JP5063495B2 - Angle detection circuit - Google Patents

Description

  The present invention relates to an angle detection circuit using a resolver that outputs a two-phase differential voltage in which an AC excitation signal is amplitude-modulated according to a rotor rotation angle.

  In the servo control system, a rotation angle sensor is necessary to detect the rotation angle and perform feedback control. Further, in AC motor control, it is necessary to energize a motor coil in accordance with the rotation angle of the motor, and therefore a rotation angle sensor is required in addition to the servo control system. Conventionally, resolvers have been widely used as rotation angle sensors because of their robustness and environmental resistance resulting from their simple structure.

  When the resolver inputs a sinusoidal AC excitation signal E1sinωt to the primary winding, the AC excitation signal is amplitude-modulated in accordance with the rotor rotation angle θ in a two-phase secondary winding having a phase difference of 90 degrees. The differential voltages Vcos and Vsin are output. When the maximum voltage of the two-phase amplitude modulation signal is E2, and the phase difference between the AC excitation signal and the amplitude modulation signal is φ, Vcos and Vsin are E2sin (ωt + φ) · cosθ and E2sin (ωt + φ) · sinθ, respectively. . If the value at ωt + φ = 90 degrees is detected in order to improve the S / N ratio against noise, Vcos = E2cosθ and Vsinθ = E2sinθ. Therefore, the rotation angle θ can be obtained by the following equation.

  Since angle detection is performed using a resolver having the above-described characteristics, a configuration as shown in FIG. 2 is conventionally used, and Japanese Patent Laid-Open No. 2000-55695 (Patent Document 1) adopts this method. ing.

  In the conventional method, the CPU 23 operated by the clock source 22 outputs a phase reference signal for determining the phase of the AC excitation signal. The excitation signal generator 24 generates a sinusoidal AC excitation signal from the phase reference signal and supplies it to the resolver.

  The CPU 23 outputs the phase reference signal, determines the timing for AD conversion of the differential voltage of the amplitude modulation signal, and activates the AD conversion circuit 28. Here, since the CPU 23 determines the phase of the AC excitation signal by itself, if the phase difference between the AC excitation signal and the amplitude modulation signal in the system is obtained in advance, the AD conversion circuit 28 can be operated with an arbitrary phase of the amplitude modulation signal. Can be activated. For example, the AD conversion circuit 28 is activated at a timing near the peak of the amplitude modulation signal, that is, ωt + α = 90 degrees, and the angle is determined by comparing the obtained AD data with the data obtained by mapping Formula 1 or Formula 1. Can be sought.

  In addition, there is a method for detecting a failure in an angle detection system including an AD converter by obtaining a square sum of Vcos and Vsin which are AD data acquired at the peak of an amplitude modulation signal. When normal, the square sum is a constant value regardless of the angle as follows.

Sum of squares = Vcos ² + Vsin ² = (E 2 · cos θ) ² + (E 2 · sin θ) ² = E ^{2 2} · (cos θ ²
+ Sinθ ² ) = E2 ²
Therefore, when the sum of squares falls outside a certain range, it can be regarded as abnormal, and a failure of the angle detection system including the AD converter can be detected.

JP 2000-55695 A

  By the way, a system for driving an AC motor using the detected angle value obtained as described above is configured to control an electric power steering, an electric brake, an electronically controlled throttle, and a steering system and a brake system in an integrated manner. When applied to x-by-Wire, which controls the behavior of vehicle bodies, high safety and failure detection functions are required.

  However, the conventional method described above cannot detect an abnormality in the clock source 22 used by the CPU 23.

  In the conventional method, the excitation signal generator 24 for outputting an AC excitation signal has the configuration shown in FIG. That is, the phase reference signal is a simple rectangular wave pulse, and the excitation signal generator 24 converts the rectangular wave pulse into a pseudo sine wave by the low-pass filter 24a, and further amplifies it by the amplifier circuit 24b and outputs it. It is. However, when the clock abnormality occurs and the period of the rectangular wave pulse fluctuates, the phase shift amount and the attenuation rate by the low-pass filter 24 change. The signal conversion accuracy of the resolver itself also depends on the frequency. For this reason, the angle detection accuracy becomes a factor.

  In order to prevent this, there is a method of generating a pseudo sine wave by using a phase reference signal as a several-bit DA output. However, also in this case, since the DA output has discrete components as digital values, it is necessary to install a low-pass filter 24a. Therefore, when the clock frequency varies, the smoothing ability of the low-pass filter 24 varies. In addition, the cost increases due to the installation of the DA conversion circuit for realizing the DA output.

  Further, in the control of the AC motor, control parameters are set according to the inductance and resistance values of the motor and the control cycle of the energization control. Therefore, when the clock cycle is lowered and the control cycle is extended, the gain of the energization control is apparently too high, and the robustness of the control is impaired, for example, an overshoot of the energization current occurs. Conversely, when the clock period rises, the CPU deviates from the operation guarantee range, and heat generation increases.

  As described above, when the clock fluctuates, the AC motor control is greatly affected. However, the conventional method cannot detect a clock fluctuation as a failure.

