JP2017163663A - Rotation direction detection device and air conditioner - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a rotation direction of a motor while suppressing an erroneous discrimination caused by influences of noise and in a configuration that does not require a high-capacity storage device.SOLUTION: In a rotation direction detection device (4), a detection section (41) detects whether inter-line induction voltages (Vun and Vwn) are matched with each other. A mean section (42) calculates phase mean calculation values (Vun_mean1 and Vwm_mean1) that are mean values of detection values of the inter-line induction voltages (Vun and Vvn) in predetermined phases during a period from the vicinity of a time in which the inter-line induction voltages (Vun and Vwn) are matched with each other to the vicinity of a time in which the inter-line induction voltages are next matched. A discrimination section (44) compares the phase mean calculation values (Vun_mean1 and Vwm_mean1) calculated by the mean section (42) with phase mean theoretical values (Vun_mean2 and Vwn_mean2) that are mean values of theoretical values of the inter-line induction voltages (Vun and Vwn) in predetermined phases in a forward rotation direction and a backward rotation direction and discriminates a direction with a smaller deviation as a rotation direction of a motor (2).SELECTED DRAWING: Figure 7

Description

  The present invention relates to a technique for detecting a rotation direction of an electric motor in an electric motor control device.

  Patent Document 1 discloses a configuration including a rotational position detection device in an electric motor control device that controls an electric motor having a permanent magnet. And in the 3rd-5th embodiment, the structure provided with the rotation direction specific | specification part which specifies the rotation direction of an electric motor in addition to the rotation position detection apparatus is disclosed.

JP 2014-45648 A

  In the third embodiment of Patent Document 1, the rotational direction specifying unit receives from the detection unit that the two-phase line-to-line induced voltages based on the minimum phase or the maximum phase coincide with each other. And the rotation direction of an electric motor is pinpointed using the voltage value of the line induced voltage when the line induced voltage of two phases corresponds mutually. However, in this configuration, for example, if the induced voltage fluctuates due to noise, the rotational direction may be erroneously determined.

  Moreover, in 5th Embodiment of patent document 1, a rotation direction specific | specification part specifies whether a rotation direction is a normal rotation direction or a reverse rotation direction using the voltage waveform of a line induced voltage. However, in this configuration, when the rotational speed of the electric motor is not known, the normal rotation estimated waveform or the reverse rotation estimated waveform cannot be calculated, and thus the normal rotation similarity or the reverse rotation similarity cannot be calculated while detecting the line induced voltage. Further, when the rotation speed of the electric motor is changing, the change in the rotation speed cannot be detected correctly, so that the normal rotation similarity or the reverse rotation similarity cannot be calculated correctly. For example, when the rotational speed is changing, in order to determine the rotational direction, it is necessary to store the value of the line induced voltage, for example, for one cycle. For this reason, a large-capacity storage device is required.

  An object of the present invention is to make it possible to detect the rotation direction of an electric motor with a configuration that suppresses erroneous determination due to the influence of noise and does not require a large-capacity storage device.

  1st invention has the field (22) containing a permanent magnet, and the armature (21) containing the winding (21u, 21v, 21w) more than three phases, The said field (22) and the said The device (4) for detecting the direction of rotation of the electric motor (2) that rotates relative to the armature (21) is intended. The device (4) uses the minimum phase and the maximum phase among the phase potentials (Vu, Vv, Vw) output from the armature (21) by induced electromotive force as a reference potential, The first line induced voltage (Vun), which is a potential difference between the phase potential (Vu) and the reference potential, and the reference potential of the second phase potential (Vw) other than the first phase potential (Vu). A detection unit (41) for detecting whether or not the second line induced voltage (Vwn) which is a potential difference with respect to the first and second line induced voltages coincides with each other, and receives the detection result of the detection unit (41), and receives the first line induced voltage (Vun) and the second line induced voltage (Vwn) when the first line induced voltage (Vun) in a predetermined period in the period from when the second line induced voltage (Vwn) coincides with each other or from the vicinity thereof to the next coincidence or the vicinity thereof. And a section average calculated value (V) that is an average value of at least one of the detected values of the second line induced voltage (Vwn). un_mean1, Vwm_mean1) and an average part (42), and at least the first line induced voltage (Vun) and the second line induced voltage (Vwn) from which the section average calculated value (Vun_mean1, Vwm_mean1) is calculated On the other hand, the theory of at least one of the first inter-line induced voltage (Vun) and the second inter-line induced voltage (Vwn) in the predetermined section when the rotation direction of the electric motor (2) is the forward direction and the reverse direction. A section average theoretical value calculation unit (44a) that calculates a section average theoretical value (Vun_mean2, Vwn_mean2) that is an average value of the values, the section average calculated values (Vun_mean1, Vwm_mean1), and the section average in the forward direction and the reverse direction A deviation calculation unit (44b) for obtaining a deviation from a theoretical value (Vun_mean2, Vwn_mean2), and a rotation direction of the section average theoretical value (Vun_mean2, Vwn_mean2) determined to be small as the deviation is a rotation direction of the electric motor (2) Rotation direction determination unit (4 4c) and a determination unit (44).

  In the first aspect of the invention, when the electric motor (2) rotates in the forward direction, the section average theoretical value (Vun_mean2, Vwn_mean2) and the section average calculated value (Vun_mean1, Vwm_mean1) in the forward direction are calculated. The deviation is smaller than the deviation between the section average theoretical value (Vun_mean2, Vwn_mean2) and the section average calculated value (Vun_mean1, Vwm_mean1) in the reverse direction. On the other hand, when the motor (2) is rotating in the reverse direction, the deviation between the section average theoretical value (Vun_mean2, Vwn_mean2) and the section average calculated value (Vun_mean1, Vwm_mean1) in the reverse direction is Is smaller than the deviation between the interval average theoretical value (Vun_mean2, Vwn_mean2) and the interval average calculated value (Vun_mean1, Vwm_mean1). Using this fact, the rotation direction determination unit (44c) determines that the deviation between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) is small in the forward direction and the reverse direction. The direction to be determined is determined as the rotation direction of the motor (2). “Near” means that when the detection unit (41) detects that the line induced voltages (Vun, Vwn) coincide with each other, the line induced voltage (Vun, Vwn) ) Means that the detection unit (41) repeatedly detects that the magnitude relationship between them has been switched (hereinafter the same).

  In the present invention, not the instantaneous detected value of at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn), but the section average calculated value that is the average value of the detected values in a predetermined section. (Vun_mean1, Vwm_mean1) is used to determine the rotation direction of the electric motor (2), so that the influence of noise superimposed on the first and second induced voltages (Vun, Vwn) is suppressed, and even if noise exists, the electric motor The rotation direction of (2) is accurately determined. Further, it is not necessary to store all values of, for example, one cycle of the first and second induced voltages (Vun, Vwn) for determining the rotation direction. Therefore, a large capacity storage device is not required.

  In a second aspect based on the first aspect, the deviation calculator (44b) calculates an absolute value of a difference between the interval average calculated value (Vun_mean1, Vwn_mean2) and the interval average theoretical value (Vun_mean2, Vwn_mean2). The difference is obtained for each of the predetermined intervals corresponding to each other, and the sum of the absolute values of the differences is used as the deviation.

  In the second invention, the deviation calculating unit (44b) calculates the absolute value of the difference between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) for each of the forward direction and the reverse direction. Find the sum as a deviation. Then, the determination direction determination unit (44c) determines, as the rotation direction of the electric motor (2), the direction determined to have a smaller sum among the normal rotation direction and the reverse rotation direction. The larger the difference between the average value of the interval (Vun_mean1, Vwm_mean1) and the average value of the interval (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction, the greater the sum of the absolute values of the difference. The smaller the difference between Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2), the smaller the sum of the absolute values of the differences. Thus, since the magnitude relationship of the deviation becomes clearer between the forward rotation direction and the reverse rotation direction, the rotation direction of the electric motor (2) can be more accurately and correctly determined.

