Speed control method and device for a two-pulse motor operating with auxiliary torque

The invention relates to a method for controlling the speed of a working with an invariant auxiliary torque two-pulse motor, and relates to a two-pulse electronically commutated motor, which operates with an invariant auxiliary torque and has a control device for controlling its speed. As an invariant auxiliary torque usually a so-called reluctance torque is used, but the invention is not limited thereto, but such an auxiliary torque could also be generated mechanically, eg through a cam curve.

Two-pulse electronically commutated reluctance torque motors are inexpensive because they can be manufactured more cheaply compared to multiphase motors of the same size and because this type of motor is simple and robust. Such motors are therefore often used to drive fans, eg for cooling in the IT sector.

A method for controlling the speed of two-pulse electronically commutated motors is known from US 4,376,261.

Due to their auxiliary torque, such motors have a direction of rotation in which they start easily, because the startup is assisted by the auxiliary torque. This direction of rotation is also referred to as preferred direction. In contrast, in the direction opposite to the preferred direction, also called reverse direction, the start-up is difficult. A four-quadrant operation of such engines also encounters difficulties because there are speed ranges where regulation is not possible for a variety of reasons, among others because, due to the auxiliary torque and its interaction with the electromagnetic drive torque, low speeds are not controllable in such an engine. The reason for this is that the size of the auxiliary torque can not be influenced by the speed controller and is quasi a fixed size.

Due to this lower speed range, in which a control is not possible, arises around the speed zero around a gap in the operating range. This results in particular difficulties when the direction of rotation is to be reversed, for example, if you want to switch from a speed of 3000 U / min in the preferred direction to 3000 rev / min in the reverse direction.

It is therefore an object of the invention to provide a new method and a new engine of the type mentioned.

According to the invention, this object is achieved by the subject matter of claim 1. Characterized in that the motor is driven below a predetermined minimum speed with independent of the desired speed current, the startup can be done in a speed window, which is defined by the minimum speed, with a not influenced by the controller high torque. Upon exiting this speed window, the motor will enter controlled operation to achieve the desired speed ("desired speed"), and this will be done by adjusting the control value as a function of the difference between desired speed and actual speed , So a speed control takes place, preferably by a PI controller.

In a very advantageous embodiment of the invention, the control value (and thus the motor current supplied) as a function of the difference between the desired speed and actual speed continuously calculated, but used for the control process only at or after reaching the minimum speed. As a result, an already calculated control value for the speed control is immediately available after leaving the mentioned speed window.

Another solution of the problem is the subject of claim 10. Such a motor has a speed window, which it goes through during startup without speed control, and only after leaving this window is the speed control. It follows at the same time that within the window speed control is not possible, which is why a desired speed, which is within this window, is interpreted as speed zero, ie as a speed that can not be displayed.

Further details and advantageous developments of the invention will become apparent from the described below and illustrated in the drawings, in no way as a limitation of the invention to be understood embodiments, and from the other dependent claims. It shows:<dl id="dl0001"><dt>Fig. 1</dt><dd>3 shows an overview circuit diagram of a preferred embodiment of a bridge circuit for a low-power, two-pulse ECM.</dd><dt>Fig. 2</dt><dd>a circuit diagram with the terminal designations of a microcontroller used in the embodiment 76,</dd><dt>Fig. 3</dt><dd>a circuit diagram of an auxiliary circuit which is connected to the input RAO of the microcontroller 76; this input RAO is the input of an A / D converter,</dd><dt>Fig. 4</dt><dd>a circuit diagram of the wiring of the input RB1, alternatively the input RB7, the microcontroller 76,</dd><dt>Fig. 5</dt><dd>a circuit diagram showing how a frequency signal is derived from the output signals of a Hall generator 6 and the input RB0 of the microcontroller 76 is supplied, where this signal is converted into Hall interrupts,</dd><dt>Fig. 6</dt><dd>an overall diagram of the hardware used; here the same reference numerals are used as in the preceding figures 1 to 5,</dd><dt>Fig. 7</dt><dd>a schematic representation of a 1-quadrant controller (1-Q controller),</dd><dt>Figure 8</dt><dd>a schematic diagram of a 2-Q controller,</dd><dt>Fig. 9</dt><dd>a schematic representation of a 4-Q controller,</dd><dt>Fig. 10</dt><dd>a representation of the course of internal signals at a transition from +3000 to -3000 rpm,</dd><dt>Fig. 11</dt><dd>a state diagram showing the change of states in different operating states,</dd><dt>Fig. 12</dt><dd>a schematic representation of the start-up process in the reverse direction,</dd><dt>Fig. 13</dt><dd>a schematic representation of a typical course of a reluctance torque,</dd><dt>Fig. 14</dt><dd>a flowchart of a routine for the comparison between the actual direction of rotation and a direction of rotation set by the user,</dd><dt>Fig. 15</dt><dd>a flow chart of a routine for the determination of the engine status,</dd><dt>FIGS. 16 &amp; 17</dt><dd>a flowchart for the discrimination between different operating states of an engine,</dd><dt>Fig. 18</dt><dd>a schematic representation of the sizes used in the commutation of the engine and their use, and</dd><dt>Fig. 19</dt><dd>a flowchart RGL, which shows calculations in the speed control of a motor.</dd></dl>

In the following description, the same reference numerals are used for the same or like parts or functions, and these are usually described only once. In the description of speeds and their changes, we follow the language, ie when the rotor rotates faster, no matter in which direction, we speak of an increase in the speed. The following abbreviations are used:<dl id="dl0002" compact="compact"><dt>Y</dt><dd>YES</dd><dt>N</dt><dd>NO</dd><dt>n</dt><dd>actual speed, actual speed; is measured as time T_HALL</dd><dt>n_should</dt><dd>desired speed, setpoint speed; is specified as time t_s</dd><dt>BW</dt><dd>Block length of a current pulse flowing in the motor 30, cf. Fig. 18. This block length represents the so-called "control value" of the speed controller and is calculated with the flowchart of FIG. 19. For BW = Max, BW is set to a value that is slightly smaller than T_HALL, for example.</dd><dt>TO YOU</dt><dd>Direction command for direction of rotation (specified from outside)</dd><dt>TO YOU_is</dt><dd>current value of the direction of rotation</dd><dt>DlR_should</dt><dd>desired value of the direction of rotation</dd><dt>inv. 180 °</dt><dd>Switching the commutation to the opposite of the current commutation. In Fig. 18 comes within the scope of 360 ° el. is designated, first a current pulse 444 (current i ') and then a current pulse 446 (current i). After switching through "inv. 180 ° "comes first a pulse 446 and then a pulse 444, so that the motor is driven in the reverse direction. By switching back to "inv. 180 ° ", the state of FIG. 18 restored. See. the equations (5), (6), (11), (12) and the explanations for FIG. 19th</dd></dl>

