Title of Invention

SENSORLESS OPERATION OF AN ELECTRONICALLY COMMUTATED DIRECT CURRENT MACHINE

Abstract

The invention relates to a method and an electric drive with an electronically commutated direct current machine (11), an inverter (12) and a control apparatus (16) for sensorless operation of the machine (11), in which the zero crossings of the voltage induced by the rotating field of the rotor in the respective phase not supplied with current are determined, wherein the zero crossings occurring are detected, in that the instant of the next zero crossing of the voltage induced by the rotating field of the rotor is calculated in advance and, a time interval (At) after the instant calculated in advance, at least the last detected zero crossing is evaluated.

Full Text

FORM 2 THE PATENT ACT 1970 (39 of 1970) The Patents Rules, 2003 COMPLETE SPECIFICATION ,See Section 10, and rule 13) TITLE OF INVENTION SENSORLESS OPERATION OF AN ELECTRONICALLY CURRENT MACHINE COMMUTATED DIRECT APPLICANT(S) a) Name b) Nationality c) Address OERLIKON TEXTILE GMBH & CO. GERMAN Company LEVERKUSER STRASSE 65, D-42897 REMSCHEID, GERMANY KG 3. PREAMBLE TO THE DESCRIPTION The following specification particularly describes the invention and the manner in which it is to be performed : - PATENTS ACT 1977 I, ASTRID TERRY, translator of 11, Bounds Oak Way, Tunbridge Wells, Kent. TN4 OUB, England, confirm that I am conversant with the English and German languages and I am a competent translator from one to the other. I declare to the best of my knowledge and belief that the attached English translation is a true and correct translation of the DE Patent No. 10 2007 040 217.3. Dated this 5th day of July 2008. The invention relates to a method for sensorless operation of an electronically commutated direct current machine, in which the zero crossings of the voltage induced by the rotating field of the rotor in the respective phase not supplied with current are determined, and to an electric drive with an electronically commutated direct current machine/ an inverter and a control apparatus, which is configured for sensorless operation of the machine and which determines the zero crossings of the voltage induced by the rotating field of the rotor in the respective phase not supplied with current. Electronically commutated direct current machines are increasingly replacing the conventional direct current machine comprising a mechanical commutator and are being used as drives in the automotive sector, in household apparatuses and industrial applications. The electronically commutated direct current machine is with regard to structure a permanently excited synchronous machine, which, in conjunction with an inverter controlled by a controller or a processor, acquires the properties of a direct current machine. In this case, the phases are fed with block-shaped currents. In one phase of a three-phase motor, a constant current flows over an electric angle of 120°, the phase is then electrically without current for 60° and a constant current follows over 120° with an inverted sign and again a 60° current gap. The three phases are electrically phase-shifted by 120°. In this manner, a rotating magnetic field is produced. Knowledge of the rotor position is required to keep the rotating field of the stator and rotor in an optimal position in relation to one another. The commutation instants are then determined from the rotor position. Sensors, for example Hall sensors can be used to determine the rotor position. However, it is also possible to operate the electronically commutated direct current machine in a sensorless manner. In this case, the position of the rotor is determined indirectly from the electrical variables of the machine. As mentioned above, there are current gaps in each phase. However, a voltage is induced in the stator phase by the rotating field of the rotor even in the current gap. The induced voltage can be used to determine the rotor position or to determine the next commutation instant. To -2- evaluate the induced voltage, its zero crossings are determined. The problem here is that owing to demagnetisation processes and to disturbances caused, for example, by static discharges, zero crossings are produced which are not relevant zero crossings of the induced voltage, because they are not induced by the rotating field of the rotor in the stator winding and therefore may not be used either for the determination of the rotor position. Each zero crossing detected therefore has to be evaluated to establish whether it is a zero crossing that is relevant for the rotor position determination, of the voltage induced by the rotating field of the rotor in the stator winding, or is a disturbance. The evaluation of the zero crossings requires processor time. The possibility now exists of being able to dimension the processor output in such a way that all zero crossings can be evaluated. The processor output is, however, connected to corresponding costs. It is therefore desirable to reduce the number of zero crossings to be evaluated. The application indications of the company STMicroelectronics "AN INTRODUCTION TO SENSORLESS BRUSHLESS DC MOTOR DRIVE APPLICATIONS WITH ST72141" published on the Internet under the address "http://www.st.com/stonline/products/literature/an/7209.pdf" show a possibility for this. Here, after a commutation, the zero crossing is firstly detected at the end of the demagnetisation and then the zero crossing of the induced voltage. Before each detection process, a dead time of 20 microseconds is inserted here to relieve the processor. In other words, no zero crossings are detected for one step for a total of 40 microseconds. This manner of proceeding is completely