation mode determination process for determining whether or not the operation mode is the low speed mode is executed as a short cycle process (step S108).

The processing unit 16 performs low-speed mode processing B (low-speed mode processing B) performed as the second processing of the low-speed mode when the operation mode is determined to be the low-speed mode in the second operation mode determination processing, that is, when “Yes” in step S108. Step S109), the interrupt count update process (step 110), and the interrupt count reset process (steps S111 and S112) are executed as short cycle processes.

Further, the processing unit 16 performs long-period processing for high-speed mode processing (high-speed mode processing) when it is determined in the lower limit threshold value determination processing that the rotation speed Speed is not equal to or less than the lower limit threshold value, that is, when “No” in step S105. (Step S115). The processing unit 16 changes the operation mode after performing steps S101 to S105 and S115 as long-period processing, or when the interrupt count count is not equal to the initial value "0" at the start of interrupt processing. The second operation mode determination process (step S108) for determining whether or not the mode is low speed, the interrupt count update process (step 110), and the interrupt count reset process (steps S111 and S112) are executed as short cycle processes. ..

Further, the processing unit 16 switches the operation mode from the low speed mode to the high speed mode when it is determined in the upper limit threshold value determination process that the rotation speed Speed is equal to or higher than the upper limit threshold value, that is, when “Yes” in step S113. The mode switching process (step S116) and the high-speed mode process (step S117) are executed as long-period processes.

The processing unit 16 performs long-period processing of steps S101 to S104, steps S113, S116, and S117, or at the start of interrupt processing, the interrupt count count is equal to the initial value "0". If not, the second operation mode determination process (step S108) for determining whether or not the operation mode is the low speed mode, the interrupt count update process (step 110), and the interrupt count reset process (steps S111 and S112) are performed. , Execute as a short cycle process.

The processing unit 16 executes parameter determination processing, parameter set processing, and second output mode switching processing as high-speed mode processing. The parameter determination process is a process of determining the timer reset value TRES2, the first level inversion threshold value Acom, and the second level inversion threshold value Bcom, which are the control parameters of the second timer 13. The parameter set process is a process of setting the timer reset value TRES2, the first level inversion threshold value Acom, and the second level inversion threshold value Bcom in the second timer 13. The second output mode switching process is a process of switching the output mode of the second timer 13 to the comparison output mode.

The timer reset value TRES2 of the second timer 13 is a value proportional to the reciprocal of the rotation speed. Therefore, the smaller the timer reset value TRES2, the higher the output frequencies of the A-phase pulse signal PA and the B-phase pulse signal PB, and the larger the timer reset value TRES2, the higher the output frequencies of the A-phase pulse signal PA and the B-phase pulse signal PB. Will be low. The processing unit 16 calculates the timer reset value TRES2 based on the following equation (14) in the parameter determination process. In the following equation (14), Tim Arr (k) is the timer reset value TRES2. Further, in the following equation (14), the KARR _DIV is a coefficient for preventing a digit loss at the time of calculation.

In the above equation (14), the occurrence of cumulative error is suppressed by performing rounding processing to carry over the remainder frac (k) generated at the time of division to the next control cycle. frac (k) is calculated based on the following equation (15). In the following equation (15), Quo (k) is the integer part (quotient) of the arithmetic expression in brackets on the right side of the above equation (14). Further, KARR is set by the following equation (16). In the following equation (16), fc is the control frequency (the reciprocal of the control cycle). Further, in the following equation (16), fsys is a clock frequency.

Here, it is preferable to set the value of KARR, which is a numerator, as large as possible in order to maintain the calculation accuracy in division. Therefore, it is preferable to set the value of KARR _DIV so that KARR is smaller than 2 ³² and as large as possible.

Further, the processing unit 16 calculates the first level inversion threshold value Acom and the second level inversion threshold value Bcom in the parameter determination process as follows. For example, in order to make the phase difference between the A-phase pulse signal PA and the B-phase pulse signal PB 90 °, the relationship of the following equation (17) may be established.

The processing unit 16 calculates the first level inversion threshold value Acom and the second level inversion threshold value Bcom so that the relationship of the above equation (17) is established. The magnitude relationship between the first level inversion threshold value Acom and the second level inversion threshold value Bcom is determined based on the relationship in Table 1 using the rotation direction (sign of rotation speed Speed) and the signal state value State.