  According to a third invention, in the first or second invention, the deviation calculating section (44b) is configured such that the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other. Of the interval average calculated value (Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2) for only at least one of the initial period and the end period of the period from or near to The deviation is obtained.

  In the third aspect, the deviation calculating unit (44b) obtains a deviation between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) only in a section where the deviation is large. And a rotation direction determination part (44c) uses the deviation which the deviation calculation part (44b) calculated | required for determination of the rotation direction of an electric motor (2). Specifically, when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other or immediately after the vicinity thereof, the first line induced voltage (Vun) and the first line induced voltage (Vun) When the next two-wire induced voltage (Vwn) coincides with or immediately before, the section average calculated value (Vun_mean1, Vwm_mean1) or the section average theoretical value between the forward direction and the reverse direction The difference between the values of (Vun_mean2, Vwn_mean2) increases. On the other hand, when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other or in the vicinity of the next coincidence or in the vicinity thereof, in the interval corresponding to the intermediate portion, The difference in the value of the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) between the forward rotation direction and the reverse rotation direction becomes small. Therefore, in the present invention, by calculating the deviation only in a section where the difference between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) is large, the amount of calculation required is reduced. Yes.

  According to a fourth invention, in any one of the first to third inventions, the determination unit (44) includes the first inter-line induced voltage (Vun) and the second inter-line induced voltage (Vwn). And a period average calculated value, which is an average value of the detected values in a period from when the two coincide with each other to the next coincidence with the next, or to the vicinity thereof, and is used to determine the rotation direction of the electric motor (2) For at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn), the period average calculated value is subtracted from the section average calculated value (Vun_mean1, Vwm_mean1) in the predetermined section. The first line-to-line induced voltage (Vun) and the second line between the second line average calculated values (Vun_mean1 ', Vwm_mean1') and the section average calculated values (Vun_mean1, Vwn_mean1) are calculated. At least one of the induced voltages (Vwn) And the first line-to-line voltage (Vun) and the second line-to-line induced voltage (Vwn) coincide with each other or from the vicinity thereof to the next coincidence or the vicinity thereof. A new value obtained by subtracting the average value of at least one of the induced voltage (Vun) and the second line induced voltage (Vwn) from the interval average theoretical value (Vun_mean2, Vwn_mean2) in the predetermined interval. It is configured to be used as theoretical values (Vun_mean2 ′, Vwn_mean2 ′).

  In the fourth aspect of the invention, the influence on the determination of the rotation direction of the electric motor (2) due to the offset superimposed on the line induced voltage (Vun, Vwn) is eliminated. Specifically, the DC component offset may be superimposed on the detected value of the line induced voltage (Vun, Vwn), and the section average calculated value (Vun_mean1, Vwm_mean1) obtained from the detected value is the same. The offset is superimposed. Compare the section average calculation value (Vun_mean1, Vwm_mean1) with this offset superimposed with the section average theoretical value (Vun_mean2, Vwn_mean2) without considering the offset to obtain the deviation, and calculate the deviation in the rotation direction of the motor (2). If used for determination, it may lead to erroneous determination of the rotational direction.

  Therefore, in the present invention, the period average calculated value (that is, the sum of the superimposed offset and the average value of the detected values of the line induced voltage (Vun, Vwn) when there is no offset) is calculated as the section average calculated value (Vun_mean1, It is subtracted from Vwm_mean1) and used as a new section average calculated value (Vun_mean1 ′, Vwm_mean1 ′). This new section average calculated value (Vun_mean1 ', Vwm_mean1') takes a value based on the average value of the detected values of the line induced voltage (Vun, Vwn) when there is no offset. In the present invention, in order to correspond to this, when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other or from the vicinity thereof to the next coincidence or the vicinity thereof, The average value of the theoretical values of the line induced voltage (Vun, Vwn) in the period is subtracted from the average value of the interval (Vun_mean2, Vwn_mean2) and used as a new interval average value (Vun_mean2 ′, Vwn_mean2 ′). This new section average theoretical value (Vun_mean2 ', Vwn_mean2') takes a value based on the average value of the theoretical values of the line induced voltage (Vun, Vwn).

  The average value of the detected value of the line induced voltage (Vun, Vwn) when there is no offset that becomes the reference of the new interval average calculated value (Vun_mean1 ', Vwn_mean1') and the new interval average theoretical value (Vun_mean2 ', Vwn_mean2) The average value of the theoretical values of the line induced voltage (Vun, Vwn), which is the standard of '), is substantially the same. Therefore, by using the deviation between the new section average calculated value (Vun_mean1 ', Vwm_mean1') and the new section average theoretical value (Vun_mean2 ', Vwn_mean2') to determine the rotation direction of the electric motor (2) The rotational direction of the electric motor (2) can be correctly determined after removing the influence of the offset superimposed on the voltage (Vun, Vwn).

  According to a fifth invention, in any one of the first to fourth inventions, a rotation speed calculation unit (43) that receives a detection result of the detection unit (41) and calculates a rotation speed of the electric motor (2). The average unit (42) includes the first line induced voltage (Vun) and the second line induced voltage (Vwn) based on the rotation speed calculated by the rotation speed calculation unit (43). The time (T) from the time when or coincides with the next time to the next coincidence or near the time (T) is obtained, and the time (T) is divided into a plurality of the intervals to calculate the interval average calculated value ( Vun_mean1, Vwm_mean1) is calculated.

  In the fifth aspect of the present invention, the time (T) from when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other or from the vicinity thereof to the next coincidence or the vicinity thereof. Is divided into a plurality of sections. Then, a section average calculated value (Vun_mean1, Vwm_mean1) is calculated in a predetermined section among the divided sections. As described above, in the method of obtaining and dividing the time (T), it is only necessary to store the section average calculated values (Vun_mean1, Vwm_mean1) as many as the number of divided sections, so that a large-capacity storage device is not required. Moreover, since only the section average calculated values (Vun_mean1, Vwm_mean1) need be calculated in each of the divided sections, the amount of calculation required is small.

  According to a sixth invention, in any one of the first to fourth inventions, a storage unit (50 having a predetermined number of storage areas (51a to 51d) for storing the detection value of the detection unit (41). ), And the average unit (42) is configured to calculate the interval average calculated value (Vun_mean1, Vwm_mean1) up to a predetermined storage upper limit number predetermined in each of the storage areas (51a to 51d). The detection values are sequentially added, and when the detection values are added to all the storage areas (51a to 51d) up to the storage upper limit number, the storage upper limit number is multiplied by an integer of 2 or more and integers are added according to the addition order. The detection values are added in increments of the storage upper limit number after doubling and the detection values are summed into a predetermined storage area (51a, 51b) and zero is stored in the remaining storage areas (51c, 51d), Up to the maximum number of storage after integer multiples for each of the storage areas (51c, 51d) storing zero Based on the stored value of the storage area (51a to 51d) at the end of the period for obtaining the section average calculated value (Vun_mean1, Vwm_mean1), the operation of adding and storing the detected value is repeated. The section average calculated values (Vun_mean1, Vwm_mean1) in the section are calculated.

  In the sixth aspect of the invention, during the period for obtaining the section average calculated values (Vun_mean1, Vwm_mean1), for example, from the vicinity when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other. Until the next match, the averaging unit (42) continues to add the detection value of the detection unit (41) to a predetermined number of storage areas (51a to 51d) up to a predetermined predetermined storage upper limit number. . Then, the averaging unit (42) multiplies the storage upper limit number by an integer of 2 or more when detection values are added to all the storage areas (51a to 51d) up to the storage upper limit number. Subsequently, the averaging unit (42) adds the detected values by the upper limit number after the integer multiple according to the order of addition and collects them as a sum in a predetermined storage area (51a, 51b), and the remaining storage areas (51c, Store zero in 51d). The averaging unit (42) adds the detected value to each of the storage areas (51c, 51d) in which zero is stored up to the upper limit of the number after the integral multiple and stores the detected value. Then, the average unit (42) calculates the section average calculated value (Vun_mean1, Vun_mean1, Vwm_mean1) based on the stored value of the storage area (51a to 51d) at the end of the period for calculating the section average calculated value (Vun_mean1, Vwm_mean1). Vwm_mean1) is calculated.