  Therefore, in the present invention, when the clock that is a source for generating the phase reference signal of the resolver AC excitation signal and the clock that is the source of the control period of the motor energization control are changed, this is detected to increase the angle detection error. And it aims at preventing the robustness reduction of motor control.

  As for the failure detection of the AD conversion circuit, the conventional failure detection method using the sum of squares includes a mode in which failure detection is not possible, such as a fixing mode in which AD conversion data does not change regardless of the input voltage. For example, if Vcos and Vsin which are AD data acquired at the peak of the amplitude modulation signal are fixed at E2 · cos45 ° and E2 · sin45 °, respectively, the motor cannot be driven normally, but the square sum Can not detect a fault as normal. When the AD conversion circuit breaks down, motor energization control is performed based on incorrect AD conversion data and detected angle values, and the motor may be rotated in reverse from the intended direction or may not be able to rotate due to electromagnetic lock.

  Therefore, in the present invention, it is possible to reliably detect a failure of the AD conversion function for AD converting the differential voltage of the amplitude modulation signal used for angle calculation, and to prevent the energization control of the motor from being performed based on erroneous angle information. With the goal.

  In addition, when the above-mentioned system for driving an AC motor is applied to the above-mentioned x-by-Wire, etc., high safety and a failure detection function are required. However, when a failure is detected, the function should be stopped. Is not necessarily safe. For example, when the function of an electric power steering for a large vehicle is stopped, the steering operation must be performed only with the driver's own power, and the steering operation requires a very large force. Therefore, it is desirable to continue the function of the electric assist while notifying the driver that a failure has occurred.

  Therefore, an object of the present invention is to extend the life of an angle detection function while reliably detecting a failure when a failure occurs in an angle detection circuit using a resolver.

  In order to achieve the first object, the first CPU that outputs the phase reference signal of the AC excitation signal and the second CPU that determines the timing for AD conversion of the differential voltage of the amplitude modulation signal are separated, and the first CPU And the second CPU are operated using different clock sources. Then, the second CPU obtains the rotation angle from the data obtained by AD converting the differential voltage of the amplitude modulation signal, and performs energization control of the motor using this. At this time, the second CPU checks the period of the excitation signal controlled by the first CPU. As a result, the clocks of the first CPU and the second CPU are mutually monitored, and both an increase in angle detection error due to clock fluctuation of the first CPU and a decrease in robustness of energization control due to clock fluctuation of the second CPU are both prevented. it can.

  Also, the analog voltage signal controlled by the first CPU is input to an AD converter that detects the differential voltage of the amplitude modulation signal, and the second CPU is transmitted with the analog signal AD conversion data and communication between the CPUs from the first CPU. A failure of the AD converter is detected by comparing the control values of the first CPU. Thereby, it is possible to prevent the energization control of the motor from being performed from erroneous AD conversion data and angle detection value.

  In addition, ECUs that control electric power steering, electric brakes, electronically controlled throttles, as well as x-by-wire, which controls the behavior of the vehicle body by controlling the steering system and brake system in an integrated manner, are highly safe. In general, two CPUs are installed in order to ensure the reliability and the failure detection function. In this case, no significant increase in cost is caused.

  In normal operation, an AC excitation signal is generated based on the phase reference signal output by the first CPU. The second CPU determines the timing for AD conversion of the differential voltage of the amplitude modulation signal, activates the AD conversion circuit, obtains the rotation angle from the obtained AD conversion data, and controls the energization of the motor using this. At this time, the first CPU and the second CPU are operated using different clock sources. The second CPU checks the period of the phase transmission signal whose period is determined by the phase reference signal output from the first CPU.

  When the period of the phase transmission signal checked from the second CPU is abnormal, it is determined that the clock of the first CPU is abnormal, and switching is performed so as to generate an AC excitation signal based on the phase reference signal output by the second CPU.

  The failure detection function of the angle detection circuit is improved, and the safety of the control system using the angle detection circuit is improved.

  The best mode for carrying out the present invention will be described below based on the description of examples.

  An embodiment of the present invention will be described below with reference to FIG.

  The COS component signal winding 1b and the SIN component signal winding 1c of the resolver 1 are magnetically coupled to the excitation winding 1a via an iron core (not shown). Therefore, when an AC excitation signal is applied to the excitation winding 1a, an amplitude modulation signal b and an amplitude modulation signal c having waveforms similar to the AC excitation signal appear in the COS component signal winding 1b and the SIN component signal winding 1c, respectively. . At this time, when the rotor angle of the resolver changes, the degree of magnetic coupling between the excitation winding 1a and the COS component signal winding 1b and between the excitation winding 1a and the SIN component signal winding 1c changes. As a result, the amplitude ratio between the AC excitation signal and the amplitude modulation signal b, and the amplitude ratio between the AC excitation signal and the amplitude modulation signal c depend on the rotor angle of the resolver.

  Further, since the COS component signal winding 1b and the SIN component signal winding 1c are magnetically orthogonal to each other, the amplitude modulation signal b and the amplitude modulation signal c have a phase difference of 90 degrees. For example, when a sinusoidal AC excitation signal E1sinωt is input to the excitation winding 1a, the differential voltage Vcos of the amplitude modulation signal b and the differential voltage Vsin of the amplitude modulation signal c are as follows according to the rotor rotation angle θ. .