The processing unit 16 sets the timer reset value TRES2, the first level inversion threshold value Acom, and the second level inversion threshold value Bcom determined as described above in the second timer 13, and sets the output mode of the second timer 13 to the comparison output mode. Switch to. As a result, as shown in FIG. 10, the second timer 13 increments the timer count value TC in synchronization with a clock signal (not shown), and each time the timer count value TC reaches the timer reset value TRES2, the timer count is counted. Reset the value TC. Further, the second timer 13 inverts the level of the A-phase pulse signal PA every time the timer count value TC reaches the first level inversion threshold value Acom, and the timer count value TC reaches the second level inversion threshold value Bcom. Each time, the level of the B-phase pulse signal PB is inverted.

As described above, according to the interrupt processing of the second embodiment, the operation mode is switched to the low speed mode when the rotation speed of the rotor is equal to or less than the lower limit, and the processing described in the first embodiment is executed as the processing of the low speed mode. As a result, the A-phase pulse signal PA and the B-phase pulse signal PB having a phase difference of 90 ° are generated while maintaining appropriate accuracy with respect to the rotation speed of the rotor. On the other hand, according to the interrupt processing of the second embodiment, when the rotation speed of the rotor is equal to or higher than the upper limit threshold value, the operation mode is switched to the high-speed mode, and the above-mentioned high-speed mode processing is executed, so that the rotation speed of the rotor is increased. A-phase pulse signal PA and B-phase pulse signal PB having a phase difference of 90 ° are generated while maintaining appropriate accuracy.
As described above, according to the interrupt processing of the second embodiment, by switching the operation mode between the low speed mode and the high speed mode according to the rotation speed of the rotor, an appropriate accuracy is maintained according to the rotation speed. Can generate an A-phase pulse signal PA and a B-phase pulse signal PB.

[Third Embodiment]
Next, a third embodiment of the present invention will be described.
FIG. 11 is a block diagram schematically showing the configuration of the signal generation device 10 according to the third embodiment of the present invention. As shown in FIG. 11, the signal generation device 10 in the third embodiment processes the A / D converter 11, the first timer 12, the second timer 13, the third timer 14, and the storage unit 15. In addition to the unit 16, a fourth timer 17 is provided.

As described above, the signal generation device 10 in the third embodiment is different from the first embodiment in that the fourth timer 17 is newly provided. Further, the third timer 14 in the third embodiment is different from the first embodiment in that it does not output the Z-phase pulse signal PZ. That is, the third timer 14 in the third embodiment counts the number of edges of the A-phase pulse signal PA and the B-phase pulse signal PB output from the second timer 13, and outputs the edge count value EC to the processing unit 16. It is consistent with the first embodiment in that it has a function and a function of resetting the edge count value EC when the edge count value EC reaches the timer reset value TRES3.

The fourth timer 17 counts the number of edges of the A-phase pulse signal PA and the B-phase pulse signal PB output from the second timer 13. The fourth timer 17 resets the count value CNT when the count value CNT of the number of edges reaches a predetermined timer reset value TRES4. The fourth timer 17 compares the count value CNT of the number of edges with a predetermined Z-phase output threshold value Zcom (fourth threshold value), and outputs a signal that becomes a high level when the count value CNT is smaller than the Z-phase output threshold value Zcom. It is output as a Z-phase pulse signal PZ indicating the origin of the absolute angle θ. The timer reset value TRES4 and the Z-phase output threshold value Zcom are set in the fourth timer 17 by the processing unit 16.

FIG. 12 is a diagram showing the relationship between the count value CNT of the fourth timer 17 and the Z-phase pulse signal PZ output from the fourth timer 17. In FIG. 12, the alternate long and short dash line indicates the edge count value EC of the third timer 14. FIG. 12 shows, as an example, a case where four pulse signals Z0, Z1, Z2 and Z3 are output as Z-phase pulse signals PZ during one rotation of the rotor. When "1535" is set as the timer reset value TRES4, the count value CNT of the fourth timer 17 increases from "0" to "1535", and when the count value CNT reaches "1535", the count value CNT increases. Is reset to "0". In this way, when the count value CNT of the fourth timer 17 is smaller than the Z phase output threshold Zcom in the period in which the count value CNT is rising from "0" to "1535", it is used as the Z phase pulse signal PZ. The pulse signal Z0 is output.

