DE2235367A1 - ELECTRIC CONTROL CIRCUIT - Google Patents

Description

■ concerning
Electric control circuit

Priorities: July 19, 1971, Great Britain, - 'No. 33767/71

9 September 1971, Great Britain, No. 42132/71

The invention relates to electrical control circuits and relates in particular to those for generating control or control units suitable for operating electric motors Supply currents or voltages.

As will be understood from the following, the invention is especially, but not exclusively, for low-inertia Electric motors applicable, such as the type of motor in which the armature is substantially plate-shaped and for example consists of one or more thin sheets of insulating material, on which according to the so-called printed circuit technology a conductor pattern is applied.

Many methods of controlling the operation of electric motors have been proposed and used to accomplish the to meet various requirements, controls for

2098 8 5/0973

variable speed, controls for constant speed, to generate variable torque and other performance characteristics. In general, the engine management is by control the voltage or the motor current supplied to the motor in a continuous manner or by means of pulse methods been.

. The invention relates to an improved form of control which is particularly advantageous for the control of pulse currents, which are fed to substantially low-inertia motors, is applicable.

The invention comprises an engine control system having a device responsive to an input signal and a device responsive to a deviation in the input signal responds beyond certain limits that determine a dead band and supplies the motor with a pulse current in one direction, that of the deviation of the input signal beyond the upper one or lower limit.

The invention also consists in features of control circuits, which are described below.

The invention is illustrated below with the aid of preferred exemplary embodiments explained in connection with the drawings; show in the drawings

1 shows a circuit diagram of an integrating arrangement with a dead band device;

Fig. 2 is a circuit diagram of a switching device or a Modulator;

3 is a circuit diagram of an integrator with a variable spread control and a dead band device;

Fig. 4 is a block diagram of a controller;

³ 223536?

Pig. 5 is a more detailed circuit diagram of the Pig control. 4;

Fig. 6 is a circuit diagram of a thyristor ignition circuit;

Pig. 7 is a circuit diagram for one embodiment of a modulator; and

'Pig. 8 is a circuit diagram for one embodiment of a high frequency power output stage.

In the embodiments of the invention described below, current pulses are supplied to the motor to be controlled. The power stage feeding the motor works with a double-directional one Switching device, such as a double-sided directional controllable silicon rectifier (triac) for 'switching the half-waves a supply. There are two types of circuits, the operating frequency and the duration of the supplied to the motor Impulses are different from each other. For example, with one type of circuit, current pulses of 10 msec can be used accordingly a frequency of 50 Hz can be used, while the other pulses of 200 usec at a frequency of 3.3 KHz can be used. This high frequency pulsed operation is possible if the controlled motor has a sufficiently small inertia; a typical motor has a low impedance Armature resistance of, for example, 1Ω and an inductance of 50 uH, which gives a time constant of 50 use.c.

In one embodiment, the present invention provides a process control in which the output information from a .ner Sequence consists of steps, each of which has maximum output power Has. The step speed is adjustable and a function of an amplifier input signal. Every step consists of a half cycle of a reference frequency that is either generated internally or from an external supply. originates. The polarity of the input signal determines whether it is positive or negative half-cycles of the reference frequency are generated.

200885/097 3

In this process control, the characteristics are for. open loop similar to a conventional process control. Before describing the transfer characteristics, it should be useful to use certain expressions below to explain.

The proportional band is the input or error signal area in which a proportional action takes place, i.e. in which the signal causes the output device by a distance proportional to the error signal at maximum speed is moved. The width of the proportional band is inversely related to the gain of the control, i.e. that a control with wide proportional band low gain Has,

The integral effect is that effect of a control in which the output device responds to the error signal proportional speed moves.

Integral time is the time taken by the output device needed to move a distance under the integral action that of that covered under the proportional action Path equivalent to isij.

Derivative effect is that effect according to which the output device adjusts itself to the rate of change of the error signal moves proportionally, or the ratio (proportional band to control gain) changes with it the rate of change of the error signal.

A discontinuous process control is a control in which a motorized device in a Series of steps, each driven with maximum output power. The integral action (defined above) is achieved by changing the speed at which these steps occur. The position that the establishment

209805/0973

_ 5 -

223538?

is always a function of the time during which the output signal was applied and not its analog value. The output device thus integrates the output signal over time. '»

A continuous control differs from the discontinuous control in that it a continuous analog output signal (voltage or current generated that can be used to control various devices. The position taken by the facility is in place always in direct relation to this output signal. The integration in a discontinuous process control makes a difference differs from that of a continuous control that it takes place in the output device. In contrast the integration takes place with the continuous control within the control.

The response behavior of the controller when the control loop is open. on a step input signal consists in changing the position of the Output device at one at the input terminal of the amplifier lying step tension. For example, she speaks control to be described on a step input signal in two different successive modes: In the first Mode, the response is proportional and the motor runs during one of the steps proportional to the size of the step signal Maximum speed period. In the second mode, the response is integral and the engine runs at one the speed proportional to the magnitude of the step voltage.

In the case of a continuously operating control, the ~ Proportional effect instantly within practical limits. The derivative effect has the consequence that the Output device as a function of a step input signal moved a maximum distance during the proportional period. This path is for a slowly changing input voltage reduced.

209885/0973

223536?