  By performing such processing, the rotational direction of the electric motor (2) can be determined without determining the rotational speed of the electric motor (2), so the processing for determining the rotational speed of the electric motor (2) is performed. There is no delay time. In addition, since the processing is performed while adding the values of the line induced voltages (Vun, Vwn) and collecting them as a sum, it is not necessary to store all the values for one cycle of the line induced voltages (Vun, Vwn), for example. Does not require large capacity storage devices.

  In the seventh invention, the air conditioner is driven by the rotation direction detecting device (4) according to any one of the first to sixth inventions, the electric motor (2), and the electric motor (2). With a fan.

  According to the present invention, the interval average calculated value (Vun_mean1, Vwm_mean1) is the average value of the detected values of the line induced voltage (Vun, Vwn), so even if noise is superimposed on the line induced voltage (Vun, Vwn) The influence can be ignored, and the rotation direction of the electric motor (2) can be correctly determined. Further, since it is not necessary to store all the values of the first and second line induced voltages (Vun, Vwn), for example, for one cycle, a large-capacity storage device can be dispensed with.

  According to the second aspect of the invention, the sum of absolute values of the difference between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) is used to determine the rotation direction of the electric motor (2). The rotation direction of the electric motor (2) can be correctly and correctly determined.

  According to the third aspect of the invention, since the deviation is obtained only in the section where the difference between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) is large, the required amount of calculation is further increased. Can be reduced.

  In addition, according to the fourth aspect of the invention, the influence of the offset superimposed on the line induced voltage (Vun, Vwn) can be removed, so that the rotation direction of the electric motor (2) can be determined more reliably and correctly. .

  In addition, according to the fifth aspect of the present invention, it is only necessary to store the section average calculated values (Vun_mean1, Vwm_mean1) as many as the number of divided sections, so that a large-capacity storage device can be dispensed with. Further, since it is only necessary to calculate the section average calculated values (Vun_mean1, Vwm_mean1) in each of the divided sections, the amount of calculation required can be reduced.

  According to the sixth aspect of the invention, the rotational direction of the electric motor (2) can be determined without obtaining the rotational speed of the electric motor (2), so that the processing for obtaining the rotational speed of the electric motor (2) is performed. The delay time can be prevented. Further, since it is not necessary to store all values of, for example, one cycle of the line induced voltage (Vun, Vwn), a large-capacity storage device can be dispensed with.

FIG. 1 is a diagram illustrating a schematic configuration of an electric motor control device including a rotation direction detection device according to the first embodiment. FIG. 2 is a diagram illustrating a change example of the induced voltage in the forward rotation direction. FIG. 3 is a diagram illustrating a change example of the line induced voltage in the normal rotation direction. FIG. 4 is a diagram illustrating a change example of the line induced voltage in the reverse rotation direction. FIG. 5 is a diagram illustrating a method of dividing the time from when the line induced voltages are matched to each other until the next matching is made into a plurality of sections. FIG. 6 is a diagram illustrating a specific configuration example of the determination unit. FIG. 7 is a diagram illustrating a change example of the section average calculated value or section average theoretical value of the line induced voltage in the forward rotation direction. FIG. 8 is a diagram illustrating a change example of the section average calculated value or section average theoretical value of the line induced voltage in the reverse rotation direction. FIG. 9 is a diagram illustrating a schematic configuration of an electric motor control device including the rotation direction detection device according to the second embodiment. FIG. 10 is a diagram illustrating a mode in which detection values of line induced voltages are stored in a plurality of storage areas.

  Embodiments of the present invention will be described in detail with reference to the drawings. The following embodiments are essentially preferable examples, and are not intended to limit the scope of the present invention, its application, or its use.

Embodiment 1 of the Invention
The first embodiment will be described. FIG. 1 is a diagram illustrating a schematic configuration of an electric motor control device including a rotation direction detection device according to the present embodiment. The motor control device shown in FIG. 1 includes a power conversion unit (1), a motor (2), a voltage detection unit (3) for detecting a line induced voltage, and a motor (2) from the detected line induced voltage. And a rotation direction detection device (4) for detecting the rotation direction of.

  The power conversion unit (1) has an input side connected to DC lines (L1, L2) and an output side connected to AC lines (Pu, Pv, Pw). A DC voltage is applied between the DC lines (L1, L2) by, for example, a converter (not shown). The converter converts, for example, an AC voltage from a commercial AC power source into a DC voltage, and applies this between DC lines (L1, L2). As such a converter, for example, a diode rectifier circuit formed by a diode bridge can be employed. A capacitor (C) provided between the DC lines (L1, L2) smoothes the DC voltage. The power converter (1) converts the DC voltage applied to the DC lines (L1, L2) into a three-phase AC voltage and applies it to the AC lines (Pu, Pv, Pw).

  In the illustration of FIG. 1, the power conversion unit (1) is a voltage source inverter. More specifically, the power converter (1) includes switching elements (Sup, Svp, Swp, Sun, Svn, Swn) and diodes (Dup, Dvp, Dwp, Dun, Dvn, Dwn). The switching elements (Sxp, Sxn (x represents u, v, w, the same applies hereinafter)) are, for example, insulated gate bipolar transistors and the like, and are connected in series between the DC lines (L1, L2). The diodes (Dxp, Dxn) are connected in antiparallel with the switching elements (Sxp, Sxn), respectively, and have an anode on the DC line (L2) side. The AC line (Px) is connected to the midpoint between the switching elements (Sxp, Sxn).

  These switching elements (Sxp, Sxn) are controlled so as to conduct exclusively. When both switching elements (Sxp, Sxn) are conducting, the DC lines (L1, L2) are short-circuited via the switching elements (Sxp, Sxn), which causes a large current to flow through the switching elements (Sxp, Sxn). It is. And by appropriately controlling these switching elements (Sxp, Sxn), the power conversion unit (1) can convert a DC voltage into an AC voltage.

  Note that the diodes (Dxp, Dxn) are not essential if the switching elements (Sxp, Sxn) are capable of reverse conduction (conduction from the DC line (L2) to the DC line (L1)). For example, when a MOS field effect transistor having a parasitic diode is employed as the switching element (Sxp, Sxn), the diode (Dxp, Dxn) is unnecessary.

  The electric motor (2) includes an armature (21) and a field (22). The armature (21) has three-phase armature windings (21u, 21v, 21w), and the armature windings (21u, 21v, 21w) are connected to AC lines (Pu, Pv, Pw), respectively. . A three-phase AC voltage from the power converter (1) is applied to the armature windings (21u, 21v, 21w). As a result, an alternating current flows through the armature winding (21u, 21v, 21w), and a rotating magnetic field is applied to the field magnet (22). The field magnet (22) has a permanent magnet and supplies a field magnetic flux to the armature (21). The field (22) receives a rotating magnetic field from the armature (21) and rotates relative to the armature (21).

  In the illustration of FIG. 1, since the electric motor (2) having three-phase armature windings (21u, 21v, 21w) is assumed, the power conversion unit (1) outputs a three-phase AC voltage. However, this is not necessarily the case. An N-phase motor larger than three phases may be employed, and an N-phase power converter may be employed as well. In the example of FIG. 1, the armature windings (21u, 21v, 21w) are connected to each other by so-called star connection, but may be connected to each other by so-called triangular connection.