unsuitable for rotational speeds of significantly more than 100,000 revolutions per minute, such as occur in rotor drives of open-end spinning machines. In the extreme case, the zero crossings of the induced voltage lie within the dead time. -3- It is therefore the object of the present invention, to improve the determination of the zero crossings of the voltage induced by the rotating field of the rotor, in particular at high rotational speeds. The object is achieved according to the invention by the characterising, features of the method claim 1 and the device claim 13. Advantageous developments of the invention are the subject of the sub-claims. To achieve the object, the zero crossings occurring are detected, the instant of the next zero crossing of the voltage induced by the rotating field of the rotor is calculated in advance and, a time interval after the instant calculated in advance, at least the last detected zero crossing is evaluated. By calculating the evaluation instant in advance, the evaluation of the last detected zero crossing is already sufficient. However, to improve the method a plurality of zero crossings detected directly before the evaluation instant can be evaluated. The observation window in which the zero crossings are evaluated can be automatically adapted to the rotational speed of the machine by the method according to the invention. In the case of high rotational speeds, in other words with small step times, it is thus achieved that only a limited number of zero crossings is evaluated and nevertheless there is no danger of the relevant zero crossing of the voltage induced by the rotating field of the rotor lying within a dead time. In the case of low rotational speeds and therefore larger step times, the observation window is displaced with the evaluation instant calculated in advance and the number of zero crossings evaluated can thus likewise be reduced compared to the prior art. The evaluation instant in this case is a time interval after the instant calculated in advance of the next zero crossing of the voltage induced by the rotating field of the rotor to ensure that the actual zero crossing, which may also lie after the instant calculated in advance, is reliably detected. -4- According to a preferred embodiment of the method according to the invention/ a step time is calculated, to evaluate a detected zero crossing, from the time difference between the detected zero crossing and the last determined zero crossing of the voltage induced by the rotating field of the rotor. For evaluation, the step time calculated from a detected zero crossing can then be compared with a minimum step time. It is, in this case, of course of equal value to compare the instant of the detected zero crossing with a lower limit of the time. In this case, the lower limit is then precisely the minimum step time after the last instant of the zero crossing of the voltage induced by the rotating field of the rotor. The first variant only has the advantage that no absolute time measurement is required and the method can be implemented with simple timers. As already mentioned above, the evaluation of the zero crossings is carried out a certain time interval after the instant of the zero crossing calculated in advance of the voltage induced by the rotating field of the rotor. Advantageously, the time interval, after the evaluation is carried out, and the minimum step time or the lower time limit described above, are determined as a function of the rotational speed or also the rotational speed change. This makes possible an optimal adaptation to the respective operating states of the electronically commutated direct current machine. According to a preferred embodiment of the method according to the invention, the instant of the zero crossing of the voltage induced by the rotating field of the rotor is equated with the instant of the last detected zero crossing if the step time determined from the last detected zero crossing is greater than the minimum step time. If the condition is fulfilled, the detected zero crossing is between a lower time limit and an upper time limit, which is determined by the evaluation instant. This provides adequate plausibility that the last detected zero crossing is also the actual zero crossing of the voltage induced by the rotating field of the rotor. In a development of the invention, two or more zero crossings detected can also be evaluated, the instant of the zero crossing of the voltage induced by the rotating field -5- of the rotor then being determined from the mean value of the step times of the zero crossings evaluated, the step times of which are greater than the minimum step time. According to the invention, the zero crossings are firstly detected and collectively evaluated after the instant calculated in advance. This additionally brings about a time advantage in comparison to the direct evaluation of the detected zero crossings. The instant of the zero crossing of the voltage induced by the rotating field of the rotor is equated with the instant calculated in advance if the step times of the evaluated zero crossings are less than the minimum step time. A current step time is advantageously calculated from the two last determined zero crossings of the voltage induced by the rotating field of the rotor. For this purpose, the two instants simply have to be subtracted from one another. If the instant of the zero crossing of the voltage induced by the rotating field of the rotor was determined by equating with the zero crossing calculated in advance, the current step time can also simply be taken over from the last step. In an advantageous development, a current step time is calculated by forming the mean value from a