In the embodiment mean:<dl id="dl0003" compact="compact"><dt>DIR = 0</dt><dd>Command: Desired run in preferred direction</dd><dt>DIR = 1</dt><dd>Command: Desired run against the preferred direction, ie in the reverse direction</dd><dt>Rt</dt><dd>Direction flag. The value of Rt is not saved until certain conditions are met and may therefore differ from DIR.</dd><dt>Rt = 0</dt><dd>preferred direction</dd><dt>Rt = 1</dt><dd>Reverse direction.</dd><dt>Rs</dt><dd>Flag that indicates whether the engine should run unregulated or regulated.</dd><dt>Rs = 0</dt><dd>Unregulated run in the desired direction of rotation with max. electricity</dd><dt>Rs = 1</dt><dd>Run in the desired direction with controlled speed</dd><dt>N_MIN</dt><dd>Minimum speed. With increasing speed lnl below | N_MIN| The motor runs unregulated with maximum current, and at speeds above it with speed control.</dd><dt>t_HALL</dt><dd>Time the rotor takes to travel through a given angle of rotation defined by two changes in a rotor position signal HALL. In the exemplary embodiment, T_HALL corresponds to the time for a quarter of a full rotor revolution. T_HALL is used as a measure for the current speed n, cf. equations (9) and (10).</dd><dt>t_s</dt><dd>Desired time, which is given to the speed controller (Fig. 19) as a desired value for the speed, and the time for a quarter of a full rotor revolution at n_should equivalent.</dd><dt>RGL_DIFF</dt><dd>Control difference. (Difference between T_HALL and t_s, see S654 in Fig. 19.)</dd><dt>RGL_P</dt><dd>Proportional factor of the speed controller</dd><dt>RGL_I</dt><dd>Integral factor of the speed controller, cf. S664 in FIG. 19.</dd></dl>

For the manipulated value BW it should be noted that this determines the length of a current pulse (i or i '), this current pulse with the length BW in turn can be composed of shorter current pulses by the PWM control, and by the operations when switching off the current pulse something is longer than BW.

The hardware

<b>Fig. 1</b> 1 shows symbolically the basic structure of a two-pulse electronically commutated motor 30, which in this preferred embodiment is shown with a full-bridge circuit 31 and therefore has only a single winding strand 32 with two terminals 34, 36, and also a permanent-magnetic rotor 38 which is shown as a four-pole rotor. When the rotor 38 rotates, the winding strand 32 is alternately traversed by current pulses i from left to right and current pulses i 'from right to left. Such a motor is called "Zweipulsig" because during a rotor rotation of 360 ° el. a pulse i and a pulse i 'flow, ie two pulses, cf. FIG. 18th It is also possible to provide a first winding strand for the pulses i and a second strand for the pulses i ', as is known to the person skilled in the art. A two-pulse motor can therefore be single or double-stranded.

For the intermediate time between two pulses i, i '(or i' and i), in which no current flows, the motor needs to its drive an auxiliary torque, which is called reluctance torque M_R designated, cf. Fig. 13, because it is caused by the interaction of the permanent magnetic rotor 38 with provided on the stator iron masses, for example with the laminated core (lamination stack) of the stator. Such engines have been manufactured in quantities of hundreds of millions, so it does not seem necessary to describe them in all their many variants.

The motor 30 is powered from a DC mains with a plus line 40 (eg + 24V) and a minus line 42 (GND). Furthermore, a voltage source with a regulated auxiliary voltage of eg +5 V is provided, the output of which is designated 43. This will be described below with reference to FIG.

The winding strand 32 four flywheel diodes 44, 46, 48, 50 are assigned. The cathodes of the diodes 44, 46 are connected to the positive line 40, the anodes of the diodes 48, 50 to the negative line 42. The anode of the diode 44 and the cathode of the diode 48 are connected to the terminal 34. The anode of the diode 46 and the cathode of the diode 50 are connected to the terminal 36.

For controlling the currents i, i ', two power op amps (OPV) 52, 54 are used in this example. In this preferred embodiment, they are connected as comparators and inverting amplifiers, respectively. The output of the OPV 52 is connected to the terminal 34, the output of the OPV 54 to the terminal 36. Naturally, the bridge circuit can also be implemented with other semiconductor elements.

From the terminal 43 (+5 V), a resistor 56 leads to the positive input 5 of the OPV 52, and from there a resistor 60 leads to the negative line 42.

Likewise, from the terminal 43 (+ 5 V), a resistor 62 leads to the positive input 64 of the OPV 54, and from there a resistor 66 leads to the negative line 42.

Preferred values ​​of the components in FIG. 1

<tables id="tabl0001" num="0001"><table frame="none"><tgroup cols="2" colsep="0" rowsep="0"><colspec colnum="1" colname="col1" colwidth="44mm" colsep="0" /><colspec colnum="2" colname="col2" colwidth="39mm" colsep="0" /><tbody><row><entry namest="col1" nameend="col1" align="left" valign="top">Resistors 56, 60, 62, 66</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 22 kOhm</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">Operational amplifier 52, 54</entry><entry namest="col2" nameend="col2" align="left" valign="top">... L2720, Manufacturer: SGS</entry></row></tbody></tgroup></table></tables>

These operational amplifiers preferably include a built-in current limiter to prevent overloading.

Since all resistances are equal, one obtains a potential of + 2.5 V at the plus inputs of both OPVs 52, 54.

The negative input 70 of the OPV 52 is supplied with a control signal which is either greater or less than 2.5 V, and the same applies to the negative input 72 of the OPV 54.

To drive either a potential of&gt; 2.5 V is applied to the input 70, so that the point 34 is connected to the negative line 42, and then to the input 72, a potential of &lt;2.5 V is applied, so be Output is connected to the positive line 40, so here with +24 V. In this case, a current i 'flows from the positive lead 40 via the OPV 54, the terminals 36, 34 and the OPV 52 to the negative lead 42. In the opposite case, the input 70 is at a potential &lt;2.5 V and the input 72 at a potential&gt; 2.5 V, so that a current i flows from the terminal 34 to the terminal 36.

To illustrate, dashed lines are shown in Fig. 1 switch. In the OPV 52 these are an upper switch HSL and a lower switch LSL. In the OPV 54, these are an upper switch HSR and a lower switch LSR. It means<tables id="tabl0002" num="0002"><table frame="none"><tgroup cols="2" colsep="0" rowsep="0"><colspec colnum="1" colname="col1" colwidth="12mm" colsep="0" /><colspec colnum="2" colname="col2" colwidth="27mm" colsep="0" /><tbody><row><entry namest="col1" nameend="col1" align="left" valign="top">HSL</entry><entry namest="col2" nameend="col2" align="left" valign="top">High Side Left</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">HSR</entry><entry namest="col2" nameend="col2" align="left" valign="top">High Side Right</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">LSL</entry><entry namest="col2" nameend="col2" align="left" valign="top">Low Side Left</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">LSR</entry><entry namest="col2" nameend="col2" align="left" valign="top">Low Side Right</entry></row></tbody></tgroup></table></tables>

When HSL and LSR are turned on, a current i flows, and when HSR and LSL are turned on, a current i 'flows.

<b>Fig. 2</b> shows the connection diagram of a microcontroller 76 of the type PIC16C711 used in the present embodiment. The connections are designated in the usual way with numbers and with letters. As shown in FIG. 6 As can be seen, the terminal 70 of the OPV 52 is connected through a resistor 78 to the output RB3 of the μC 76, and the terminal 72 via a resistor 80 to the output RB2, ie the full bridge circuit 31 of FIG. 1 is controlled directly by the μC 76 and its control program. The latter will be described below under "Software". The resistors 78 and 80 can eg each 10 kOhm have.

<b>Fig. 3</b> shows the generation of the signal for the desired speed (target speed n_should). The μC 76 has a multi-channel 8-bit A / D conversion, ie It converts a supplied analogue signal into an 8-bit digital signal. From this A / D conversion, a channel is used at the input RAO. Since it is common in the industry to use such signals either as DC voltage with a voltage range of 0 ... 10 V, or as a frequency signal with a duty cycle of 0 to 100%, a circuit according to FIG. 3 used. This has a connection 84 for the externally supplied signal, ie either a DC voltage U = or a PWM signal PWM. From the terminal 84, a resistor 86 leads to a node 88, and from there a resistor 90 to the negative line 42.

From node 88, a resistor 92 leads to terminal RAO ​​of μC 76, and from RAO to negative lead 42, a capacitor 94. In this way, the voltage or PWM signal at input 84 is converted to a voltage range which is digitized at input RAO can.