  The electric motor (2) is used for a blower such as a fan or a blower, for example. For example, the electric motor (2) may drive a fan or a compressor mounted on a heat pump (such as an air conditioner or a hot water supply device). For example, when driving a fan mounted on an outdoor unit placed outdoors, the flow of outdoor air (1) even when the power converter (1) does not output AC voltage to the motor (2) ( Rotate by wind). Therefore, when starting such an electric motor (2), it is necessary to detect the relative rotation direction of the armature (21) and the field (22) (the rotation direction of the electric motor (2)). Of course, even if the compressor or the fan is not rotated by an external force, the compressor or the fan rotates due to inertia. Therefore, even when the compressor or the fan is rotated again, it is necessary to detect the rotation direction.

  When the electric motor (2) rotates, the magnetic flux passing through the armature winding (21u, 21v, 21w) changes based on the rotation. Along with this, an induced electromotive force based on the rotation is generated in each of the armature windings (21u, 21v, 21w), and the electric motor (2) has a phase potential (hereinafter, referred to as “phase potential”) on the AC line (Pu, Pv, Pw). (Also called induced voltage) (Vu, Vv, Vw).

  FIG. 2 shows an example of changes in the induced voltage (Vu, Vv, Vw). As illustrated in FIG. 2, the induced voltage (Vu, Vv, Vw) takes a substantially sinusoidal shape that varies depending on the rotational position (electrical angle) of the electric motor (2). In addition, in FIG. 2, the induced voltage (Vu, Vv, Vw) when an electric motor (2) rotates to a normal rotation direction is illustrated. In the forward rotation direction, the induced voltage (Vv, Vw) advances 120 degrees with respect to the induced voltage (Vu, Vv). In other words, such a rotation direction is defined as a normal rotation direction.

  The voltage detector (3) detects the line induced voltage (Vun, Vwn). The line-to-line induced voltage (Vun, Vwn) refers to a potential difference between the induced voltage (Vu, Vw) and the reference potential. Here, the induced voltage of the minimum phase among the induced voltages (Vu, Vv, Vw) is adopted as the reference potential of the line induced voltage (Vun, Vwn).

  For example, in the forward rotation direction shown in FIG. 2, the induced voltage (Vu) takes the minimum value in the range where the rotational position is 30 to 150 degrees, and thus the induced voltage (Vu) of the minimum phase is the induced voltage (Vu) in this range. Therefore, in this range, the line induced voltage (Vun) is zero. In addition, since the induced voltage (Vv) is the minimum phase induced voltage in the range of the rotational position of 150 to 270 degrees, the line induced voltage (Vun) is the induced voltage (Vu) and the minimum phase induced voltage (Vu) in this range. Vv).

  FIG. 3 shows a change in the line induced voltage when the rotation direction is the forward rotation direction. As shown in FIG. 3, the line-to-line induced voltage (Vun) becomes zero when the rotational position is in the range of 30 to 150 degrees and becomes a potential difference (Vu−Vv) when the rotational position is in the range of 150 to 270 degrees. The potential difference (Vu−Vw) is in the range of 270 to 360 degrees and 0 to 30 degrees. On the other hand, the line-to-line induced voltage (Vwn) becomes zero when the rotational position is in the range of 240 to 360 degrees and 0 to 30 degrees, and becomes a potential difference (Vw−Vu) when the rotational position is in the range of 30 to 150 degrees. The potential difference (Vw−Vv) is in the range of 150 to 270 degrees.

  FIG. 4 shows a change in the induced voltage between the lines when the rotation direction is the reverse direction. In the reverse direction, the induced voltages (Vv, Vw) are delayed by 120 degrees with respect to the induced voltages (Vu, Vv), respectively. Therefore, the line induced voltage (Vun, Vwn) at this time has a waveform as shown in FIG.

  The voltage detection unit (3) has a path (31, 32) for connecting each of the AC lines (Pu, Pv) to which the induced voltages (Vu, Vw) are applied and the DC line (L2), In the path (31, 32), the voltage between each of the AC lines (Pu, Pv) and the DC line (L2) is detected as a line induced voltage (Vun, Vwn).

  In addition, when the power converter (1) is controlled and AC voltage is being output to the AC line (Pu, Pv, Pw), the voltage detector (3) properly generates the line-to-line induced voltage (Vun, Vwn). Since it cannot detect, a voltage detection part (3) detects a line induced voltage in the state in which a power converter (1) does not output an alternating voltage. That is, the line induced voltage is detected in a state where all the switching elements (Sxp, Sxn) are controlled to be non-conductive.

  In the example of FIG. 1, the voltage detection unit (3) includes voltage dividing resistors (R11, R12, R21, R22). The voltage dividing resistors (R11, R12) are connected in series in the path (31). The voltage dividing resistors (R21, R22) are connected in series in the path (32).

  In such a voltage detector (3), during the period in which the induced voltage (Vv) is the minimum phase induced voltage, the line induced voltage (Vun) is represented by the path (31), the DC line (L2), and the diode (Dvn). ) Between the AC lines (Pu, Pv) via At this time, since a voltage is applied to the diode (Dvn) in the forward direction, the voltage is almost zero. Therefore, the voltage across the entire voltage dividing resistor (R11, R12) substantially matches the potential difference between the induced voltage (Vu) and the induced voltage (Vv) of the minimum phase, that is, the line induced voltage (Vun). Similarly, the voltage across the voltage dividing resistors (R21, R22) substantially matches the potential difference between the induced voltage (Vw) and the induced voltage (Vv) of the minimum phase, that is, the line induced voltage (Vwn).

  In the period when the induced voltage (Vw) is the minimum phase induced voltage, the line induced voltage (Vun) is the AC line (Pu, Pw) via the path (31), DC line (L2) and diode (Dwn). Applied between At this time, since a voltage is applied to the diode (Dwn) in the forward direction, the voltage is almost zero. Therefore, at this time, the voltage across the entire voltage dividing resistor (R11, R12) substantially coincides with the line induced voltage (Vun). On the other hand, since the potentials of the DC line (L2) and the AC line (Pw) are substantially equal to each other, the voltage across the voltage dividing resistors (R21, R22) is almost zero. Since the line induced voltage (Vwn) is zero during this period, the voltage across the voltage dividing resistors (R21, R22) substantially matches the line induced voltage (Vwn).

  During the period when the induced voltage (Vu) is the minimum phase induced voltage, the line induced voltage (Vwn) is the path (32), the DC line (L2) and the AC line (Pu, Pw) via the diode (Dun). Applied between. At this time, since a voltage is applied to the diode (Dun) in the forward direction, the voltage is almost zero. Therefore, at this time, the voltage across the voltage dividing resistor (R21, R22) is almost equal to the line induced voltage (Vwn), and the voltage across the voltage dividing resistor (R11, R12) is the line induced voltage (Vun). Almost matches. Therefore, the voltage applied to the voltage dividing resistor (R11, R12) corresponds to the line induced voltage (Vun), and the voltage applied to the voltage dividing resistor (R21, R22) corresponds to the line induced voltage (Vwn). . Therefore, the line induced voltage (Vun, Vwn) can be detected by detecting these voltages.

  In the example of FIG. 1, the voltage detection unit (3) detects the voltage of the voltage dividing resistor (for example, the voltage dividing resistor (R12) on the DC line (L2) side) in the path (31) as the line induced voltage (Vun). The voltage of the voltage dividing resistor (for example, the voltage dividing resistor (R22) on the DC line (L2) side) in the path (32) is detected as a line induced voltage (Vwn). This makes it possible to detect the line induced voltage (Vun, Vwn) with a smaller voltage value. The difference between the voltage dividing ratio of the voltage dividing resistors (R11, R12) and the voltage dividing ratio of the voltage dividing resistors (R21, R22) is desirably small. This is because when these differences occur, a difference occurs between the time when the magnitude relationship of the divided resistance voltage is switched and the time when the magnitude relationship between the line induced voltages (Vun, Vwn) is switched.