Vcos = E2sin (ωt + φ) ・ cosθ
Vsin = E2sin (ωt + φ) · sinθ
(Φ is the phase difference between the AC excitation signal and the amplitude modulation signal)
As a result, the rotor rotation angle θ can be calculated backward from Vcos and Vsin.

  In the present invention, the angle detection is performed using the resolver having the above-described characteristics. However, instead of the complete sine wave AC excitation signal, the first clock source 2, the first CPU 3 and the excitation signal shown in FIG. The generation circuit 4 generates a pseudo sine wave and supplies an AC excitation signal. The process of generating the AC excitation signal is shown in FIG. 4, elements indicated by the same reference numerals as those in FIG. 1 have the same configuration and operation as those in FIG.

  The first CPU 3 outputs from the IO port a 20 kHz rectangular wave in which 0V / 5V is switched every 25 μs. In the excitation signal generation circuit 4 to which the rectangular wave is inputted, first, it passes through the low-pass filter 4a and becomes a pseudo sine wave. At this time, the amplitude of the pseudo sine wave is greatly attenuated from the amplitude 0-5V of the rectangular wave. Then, the amplitude of the amplitude modulation signal after passing through the resolver also becomes small, and the SN ratio is deteriorated. Therefore, the amplification circuit 4b amplifies the amplitude of the pseudo sine wave in the normal phase and the reverse phase, and supplies the amplified pseudo sine wave as a differential voltage to the resolver to obtain an AC excitation signal.

  The low-pass filter 4a is preferably a secondary or higher-order low-pass filter in order to approximate a rectangular wave to a sine wave. However, the low-pass filter 4a can be changed depending on the angle detection accuracy required by the system, the noise environment, and the like. In addition, since it is only necessary to supply a waveform in which a rectangular wave is smoothed to a pseudo sine wave as an AC excitation signal, it is possible to arbitrarily set which of the low-pass filter and the amplifier circuit is the previous stage and the presence or absence of the amplifier circuit.

  By supplying the AC excitation signal generated as described above to the resolver, the amplitude modulation signal b and the amplitude modulation signal c are output from the COS component signal winding 1b and the SIN component signal winding 1c. Here, in order to detect the rotor rotation angle, it is necessary to detect the differential voltage of the amplitude modulation signal b and the amplitude modulation signal c. In this embodiment, as shown in FIG. 1, the amplitude modulation signal b and the amplitude modulation signal c are input to the first differential amplification circuit 9 and the second differential amplification circuit 10 respectively, and the outputs thereof are converted into an AD converter 8. To enter.

  The AD converter 8 shown in FIG. 1 starts AD conversion in response to an AD activation command from the second CPU 7. After the conversion operation of the AD converter 8 is completed, the second CPU 7 performs AD conversion data: Vcos, which is the differential voltage AD conversion result of the amplitude modulation signal b, and AD conversion data, which is the differential voltage AD conversion result of the amplitude modulation signal c. : Vsin can be read out. When the AD converter 8 has a differential voltage detection function, the first differential amplifier circuit 9 and the second differential amplifier circuit 10 may be omitted.

  Further, the excitation signal generation circuit 4 of the present embodiment described above has a phase transmission signal generation circuit 4c as shown in FIG. The phase transmission signal generation circuit 4c is a circuit that amplifies the differential voltage of the AC excitation signal by shifting the phase by 90 °. At this time, by setting the amplification gain sufficiently large, the output is clamped by the + power supply of the operational amplifier and the GND potential. As a result, the output of the phase transfer signal generation circuit 4c becomes a phase transfer signal in which a falling edge is generated at the timing when the differential voltage of the AC excitation signal is maximum and a rising edge is generated at the minimum timing.

  As shown in FIG. 1, the phase transfer signal is input to an external interrupt terminal of the second CPU 7. Since the phase reference signal output from the first CPU 3 is switched between High / Low every 25 μs, the rising edge and the falling edge of the phase transmission signal are respectively input about every 50 μs.

  However, the second CPU 7 may or may not accept an interrupt based on the phase transmission signal. An operation flowchart of the second CPU 7 is shown in FIG.

  Since the second CPU 7 controls the AC motor by changing the PWM duty every 250 μs, it has a time counter that is cleared every 250 μs. Here, assuming that the relative difference between the clocks of the first CPU 3 and the second CPU 7 is 1%, the falling edge / rising edge of the phase transfer signal while the time counter of the second CPU 7 changes to a value corresponding to 0 to 50.5 μs. Are entered at least once. Therefore, the second CPU 7 masks the interruption due to the phase transmission signal when the time counter is not a value corresponding to 0 to 50.5 μs (according to 101 and 121 in FIG. 5).

  As a result, if the time counter performs AC motor control between 50.5 μs and 250 μs, it is possible to avoid interruption during motor control. In this embodiment, the time counter calculates the duty of each phase for AC motor control (128 in FIG. 5) at a timing corresponding to 150 μs, and reflects it to the actual output duty (144 in FIG. 5) at a timing corresponding to 250 μs. )I do.