plurality of last determined zero crossings of the voltage induced by the rotating field of the rotor. According to a development of the invention, the sampling frequency of the induced voltage is adapted as a function of the current step time. The sampling frequency can thus be adapted to a changing rotational speed. In the case of large step times, the sampling frequency can be correspondingly reduced, so processor output can be saved. In the case of highy rotational speeds, in other words with small step times, the sampling frequency has to be correspondingly increased to obtain adequate precision. Advantageously, the next commutation instant is calculated from the current step time. Furthermore, the actual rotational speed of the machine, for example for a -6- superordinate rotational speed control loop can be calculated from the current step time. The current step time can then also be used in the next step to calculate the zero crossing of the voltage induced by the rotating field of the rotor in advance. To achieve the object, an electric drive with an electronically commutated direct current machine, an inverter and a control apparatus are furthermore proposed, with the control apparatus detecting the zero crossings occurring, calculating in advance the instant of the next zero crossing of the voltage induced by the rotating field of the rotor and, a time interval after the instant calculated in advance, evaluating at least the last detected zero crossing. The control apparatus may have a micro controller or a comparable type of processor, in which the evaluation routines and the control and regulating algorithms are implemented as software. A component of the microcontroller is a timer, which triggers the evaluation at least of the last detected zero crossing by an interrupt. In the device according to the invention, only one interrupt is therefore triggered per step interval. This produces a time interval and therefore lower loading of the microcontroller in comparison to the triggering of an interrupt at each detected zero crossing, in order to evaluate it directly. To detect the zero crossings, the control apparatus may have a comparator. The comparator is also in a position to recognise whether the voltage changes from a negative to a positive value or vice versa. Accordingly, the comparator signal exhibits a positive or negative edge. According to an embodiment of the device according to the invention, the microcontroller, to evaluate a detected zero crossing, calculates a step time from the time difference between the detected zero crossing and the last determined zero crossing of the voltage induced by the rotating field of the rotor. In this case, the -7- microcontroller, for evaluation, may compare the step time calculated from a detected zero crossing with a minimum step time. According to a development of the invention, the microcontroller determines the time interval, after which the evaluation is carried out, and the minimum step time as a function of the rotational speed or the rotational speed change or as a function of the two variables. According to an embodiment, the microcontroller equates the instant of the zero crossing of the voltage induced by the rotating field of the rotor with the instant of the last detected zero crossing if the step time determined from the last detected zero crossing is greater than the minimum step time. According to a further development, the microcontroller determines the instant of the zero crossing of the voltage induced by the rotating field of the rotor from the mean value of the step times of at least two evaluated zero crossings, the step times of which are greater than the minimum step time. The microcontroller may advantageously equate the instant of the zero crossing of the voltage induced by the rotating field of the rotor with the instant calculated in advance if the step times of the evaluated zero crossings are shorter than the minimum step time. According to an advantageous embodiment of the device according to the invention, the microcontroller calculates a current step time from the two last determined zero crossings of the voltage induced by the rotating field of the rotor. According to a development, the microcontroller calculates a current step time by forming the mean value from a plurality of last determined zero crossing of the voltage induced by the rotating field of the rotor. -8- Advantageously, the microcontroller can adapt the sampling frequency of the induced voltage as a function of the current step time. The microcontroller can calculate the next commutation instant and the rotational speed of the machine from the current step time. The invention will be described in more detail below with the aid of an embodiment shown in the drawings, in which: Fig. 1 shows the schematic structure of an electric drive according to the invention; Fig. 2 shows the schematic view of an inverter used; Fig. 3 shows the time course of the phase voltages, the comparator signals and the switching states of the inverter; Fig. 4 shows an enlarged detail of the comparator signal from Fig. 3; Fig. 5 shows a simplified view of the comparator signal with the zero crossings of the induced voltage and the instant of the interrupt; Fig. 6 shows a flow chart with the sequence of the commutation. Fig. 1 shows the schematic view of an electric drive according to the invention with an electronically commutated direct current machine 11. The machine 11 consists of a three-phase current winding in the stator and a permanently excited rotor. The connections of the three phases A, B and C are connected to an inverter 12. The inverter is connected to a direct voltage source. The direct voltage source is symbolised