When "2559" is set as the timer reset value TRES4, the count value CNT of the fourth timer 17 rises from "0" to "2559", and when the count value CNT reaches "2559", the count value CNT Is reset to "0". In this way, when the count value CNT of the fourth timer 17 is smaller than the Z phase output threshold Zcom in the period in which the count value CNT is rising from "0" to "2559", it is used as the Z phase pulse signal PZ. The pulse signal Z1 is output.

When "1023" is set as the timer reset value TRES4, the count value CNT of the fourth timer 17 increases from "0" to "1023", and when the count value CNT reaches "1023", the count value CNT Is reset to "0". As described above, when the count value CNT of the fourth timer 17 is smaller than the Z phase output threshold Zcom in the period in which the count value CNT is rising from "0" to "1023", it is used as the Z phase pulse signal PZ. The pulse signal Z2 is output.

When "3071" is set as the timer reset value TRES4, the count value CNT of the fourth timer 17 increases from "0" to "3071", and when the count value CNT reaches "3071", the count value CNT increases. Is reset to "0". In this way, when the count value CNT of the fourth timer 17 is smaller than the Z-phase output threshold Zcom in the period in which the count value CNT is rising from "0" to "3071", it is used as the Z-phase pulse signal PZ. The pulse signal Z3 is output.

Hereinafter, the section from the rising edge of the pulse signal Z0 to the rising edge of the pulse signal Z1 is referred to as a “Z0 section”. Further, the section from the rising edge of the pulse signal Z1 to the rising edge of the pulse signal Z2 is referred to as a "Z1 section". Further, the section from the rising edge of the pulse signal Z2 to the rising edge of the pulse signal Z3 is referred to as a "Z2 section". Further, the section from the rising edge of the pulse signal Z3 to the rising edge of the pulse signal Z0 is referred to as a "Z3 section".

While counting in the Z0 section, the processing unit 16 sets the timer reset value TRES4 "2559" used in the next Z1 section in the shadow register, which is one of the internal registers of the fourth timer 17. When the Z0 section ends, that is, when the count value CNT of the 4th timer 17 reaches "1535" and is reset to "0", the auto reload register, which is one of the internal registers of the 4th timer 17, is set. By transferring the value of the shadow register (“2559”), the timer reset value TRES4 in the Z1 section is set.

While counting in the Z1 section, the processing unit 16 sets the timer reset value TRES4 "1023" used in the next Z2 section in the shadow register of the fourth timer 17. When the Z1 section ends, that is, when the count value CNT of the fourth timer 17 reaches "2559" and is reset to "0", the value of the shadow register ("1023") is transferred to the auto reload register. As a result, the timer reset value TRES4 in the Z2 section is set.

While counting in the Z2 section, the processing unit 16 sets the timer reset value TRES4 "3071" used in the next Z3 section in the shadow register of the fourth timer 17. When the Z2 section ends, that is, when the count value CNT of the fourth timer 17 reaches "1023" and is reset to "0", the value of the shadow register ("3071") is transferred to the auto reload register. As a result, the timer reset value TRES4 in the Z3 section is set.

While counting in the Z3 section, the processing unit 16 sets the timer reset value TRES4 "1535" used in the next Z0 section in the shadow register of the fourth timer 17. When the Z3 section ends, that is, when the count value CNT of the fourth timer 17 reaches "3071" and is reset to "0", the value of the shadow register ("1535") is transferred to the auto reload register. As a result, the timer reset value TRES4 in the Z0 section is set.

Similar to the third timer 14, when the signal generation device 10 is switched from the power off state to the power on state, the count value CNT of the fourth timer 17 is reset to “0”. Therefore, it is necessary to set the initial value of the count value CNT of the fourth timer 17 at the time of starting the signal generation device 10. As shown in FIG. 12, for example, when the value of TIM_CNT calculated at the time of starting the signal generation device 10 is "2048", the processing unit 16 performs the initial position at the time of starting based on the table data shown in Table 2. Is determined to be in the Z1 section. The table data shown in Table 2 is stored in the storage unit 15 in advance.