Preferred values ​​for FIG. 3

<tables id="tabl0003" num="0003"><table frame="none"><tgroup cols="2" colsep="0" rowsep="0"><colspec colnum="1" colname="col1" colwidth="32mm" colsep="0" /><colspec colnum="2" colname="col2" colwidth="23mm" colsep="0" /><tbody><row><entry namest="col1" nameend="col1" align="left" valign="top">Resistors 86, 90</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 2,7 kOhm</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">Resistor 92</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 680 kOhm</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">Capacitor 94</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 68 nF</entry></row></tbody></tgroup></table></tables>

<b>Fig. 4</b> shows the generation of a signal for the desired direction of rotation which is fed to the input RB1 of the μC 76. From an input 98 for an externally supplied logic level (control signal) DIR (either DIR = 0 or DIR = 1), a resistor 100 leads to input RB1, and from there a capacitor 102 to the negative lead 42.

Preferred values:

<tables id="tabl0004" num="0004"><table frame="none"><tgroup cols="2" colsep="0" rowsep="0"><colspec colnum="1" colname="col1" colwidth="29mm" colsep="0" /><colspec colnum="2" colname="col2" colwidth="21mm" colsep="0" /><tbody><row><entry namest="col1" nameend="col1" align="left" valign="top">Resistance 100</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 2,7 kOhm</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">Capacitor 102</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 10 nF</entry></row></tbody></tgroup></table></tables>

If the desired speed is to be supplied to an input 98 'in the form of a frequency, the circuit according to FIG. 4 is also used, but with different values, for which reason the corresponding components are denoted by reference numbers in parentheses. The resistor 100 'here also preferably has 2.7 kOhm, the capacitor 102' is 4.7 nF, and the signal at the output is supplied to the input RB7 of the μC 76.

<b>Fig. 5</b> shows a circuit for detecting a frequency value which is proportional to the rotational speed of the rotor 38.

For this purpose, a Hall generator 106 is provided in the region of the rotor 38, as shown in FIG. 1 symbolically indicated. Its current inputs are connected via a resistor 108 to the terminal 43 (+5 V) and via a resistor 110 to the negative line 42. Its output signals are fed to the negative input 112 and the positive input 114 of an OPV 116, whose output 118 is connected via a feedback resistor 120 to the negative input 112 and directly to the input RB0 of the μC 76. Input RB0 has an interrupt on change function, ie When the signal at the Hall generator 106 rises rapidly or drops rapidly, a so-called Hall interrupt is generated in the μC 76. This is used to control the commutation of the motor 30 and to detect its speed.

Preferred values ​​in FIG. 5:

<tables id="tabl0005" num="0005"><table frame="none"><tgroup cols="2" colsep="0" rowsep="0"><colspec colnum="1" colname="col1" colwidth="35mm" colsep="0" /><colspec colnum="2" colname="col2" colwidth="25mm" colsep="0" /><tbody><row><entry namest="col1" nameend="col1" align="left" valign="top">Resistors 108, 110</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 470 ohms</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">Resistance 120</entry><entry namest="col2" nameend="col2" align="left" valign="top">... 100 kOhm</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">Hall generator 106</entry><entry namest="col2" nameend="col2" align="left" valign="top">... HW301A FU</entry></row><row><entry namest="col1" nameend="col1" align="left" valign="top">OPV 116</entry><entry namest="col2" nameend="col2" align="left" valign="top">... LM29030</entry></row></tbody></tgroup></table></tables>

<b>Fig. 6</b> shows a preferred circuit according to FIGS. 1 to 5 in the synopsis. For identical or equivalent parts, the same reference numerals are used in all figures, and these parts are usually described only once.

The positive line 40 is connected via a diode 128 to a terminal 130, which may be connected to a battery 132, for example. The diode 128 prevents the circuit from being broken when the battery 132 is connected with the wrong polarity. A capacitor 134 (eg 10 μF) is connected between the lines 40 and 42 and serves to absorb energy which is fed back to the line 40 via the freewheeling diodes 44 to 50.

From the terminal 130, a resistor 134 leads via a node to a Zener diode 138 (5.6 V) and from this to the negative line 43, to which the emitter of an npn transistor 140 is connected, whose base with the node 136 and its collector via a resistor 142 is connected to the terminal 130. The transistor 140 operates as emitter follower and provides at its emitter and thereby on the line 43 a regulated voltage of eg + 5V, which is filtered through a capacitor 144 (10 nF).

The terminal Vss of the μC 76 is connected to the negative line 42, the terminal VDO to the line 43. To the terminals OSC1 and OSC2 a quartz oscillator 146 (1 MHz) is connected, to which a series circuit of two capacitors 148, 150 is connected in parallel, whose connection point is connected to the line 42. The unused ports RA1, RA2, RA3, RA4, MLCR /, RB4, RB5, RB6 are each connected via a resistor R (eg 10 kOhm) connected to the line 43 and are thus at a defined potential.

The rest of the circuit according to FIG. 6 has already been described in detail in FIGS. 1 to 5.

Explanation of four-quadrant operation (4-Q operation)

In Figs. 7 to 9, the four quadrants are designated by I to IV.

<b>Fig. 7</b> shows a diagram for a 1-Q operation, eg operation of the motor 30 in its preferred direction. On the horizontal axis, the speed n is plotted, on the vertical axis of the driving or braking torque M.

In a region 150 around the zero point no controlled operation is possible in this type of engine, but only in a region 152. In area 152, the engine 30 is eg accelerated to a regulated speed + N. If the speed + N is exceeded, so that a higher speed value + N1 is reached, the motor 30 is temporarily no or little power, so that its speed along a curve 154 due to friction, etc. again sinks. An active braking does not take place, but the positive torque M is reduced, if necessary to zero.

<b>Fig. 8</b> shows a 2-Q operation, ie a two-quadrant operation in which the motor 30 can also run backwards. When reversing it is driven with a negative torque and thus reaches the regulated speed -N. If this speed increases even higher to -N1 due to external influences, then the negative torque is reduced in its magnitude (curve 156), possibly to 0, so that the speed decreases again to the value -N due to friction, etc.

<b>Fig. 9</b> shows the representation of a 4-Q operation, so a four-quadrant controller. It is assumed that when driving in the forward direction to a speed + N is regulated, corresponding to the point 160th

If, due to external influences, the speed is increased to a value + N1 corresponding to a point 162 which is above the desired speed, the controller generates a negative (braking) torque, ie the motor goes to a point 164 and there a brake torque is generated by the controller so that the speed goes to a point 166 which corresponds to the regulated speed + N. There it is switched back to a positive torque, so that the controller goes back to point 160, corresponding to the desired speed + N.

The braking torque is generated in this type of motor in a very simple manner that the motor 30 is operated with a 180 ° el. Shifted commutation, ie with the same commutation as in reverse. This will be described below with reference to FIG. 19.

When motor 30 is reversing at speed -N, it operates at point 168. If it is driven there by external influences, then its speed can increase to -N1, that is to the point 170th There, the 4-Q controller switches to a positive torque, so to point 172, and by the positive torque, the speed drops back to -N, ie to point 174, where the torque is switched back to drive in the reverse direction, so that you are back at the starting point 168, ie the desired speed.

In the same way, a transition from + N (point 160) to -N (point 168) is possible. For this purpose, a negative torque is generated by a command DIR = 1 (at the input 98 of FIG. 6) at the speed + N, ie the motor goes from point 160 to point 166, from there to speed 0, and then - with the same , now driving torque, which remains constantly on - the speed -N_MIN at the left end of the window 150. There is switched to speed control, and the speed is controlled to the value -N, corresponding to the point 168. The motor then runs in the reverse direction with the speed -N.