in Fig. 1 by the capacitor 15 and connected to the terminals 13 and 14 of the inverter. The inverter is activated by a control unit 16, the inverter and the control unit being connected to one another by means of a control line 17. The -9- measurement lines 18 connect the control apparatus 16 to the three phases A, B and C to measure the induced voltage. Fig. 2 shows the structure in principle of the inverter 12 with the direct voltage connections 13 and 14 and the phase connections A, B and C. The inverter is a three-phase bridge circuit with the switches TO to T5. The switches are semiconductor switches, for example configured as transistors with anti-parallel diodes. In Fig. 3, the curves 21, 22 and 23 show the course in principle of the phase voltage A, B and C. The curves 31, 32 and 33 in each case show the output of a comparator, which detects the zero crossings of the phase voltages A, B or C. Moreover, the switching states of the switches TO to T5 are shown. The steps are plotted on the abscissa, one step corresponding to an electric angle of 60°. In the electronically commutated direct current machine, two phases are always provided with current and a third is without current. After each step, the current commutates from a phase provided with current to a phase without current. In the first step, the phases A and B are provided with current and the voltage induced in phase C is evaluated to determine the rotor position. As can be seen in curve 33, the comparator detects with the negative edge the zero crossing of the induced voltage. As can be seen from the graph, the switches TO and T3 are switched on during the first step. The other switches are switched off. When observing the switching states in Fig. 3, it can be seen that the signals for the switches TO, T2 and T4 are shown on one side and Tl, T3 and T4 on the other side in an inverse logic. A low level for the upper switches of the half-bridge (TO, T2, T4) means that the switch is switched on, and a high level means that the switch is switched off. It is the reverse for the lower switches of the half-bridge (Tl, T3, T5). From the first to the second step, the current changes from phase B to phase C. For this purpose, the switch T3 is switched off and T5 is switched on. In this case, phase C is placed at the potential of the connection point 14. The current through phase B cannot be abruptly interrupted with the opening of the switch T3 and flows on via -10- the anti-parallel diode of the switch T2 so until the current decays, phase B briefly takes on the potential of the connection point 13. In this case, the magnetisation energy stored in the phase winding is reduced- Therefore, demagnetisation is also referred to here. The further switching states and voltage courses can be seen in the drawing. The marking 34 identifies a region of the curve 33, which represents the comparator signal associated with phase C. The region in the marking 34 is shown enlarged in Fig. 4. As can be seen, the comparator with the commutation K detects a zero crossing of the voltage. A further zero crossing is detected at the end of the demagnetisation E. The zero crossing of the voltage Z induced by the rotating field of the rotor, relevant for determining the rotor position then follows in the drawing. The results are repeated correspondingly. In practice, the comparator detects still further zero crossings caused by disruption, not shown here. In Fig. 5, the comparator signal is only shown with the relevant zero crossings of the voltage Z induced by the rotating field of the rotor and specifically regardless of in which phase the zero crossing occurs. The zero crossings occur at the instants t1, t2 and t3. The difference between the instants t2 and ft is precisely the step time T. The next zero crossing to be expected can therefore be calculated in advance from the addition of the step time T and the time t2. An estimated value for the instant t3 is therefore obtained. By adding the rotational speed-dependent time interval Δt, the instant t4 is obtained, which is shortly after the instant t3. At the instant t4, an interrupt is then triggered by means of a timer and the zero crossing last detected by the comparator or a plurality of zero crossings detected directly before the interrupt are evaluated. From this, the actual instant t3 can be established at which the voltage induced by the rotating field of the rotor has a zero crossing. In this case, a step time is calculated for the detected zero crossings to be evaluated by subtracting the instant of the zero crossing and the last determined zero crossing of the voltage induced by the rotating field of the rotor. In this case, in the embodiment, the instant t2 is the instant of the last determined zero crossing of the voltage induced by the -11- rotating field of the rotor. The detected zero crossings to be evaluated are not shown for reasons of clarity. The step times thus determined are then compared with a minimum step time Tmin. Fig. 6 shows a flow chart of a possible commutation process. In the embodiment shown, only the last detected zero crossing is evaluated at the evaluation instant or with the interrupt. With the triggering of the interrupt, the programme implemented in the control apparatus jumps into the terminator 41. The last detected zero crossing is fetched from the memory in the statement block 42. In the statement block 43, the current step time is then calculated from the last detected zero crossing and the last determined zero crossing of the voltage induced by the rotating field of the rotor. In the branch 44, the current step time is compared with the minimum step time Tmin. If the current step time is smaller or equal to the minimum step