The processing unit 16 acquires the timer reset value TRES4 (“1535”) of the Z0 section, which is the Z section immediately before the Z1 section determined as described above, from the table data shown in Table 2, and then obtains the following (18). Based on the equation, the initial value of the count value CNT of the fourth timer 17 is set. When the value of TIM_CNT calculated at the time of starting the signal generation device 10 is "2048", the initial value of the count value CNT of the fourth timer 17 is set to "512" by the following equation (18).
Initial value of count value CNT = TIM_CNT- (TRES4 + 1 in Z0 section) ... (18)

Subsequently, with reference to FIG. 13, processing when the rotor reverses during counting by the fourth timer 17 will be described. As shown in FIG. 13, when the count value CNT of the 4th timer 17 underflows while "1023" is set in the auto reload register of the 4th timer 17 as the timer reset value TRES4, as shown by the dotted line 100. As the count value CNT changes, an error occurs in the Z section (in this case, the Z0 section).

Therefore, the processing unit 16 performs the following comparison processing in the periodic processing (that is, short-period processing) of the first timer 12. In the following comparison process, i represents the index of the current Z section, and TRES4 [i] represents the timer reset value of the fourth timer 17 used in the current Z section.
(1) When CNT <TRES4 [i] / 2, TRES4 [i-1] is set in the shadow register.
(2) When CNT> TRES4 [i] / 2, TRES4 [i + 1] is set in the shadow register.

For example, in FIG. 13, when the count value CNT in the Z1 section is smaller than TRES4 [1] / 2, the processing unit 16 sets TRES4 [0] in the shadow register. On the other hand, in FIG. 13, when the count value CNT in the Z1 section is larger than TRES4 [1] / 2, the processing unit 16 sets TRES4 [2] in the shadow register. As a result, even if the rotor reverses during counting in the Z1 section, it is possible to prevent an error from occurring in the Z section.

As described above, according to the third embodiment, a plurality of Z-phase pulse signals PZ that serve as a reference for the absolute angle θ can be output from the fourth timer 17 while the rotor makes one rotation. On the other hand, in the first embodiment, the timer reset value TRES3 of the third timer 14 is set to the value of "ENC _res -1", so that the third timer 14 to the Z phase are set during one rotation of the rotor. Only one pulse signal PZ can be output. Therefore, according to the third embodiment, it is possible to meet the customer's request to detect the absolute position required for the origin adjustment in less than one rotation.

[Modification example]
The present invention is not limited to the above-described embodiment, and the configurations described in the present specification can be appropriately combined within a range that does not contradict each other.

For example, in the second embodiment, when the timer reset value TRES2 of the second timer 13 is set by using the auto reload register, the timing for updating the timer reset value TRES2 is originally updated as shown in FIG. It will always be behind the timing that should be (the timing that is the same as the control cycle). Therefore, a cumulative error may occur in the method of simply calculating the timer reset value TRES2 (TimArr) from the rotation speed Speed. An algorithm for suppressing the occurrence of such a cumulative error may be added as one of the high-speed mode processing.

Further, for example, in the first to third embodiments, the case where the rotating body is a rotor of a three-phase brushless DC motor has been described, but the rotating body is not limited to the rotor.
Further, in the first to third embodiments, the case where the sensor unit 20 has three magnetic sensors 21, 22 and 23 is illustrated, but depending on the type of the rotating body or the content of the absolute angle calculation algorithm, the magnetism The type, number, or arrangement of sensors may be changed as appropriate.
Further, in the first to third embodiments, the predetermined phase difference is not limited to 90 ° and can be set to any value.

10 ... Signal generator, 11 ... A / D converter, 12 ... 1st timer, 13 ... 2nd timer, 14 ... 3rd timer, 15 ... Storage unit, 16 ... Processing unit, 17 ... 4th timer, 20 ... Sensor unit, 21, 22, 23 ... Magnetic sensor, INT ... Interruption signal, PA ... A phase pulse signal (first pulse signal), PB ... B phase pulse signal (second pulse signal), PZ ... Z phase pulse signal ( Origin pulse signal)

Claims (11)