If you want to get from point 168 to point 160, ie the speed goes through the value zero and the motor reverses its direction of rotation, the course goes from point 168 to point 174 (positive torque is generated), to speed 0, and then - with the same positive torque - to the speed + N_MIN and then - with speed control - to point 160 with the speed + N.

The advantage of 4-Q operation is that speed changes run very fast, so the drive is very dynamic, and that eg also the drive of a shutter is possible because when closing the shutter, where this pulls down and thereby drives the motor 30, the latter automatically goes into braking mode and thereby maintains the set speed. In this type of engine also the switching of the torque from positive to negative or vice versa is extremely simple and easy.

software

This is on <b>Fig. 10</b> Referenced. This shows as an example a speed range from -3000 to +3000. In this area is the window 150, which is from a lower minimum speed -N_MIN up to an upper minimum speed + N_MIN goes, for example, as shown, from -1000 to +1000.

In this window 150, as described above, in this type of engine, a speed control is not possible. For example, if the user sets a speed of +800 rpm, this is interpreted by the μC 76 as if the speed 0 were set, ie, the motor 30 does not start, or it is turned off when it is running.

As long as the current speed n is within the window 150, a flag Rs = 0 is set, which is interpreted by the computer as "engine is", although the speed may be, for example, between -999 and +999 U / min.

When leaving the window 150, this flag is changed to Rs = 1, that is, "engine is running", and in this region where Rs = 1, the engine speed is controlled by a speed controller.

Further, a direction flag Rt is used. Only when at start-up or a change of direction does the rotational speed leave the window 150, the direction flag Rt is switched to the value of the instantaneous user commanded direction DIR, which is supplied at the input 98 (FIG. 6). (DIR can be either 0 or 1).

If eg in Fig. 10 the motor 30 operates at point A, that is, at a speed of 3000, where Rt = 0, and the command DIR = 1 is given, that is the command that the motor 30 should reverse and go to point B, ie Speed ​​value -3000, so from point A, the motor 30 is energized with maximum braking current (ie no longer regulated) and thus falls below + NMIN rapidly, where Rs is switched from 1 to 0. The information Rs = 0 instructs the computer to continue to work with maximum braking current, so that first the speed 0 is reached, where the braking current - due to the changed direction of rotation - automatically becomes the drive current, so that the speed increases in the negative direction. During this entire time, the "old" direction of rotation Rt = 0 remains stored. The speed 0 is passed through easily and continuously.

From reaching the speed -N_MIN the result is a change from Rs = 0 to Rs = 1, ie from here on, the speed control resumes and the new directional value DIR = 1 given at the input 98 is stored as Rt = 1, so that the motor starts at -N_MIN regulated up to the speed -3000 starts up. The direction flag Rt is thus not changed when passing through the speed 0, but only when leaving the window 150 of FIG. 10.

In the opposite direction (from B to A), the sequence is mirror-invariably exactly the same, ie, the speed 0 is easily traversed, because in this type of motor applies: braking in one direction = drive in the opposite direction. This property is exploited in the present invention in an advantageous manner.

Within the window 150 is not regulated, but the motor 30 runs there unregulated with maximum control value, ie maximum current, depending on the motor type, the motor current may possibly be limited by a current limit, as described in Fig. 1. In a preferred manner, the control value BW (FIGS. 18, 19) for the control is already continuously calculated in the region of the window 150 so that it is immediately available when switching from Rs = 0 to Rs = 1.

In the case of the operational amplifiers 52 and 54 of the type L2720 described in connection with FIGS. 1 and 6, an internal current limitation is provided. Alternatively, especially at higher powers, an external current limiting can be provided.

<b>Fig. 11</b> shows a state diagram for the various processes that occur in 4-Q operation.

The "normal" start in the forward direction

If in state 181 (standstill of the motor 30) by a signal DIR = 0 at the input 98 (Fig. 4 and 6), the direction of rotation is set to the right, so preferred direction, and the speed setpoint n_should at input 86 or 98 'greater than the minimum speed + N_MIN is, the controller goes to state 182, ie the motor 30 starts with maximum control value BW, ie maximum current, in the preferred direction (DIR = 0). The start in the preferred direction is easy with these engines. At this time, the flag Rs is set to Rs = 0 because the actual speed n is smaller than N_MIN is.

If the actual speed n is + N_MIN has exceeded, the motor 30 goes from state 182 to state 183, that is, its operating range is reached, and the speed n is now at the predetermined value n_should regulated by the controller of FIG. 19, which is indicated symbolically in FIG. 11 by "BW via RGL". The motor status Rs is switched to Rs = 1 because the actual speed n is greater than N_MIN has been changed, and the controller regulates the speed by changing the block length BW (Fig. 18, 19) to the desired value. If, for example, the motor 30 is to run faster, the control value BW is increased within the permissible limits, ie the motor 30 then operates with a higher control value BW and consequently a higher current.

Start in the reverse direction

This type of motor is actually designed only for starting in a preferred direction DIR = 0. Therefore, for a reliable start in the reverse direction (DIR = 1) "electronic help" is necessary. In this case, the motor 30 initially runs from the status 181 in the preferred direction, ie counter to the commanded direction of rotation DIR_should,

If, in state 181, the direction of rotation DIR = 1, ie, left (reverse rotation, rotation in the reverse direction) is specified, and the predetermined setpoint value n_should more negative than the (negative) minimum speed -N_MIN If the motor 30 is initially accelerated in the preferred direction, ie with DIR = 0, the time T_HALL between two Hall changes, ie between two Hall interrupts, until T_HALL is less than a predetermined value T, is measured_G and then energized again during a time (T._G - EW), so that the motor 30 can store magnetic energy for the reverse start, then reverses from this state, and accelerates to -N_MIN with maximum control value BW = MAX in the reverse direction. This is the state 184. In this case, the value EW is an empirically determined constant for the relevant motor 30 in its specific application, ie installed in a fan.

After reaching -N_MIN Rs = 1 is set and the engine transitions to state 183. There, the speed is controlled by the speed controller (Fig. 19) to the desired value ("BW via RGL"), and the set target direction (DIR = 1) is now stored as a flag Rt = 1.

The μC 76 has in this example for the detection of the speed a counter that can count only up to 64,000 μs, ie up to 64 ms. This means that you can only measure the speed n when the time between two Hall interrupts is less than T_G = 64 ms has become. This time corresponds to a rotor rotation of 180 ° el. If the motor 30 is still too slow in the measurement, an overflow occurs, and therefore this measurement can not yet be used. Therefore, the motor 30 is driven in the preferred direction (DIR = 0) until the time between two Hall interrupts is less than T_G = 64 ms has reached. If the engine 30 is a small GD² has reached this speed n between the first two Hall interrupts, but with a fan with a large fan it takes longer, until this speed is reached.

If the speed n has become so high that between two Hall interrupts the time T_HALL for 180 ° el. Is eg 63.8 ms, which corresponds to a speed of 235 rpm, namely 60 / (4 x 0.0638) , one deducts from the time T_Hall thus determined the empirically determined value EW, thus carrying out the operation <maths id="math0001" num="(1)"><math display="block"><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mi mathvariant="normal">e</mi><mi mathvariant="normal">W</mi></mrow><mo>)</mo></mrow></mrow></math><img file="EP1535388B1_D0001.tif" /></maths>through, and from the next Hall interrupt measures this time (T_HALL - EW) and switches after its expiry on commutation in the opposite direction, ie only from this point on the rotor 38 is driven in the desired direction. At this moment, the value T_HALL practically corresponds to the value T_G,

Is in <b>Fig. 12</b> for a four-pole motor 38 shown. 12 shows two Hall interrupts 190, 192. Between these, a time T_HALL which is smaller than T is measured for the first time_G = 64 ms. From this time the value EW is subtracted, eg 5 ms. From the Hall interrupt 192, the calculated time (T_HALL-EW) is now measured, and at time 194 DIR = 0 is switched over to energization in the desired setpoint direction, that is to say with DIR = 1. This means the drive in the reverse direction.