time, the current step time is replaced in the statement block 45 by a step time determined earlier or recalculated from zero crossings determined earlier. The sequence is then continued in the statement block 46. If the current step time is greater than the minimum step time, the operation 46, which calculates the next commutation instant, is directly triggered. It is queried in the branch 47 whether the clock frequency of the timer, with which the comparator signal is sampled, has to be adapted to a higher rotational speed. If necessary, the adaptation is carried out in the operation block 48. In the statement block 49, the interrupt instant for the next commutation, or to detect the next zero crossing of the induced voltage, is determined. In the next step 50, the minimum step time Tmin and the time interval At are recalculated for the next commutation to carry out an adaptation to the rotational speed conditions. In the block 51, the rotational speed is then calculated from the last step times. The interrupt routine is ended with the terminator 52, and the superordinate programme is continued. Alternatively, the step times determined from the zero crossings detected, which are greater than the minimum step time, can be averaged to determine or establish the instant t3 at which the voltage induced by the rotating field of the rotor has a zero -12- crossing. Furthermore, the step times can be averaged over a plurality of steps to compensate lack of symmetry. In this case, averaging is preferably over six steps per phase spacing, corresponding to a revolution of the rotor. -13- WE CLAIM: 1. Method for sensorless operation of an electronically commutated direct current machine (11), in which the zero crossings of the voltage induced by the rotating field of the rotor in the respective phase not supplied with current are determined, characterised in that the zero crossings occurring are detected, in that the instant of the next zero crossing of the voltage induced by the rotating field of the rotor is calculated in advance and in that, a time interval (At) after the instant calculated in advance, at least the last detected zero crossing is evaluated. 2. Method according to claim 1, characterised in that to evaluate a detected zero crossing, a step time is calculated from the time difference between the detected zero crossing and' the fast determined zero crossing of the voltage induced by the rotating field of the rotor. 3. Method according to claim 2, characterised in that the step time calculated from a detected zero crossing is compared with a minimum step time (Tmin) for evaluation. 4. Method according to claim 3, characterised in that the time interval (At), after which the evaluation is carried out, and the minimum step time (Tmin) are determined as a function of the rotational speed and/or the rotational speed change. 5. Method according to claim 4, characterised in that the instant of the zero crossing of the voltage induced by the rotating field of the rotor is equated with the instant of the last detected zero crossing if the step time determined from the last detected zero crossing is greater than the minimum step time. -14- 6. Method according to claim 4, characterised in that the instant of the zero crossing of the voltage induced by the rotating field of the rotor is determined from the mean value of the step times of at least two evaluated zero crossings, the step times of which are greater than the minimum step time. 7. Method according to claim 4, characterised in that the instant of the zero crossing of the voltage induced by the rotating field of the rotor is equated with the instant calculated in advance if the step times of the evaluated zero crossings are smaller than the minimum step time. 8. Method according to any one of claims 5 to 7, characterised in that a current step time is calculated from the two last determined zero crossings of the voltage, induced by the rotating field of the rotor. 9. Method according to any one of claims 5 to 7, characterised in that a current step time is calculated by forming the mean value of a plurality of last determined zero crossings of the voltage induced by the rotating field of the rotor. 10. Method according to either of claims 8 or 9, characterised in that the sampling frequency of the induced voltage is adapted as a function of the current step time. 11. Method according to either of claims 8 or 9, characterised in that the next commutation instant is calculated from the current step time. 12. Method according to either of claims 8 or 9, characterised in that the rotational speed of the machine is calculated from the current step time. 13. Electric drive with an electronically commuted direct current machine (11), an inverter (12) and a control apparatus (16), which is configured to carry out1 -15- the method for sensorless operation of the machine (11) according to any one of claims 1 to -12 and which determines the zero crossings of the voltage induced by the rotating field of the rotor in the respective phase not supplied with current, characterised in that the control apparatus (16) is configured to detect the zero crossings occurring, to calculate in advance the instant of the next zero crossing of the voltage induced by the rotating field of the rotor and, a time interval (At) after the instant calculated in advance, to evaluate at least the last detected zero crossing. 14. Device according to claim 13, characterised in that the control apparatus (16) has a microcontroller. 15. Device according to claim 14, characterised in that the microcontroller has a timer, which triggers the evaluation at feast of the last detected zero crossing by an interrupt. -16- Dated this 4th day of August, 2008