  When the second CPU 7 detects the rising edge or the falling edge of the phase transmission signal while the time counter corresponds to 0 to 50.5 μs, it starts the AD converter 8 (104 and 111 in FIG. 5).

  The second CPU 7 reads whether the phase transmission signal is at the high level or the low level after the AD converter 8 is activated. When the phase transmission signal is at a high level, it can be determined that a rising edge has been input, and conversely, when it has a low level, it can be determined that a falling edge has been input. Therefore, the timing at which the differential voltage of the AC excitation signal becomes maximum / minimum can be detected.

  When the phase transmission signal is at the low level (106 and 113 in FIG. 5), the second CPU 7 stores the time counter value at that time as τ1 (107a in FIG. 5). Further, AD conversion data: Vcos at that time is stored as Vcos1, and AD conversion data: Vsin is stored as Vsin1 (109a in FIG. 5).

  On the other hand, when the phase transmission signal is at a high level, the time counter value at that time is stored as τ2 (107b in FIG. 5). Further, AD conversion data: Vcos at that time is stored as Vcos2, and AD conversion data: Vsin is stored as Vsin2 (109b in FIG. 5).

  The second CPU 7 first confirms whether Vcos, Vsin and τ1, τ2 are all updated at the timing when the time counter reaches 150 μs (124 in FIG. 5). When the clock of the first CPU 3 or the second CPU 7 is abnormal, at least one of the rising edge and the falling edge is not input during 0 to 50.5 μs, and the above-described data may not be updated. Therefore, if it has not been updated, it is detected as abnormal (126 in FIG. 5). If all of them are updated, the angle is calculated by the following method from the time and AD conversion data acquired by the above method.

  In calculating the angle, first, center values of the amplitude modulation signal b and the amplitude modulation signal c: OFSETcos and OFSETsin are obtained by the following equations (128a in FIG. 5).

OFSETcos = Vcos1 + Vcos2
OFSETsin = Vsin1 + Vsin2
Next, it is confirmed whether the center value of the amplitude modulation signal is within the variation range of the theoretical value. When noise is mixed in the amplitude modulation signal, the AD converter 8 or the like, the center value obtained by the above formula may exceed the variation range of the theoretical value.

  If the center value is within the variation range of the theoretical location, the angle is calculated by the following formula (128c in FIG. 5).

  The angle obtained as described above is obtained from the average of the AD conversion data at time: τ1 and the AD conversion data at time: τ2, and in other words, the time obtained at I can say that.

  Therefore, in this embodiment, the τ is calculated and stored as the detection time.

  On the other hand, when the center value exceeds the variation range of the theoretical value, the angle is calculated by the following formula (128d in FIG. 5).

  In this embodiment, OFSETcos0 and OFSETsin0 are theoretical values of center values, but can be average values of center values calculated so far.

  At this time, since the obtained angle is obtained from the AD conversion data at time: τ1, the detection time: τ is as follows.

τ = τ1
In the present embodiment, as described above, the duty of each phase for AC motor control is actually reflected at the timing when the time counter is equivalent to 250 μs (143 in FIG. 5). Therefore, if the motor is rotating, the angle at which the duty is actually reflected deviates from the angle detected by AD conversion. Therefore, the detection angle is corrected using the detection time stored by the method described above and the time when the duty is actually reflected.

  The time when the output duty is actually reflected is τc (= 250 μs). Further, the detection angle one cycle before the time counter is cleared is θa, and the detection time is τa. Since the current detection angle is θ and the current detection time is τ, the angle θ ′ when the output duty is actually reflected can be obtained by the following equation (128e in FIG. 5).

  If the AC motor is controlled with the output duty calculated from the more accurate angle information θ ′ obtained in this way, more stable motor control can be realized.

  By the way, when the clock of the first CPU 3 is abnormal, the phase of the phase reference signal varies. The excitation signal generation circuit 4 is set to output an appropriate AC excitation signal amplitude and phase transmission signal timing with respect to a 20 kHz rectangular wave pulse. Therefore, when the clock of the first CPU 3 is abnormal, angle detection is performed. Impedes accuracy.

  Further, when the clock of the second CPU 7 is abnormal, the control cycle of the AC motor varies. Since parameters such as gain in AC motor control are set according to the inductance and resistance value of the motor and the control cycle, when the clock cycle is lowered and the control cycle is extended, apparently the gain of the energization control is increased. This is too high, and the robustness of the control is impaired, for example, an overshoot of the conduction current occurs. On the contrary, when the clock cycle rises, the operation deviates from the operation guarantee range of the second CPU 7 and heat generation increases. Therefore, when the clock fluctuation of the first CPU 3 and the second CPU 7 occurs, it is necessary to detect this abnormality. In the present invention, since τ1 and τ2 are stored as described above, the time difference is obtained by the following equation (125 in FIG. 5).