  As described above, the voltage detection unit (3) has the paths (31, 32) for connecting the DC lines (L2) and the AC lines (Pu, Pw). , 32) can be detected as line induced voltage (Vun, Vwn).

  According to such a voltage detector (3), for example, the line induced voltage (Vun, Vwn) can be easily obtained as compared with the following case. That is, the induced voltage (Vu, Vv, Vw) is detected, the induced voltage of the minimum phase is extracted from the detected induced voltage (Vu, Vv, Vw), and the minimum phase is detected from the detected induced voltage (Vu, Vw). Compared to the case where the induced voltage (Vun, Vwn) is calculated by subtracting the induced voltage, the induced voltage (Vun, Vwn) can be easily obtained.

  When the line induced voltage (Vup, Vwp) based on the maximum phase is adopted, the voltage detector (3) is connected between each AC line (Pu, Pw) and the DC line (L1). What is necessary is just to detect a voltage.

<Rotation direction detector (4)>
The rotation direction detection unit (4) detects the rotation direction (D) of the electric motor (2) using the line induced voltage (Vun, Vwn) detected by the voltage detection unit (3). The rotation direction detection unit (4) includes an analog / digital conversion unit (40a, 40b), a detection unit (41), a rotation speed calculation unit (43), an average unit (42), and a determination unit (44). It has. Here, the rotation direction detection unit (4) includes a microcomputer and a storage device. The microcomputer executes each processing step described in the program. The storage device can be composed of one or more of various memories such as RAM (Random-Access-Memory) and various storage devices such as a hard disk device. The storage device stores various information, data, and the like, stores a program executed by the microcomputer, and provides a work area for executing the program. It can be understood that the microcomputer functions as various means corresponding to each processing step described in the program, or can realize that various functions corresponding to each processing step are realized. In addition, various procedures executed by the rotation direction detection unit (4), or various means or various functions to be realized may be realized by hardware.

  The analog / digital conversion units (40a, 40b) convert the line-to-line induced voltages (Vun, Vwn) into digital signals, respectively.

  The detection unit (41) receives the line induced voltage (Vun, Vwn) converted into a digital signal and detects whether the line induced voltages (Vun, Vwn) match each other. The detection here can be realized, for example, by using a known comparison unit that compares the magnitudes of the line induced voltages (Vun, Vwn). More specifically, when the magnitude relation between the line induced voltages (Vun, Vwn) is switched, it can be determined that the line induced voltages (Vun, Vwn) coincide with each other. For example, in the example of FIGS. 3 and 4, when the rotational position is 30 degrees, the line induced voltage (Vwn) exceeds the line induced voltage (Vvn), and the state A (Vun> Vwn) to the state B (Vun) <Vwn). At this time, it is determined that the line induced voltages (Vun, Vwn) coincide with each other. When the rotational position is 210 degrees, the line induced voltage (Vun) exceeds the line induced voltage (Vwn), and the state B (Vun <Vwn) is changed to the state A (Vun> Vwn). . At this time, it is determined that the line induced voltages (Vun, Vwn) coincide with each other.

  The rotation speed calculation unit (43) calculates the rotation speed of the electric motor (2) using the output from the detection unit (41). The rotation speed calculation unit (43) calculates the absolute value of the rotation speed. That is, the time from when the line induced voltages (Vun, Vwn) match each other until the next line induced voltages (Vun, Vwn) match each other corresponds to the electrical angle half cycle of the electric motor (2). . Therefore, the rotation speed calculation unit (43) is notified from the detection unit (41) that the line induced voltages (Vun, Vwn) coincide with each other, and then is detected from the detection unit (41). The time until it is notified that (Vun, Vwn) match each other is measured, and this time is acquired as the electrical angle half cycle of the electric motor (2). Of course, using the output from the detection section (41), the time of one cycle of the electrical angle of the electric motor (2) or the time of two or more cycles may be measured. Then, the rotation speed calculation unit (43) outputs the rotation speed obtained from the measured time to the determination unit (44). Alternatively, the time (T) of the electrical angle half cycle of the electric motor (2) may be output to the determination unit (44) as data that indirectly represents the rotation speed. The rotation speed calculation unit (43) outputs the electrical angle half cycle time (T) of the electric motor (2) to the averaging unit (42).

  The average part (42) is the line induced voltage (Vun, Vwn) of a predetermined section in the period from the vicinity when the line induced voltages (Vun, Vwn) match each other to the vicinity when the line coincides next. A section average calculated value (Vun_mean1, Vwm_mean1) that is an average value of the detected values is calculated. The average unit (42) acquires the time (T) of the electrical angle half cycle of the electric motor (2) from the rotation speed calculation unit (43). As shown in FIG. 5, the average unit (42) divides the acquired electrical angle half-period time (T) of the electric motor (2) into a plurality of sections (six sections in the present embodiment). Based on the line induced voltage (Vun, Vwn) converted into a digital signal, the average unit (42) calculates the section average calculated value (Vun_mean1, Vwm_mean1) of the line induced voltage (Vun, Vwn) in each section. calculate. The section average calculated values (Vun_mean1, Vwm_mean1) calculated in this way take values as shown in FIG. 7, for example, in the range of 30 to 210 degrees in the forward rotation direction. On the other hand, the section average calculated values (Vun_mean1, Vwm_mean1) take values as shown in FIG. 8, for example, in the range of 30 (390) to 210 degrees in the reverse direction. In addition, it is not necessary to divide | segment the time (T) of the electrical angle half cycle of an electric motor (2) into six, You may divide into 5 or less, and may divide into 7 or more.

  As shown in FIG. 6, the determination unit (44) includes a section average theoretical value calculation unit (44a), a deviation calculation unit (44b), and a rotation direction determination unit (44c). 42) The rotation direction of the electric motor (2) is obtained using the section average calculated values (Vun_mean1, Vwm_mean1) of the line induced voltages (Vun, Vwn) calculated in (42).

  The section average theoretical value calculation unit (44a), when the detection unit (41) detects that the line-to-line induced voltages (Vun, Vwn) match each other, from the rotation speed calculation unit (43) to the electric motor (2) The rotation speed of the or data representing the rotation speed is acquired. Then, the section average theoretical value calculation unit (44a) corresponds to the section average calculation value (Vun_mean1, Vwm_mean1) of the line induced voltage (Vun, Vwn) calculated by the average unit (42) based on this rotational speed. The interval average theoretical value (Vun_mean2, Vwn_mean2) is calculated. The section average theoretical value calculation unit (44) does not calculate the section average theoretical value (Vun_mean2, Vwn_mean2) of the line induced voltage (Vun, Vwn), but the section average theoretical value corresponding to each rotation speed ( Vun_mean2, Vwn_mean2) data may be stored in advance.

  The deviation calculating unit (44b) is configured to obtain a deviation between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction. The section average theoretical value (Vun_mean2, Vwn_mean2) in the forward direction takes a value as shown in FIG. 7, for example, in the range of 30 to 210 degrees. On the other hand, the interval average theoretical value (Vun_mean2, Vwn_mean2) in the reverse direction takes a value as shown in FIG. 8, for example, in the range of 30 (390) to 210 degrees.

  The rotation direction determination unit (44c) is configured to determine the rotation direction of the electric motor (2) based on the deviation obtained by the deviation calculation unit (44b).

  Here, the rotation direction determination method in the present embodiment will be described.