Documents:

14925Assignment.pdf

14925Form-6.pdf

14925Relevant Documents.pdf

1656-MUM-2008-ABSTRACT(11-9-2013).pdf

1656-MUM-2008-ABSTRACT(5-11-2014).pdf

1656-mum-2008-abstract.doc

1656-mum-2008-abstract.pdf

1656-MUM-2008-CANCELLED PAGE(11-9-2013).pdf

1656-MUM-2008-CANCELLED PAGE(12-11-2012).pdf

1656-MUM-2008-CLAIMS(AMENDED)-(11-9-2013).pdf

1656-MUM-2008-CLAIMS(AMENDED)-(5-11-2014).pdf

1656-mum-2008-claims.doc

1656-mum-2008-claims.pdf

1656-MUM-2008-CORRESPONDENCE(01-09-2008).pdf

1656-MUM-2008-CORRESPONDENCE(12-11-2012).pdf

1656-MUM-2008-CORRESPONDENCE(7-5-2014).pdf

1656-mum-2008-correspondence.pdf

1656-mum-2008-description(complete).doc

1656-mum-2008-description(complete).pdf

1656-mum-2008-drawing.pdf

1656-MUM-2008-FORM 1(01-09-2008).pdf

1656-MUM-2008-FORM 1(7-5-2014).pdf

1656-mum-2008-form 1.pdf

1656-mum-2008-form 18.pdf

1656-MUM-2008-FORM 2(TITLE PAGE)-(7-5-2014).pdf

1656-mum-2008-form 2(tittle page).pdf

1656-mum-2008-form 2.doc

1656-mum-2008-form 2.pdf

1656-MUM-2008-FORM 26(5-11-2014).pdf

1656-MUM-2008-FORM 3(11-9-2013).pdf

1656-MUM-2008-FORM 3(12-11-2012).pdf

1656-MUM-2008-FORM 3(7-5-2014).pdf

1656-mum-2008-form 3.pdf

1656-MUM-2008-FORM 5(7-5-2014).pdf

1656-mum-2008-form 5.pdf

1656-MUM-2008-GENERAL POWER OF ATTORNEY(7-5-2014).pdf

1656-mum-2008-general power of attorney.pdf

1656-MUM-2008-MARKED COPY(5-11-2014).pdf

1656-MUM-2008-OTHER DOCUMENT(5-11-2014).pdf

1656-MUM-2008-PETITION UNDER RULE-137(12-11-2012).pdf

1656-MUM-2008-REPLY TO EXAMINATION REPORT(11-9-2013).pdf

1656-MUM-2008-REPLY TO HEARING(5-11-2014).pdf

1656-MUM-2008-SPECIFICATION(AMENDED)-(11-9-2013).pdf

1656-MUM-2008-SPECIFICATION(AMENDED)-(5-11-2014).pdf

abstract1.jpg