The motor 38 continues in this process by its stored energy initially by a small rotation angle further in the preferred direction, and EW is empirically determined so that the motor continues to a rotational position, where he has stored in his magnetic circuit enough magnetic energy to To support and facilitate the start in the reverse direction. It can be said that the rotor 38 at its initial start in the preferred direction, ie the "wrong" direction of rotation, "momentum" stores, so magnetic energy for the start in the reverse direction, and by this "momentum", the motor then runs from this empirically determined starting point from easily in the reverse direction.

For this purpose, reference is made to the diagram of FIG. 13, where the course of the reluctance torque M_R is shown schematically. At the point 196, the rotor 38 has a through the reluctance M_R defined rest position, because there M_R = 0, and when it starts in the preferred direction (DIR = 0), the braking reluctance torque 198 increases only relatively slowly, so that a start-up is easily possible. However, if the rotor 38 starts in the reverse direction (DIR = 1), a steeply increasing braking reluctance torque 200 occurs immediately after the start at 196, which can prevent startup there. Therefore, in this case, the rotor 38 is energized by energizing in the direction DIR = 0 to about a point 202 (FIG. 13), from which he can easily start in the direction DIR = 1.

Transition from state 183 to the braking state

There are several possibilities here:<ul id="ul0001" list-style="none" compact="compact"><li>a) The speed n should be reduced.</li><li>b) The motor 30 should be stopped.</li><li>c) The engine should be reversed.</li></ul>

a) The speed should be reduced.

If only the rotational speed n is reduced without the direction of rotation being changed, the nominal value is n_should smaller than the actual value n, and the controller goes into the state 185, where the motor is energized by the corresponding (regulated) current offset by 180 ° el., which is indicated symbolically by "inv 180 °". The order of the current pulses i, i 'is inverted, ie vice versa. The flag Rs retains the value 1.

Is the setpoint n_should again greater than or equal to the actual value n, the controller goes from 185 back to the state 183, ie the 180 ° el.versetzte energization is switched back to a normal current supply. The new speed is thus achieved by a "metered braking" by means of inverse energization, which allows very fast control operations.

b) The motor 30 should be stopped.

Is the setpoint n_should less than | N_MIN| (including the setpoint 0), so is also braked by the current flow is shifted by 180 ° el., And the controller goes - via the state 185 and the path 185 '- in the state 181, ie Rs = 0, BW = 0 and standstill because below | N_MIN| For physical reasons, a speed control is not possible and therefore such a setpoint n_should is interpreted as desired speed 0. The control value BW is set to 0, ie the current to the motor 30 is switched off.

c) The engine should be reversed.

If the motor 30 is to be reversed out of the state 183, then according to 183 ' Desired direction ≠ Istrichtung, and the controller is in the state 186. There, the motor 30 is energized with maximum control value BW in the braking direction, ie with "inv. 180 °", while | N falls below_MIN|, and enters the state 188, where Rs is switched to 0, as long as the rotational speed n in the window 150 (FIG. 10), ie between + N_MIN and -N_MIN, lies. Thus, the status 188 (Rs = 0) does not mean that the engine 30 is actually stopped, only that its actual speed n is below the predetermined values. This is shown in FIG. 10. Within the entire window 150, Rs = 0, and this gives the μC 76 the information to drive there with the maximum manipulated variable, that is, the maximum current, in accordance with the state 186. (The change in Rs = 0 in State 188 is a state change and is therefore shown separately.)

When reversing the direction of rotation of the motor 30 goes in Fig. 10, for example, from A to B, ie from state 183 to state 186. Although in this case the predetermined desired direction of rotation DIR_should has the value 1, the previous value Rt = 0 for the direction of rotation remains initially stored, and because as the new direction of rotation DIR = 1 is given and the new speed outside the window 150 (Fig. 10), the μC 76 receives the information that the motor 30 must be energized with full control value in the reverse direction of rotation (state 186). In the area of ​​the previous direction of rotation, this means a braking operation with high current, and this type of energization with high current is continued unchanged between A and B up to the speed - N_MIN, ie until it leaves the window 150, but starting at the speed 0, the voltage induced by the rotor 38 in the winding 32 reduces the amplitude of the current pulses i or i '. Here, the speed 0 is passed through without any problems continuously.

From -N_MIN μC 76 goes to state 187, where Rs changes from 0 to 1, ie, from there, the speed control (FIG. 19) starts, and the flag Rt = 1 for the new direction value DIR = 1 (at input 98 of FIG. 6) is now stored (state 187) so that the motor 30 is off -N_MIN regulated up to the desired speed n_should starts up, according to state 183 of FIG. 11.

In the opposite direction, that is, in Fig. 10 from B to A, the process is mirror-image just the same, ie the speed 0 is easily traversed, because in this type of engine: Brakes in one direction = drive in the opposite direction.

Within the window 150 around the speed 0 around is thus not regulated, but the motor 30 runs there unregulated with maximum control value, because in this area a control is not possible, but still there is the control value BW for the speed control preferably continuously calculated, so that it is immediately available when Rs is switched from 0 to 1 and must use the speed control.

<b>Fig. 14</b> shows the flowchart of a routine S204 for the comparison of the desired direction DIR_should with the actual direction DIR as it takes place at position 183 'in FIG. 11.

In step S206, it is checked if DIRist = DIR_should is. If so, the program proceeds to S208 (normal operation), that is, in Fig. 11, the state 183 remains unchanged.

If the answer is NO in S206, the current is set to a maximum value in S210, and in S212, it is switched to a 180 ° el. Phase-shifted commutation. This corresponds to the state 186 in FIG. 11.

Thereafter, the routine S204 goes to step S214 "Return".

<b>Fig. 15</b> shows a routine S220 labeled "Motor Status". In S222, the current value of Rs is stored in register Rsalt. It is then checked in S224 whether the amount | n | the instantaneous speed is greater than the amount | N_MIN| the minimum speed, ie whether the speed n is outside the window 150. If YES, the flag Rs is set to 1 in S226, which means that the engine 30 has an Inl&gt; | N_MIN| runs and a regulation of the speed must take place. If NO, Rs is set to 0 in S228, which means that the speed is within the window 150 (FIG. 10), so speed control is not possible.

Subsequently, it is checked in step S230 whether Rs is still the value of S222 or has changed. The latter means that the speed threshold, ie | N_MIN| in which one or the other direction has been exceeded, which is why it is then necessary to check in which direction this threshold was exceeded. If the answer is YES in S230, the routine goes to its end, S232, Return. If NO, the value of Rs has changed since step S222, and it is checked in S234 if Rsalt = 0. If NO, this means that Rsalt has the value 1 and has changed to Rs = 0, ie the current speed n now has a value within the window 150 and the routine S220 is exited via S232.

If the answer is YES in S234, it means that the rotational speed of the engine 30 has increased, the threshold | N_MIN| Now, in S236, the flag Rt now adopts the new target direction value DIR set by the user at the input 98, and in S238, goes to a speed control RGL, so that the speed controller intervenes and the speed to the user-specified value n_should is regulated. Thereafter, this routine S220 is exited via S232.

The <b>FIGS. 16 and 17</b> 12 show a discrimination function S250 which serves to control the operations at startup, a speed change, or a rotational direction change of the motor 30.