Δτ = | τ1-τ2 |
Originally, Δτ is equivalent to 25 μs. Therefore, when Δτ exceeds the variation range of the theoretical value, it can be detected that at least one of the clocks of the first CPU 3 or the second CPU 7 is abnormal (129 in FIG. 5). ).

  If the AD converter 8 is abnormal, the motor may be controlled based on incorrect angle information. In this case, the motor rotates in the opposite direction to the direction of rotation, or cannot be turned even if an external force is applied in an electromagnetic lock state. Therefore, when the AD converter 8 is abnormal, it is necessary to detect the abnormality.

  In the present embodiment, an analog voltage signal controlled by the first CPU 3 is input to the AD converter 8 that is detecting the amplitude modulation signal and confirmed. The first CPU 3 outputs a 2-bit 0V / 5V signal to the DA converter circuit 5. The DA conversion circuit 5 can output four types of voltages of 0V-1V-3V-4V corresponding to the 2-bit signal from the first CPU 3 to the AD converter 8.

  The first CPU 3 uses the 2-bit digital signal to change the output of the DA converter 5 to 0V-1V-3V-4V every 250 μs. The second CPU 7 can know the output state of the 2-bit digital signal of the first CPU 3 by the communication function (122c in FIG. 5). Further, the second CPU 7 activates analog signal AD conversion while the time counter is 50.5 to 150 μs so as not to interfere with AD conversion of the amplitude modulation signal and motor control (122a in FIG. 5), and converts the conversion result. The output state of the 2-bit digital signal of the first CPU 3 is compared (122d in FIG. 5). Abnormality can be detected when the AD conversion result and the output state of the 2-bit digital signal cannot be matched (122f in FIG. 5).

  According to the present embodiment, since the first clock source that is the source of the resolver excitation cycle is different from the second clock source that measures the resolver excitation cycle, either the first clock source or the second clock source is abnormal. In such a case, a failure can be detected. Thereby, it is possible to prevent the motor control from being performed using the angle detection value whose accuracy has deteriorated due to the fluctuation of the excitation frequency. Further, it is possible to prevent the motor control from being continued in a state where the robustness is lowered due to the fluctuation of the motor control frequency.

  Further, it is possible to detect a failure of a converter that performs AD conversion on data for calculating an angle. As a result, it is possible to prevent reverse rotation of the motor and electromagnetic locking caused by driving the motor with incorrect angle information.

  In addition, by having means for storing the time when AD conversion is performed on the data for calculating the angle, the angle at the timing at which the PWM duty is actually reflected can be estimated, and more precise motor control can be performed. it can.

  In addition, by performing AD conversion twice per excitation period of the resolver, it is possible to eliminate the influence of errors due to fluctuations in the center value. Furthermore, if noise is mixed, the motor control can be continued even if noise is mixed by using only one data of the two samplings.

  In the configuration of FIG. 1, it is conceivable to stop the control of the AC motor itself when a clock failure of the first CPU 3 is detected. However, the angle detection circuit according to the present embodiment can extend the life of the angle detection function by switching the operation mode while performing reliable failure detection when the above-described failure occurs. Hereinafter, the present embodiment will be described based on the drawings.

  In FIG. 6, elements having the same reference numerals as those in FIG. 1 have the same configuration and operation as those in FIG. The difference from the first embodiment is that the angle detection is performed using the resolver having the characteristics described with reference to FIG. 1, but the first CPU 200 shown in FIG. 6 outputs the first CPU 200 instead of the complete sinusoidal AC excitation signal. A pseudo sine wave is generated by the excitation signal generation circuit 4 from the one-phase reference signal or the second phase reference signal output from the second CPU 202, and an AC excitation signal is supplied. The process of generating the AC excitation signal is shown in FIG.

  In normal time, the first CPU 200 outputs a 20 kHz rectangular wave in which 0V / 5V is switched every 25 μs as a first phase reference signal from the IO port. At this time, the second phase reference signal terminal of the second CPU 202 is set as an input port and is set to high impedance. In the excitation signal generation circuit 4 to which the first phase reference signal is input, first, the rectangular wave is transmitted through the low-pass filter 4a to be a pseudo sine wave. At this time, the amplitude of the pseudo sine wave is greatly attenuated from the amplitude 0-5V of the rectangular wave. Then, the amplitude of the amplitude modulation signal after passing through the resolver also becomes small, and the SN ratio is deteriorated. Therefore, the amplification circuit 4b amplifies the amplitude of the pseudo sine wave in the normal phase and the reverse phase, and supplies the amplified pseudo sine wave as a differential voltage to the resolver to obtain an AC excitation signal.

  The low-pass filter 4a is preferably a secondary or higher-order low-pass filter in order to approximate a rectangular wave to a sine wave. However, the low-pass filter 4a can be changed depending on the angle detection accuracy required by the system, the noise environment, and the like. In addition, since it is only necessary to supply a waveform in which a rectangular wave is smoothed to a pseudo sine wave as an AC excitation signal, it is possible to arbitrarily set which of the low-pass filter and the amplifier circuit is the previous stage and the presence or absence of the amplifier circuit.