  As shown in FIG. 3, in the forward rotation direction, the line induced voltage (Vun, Vwn) at a rotational position of 30 degrees (when the line induced voltage Vwn exceeds the line induced voltage Vun) is a relatively small value (for example, zero). Take. On the other hand, the line induced voltage (Vun, Vwn) at the rotational position 210 degrees (when the line induced voltage Vun exceeds the line induced voltage Vwn) takes a relatively large value (ΔV in FIG. 3). On the other hand, as shown in FIG. 4, in the reverse rotation direction, the line induced voltage (Vun, Vwn) takes a relatively large value (ΔV in FIG. 4) at the rotational position of 30 degrees. On the other hand, at the rotation position of 210 degrees, the line induced voltage (Vun, Vwn) takes a relatively small value (for example, zero).

  Due to such characteristics, the rotational direction of the electric motor (2) can be detected by comparing the line induced voltages (Vun, Vwn) at the rotational positions of 30 degrees and 210 degrees. That is, when the line induced voltage (Vun, Vwn) at the rotational position 30 degrees is smaller than the line induced voltage (Vun, Vwn) at the rotational position 210 degrees, the forward rotation direction can be determined. When the line induced voltage (Vun, Vwn) at a degree is greater than the line induced voltage (Vun, Vwn) at the rotational position 210 degrees, it can be determined that the direction of reverse rotation.

  However, in the above determination method, there is a possibility of erroneous determination due to the influence of noise. That is, since the magnitude comparison is performed using the detected value of the line induced voltage at the time when the two-phase line induced voltages at the rotational positions of 30 degrees and 210 degrees coincide with each other, If the detected voltage value fluctuates, an error may occur in the size comparison. As a cause of noise, for example, noise caused by the switching operation of another power conversion unit disposed on or near the same substrate can be cited. For example, when the motor (2) is rotating in the forward rotation direction, noise is added to the line induced voltage (Vun, Vwn) at the rotational position of 30 degrees, and the line induced voltage (Vun at the rotational position of 210 degrees). , Vwn), it is erroneously determined as the reverse direction.

  Therefore, in the present embodiment, the deviation calculating unit (44b) calculates the section average in a predetermined section of the period from the vicinity when the line induced voltages (Vun, Vwn) coincide with each other to the vicinity when the next coincidence occurs. The values (Vun_mean1, Vwm_mean1) are compared with the interval average theoretical values (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction, and their deviations are obtained. In the present embodiment, the deviation calculation unit (44b) is configured to equally divide each section of 30 to 60 degrees, 60 to 90 degrees, 90 to 120 degrees, 120 to 150 degrees, 150 to 180 degrees, and 180 to 210 degrees. , The deviation between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction is obtained.

  Here, when the motor (2) rotates in the forward direction, ideally, the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward direction in each section Matches. Therefore, when the sum of the absolute values of the difference between the interval average calculated value (Vun_mean1, Vwm_mean1) in each interval and the interval average theoretical value (Vun_mean2, Vwn_mean2) in the forward rotation direction is obtained as a deviation, the value becomes substantially zero. . In addition, when the motor (2) is rotating in the forward direction, a predetermined difference between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the reverse direction in each section Occurs. In the example shown in FIGS. 7 and 8, the difference occurs in each section of 30 to 90 degrees and 150 to 210 degrees. Therefore, if the sum of the absolute values of the difference between the section average calculated value (Vun_mean1, Vwm_mean1) in each section and the section average theoretical value (Vun_mean2, Vwn_mean2) in the reverse direction is calculated as a deviation, the value is relatively large. Take.

  From the above, when the rotational direction determination unit (44c) has a small sum of absolute values of differences between the average value of the interval (Vun_mean1, Vwm_mean1) and the average value of the interval (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction, The determined direction (in the above-described example, the normal rotation direction) can be determined as the rotation direction of the electric motor (2). Even when the electric motor (2) rotates in the reverse direction, the rotation direction can be determined in the same manner.

-Effect of Embodiment 1-
In the rotation direction detection device (4) of the present embodiment, the rotation direction determination unit (44c) of the determination unit (44) calculates the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value in the forward direction and the reverse direction. The direction in which the sum of absolute values of differences from (Vun_mean2, Vwn_mean2) is judged to be smaller is determined as the rotation direction of the electric motor (2). Thus, instead of the instantaneous detection value of the line-to-line induced voltage (Vun, Vwn), the section average calculated value (Vun_mean1, Vwm_mean1), which is the average value of the detection values in the predetermined section, is the rotation direction of the motor (2). Therefore, the influence of noise superimposed on the line-to-line induced voltage (Vun, Vwn) is suppressed, and the rotational direction of the electric motor (2) can be accurately determined even if noise is present. In addition, the larger the difference between the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward rotation direction and the reverse rotation direction, the greater the sum of the absolute values of the difference, and the section average calculation The smaller the difference between the value (Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2), the smaller the sum of the absolute values of the differences. Thus, since the magnitude relationship of the sum of the absolute values of the difference between the forward rotation direction and the reverse rotation direction is clear, the rotation direction of the electric motor (2) can be determined more reliably and correctly. Further, since it is not necessary to store all the values of the first and second line induced voltages (Vun, Vwn), for example, for one cycle, a large-capacity storage device can be dispensed with.

  In addition, the average part (42) divides the time (T) from the vicinity when the line induced voltages (Vun, Vwn) coincide with each other to the vicinity when the next coincidence coincides with each other, and divides it into a plurality of times. A section average calculated value (Vun_mean1, Vwm_mean1) is calculated in a predetermined section among the sections. Therefore, since it is only necessary to store the section average calculated values (Vun_mean1, Vwm_mean1) as many as the number of divided sections, a large-capacity storage device can be dispensed with. Further, since it is only necessary to calculate the section average calculated values (Vun_mean1, Vwm_mean1) in each of the divided sections, the amount of calculation required can be reduced.

<< Embodiment 2 of the Invention >>
Embodiment 2 will be described. In the present embodiment, the processing content of the average unit (42) is different from that in the first embodiment. In the following, differences from the first embodiment will be mainly described.

  As shown in FIG. 9, in this embodiment, information on the time (T) of the electrical angle half cycle of the electric motor (2) is not output from the rotation speed calculation unit (43) to the average unit (42). Then, the average part (42) does not use the information on the rotation speed of the electric motor (2) or the time of the electrical angle half cycle (T), and calculates the section average calculated value (Vun_mean1, Vun, Vwn) of the line induced voltage (Vun, Vwn). Vwm_mean1) is calculated.

  The average part (42) of the rotation direction detection part (4) of this embodiment has the memory | storage part (50) shown in FIG. One storage unit (50) is provided for each line induced voltage (Vun, Vwn) (two in this embodiment). The storage unit (50) has a predetermined number (four in this embodiment) of storage areas (51a to 51d) for storing the detection values of the detection unit (41). The first to fourth storage areas (51a to 51d) are composed of, for example, a RAM (Random-Access-Memory). The storage unit (50) may be provided as a component different from the average unit (42). Further, the number of storage areas (51a to 51d) is not limited to four, and may be a multiple of the integer (however, excluding the integer) as long as the storage upper limit number is multiplied by an integer. For example, in the configuration in which the storage upper limit number is doubled, it may be any even number of 4 or more (for example, 4, 6, 8, 10). For example, in the configuration in which the storage upper limit number is tripled, it is a multiple of 3 (for example, 6). , 9, 12, 15).

  The average part (42) is the first to fourth storage areas from the vicinity when the line induced voltages (Vun, Vwn) coincide with each other (at the start of the period for obtaining the section average calculated values (Vun_mean1, Vwm_mean1)). The values of the line induced voltages (Vun, Vwn), which are detection values of the detection unit (41), are sequentially added to a predetermined upper limit number of storage (1 in this example) to each of (51a to 51d) ( (Corresponding to calculation cycles 1 to 4). Here, the symbols in FIG. 10 circled with numbers represent the line induced voltage (Vun, Vwn) that is the detection value of the detection unit (41) in the order of the numbers. For example, the symbol circled 1 is the first detected line induced voltage (Vun, Vwn), and the symbol circled 2 is the second detected line induced voltage (Vun, Vwn). Represents.