In step S252, it is checked whether a new value for the target speed n_should is present. In the present exemplary embodiment, this value is specified as a time t_s which the motor 30 is required, for example, for a rotation of 180 ° el. If YES, in S254 the new value for n_should (ie, t_s), and in S256, a flag called "n_should NEWFLG "is set to 1 to indicate a new value.

If the answer is NO in S252, the program goes to S258, where it is checked whether there is a new value for the actual speed n. (In the embodiment, the value of the actual speed T_HALL is used, the rotor 38 for a rotation of 180 ° el. needed). If NO, the program goes back to S252. If YES, the new value for n is stored in S260, and at S262, a flag is set to 1 that is labeled "n NEWFLG" to indicate that there is a new value for n. As described below, if there are two new values, a calculation is made with these values, and then both flags are set to 0 at S268 to indicate that these values ​​are "processed" 11.

Subsequent to S256 and S262, it is checked in S264 whether the two flags of S256 and S262 have the value "1". If NO, the program goes back to S252. If YES, in S266, the two new values ​​for n and n are used_should, ie from T_HALL and t_s, a new control value BW is calculated. This value indicates the length of a current block in the so-called block control, which is explained in FIG. 18. (The calculation of BW is described in Fig. 19.) Thereafter, in S268, the two flags "n NEWFLG" and "n_should NEWFLG "is set to" 0 "again to indicate that the two current values ​​have been processed to a new control value BW.

In step S272 (Fig. 17), it is checked whether the target rotational speed n_should inside or outside of the window 150 (FIG. 10), ie whether <maths id="math0002" num="(2)"><math display="block"><mrow><mrow><mo>|</mo><mrow><msub><mi mathvariant="normal">n</mi><mrow><mi mathvariant="normal">should</mi></mrow></msub></mrow><mo>|</mo></mrow><mo>≥</mo><mrow><mo>|</mo><mrow><msub><mi mathvariant="normal">N</mi><mrow><mi mathvariant="normal">MIN</mi></mrow></msub></mrow><mo>|</mo></mrow><mo>?</mo></mrow></math><img file="EP1535388B1_D0002.tif" /></maths>

If NO ("N"), it should be interpreted that the desired speed is zero, and it is then checked in S274 whether the actual speed n is within the window 150, that is, whether <maths id="math0003" num="(3)"><math display="block"><mrow><mrow><mo>|</mo><mi mathvariant="normal">n</mi><mo>|</mo></mrow><mo>≥</mo><mrow><mo>|</mo><mrow><msub><mi mathvariant="normal">N</mi><mrow><mi mathvariant="normal">MIN</mi></mrow></msub></mrow><mo>|</mo></mrow><mo>?</mo></mrow></math><img file="EP1535388B1_D0003.tif" /></maths>

If the speed n is outside the window 150 (answer: YES), the routine goes to S275, and there, the motor 30 is decelerated with the calculated manipulated value BW. If the speed n lies within the window 150, in the example in the range between -1000 and +1000 rpm, set Rs = 0 in S276 and zero in S277 BW = 0, ie motor 30, and the routine goes to S278 (Return).

If, in S272, the answer is Y (YES), that is, the desired speed n_should is outside the window 150, it is checked in S280 in which direction the motor 30 should start. The default DIR = 0 at the input 98 means that the motor 30 should start in the preferred direction. DIR = 1 means start against the preferred direction.

If in S280 the answer is yes, ie start in preferred direction, the usual commutation in preferred direction "0" occurs at maximum control value BW, ie maximum current, in order to obtain a rapid start, ie no speed control takes place within window 150.

Subsequently, the routine goes to step S284, where it is checked if the actual rotational speed n is out of the window 150. If NO, the routine goes back to S272. If YES, the routine goes to S286, where now the flag Rt receives the current value of DIR, in this case Rt = 0, since the motor 30 is running in the preferred direction "0".

Subsequently, Rs = 1 is set in S288 to indicate that the current speed n has reached a value outside the window 150, and therefore that the motor 30 now has to be operated with speed control. This flag Rs is evaluated accordingly in the μC76, cf. Fig. 15, and turns on the speed control n_CTL, so that in S289, the motor 30 with the calculated control value BW, so regulated, is accelerated.

If DIR = 1 in S280, this means a start against the preferred direction, as already explained in FIG. In this case, the routine goes to S290, where the motor 30 - contrary to the command DIR = 1 - in the preferred direction "0", that is with DIR = 0, is started with maximum control value BW, as well as in S282.

It is then checked in S292 whether the time T_HALL (FIG. 12), ie the time between two successive Hall interrupts, is less than a time T_G which is 64 ms in this example, as already explained. (The time T_HALL corresponds to the time for one quarter revolution of the four-pole rotor 38. In S292, in this example, it is indirectly queried whether the rotor 38 has reached a speed of at least 234 rpm). If the answer is NO, the routine goes back to S290, and further commutation with maximum manipulated value BW, that is, maximum current, in the preferred direction until the condition of S292 is satisfied.

Most motors reach a speed of 234 rpm after only a quarter turn, because the start is at maximum current. For some drives, however, a larger angle of rotation may be needed to achieve a value for T_HALL that is less than T_G is.

If in S292 the answer is yes, that is, the speed has reached or exceeded the value of 234 rpm, in S294, in a TIMER, the value (T_G - EW), and this TIMER is started. Subsequently, in S296, it is checked if TIMER = 0. If NO, so if the time (T_G EW) has not yet expired, the routine goes back to S294.

If the answer is yes in S296, the routine goes to S298, where now - according to the command DIR = 1 at input 98 - the commutation is switched to commutation in the reverse direction "1" at maximum control value BW, so that the speed of the reached +234 U / min via the speed 0 to the desired negative speed, eg -1205 U / min, running up, whereby the speed 0, ie the zone of the standstill, is run through without problems.

Subsequent to S298, the program goes to S284, where it is checked whether the actual rotational speed n is already outside the window 150. If NO, the routine goes back to S272 as already described, and if YES, it goes to S286, where now Rt = 1 is set, ie the direction of rotation DIR = 1 set by the user at the input 98 is also stored in the flag Rt in the μC76 so that the latter assumes the value Rt = 1, and then Rs = 1 is set in S288 to indicate that now the speed control is activated, and at S289, the engine 30 is switched to operation with the control value BW calculated in the routine S266, that is, operation with speed control. At S278, routine S250 is exited.

<b>Fig. 18</b> serves to explain operations in the commutation of the motor 30th

<b>Fig. 18a</b> shows a signal PWM which is generated continuously and eg has a frequency of 20 kHz. Its duty cycle pwm is programmatically adjustable. It should be understood that the present invention may be practiced without the use of such a PWM signal. With this duty cycle, the amplitude of the current pulses i, i '(FIG. 1) which are supplied to the motor 30 can be controlled.

<b>Fig. 18b</b> shows the signal HALL. This has rising edges 370 in the transition from LOW to HIGH, and it has falling edges 372 in the transition from HIGH to LOW.

The times t1, t2, etc. at which these edges occur are measured by a ring counter TIMER1 and stored in a temporary variable t_TEMP. As shown in FIG. 18, the rising flanks 3 70 govern the switching on of the transistors HSL and LSR, ie of the current i of FIG. 1. Conversely, the falling flanks 372 govern the switching on of the transistors HSR and LSL, ie of the current i 'of FIG . 1.