  By supplying the AC excitation signal generated as described above to the resolver, the amplitude modulation signal b and the amplitude modulation signal c are output from the COS component signal winding 1b and the SIN component signal winding 1c. Here, in order to detect the rotor rotation angle, it is necessary to detect the differential voltage of the amplitude modulation signal b and the amplitude modulation signal c. In this embodiment, the amplitude modulation signal b and the amplitude modulation signal c are input to the first differential amplifier circuit 9 and the second differential amplifier circuit 10, respectively, and the outputs thereof are input to the AD converter 8.

  The AD converter 8 starts AD conversion in response to an AD activation command from the second CPU 202. After the conversion operation of the AD converter 8 is completed, the second CPU 202 performs AD conversion data: Vcos, which is the differential voltage AD conversion result of the amplitude modulation signal b, and AD conversion data, which is the differential voltage AD conversion result of the amplitude modulation signal c. : Vsin can be read out. When the AD converter 8 has a differential voltage detection function, the first differential amplifier circuit 9 and the second differential amplifier circuit 10 may be omitted.

  Further, the excitation signal generation circuit 4 of the present embodiment described above has a phase transmission signal generation unit 4c as shown in FIG. The phase transmission signal generator 4c is a circuit that amplifies the differential voltage of the AC excitation signal by shifting the phase by 90 °. At this time, by setting the amplification gain sufficiently large, the output is clamped by the + power supply of the operational amplifier and the GND potential. As a result, the output of the phase transfer signal generator 4c is a phase transfer signal in which a falling edge is generated at the timing when the differential voltage of the AC excitation signal is maximum and a rising edge is generated at the minimum timing.

  The phase transfer signal is input to the external interrupt terminal of the second CPU 202. Since the phase reference signal output from the first CPU 200 is switched between High and Low every 25 μs, the rising edge and the falling edge of the phase transmission signal are input approximately every 50 μs.

  However, the second CPU 202 may or may not accept an interrupt based on the phase transmission signal. An operation flowchart of the second CPU 202 is shown in FIG.

  Since the second CPU 202 controls the AC motor by changing the PWM duty every 250 μs, it has a time counter that is cleared every 250 μs. Here, assuming that the relative difference between the clocks of the first CPU 200 and the second CPU 202 is 1%, the falling edge / rising edge of the phase transfer signal while the time counter of the second CPU 202 changes to a value corresponding to 0 to 50.5 μs. Are entered at least once. Therefore, the second CPU 202 masks the interrupt due to the phase transmission signal when the time counter is not a value corresponding to 0 to 50.5 μs (according to 101 and 121 in FIG. 8).

  As a result, if the time counter performs AC motor control between 50.5 μs and 250 μs, it is possible to avoid interruption during motor control. In this embodiment, the time counter calculates the duty of each phase for AC motor control (128 in FIG. 8) at a timing corresponding to 150 μs, and reflects it to the actual output duty (144 in FIG. 8) at a timing corresponding to 250 μs. )I do.

  The second CPU 202 activates the AD converter 8 when detecting the rising edge or the falling edge of the phase transmission signal while the time counter corresponds to 0 to 50.5 μs (104 and 111 in FIG. 8).

  The second CPU 202 reads whether the phase transmission signal is high level or low level after the AD converter 8 is activated. When the phase transmission signal is at a high level, it can be determined that a rising edge has been input, and conversely, when it has a low level, it can be determined that a falling edge has been input. Therefore, the timing at which the differential voltage of the AC excitation signal becomes maximum / minimum can be detected.

  When the phase transmission signal is at the low level (106 and 113 in FIG. 8), the second CPU 202 stores the time counter value at that time as τ1 (107a in FIG. 8). Further, AD conversion data: Vcos at that time is stored as Vcos1, and AD conversion data: Vsin is stored as Vsin1 (109a in FIG. 8).

  On the other hand, when the phase transmission signal is at the high level, the time counter value at that time is stored as τ2 (107b in FIG. 8). Further, AD conversion data: Vcos at that time is stored as Vcos2, and AD conversion data: Vsin is stored as Vsin2 (109b in FIG. 8).

  The second CPU 202 first checks whether all of Vcos1, Vsin1, Vcos2, Vsin2, and τ1, τ2 have been updated at the timing when the time counter reaches 150 μs (124 in FIG. 8). When the clock of the first CPU 200 or the second CPU 202 is abnormal, at least one of the rising edge or the falling edge is not input during 0 to 50.5 μs, and the above-described data may not be updated. Therefore, if it has not been updated, it is determined that the clock of the first CPU 200 is abnormal (126 in FIG. 8).

  If all of them are updated, the angle is calculated by the following method from the time and AD conversion data acquired by the above method.

  In calculating the angle, first, center values of the amplitude modulation signal b and the amplitude modulation signal c: OFSETcos and OFSETsin are obtained by the following equations (128a in FIG. 8).

OFSETcos = Vcos1 + Vcos2
OFSETsin = Vsin1 + Vsin2
Next, it is confirmed whether the center value of the amplitude modulation signal is within the variation range of the theoretical value. When noise is mixed in the amplitude modulation signal, the AD converter 8 or the like, the center value obtained by the above formula may exceed the variation range of the theoretical value.