The first timer that outputs an interrupt signal at a predetermined cycle,
A second timer that outputs a first pulse signal and a second pulse signal having a predetermined phase difference, and
A third timer that counts the number of edges of the first pulse signal and the second pulse signal, and
A processing unit that performs interrupt processing that controls the output mode of the second timer each time the interrupt signal is generated.
Equipped with
When the number of interrupts, which is the number of times the interrupt signal is generated, is equal to the initial value at the start of the interrupt processing, the processing unit performs a predetermined long-period processing and then a predetermined short-period processing, while the number of interrupts is If it is not equal to the initial value, the short-period processing is performed without performing the long-period processing, and the short-period processing is performed.
The processing unit performs the long-period processing.
Angle acquisition process to acquire the current value of the absolute angle of the rotating body,
A function calculation process that calculates an absolute angle function that expresses the absolute angle as a linear function of time based on the current value and the previous value of the absolute angle.
And run
The processing unit performs the short cycle processing.
Angle estimation value calculation processing that calculates the estimated value of the absolute angle as an angle estimation value based on the absolute angle function calculated by the function calculation processing, and
A count value acquisition process for acquiring the count value of the number of edges from the third timer, and
Angle current value calculation processing that calculates the current value of the absolute angle as the current angle value based on the count value, and
A signal state value calculation process for calculating a signal state value representing the relationship between the level of the first pulse signal and the level of the second pulse signal based on the count value, and
A timing determination process for determining whether or not it is a timing to change the output mode of the second timer based on the angle estimation value and the angle current value.
When it is determined in the timing determination process that the timing is to change the output mode, the output mode of one of the first pulse signal and the second pulse signal is set to a high level based on the signal state value. The first output mode switching process that changes to one of the output mode and the low level output mode,
The interrupt count update process for updating the interrupt count and
When the number of interrupts after the update is equal to the predetermined first threshold value, the interrupt count reset process for resetting the interrupt count to the initial value, and the interrupt count reset process.
To execute,
Signal generator. The processing unit is in the function calculation process.
The intercept calculation process for calculating the intercept of the absolute angle function by applying delay compensation to the current value of the absolute angle, and
The slope calculation process that calculates the slope of the absolute angle function by subtracting the previous value from the current value of the absolute angle,
To execute,
The signal generator according to claim 1. When the absolute value of the difference value obtained by subtracting the current angle value from the estimated angle value in the timing determination process is larger than the predetermined second threshold value, the processing unit said the second timer. Judging that it is the timing to change the output mode,
The signal generator according to claim 1 or 2. The third timer resets the count value when the count value of the number of edges reaches a predetermined timer reset value.
The signal generator according to any one of claims 1 to 3. The third timer compares the count value of the number of edges with a predetermined third threshold value, and sends a signal that becomes a high level when the count value is smaller than the third threshold value to the origin indicating the origin of the absolute angle. Output as a pulse signal,
The signal generator according to claim 4. A fourth timer for counting the number of edges of the first pulse signal and the second pulse signal output from the second timer is provided.
The fourth timer is
When the count value of the number of edges reaches a predetermined timer reset value, the count value is reset.
The count value of the number of edges is compared with a predetermined fourth threshold value, and a signal that becomes a high level when the count value is smaller than the fourth threshold value is output as an origin pulse signal indicating the origin of the absolute angle.
The signal generator according to any one of claims 1 to 3. When the number of interrupts is equal to the initial value at the start of the interrupt processing, the processing unit performs the long-period processing.
The angle acquisition process and
A speed calculation process for calculating the rotation speed of the rotating body based on the current value and the previous value of the absolute angle, and
The first operation mode determination process for determining whether or not the operation mode is a high-speed mode, and
When the operation mode is determined to be the high-speed mode in the first operation mode determination process, the lower limit threshold value determination process for determining whether or not the rotation speed is equal to or less than a predetermined lower limit threshold value,
The first mode switching process for switching the operation mode from the high speed mode to the low speed mode when the rotation speed is determined to be equal to or lower than the lower limit threshold value in the lower limit threshold value determination process.
The function calculation process performed as the first process of the low-speed mode and
And run
The processing unit performs the short-period processing after performing the long-period processing or when the number of interrupts is not equal to the initial value at the start of the interrupt processing.
A second operation mode determination process for determining whether or not the operation mode is the low speed mode, and
The second process of the low speed mode, which is performed when the operation mode is determined to be the low speed mode in the second operation mode determination process,