The duration T_HALL between two edges results as <maths id="math0004" num="(4)"><math display="block"><mrow><mi mathvariant="normal">t_HALL</mi><mo>=</mo><mi mathvariant="normal">t</mi><mn>2</mn><mo>-</mo><mi mathvariant="normal">t</mi><mn>1</mn></mrow></math><img file="EP1535388B1_D0004.tif" /></maths>

This period of time is a measure of the instantaneous speed n of the rotor 38 and corresponds to the time it takes for a rotation of 180 ° el. needed. Of course, this time can be measured in many ways. Preferably, as soon as the rotational speed is high enough, the time is measured for a larger angle of rotation, in particular for one or more full revolutions of the rotor 38, which in the exemplary embodiment is a rotation angle of 720 ° el. (or a multiple thereof). - The measurement over a full turn is particularly accurate, but requires that a certain minimum speed is reached.

<b>Fig. 18c</b> and <b>Fig. 18d</b> FIG. 18c shows the signals HSR, LSL for switching on the current i '. FIG. FIG. 18d shows the signals HSL, LSR for switching on the current i.

The beginning of one of the pulses 444 of FIG. 18c is calculated from the falling edge 372 of an associated signal HALL, symbolized by the arrow 445, and the beginning of one of the pulses 446 of FIG. 18d is calculated from the rising edge 370 of the associated signal HALL, as symbolized by the arrow 447. The edges 370, 372 of the signal HALL correspond to predefined, measurable rotational positions of the rotor 38. When calculating commutation operations, these are the only rotational positions that are known exactly, and therefore the calculations refer to these "fixed points".

Assuming that the control signals 444, 446 are symmetrical to the pulses of the signal HALL, the value is given for the time t3 at which a signal 446 starts <maths id="math0005" num="(5)"><math display="block"><mrow><mi mathvariant="normal">t</mi><mn>3</mn><mo>=</mo><mi mathvariant="normal">t</mi><mn>1</mn><mo>+</mo><mi mathvariant="normal">t_HALL</mi><mo>+</mo><mrow><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mi mathvariant="normal">B</mi><mi mathvariant="normal">W</mi></mrow><mo>)</mo></mrow></mrow><mo>/</mo><mn>2</mn></mrow></mrow></math><img file="EP1535388B1_D0005.tif" /></maths>

This means BW = block length of the signals 444, 446. This block length is calculated by the speed controller RGL, which is described below with reference to FIG. 19, and is therefore also referred to as the manipulated variable, as usual.

Analogously, the value results for the time t4 at which the control signal 444 is to begin <maths id="math0006" num="(6)"><math display="block"><mrow><mi mathvariant="normal">t</mi><mn>4</mn><mo>=</mo><mi mathvariant="normal">t</mi><mn>2</mn><mo>+</mo><mi mathvariant="normal">t_HALL</mi><mo>+</mo><mrow><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mi mathvariant="normal">B</mi><mi mathvariant="normal">W</mi></mrow><mo>)</mo></mrow></mrow><mo>/</mo><mn>2</mn></mrow></mrow></math><img file="EP1535388B1_D0006.tif" /></maths>

It should be noted that, for example, the time t3 is not calculated from the time t2 (the immediately preceding edge 372 of the signal HALL) which is closest to t3, but from an earlier time t1, namely from the preceding edge 370th The reason is that if BW = T_HALL, the time t2 would coincide with the time t3, which does not work, since computational steps must be performed between t2 and t3.

If a so-called ignition angle shift, for example by a fixed value VZ, is used, the above formulas change as follows: <maths id="math0007" num="(7)"><math display="block"><mrow><mi mathvariant="normal">t</mi><msup><mn>3</mn><mo>'</mo></msup><mo>=</mo><mi mathvariant="normal">t</mi><mn>1</mn><mo>+</mo><mi mathvariant="normal">t_HALL</mi><mo>+</mo><mrow><mo>(</mo><mrow><mrow><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mi mathvariant="normal">B</mi><mi mathvariant="normal">W</mi></mrow><mo>)</mo></mrow></mrow><mo>/</mo><mn>2</mn></mrow></mrow><mo>)</mo></mrow><mo>-</mo><mi mathvariant="normal">V</mi><mi mathvariant="normal">Z</mi></mrow></math><img file="EP1535388B1_D0007.tif" /></maths><maths id="math0008" num="(8)"><math display="block"><mrow><mi mathvariant="normal">t</mi><msup><mn>4</mn><mo>'</mo></msup><mo>=</mo><mi mathvariant="normal">t</mi><mn>2</mn><mo>+</mo><mi mathvariant="normal">t_HALL</mi><mo>+</mo><mrow><mo>(</mo><mrow><mrow><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mi mathvariant="normal">B</mi><mi mathvariant="normal">W</mi></mrow><mo>)</mo></mrow></mrow><mo>/</mo><mn>2</mn></mrow></mrow><mo>)</mo></mrow><mo>-</mo><mi mathvariant="normal">V</mi><mi mathvariant="normal">Z</mi></mrow></math><img file="EP1535388B1_D0008.tif" /></maths>

In this case, the times t3 'and t4' are further to the left by the size VZ, as shown in FIG. 18d for t3 indicated, which means a slightly earlier switching on the currents i and i 'and can cause an improvement in the efficiency. It can also be seen that t3 'in this case lies ahead of time t2, which is only possible because the reference time RefTime for the calculation of t3' is not the time t2, ie the falling edge 372, but the time t1, ie the rising Hall edge 370, as symbolically represented by the arrow 447. VZ is usually a constant, but can also be a speed-dependent function or be constantly optimized by separate, not shown program parts.

<b>Fig. 19</b> shows the routine RGL S266 for the speed control. This is based on a comparison of the reverberation time T_HALL with the desired time t_s, which latter of the desired speed n_should corresponds and at the input 84 in the form of a PWM signal or a DC voltage can be specified. The controller according to the embodiment thus does not work directly with speeds, but with times that the rotor 38 is needed or needed for a certain angle of rotation. The reverberation time T_HALL corresponds to the time that the rotor 38 requires for a rotation of 180 ° el. If the rotor 38 is four-pole and rotates at 3000 rpm, then<maths id="math0009" num="(9)"><math display="block"><mrow><mi mathvariant="normal">t_HALL</mi><mo>=</mo><mrow><mrow><mn>60</mn></mrow><mo>/</mo><mrow><mrow><mo>(</mo><mrow><mn>3000</mn><mo>*</mo><mn>4</mn></mrow><mo>)</mo></mrow></mrow></mrow><mo>=</mo><mn>0005</mn><mi mathvariant="normal"> </mi><mi mathvariant="normal">s</mi><mo>=</mo><mn>5</mn><mi mathvariant="normal"> </mi><mi mathvariant="normal">ms</mi></mrow></math><img file="EP1535388B1_D0009.tif" /></maths>

Analogously, this time is 1000 rpm <maths id="math0010" num="(10)"><math display="block"><mrow><mi mathvariant="normal">t_HALL</mi><mo>=</mo><mrow><mrow><mn>60</mn></mrow><mo>/</mo><mrow><mrow><mo>(</mo><mrow><mn>1000</mn><mo>*</mo><mn>4</mn></mrow><mo>)</mo></mrow></mrow></mrow><mo>=</mo><mn>0</mn><mo>.</mo><mn>015</mn><mi mathvariant="normal"> </mi><mi mathvariant="normal">s</mi><mo>=</mo><mn>15</mn><mi mathvariant="normal"> </mi><mi mathvariant="normal">ms</mi></mrow></math><img file="EP1535388B1_D0010.tif" /></maths>

At low speeds, therefore, the actual value T_HALL becomes very large and is then significantly greater than the setpoint t_s, which is 5 ms, for example, for 3000 rpm. For this reason, the control difference RGL_DIFF is preferably formed as a difference (T_HALL -t_s) according to step S654, so that a positive result of the difference formation is obtained.

At S656, it is checked whether the control difference is greater than a permissible positive maximum value RGL_DIFF_MAX. If so, at S658 the control difference is set to this positive maximum value. This is especially important at startup, because otherwise the rule difference would be very large.