  When the center value is within the variation range of the theoretical value, the angle is calculated by the following formula (128c in FIG. 8).

  The angle obtained as described above is obtained from the average of the AD conversion data at time: τ1 and the AD conversion data at time: τ2, and in other words, the time obtained at I can say that.

  Therefore, in this embodiment, the τ is calculated and stored as the detection time.

  On the other hand, when the center value exceeds the variation range of the theoretical value, the angle is calculated by the following formula (128d in FIG. 8).

  In this embodiment, OFSETcos0 and OFSETsin0 are theoretical values of center values, but can be average values of center values calculated so far.

  At this time, since the obtained angle is obtained from the AD conversion data at time: τ1, the detection time: τ is as follows.

τ = τ1
In the present embodiment, as described above, the duty of each phase for AC motor control is actually reflected at the timing when the time counter is equivalent to 250 μs (143 in FIG. 8). Therefore, if the motor is rotating, the angle at which the duty is actually reflected deviates from the angle detected by AD conversion. Therefore, the detection angle is corrected using the detection time stored by the method described above and the time when the duty is actually reflected.

  The time when the output duty is actually reflected is τc (= 250 μs). Further, the detection angle one cycle before the time counter is cleared is θa, and the detection time is τa. Since the current detection angle is θ and the current detection time is τ, the angle θ ′ when the output duty is actually reflected can be obtained by the following equation (128e in FIG. 8).

  If the AC motor is controlled with the output duty calculated from the more accurate angle information θ ′ obtained in this way, more stable motor control can be realized.

  By the way, if the clock of the first CPU 200 is abnormal, the cycle of the first phase reference signal varies and the cycle of the phase transmission signal also varies.

  In the present invention, since τ1 and τ2 are stored as described above, the clock cycle of the first CPU can be checked by obtaining the time difference by the following equation (125 in FIG. 8).

Δτ = | τ1-τ2 |
Since Δτ is essentially equivalent to 25 μs, when Δτ exceeds the variation range of the theoretical value, it is determined that the clock of the first CPU 200 is abnormal (129 in FIG. 8).

  Therefore, when any one of Vcos1, Vsin1, Vcos2, Vsin2, τ1, and τ2 has not been updated (126 in FIG. 8) and Δτ is outside the theoretical value variation range (129 in FIG. 8), the first CPU 200 Is determined to be abnormal. The excitation signal generation circuit 4 is set to output an appropriate AC excitation signal amplitude and phase transmission signal timing for a 20 kHz rectangular wave pulse. Impedes accuracy.

  Therefore, when it is detected that the clock of the first CPU 200 is abnormal, the system is switched to a method of performing angle detection without using the first CPU. Hereinafter, the switching method will be described in order.

  First, the second CPU 202 outputs a reset command signal to the first CPU 200. When the first CPU 200 enters the reset mode, the output of the IO port becomes high impedance. Therefore, the first phase reference signal terminal also becomes high impedance.

  Next, the second CPU 202 outputs a 20 kHz rectangular wave in which 0 V / 5 V is switched every 25 μs from the second phase reference signal terminal.

  Then, compared with the normal time, only the relationship between the first phase reference signal and the second phase reference signal is switched, and the AC excitation signal and the phase transmission signal after passing through the excitation signal generation circuit 4 are completely different from the normal time. Absent. Therefore, the angle detection can be continued in the same procedure as in the normal time.

  Next, the motor control device 204 and the motor 206 of the control device for electric power steering to which the embodiment of Example 1 or Example 2 is applied will be described with reference to FIG.

  The motor control device 204 incorporates the circuit components shown in FIG. 1 or FIG. 6 (except for the resolver 1). The motor 206 includes the resolver 1 and transmits the amplitude modulation signal b and the amplitude modulation signal c to the motor control device 204.

  The metal casing 240 of the motor control device 204 is mechanically connected to the motor 206 by screws 208. Next, the bus bar 231 that is electrically connected to the circuit components incorporated in the motor control device 204 is electrically connected to the three-phase input portion of the motor 100 by screws. The connecting portion of the bus bar 231B is closed with a metal shield cover 250M. Finally, the shield cover 250M is fixed to the metal housing 240 with at least one screw. As described above, the motor control device 204 and the motor 100 are assembled.

  Next, a system configuration using the electric power steering control device according to the present embodiment will be described with reference to FIG. FIG. 10 is a system configuration diagram showing the configuration of the electric power steering using the electric power steering control device according to the embodiment of the present invention.

  When the steering ST is rotated, the rotational driving force is decelerated by the manual steering gear STG via the rod RO, is transmitted to the left and right tie rods TR1, TR2, and is transmitted to the left and right wheels WH1, WH2. Steer the wheels WH1, WH2.

  The motor 206 according to the present embodiment is attached in the vicinity of the rod RO, and transmits the driving force to the manual steering gear STG via the gear GE. A torque sensor TS is attached to the rod RO, and a rotational driving force (torque) applied to the steering ST is detected.