The interrupt count update process and the interrupt count reset process,
And run
The second process of the low speed mode includes the angle estimation value calculation process, the count value acquisition process, the angle current value calculation process, the signal state value calculation process, the timing determination process, and the first output mode switching process. including,
The signal generator according to any one of claims 1 to 6. When the number of interrupts is equal to the initial value at the start of the interrupt processing, the processing unit performs the long-period processing.
The angle acquisition process and
A speed calculation process for calculating the rotation speed of the rotating body based on the current value and the previous value of the absolute angle, and
The first operation mode determination process for determining whether or not the operation mode is a high-speed mode, and
When the operation mode is determined not to be the high-speed mode in the first operation mode determination process, the upper limit threshold value determination process for determining whether or not the rotation speed is equal to or higher than a predetermined upper limit threshold value.
When it is determined in the upper limit threshold value determination process that the rotation speed is not equal to or higher than the upper limit threshold value, the function calculation process performed as the first process of the low speed mode and the function calculation process.
And run
The processing unit performs the short-period processing after performing the long-period processing or when the number of interrupts is not equal to the initial value at the start of the interrupt processing.
A second operation mode determination process for determining whether or not the operation mode is the low speed mode, and
The second process of the low speed mode, which is performed when the operation mode is determined to be the low speed mode in the second operation mode determination process,
The interrupt count update process and the interrupt count reset process,
And run
The second process of the low speed mode includes the angle estimation value calculation process, the count value acquisition process, the angle current value calculation process, the signal state value calculation process, the timing determination process, and the first output mode switching process. including,
The signal generator according to any one of claims 1 to 6. When the number of interrupts is equal to the initial value at the start of the interrupt processing, the processing unit performs the long-period processing.
The angle acquisition process and
A speed calculation process for calculating the rotation speed of the rotating body based on the current value and the previous value of the absolute angle, and
The first operation mode determination process for determining whether or not the operation mode is a high-speed mode, and
When the operation mode is determined to be the high-speed mode in the first operation mode determination process, the lower limit threshold value determination process for determining whether or not the rotation speed is equal to or less than a predetermined lower limit threshold value,
The high-speed mode processing performed when the rotation speed is determined not to be equal to or lower than the lower limit threshold value in the lower limit threshold value determination processing.
And run
The processing unit performs the short-period processing after performing the long-period processing or when the number of interrupts is not equal to the initial value at the start of the interrupt processing.
The second operation mode determination process for determining whether or not the operation mode is the low speed mode, and
The interrupt count update process and the interrupt count reset process,
To execute,
The signal generator according to any one of claims 1 to 6. When the number of interrupts is equal to the initial value at the start of the interrupt processing, the processing unit performs the long-period processing.
The angle acquisition process and
A speed calculation process for calculating the rotation speed of the rotating body based on the current value and the previous value of the absolute angle, and
The first operation mode determination process for determining whether or not the operation mode is a high-speed mode, and
When the operation mode is determined not to be the high-speed mode in the first operation mode determination process, the upper limit threshold value determination process for determining whether or not the rotation speed is equal to or higher than a predetermined upper limit threshold value.
A second mode switching process for switching the operation mode from the low speed mode to the high speed mode when the rotation speed is determined to be equal to or higher than the upper limit threshold value in the upper limit threshold value determination process.
The high-speed mode processing and
And run
The processing unit performs the short-period processing after performing the long-period processing or when the number of interrupts is not equal to the initial value at the start of the interrupt processing.
A second operation mode determination process for determining whether or not the operation mode is the low speed mode, and
The interrupt count update process and the interrupt count reset process,
To execute,
The signal generator according to any one of claims 1 to 6. The processing unit is used as processing in the high-speed mode.
The parameter determination process for determining the timer reset value, the first level inversion threshold value, and the second level inversion threshold value, which are the control parameters of the second timer,
Parameter set processing for setting the timer reset value, the first level inversion threshold value, and the second level inversion threshold value in the second timer, and
The second output mode switching process for switching the output mode of the second timer to the comparison output mode, and
And run
The second timer is
Every time the automatically updated timer count value reaches the timer reset value, the timer count value is reset.
Every time the timer count value reaches the first level inversion threshold value, the level of the first pulse signal is inverted.
Every time the timer count value reaches the second level inversion threshold value, the level of the second pulse signal is inverted.
The signal generator according to claim 9 or 10.

Images