  The average unit (42) stores the upper limit number of storages when the detection values are added to all the first to fourth storage areas (51a to 51d) by the upper limit number of storages (one by one in the calculation cycles 1 to 4). Is doubled (in the calculation cycle 5, the upper limit number of storage is 2), and the detected values are added by the upper limit of the number of storage after doubling (2 in the calculation cycle 5) in accordance with the stored order. Are summed as a sum in the first and second storage areas (51a, 51b), and zero is stored in the third and fourth storage areas (51c, 51d). The average unit (42) is a detection unit up to the upper limit number of storage after double (two in the calculation cycle 5 to 8) for each of the third and fourth storage areas (51c, 51d) storing zero. The value of the line induced voltage (Vun, Vwn), which is the detected value of (41), is added and stored. In FIG. 10, when N detection values are included in each storage area (51a to 51d), it means that the sum of the N detection values is stored. For example, the sum of two detection values is stored in the first and second storage areas (51a, 51b) in the calculation cycle 5. Note that when the detected values are added to all the first to fourth storage areas (51a to 51d) by the storage upper limit number, it is not always necessary to double the storage upper limit number, and the storage upper limit number is an integer of 2 or more. It may be doubled.

  The averaging unit (42) again stores the upper limit of storage at the time when the detection values are added to all the first to fourth storage areas (51a to 51d) up to the upper limit of storage (two in the calculation cycle 5 to 8). The number is doubled (in the calculation cycle 9, the storage upper limit number is set to 4), and the detection value is added by the detected upper limit number (4 in the calculation cycle 9) in accordance with the storage order. The values are collected as a sum in the first and second storage areas (51a, 51b), and zero is stored in the third and fourth storage areas (51c, 51d). The average unit (42) is a detection unit for the third and fourth storage areas (51c, 51d) storing zero and the storage upper limit number after double (four in the calculation cycles 9 to 13). The value of the line induced voltage (Vun, Vwn), which is the detected value of (41), is added and stored.

  Thereafter, the averaging unit (42) performs the above-described operation until the vicinity when the line-induced voltages (Vun, Vwn) coincide with each other (at the end of the period for obtaining the section average calculated values (Vun_mean1, Vwm_mean1)). Go repeatedly.

  The average part (42) is a section in a predetermined section based on the stored values of the first to fourth storage areas (51a to 51d) in the vicinity when the line induced voltages (Vun, Vwn) are next matched. The average calculated value (Vun_mean1, Vwm_mean1) is calculated. For example, when calculating the section average calculated value (Vun_mean1, Vwm_mean1) based on the stored value in the calculation cycle 8, when the line induced voltages (Vun, Vwn) match next from the vicinity when they match each other The period until the vicinity of is divided into four sections, and the section average calculated values (Vun_mean1, Vwm_mean1) in each section are calculated.

  In this case, the average section (42) divides the stored value of each storage area (51a to 51d) by 2, which is the number of detection values stored in each storage area (51a to 51d), thereby calculating the section average calculated value (Vun_mean1, Vwm_mean1) is calculated, and the deviation calculation unit (44b) of the determination unit (44) calculates the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward and reverse directions The deviation may be obtained. Instead, the deviation calculation unit (44b) multiplies the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction by 2 and the stored value of each storage area (51a to 51d). In this case, the deviation between the interval average calculated value (Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2) in the forward and reverse directions can be substantially calculated. Can be sought.

  For example, when calculating the section average calculated value (Vun_mean1, Vwm_mean1) based on the stored value in the calculation cycle 12, the line-to-line induced voltages (Vun, Vwn) match next from the vicinity when they match each other. The period until the vicinity of time is divided into three sections, and the section average calculated values (Vun_mean1, Vwm_mean1) in each section are calculated.

  In this case, the average section (42) divides the stored value of each storage area (51a to 51c) by 4 which is the number of detection values stored in each storage area (51a to 51c), thereby calculating the average value of the section. (Vun_mean1, Vwm_mean1) is calculated, and the deviation calculation unit (44b) of the determination unit (44) calculates the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward and reverse directions The deviation may be obtained. Instead of this, the deviation calculation unit (44b) multiplies the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction by 4 and the stored value of each storage area (51a to 51c). In this case, the deviation between the interval average calculated value (Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2) in the forward and reverse directions can be substantially calculated. Can be sought.

  The method for determining the direction of rotation of the electric motor (2) is the same as that in the first embodiment, and a description thereof will be omitted here.

-Effect of Embodiment 2-
In the rotation direction detection device (4) of the present embodiment, the section average calculated values (Vun_mean1, Vwm_mean1) can be calculated without using information regarding the rotation speed of the electric motor (2). Accordingly, there is no delay time for performing the process of obtaining the rotation speed of the electric motor (2). In addition, since the processing is performed while adding the values of the line induced voltages (Vun, Vwn) and collecting them as a sum, it is not necessary to store all the values for one cycle of the line induced voltages (Vun, Vwn), for example. A large-capacity storage device can be dispensed with.

<< Other Embodiments >>
In each of the above embodiments, the deviation calculating unit (44b) calculates the section average calculated value (Vun_mean1) for the period from the vicinity when the line induced voltages (Vun, Vwn) coincide with each other to the vicinity when the line coincides next. , Vwm_mean1) and the mean deviation of the average value in the forward and reverse directions (Vun_mean2, Vwn_mean2). However, the deviation calculation unit (44b) determines the interval average calculated value (Vun_mean1, Vwm_mean1), the normal rotation direction, and the period from the time when the line induced voltages (Vun, Vwn) match each other until the next match. You may comprise so that the deviation with the area average theoretical value (Vun_mean2, Vwn_mean2) in a reverse direction may be calculated | required.

  In each of the above embodiments, the deviation calculation unit (44b) calculates the section average calculated value over the entire section from the vicinity when the line induced voltages (Vun, Vwn) match each other to the vicinity when the next matching occurs. A difference between (Vun_mean1, Vwm_mean1) and a section average theoretical value (Vun_mean2, Vwn_mean2) in the forward rotation direction and the reverse rotation direction is obtained. However, the deviation calculator (44b) calculates the section average for at least one of the initial period and the final period in the period from the vicinity when the line induced voltages (Vun, Vwn) match each other to the vicinity when the next matching occurs. The calculated value (Vun_mean1, Vwm_mean1) may be configured to obtain a deviation between the normal average value (Vun_mean2, Vwn_mean2) in the forward direction and the reverse direction. For example, in the example shown in FIGS. 7 and 8, the deviation calculation unit (44b) calculates the section average calculated value (Vun_mean1, Vwm_mean1) and the normal rotation direction for at least one of the sections of 30 to 60 degrees and 180 to 210 degrees. You may comprise so that the deviation with the area average theoretical value (Vun_mean2, Vwn_mean2) in a reverse direction may be calculated | required. In these sections, the difference between the section average theoretical value in the forward direction (Vun_mean2, Vwn_mean2) and the section average theoretical value in the reverse direction (Vun_mean2, Vwn_mean2) is large, and the section average calculation is limited to such sections. By calculating the deviation between the value (Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2), it is possible to reduce the amount of calculation while correctly determining the rotation direction of the motor (2). .

  In each of the above embodiments, both of the two line induced voltages (Vun, Vwn) are used for determining the rotation direction of the electric motor (2). Of the two line induced voltages (Vun, Vwn), Either one may be used for determination of the rotation direction of the electric motor (2).