  The motor control device 206 calculates a target torque of the motor 206 based on the output of the torque sensor TS and the output of a vehicle speed sensor (not shown) of the vehicle. In this calculation, the target torque of the motor 206 may be calculated in consideration of the rotation speed and the rotation acceleration of the motor 100, which enables more optimal control or excellent control of feeling. Further, the temperature and current value of the motor 206 are detected from the safety aspect, and the energization current to the motor 100 is controlled so that the current value that is the output torque of the motor 100 becomes a current value corresponding to the target torque. The motor controller 206 and the motor 206 are supplied with power from the battery BA.

  In the above configuration, the torque sensor and the motor 206 for assisting torque are placed in the portion of the steering column immediately below the steering wheel, but the rack type power steering equipped with the motor 206 in the vicinity of the rack and pinion gear. However, the motor control device 206 including the motor 206 of the present embodiment and the above-described inverter can be used as it is.

1 is a block diagram of a circuit configuration according to Embodiment 1. FIG. It is an angle detection circuit configuration block diagram using a conventional resolver. It is a block diagram of a conventional excitation signal generator. 3 is a block diagram of an excitation signal generator according to Embodiment 1. FIG. 3 is a flowchart of angle detection according to the first embodiment. FIG. 6 is a block diagram of a circuit configuration according to a second embodiment. 3 is a block diagram of an excitation signal generator according to Embodiment 1. FIG. 10 is a flowchart of angle detection according to the second embodiment. It is a perspective view which shows the motor control apparatus 204 and the motor 206 of the control apparatus for electric power steering. 1 is a system configuration diagram showing a configuration of an electric power steering using an electric power steering control device according to an embodiment of the present invention.

Explanation of symbols

1 Resolver 2 First Clock Source 3,200 First CPU
4 Excitation signal generation circuit 5 DA conversion circuit 6 Second clock source 7, 202 Second CPU
8 AD converter 9 1st differential amplifier circuit 10 2nd differential amplifier circuit

Claims (9)

A first clock source;
A first CPU that operates based on a signal from the first clock source;
An excitation signal generator for outputting an AC excitation signal based on the phase reference signal output by the first CPU;
A second clock source;
A second CPU that operates based on a signal from the second clock source;
An AD converter for acquiring a two-phase differential voltage in which the AC excitation signal is amplitude-modulated from a resolver, and AD converting the differential voltage at a timing designated by the second CPU;
Have a, an angle calculator for calculating rotor angle of the resolver from the AD conversion data of the amplitude modulation signal differential voltage of two phases obtained by the AD converter,
The angle detection circuit , wherein the second CPU includes an excitation cycle diagnosis unit that checks a cycle of an excitation signal output from the excitation signal generation unit . The angle detection circuit according to claim 1.
A first analog voltage signal output from the first CPU and a second analog voltage signal output from the first CPU via the AD converter at a timing designated by the second CPU are acquired, and the first analog voltage is acquired. An angle detection circuit comprising: an AD circuit diagnosis unit that compares a signal with the second analog voltage signal and diagnoses a failure of the AD converter. The angle detection circuit according to claim 1.
An angle detection circuit comprising means for storing a time when the AD converter is activated at a timing designated by the second CPU. The angle detection circuit according to claim 1,
The timing at which the second CPU activates the AD converter circuit is such that the differential voltage of the AC excitation signal is near the maximum value and the differential voltage of the AC excitation signal is near the minimum value. . The angle detection circuit according to claim 4 ,
At the timing when the differential voltage of the AC excitation signal becomes the maximum value / minimum value, the first CPU generates a phase transmission signal that generates an edge signal,
The angle detection circuit characterized in that the timing at which the second CPU activates the AD converter is determined by the phase transmission signal. A first clock source;
A first CPU that operates based on a signal from the first clock source;
A second clock source;
A second CPU that operates based on a signal from the second clock source;
An excitation signal generator that outputs an AC excitation signal based on either the first phase reference signal output by the first CPU or the second phase reference signal output by the second CPU;
An AD converter for acquiring a two-phase differential voltage in which the AC excitation signal is amplitude-modulated from a resolver, and AD converting the differential voltage at a timing designated by the second CPU;
An angle calculation unit that calculates a rotor angle of the resolver from AD conversion data of a two-phase amplitude modulation signal differential voltage acquired by the AD converter;
The second CPU determines which of the first phase reference signal and the second phase reference signal the excitation signal generator refers to,
The angle detection circuit , wherein the second CPU includes an excitation cycle diagnosis unit that checks a cycle of an excitation signal output from the excitation signal generation unit . The angle detection circuit according to claim 6 .
An angle detection circuit comprising means for storing a time when the AD conversion circuit is activated at a timing designated by the second CPU. The angle detection circuit according to claim 6 .
The timing at which the second CPU activates the AD converter circuit is such that the differential voltage of the AC excitation signal is near the maximum value and the differential voltage of the AC excitation signal is near the minimum value. The angle detection circuit according to claim 8 .
At the timing when the differential voltage of the AC excitation signal reaches the maximum value / minimum value, a phase transmission signal that generates an edge signal in the opposite direction
The angle detection circuit characterized in that the timing at which the second CPU activates the AD conversion circuit is determined by the phase transmission signal.

Images