If the answer is NO at S656, the program goes to step S660 and checks there as to whether the control difference is smaller than an allowable maximum negative value -RGL_DIFF_MAX. If YES, the control deviation is set to this negative maximum value at S662. (This applies to the case that the engine is faster than the desired speed).

Steps S658, S660 or S662 are followed by S664, where the computational steps of a PI controller are executed. For this purpose, the control difference is multiplied by a proportional factor RGL_P, which can have the value 2, for example, and the proportional component RGL_PROP is obtained.

Likewise, the control difference is multiplied by an integral factor RGL_I, which can amount to, for example, 0.0625, and is then added to the old integral component RGL_INT, so that a new integral component is obtained.

Finally, the length BW of a current block 444 or 446 (FIG. 18) is calculated as the sum of the new proportional component and the new integral component.

Proportional factor RGL_P and integral factor RGL_I are determined empirically, depending mainly on the size of the motor 30 and the moment of inertia of the load to be driven.

The PI controller S664 can supply a control value BW whose magnitude | BW | is too big. Since this amount must not be greater than the time T_HALL that the rotor 38 requires to travel 180 ° el., It is checked in the next step S666 if BW is too large, and if necessary, the block length is limited in step S668. eg to the current value T_HALL or a slightly smaller value.

If the answer is NO in S666, the routine S266 proceeds to step S670, where it is checked if BW is smaller than -T_HALL. If YES, it is set in S672 BW: = -T_HALL, ie BW is limited to -T_HALL (or a slightly smaller value). The sign (+ or -) calculated in S664 remains.

The command "inv. 180 °" thus always switches the current to the motor from the current direction to the opposite direction.

example 1

The motor runs according to the DIR = 0 command in the preferred direction. The command DIR = 1 is given. The energization is then switched by "inv. 180 °" so that the motor is braked and driven in the reverse direction.

Example 2

At the end of Example 1, the motor with DIR = 1 runs in the reverse direction. The motor now receives the DIR = 0 command. The current is also switched over by an "inv. 180 °" command, in such a way that the motor is braked and then driven again in the preferred direction.

For the switching of the current supply results in the embodiment of Fig. 6 is a very simple possibility, because there the OPV 52 is controlled by the output RB3 of the μC 76, and the OP 54 is controlled by the output RB2 of the μC 76.

In the case of a reversal of the current supply, a changeover is performed in the μC 76 in such a way that the commutation pulses which were previously supplied to the output RB3 are now supplied to the output RB2, and the commutation pulses which were previously supplied to the output RB2 are now connected to the output RB3 be supplied.

For motors with more complicated circuits, the signals at the RB2 and RB3 outputs can also be calculated according to a slightly different scheme, according to the following formulas <maths id="math0011" num="(11)"><math display="block"><mrow><mi mathvariant="normal">t</mi><mn>3</mn><mo>=</mo><mi mathvariant="normal">t</mi><mn>2</mn><mo>+</mo><mi mathvariant="normal">t_HALL</mi><mo>+</mo><mrow><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mrow><mo>|</mo><mrow><mi mathvariant="normal">B</mi><mi mathvariant="normal">W</mi></mrow><mo>|</mo></mrow></mrow><mo>)</mo></mrow></mrow><mo>/</mo><mn>2</mn></mrow></mrow></math><img file="EP1535388B1_D0011.tif" /></maths><maths id="math0012" num="(12)"><math display="block"><mrow><mi mathvariant="normal">t</mi><mn>4</mn><mo>=</mo><mi mathvariant="normal">t</mi><mn>1</mn><mo>+</mo><mi mathvariant="normal">t_HALL</mi><mo>+</mo><mrow><mrow><mrow><mo>(</mo><mrow><mi mathvariant="normal">t_HALL</mi><mo>-</mo><mrow><mo>|</mo><mrow><mi mathvariant="normal">B</mi><mi mathvariant="normal">W</mi></mrow><mo>|</mo></mrow></mrow><mo>)</mo></mrow></mrow><mo>/</mo><mn>2</mn></mrow></mrow></math><img file="EP1535388B1_D0012.tif" /></maths>

When "inv. 180 °" is then switched either from the formula (7) to the formula (11), or conversely from the formula (11) to the formula (7), and exactly analogous to the formula (8) on the Formula (12), or vice versa from the formula (12) to the formula (8).

The shutdown occurs in all cases at (t3 + | BW |) or at (t4 + | BW |), so that the current during each time | BW | flows. A preignition VZ can, if desired, be taken into account analogously to equations (7) and (8).

After S670 the program goes to S674. There it is checked which type of command DIR is present. This command can of course be different from the current state of the engine. For example, the engine may have just received the command DIR = 0 at the moment, ie forward, but the engine is still running in the reverse direction because the previous command DIR = 1.

If DIR = 0, it means that the instruction is for the preferential direction, and the program goes to S676, where the sign of the manipulated value BW is checked, which is done by examining its most significant bit MSB having the value 0, if the control value BW is positive.

If the control value BW calculated in S664 is positive, this means that the power is supplied in the correct manner to drive the motor in the preferred direction, and that the speed does not become too high, and therefore the control value BW is supplied unchanged to the motor, as symbolically indicated in S678. There it is symbolically expressed as | BW | -&gt; MOT.

If it is determined in S676 that the control value has a negative sign, this means that either the speed of the motor is too high, but with the motor rotating in the preferred direction, or that the motor is currently being driven in the reverse direction. In both cases, the program goes to S680 where "inv. 180 ° ", the direction of the current supply is switched, wherein the pulse length is determined by the size of the control value BW. As a result, either the speed of the motor (when rotating in the preferred direction) is reduced until BW becomes positive again, or the direction of rotation is reversed, which then BW also receives a positive sign.

If the answer is NO in S674, this means the command DIR = 1, ie running in the reverse direction.

In this case, the sign of the calculated manipulated variable BW is checked in S682. If this sign is negative, this means that the command DIR = 1 coincides with the instantaneous direction of rotation of the motor, ie the motor is running in the reverse direction, and the program proceeds to S684, where the motor is energized unchanged with the calculated control value BW in the reverse direction ,

If the answer is NO in S682, this means that the control value BW has a positive sign, ie either that the speed of the motor is too high, with the motor rotating in the reverse direction, or that the motor is currently energized in the preferred direction , ie opposite to the command DIR = 1.

In both cases, the program goes to S686, where the direction of the current is inverted. As a result, the motor is now energized in a direction opposite to the previous energization, whereby either its speed decreases until the calculated control value BW is negative again, or the direction of rotation of the motor is reversed so that it runs at the desired speed in the reverse direction.

The right half of the flowchart of Fig. 19 (steps S676, 678, 680) thus concerns the case that the engine has the command DIR = 0 for the run in the preferential direction, and the left half (steps S682, 684, 686) means that the engine has the command DIR to run in the reverse direction.

Subsequent to steps S678, S680, S684 and S686, the routine goes to S688 (return).

The routine of FIG. 19 thus gives a control value BW, the magnitude of which decreases the more the actual speed becomes the desired value n_should approaches. The sign of the calculated control value indicates whether the speed is too high or too low. If the value n is exceeded_should the commutation is switched to commutation against the set direction of rotation DIR, so that then the motor 30 is controlled to the desired speed n_should is braked, provided that the latter is greater in amount than the minimum speed NMIN.

Particularly suitable is the invention for the drive of fans whose direction of rotation must be switched and in which the speed must be maintained even if, for example, by wind gusts the speed of the fan is increased. Another application is drives of shutters, blinds and the like. A major advantage of the invention is that both for the engine and for the electronics, the costs are lower than in previous drives with similar functions.