  In addition, the determination unit (44) starts from the vicinity when the line induced voltages (Vun, Vwn) coincide with each other with respect to the line induced voltages (Vun, Vwn) used for determining the rotation direction of the electric motor (2). Calculate the period average calculated value that is the average value of the detected values in the period up to the vicinity when it coincides with, and subtract the period average calculated value from the section average calculated value (Vun_mean1, Vwm_mean1) to obtain a new section average calculation You may be comprised so that it may use as a value (Vun_mean1 ', Vwm_mean1'). At the same time, the determination unit (44) calculates the interval average calculated values (Vun_mean1, Vwn_mean1) from the vicinity when the line induced voltages (Vun, Vwn) match each other. Subtracting the average of the theoretical values of the line induced voltage (Vun.Vwn) from the interval average theoretical value (Vun_mean2, Vwn_mean2) in the period until the next matching point to the new interval average theoretical value (Vun_mean2 ' , Vwn_mean2 ′). As a result, even when the DC component offset is superimposed on the line induced voltage (Vun, Vwn), the rotational direction of the electric motor (2) can be reliably determined.

  Further, in each of the above embodiments, the line induced voltage (Vun, Vwn) is used as the two line induced voltages, but the line induced voltage (Vun, Vvn) may be used or the line-to-line induced voltage may be used. An induced voltage (Vvn, Vwn) may be used.

  Further, the rotation direction detection device (4) disclosed in each of the embodiments can be applied to an air conditioner including an electric motor (2) and a fan driven by the electric motor (2).

  INDUSTRIAL APPLICABILITY The present invention is useful for detecting the rotation direction of an electric motor by a configuration that is not affected by noise and does not require a large-capacity storage device.

2 Electric motor
4 Rotation direction detector
21 Armature
21u winding
21v winding
21w winding
22 Field
41 Detector
42 Average part
43 Rotational speed calculator
44 Judgment part
44a Interval average theoretical value calculator
44b Deviation calculator
44c Rotation direction detector
50 storage
51a First storage area (storage area)
51b Second storage area (storage area)
51c Third storage area (storage area)
51d 4th storage area (storage area)
Vu phase potential
Vun Line induced voltage (First line induced voltage)
Vun_mean1 section average calculation value
Vun_mean2 Interval average theoretical value
Vv phase potential
Vw phase potential
Vwn Line induced voltage (second line induced voltage)
Vwn_mean1 section average calculated value
Vwn_mean2 Theoretical value of interval average

Claims (7)

A field magnet (22) including a permanent magnet, and an armature (21) including windings (21u, 21v, 21w) of three or more phases, the field (22) and the armature (21), Is a device (4) for detecting the direction of rotation of the electric motor (2) that rotates relatively,
Among the phase potentials (Vu, Vv, Vw) output from the armature (21) by the induced electromotive force, one of the minimum phase and the maximum phase is set as a reference potential, and the first phase potential (Vu) A first line induced voltage (Vun) that is a potential difference with respect to the reference potential and a second line that is a potential difference between the second phase potential (Vw) other than the first phase potential (Vu) and the reference potential. A detector (41) for detecting whether or not the induced voltage (Vwn) between the two coincides;
In response to the detection result of the detection unit (41), when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other or when they coincide with each other from the vicinity or A section average calculated value (Vun_mean1, Vwm_mean1) that is an average value of at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn) in a predetermined section in a period up to the vicinity. An average part (42) for calculating
The rotation direction of the electric motor (2) is normal rotation with respect to at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn) for which the section average calculated values (Vun_mean1, Vwm_mean1) are calculated. Section average theoretical value (Vun_mean2, Vwn_mean2) which is an average value of at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn) in the predetermined section in the direction and the reverse direction ) To calculate the deviation between the section average theoretical value calculation unit (44a) and the section average calculated value (Vun_mean1, Vwm_mean1) and the section average theoretical value (Vun_mean2, Vwn_mean2) in the forward and reverse directions And a rotation direction determination unit (44c) that determines the rotation direction of the section average theoretical value (Vun_mean2, Vwn_mean2) determined to have a small deviation as the rotation direction of the electric motor (2). Part (44) Rotation direction detection apparatus according to claim and. In claim 1,
The deviation calculating unit (44b) obtains an absolute value of a difference between the section average calculated value (Vun_mean1, Vwn_mean2) and the section average theoretical value (Vun_mean2, Vwn_mean2) for each of the predetermined sections corresponding to each other. The rotational direction detection device is configured to set the sum of absolute values of the differences as the deviation. In claim 1 or 2,
The deviation calculating unit (44b) is configured to detect when the first inter-line induced voltage (Vun) and the second inter-line induced voltage (Vwn) coincide with each other or from the next to the next coincidence or the vicinity thereof. Rotation configured to obtain the deviation between the interval average calculated value (Vun_mean1, Vwm_mean1) and the interval average theoretical value (Vun_mean2, Vwn_mean2) for at least one of the initial period and the final period of the period Direction detection device. In any one of Claims 1-3,
The determination unit (44)
The average value of the detected values in the period from when the first line induced voltage (Vun) and the second line induced voltage (Vwn) coincide with each other or from the vicinity thereof to the next coincidence or the vicinity thereof. Calculate the average value for a certain period,
The section average calculated value (Vun_mean1) in the predetermined section for at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn) used for determining the rotation direction of the electric motor (2). , Vwm_mean1) minus the period average calculated value is used as the new section average calculated value (Vun_mean1 ′, Vwm_mean1 ′),
For at least one of the first line induced voltage (Vun) and the second line induced voltage (Vwn) for which the section average calculated values (Vun_mean1, Vwn_mean1) are calculated, the first line induced voltage (Vun) The first inter-line induced voltage (Vun) and the second inter-line induced voltage (Vun) during the period from when the second inter-line induced voltage (Vwn) coincides with each other or from the vicinity thereof to the next coincidence or the vicinity thereof. Vwn) is obtained by subtracting the average value of at least one of the theoretical values from the interval average theoretical value (Vun_mean2, Vwn_mean2) in the predetermined interval as the new interval average theoretical value (Vun_mean2 ', Vwn_mean2'). It is comprised in the rotation direction detection apparatus characterized by the above-mentioned. In any one of Claims 1-4,
A rotation speed calculation unit (43) that receives the detection result of the detection unit (41) and calculates the rotation speed of the electric motor (2),
When the average line (42) matches the first line induced voltage (Vun) and the second line induced voltage (Vwn) based on the rotation speed calculated by the rotation speed calculation unit (43) Alternatively, the time (T) from the vicinity to the next match or the vicinity thereof is obtained, and the time (T) is divided into a plurality of the sections, and the section average calculated values (Vun_mean1, Vwm_mean1) in the predetermined section A rotation direction detection device configured to calculate In any one of Claims 1-4,
A storage unit (50) having a predetermined number of storage areas (51a to 51d) for storing the detection value of the detection unit (41);
The average part (42)
During the period for obtaining the section average calculated values (Vun_mean1, Vwm_mean1), the detection values are sequentially added to each of the storage areas (51a to 51d) up to a predetermined upper limit number of storage, and all the storage areas (51a to 51d) When the detection value is added up to the storage upper limit number, the storage upper limit number is multiplied by an integer of 2 or more, and the detection upper limit number is multiplied by the integer multiple according to the added order. The detected values are summed into the predetermined storage area (51a, 51b) as a sum and zero is stored in the remaining storage areas (51c, 51d), and the storage area (51c, 51d) in which zero is stored Repeating the operation of adding and storing the detected value up to the upper limit number of storage after an integer multiple for each of
The section average calculated value (Vun_mean1, Vwm_mean1) in the predetermined section is calculated based on the stored value of the storage area (51a to 51d) at the end of the period for calculating the section average calculated value (Vun_mean1, Vwm_mean1). A rotation direction detecting device configured as described above. The rotational direction detection device (4) according to any one of claims 1 to 6,
The electric motor (2);
An air conditioner comprising a fan driven by the electric motor (2).

Images