Negation is the process of reducing the size of the signal at the input of an amplifier or other device by the action of a negative feedback voltage or a negative feedback current. ·

In the following, before describing the system as a whole, individual components of a complete System.

Fig. 1 shows a block diagram of an arrangement that has a Function amplifier 10 includes negative feedback. The negative feedback is provided by a second amplifier 11 formed, which together with a feedback capacitor 12 works as an integrator. The amplifier 10 is provided with an input resistor 13 and with feedback resistors 14 and 15 provided. The amplifier 10 gives when a step input signal is applied an output signal that persists until the integrator has charged and the input signal negated.

The period of time during which the output signal exists, is therefore a function of the integrator time constant and the. Step input signal size; suppose the integrator generates a linearly changing integration output signal, the duration of the output signal with a fixed integrator time constant is proportional to the size of the step input signal.

In a system implemented in practice, there is a dead band with an input signal within the dead band no change in the output signal generated. In one of those In this case, an integration only occurs when the output signal of the amplifier determines the dead band exceeds positive or negative levels, and integration stops when the step input negates so much that it falls into this area. The predetermined levels

209885/0973

are in Pig. 1 as a positive or negative response switch 16 and 17 shown in the feedback loop.

If the integrator is perfect and the integrated voltage is not reduced by scattering, the system remains infinite long in this state and with this value of the output signal. However, if there is a scattering path through the integrator, like him in Pig. 1 is represented by a resistor 18, the integrator will listen for a signal that is dependent on the spread Time interval to negate the input signal, the amplifier output signal exceeds the specified level and the integration begins again. In this way, the Pig system repeats. 1 puncture repeatedly with a fixed rate determined by the various fixed constants in the circle including the rate of integration, the spread, the gain and the switching hysteresis will. The negative and positive switching hysteresis is important; without vsi2, the integrator only needed a very small ' To change tension in order to resume integration, so ' that a high frequency vibrational state would be caused.

The switching devices used in the control can advantageously be triacs that can be controlled with positive or negative pulses. In Pig. 2 are two Triacs 19 and 20 are shown with a common resistor 21, one of which has a positive voltage across a diode 22 and the other controlled by a negative voltage via a diode 23, the negatively controlled triac 20 via a diode 24 from the negative half-wave of a 50 Hz supply and the positive controlled -Triac 19 is fed via a diode 25 from the positive half-wave of this supply. The polarity of the The control signal determines which triac conducts. The output signal of the demodulator is obtained from the voltage across the resistor 21. It is very convenient that a triac is switched on at the beginning of the respective half-wave. This can be traced back to Pig. 2 way shown thereby

209885/0973

223536?

achieve that the control or gate signal outside the required Control period of the supply frequency, i.e. between 0 and 10 or between 180 and 190, short-circuited to zero will. The gate electrodes are short-circuited by transistors £ 6, 27, the base electrodes of which are switched by that they are supplied with the AC supply voltage via rectifiers 28, 29 with capacitor-resistor filters 30, 31 will,.

If the charging speed of the capacitor 12 of the Integrator changed during this initial proportional period without affecting the integral action, so is a third term has been introduced which enables a practical determination of how long after a step input signal has been applied the exit device at maximum speed runs. This integration acceleration possibility can be achieved in the manner shown in FIG. 3 by that two zener diodes 32, 33 connected against one another vjrv / en -'- det selected in such a way that they do not end the power line until the negation signal is in the vicinity of the one Value that is required for complete negation. The Zener diodes connected in opposition and in series 32, 33 are parallel to the output of the amplifier 10 and in series with a potentiometer 34, the tapped on it Voltage is supplied to the input of the amplifier 11. 3 shows a further amplifier 35, which may expediently be inserted.

During the conductive period, the charging rate of the integrator is increased by the signal from the Potentiometer 34 is fed to the summation point of amplifier 11; the potentiometer '34 allows this to be set additional charging current to the required value,

An important property of the integrator is that the speed of the voltage that occurs at the output

209885/0973

change is constant and not a function of the setting of the leakage resistance 18. This means that there is always a half period the supply frequency is sufficient to accommodate the integrator negations to pass voltage through the hysteresis of the triac gate conditions; so only a half period is generated, before the triac switching voltage is negated so much that it falls into the dead band.

In Pig. 4 shows a block diagram of a process control suitable for low-frequency operation. Error input signals are on terminals 40 and are fed to a summing amplifier 41 with a gain factor of 1. The amplifier may include a noise filter capacitor. The amplifier output voltage is fed to two independent variable gain amplifiers 42, 43, one of which is the amplifier 42 forms a transient switching amplifier.

The output of amplifier 42 becomes positive and negative error detectors 44, 45 supplied. If the output signals exceed a certain value, a transition switchover relay is activated 46 actuated depending on the output polarity. The effect of this relay is to switch off, output signal of the controller to a reduced voltage supply, if the fault in the system under falls a certain level.

The other amplifier 43 is a proportional band amplifier, which is also fed from the summing amplifier. The gain factor of this amplifier determines the proportional band of the system. The output of the amplifier 43 is fed to the inverting input of an error and integration sum amplifier 47.

The output of amplifier 47 is not due to this inverting input of a control amplifier 48 for a modulator 49 as well as two counter-connected Zener

20988B / 0973

diodes 50, 51. The diodes are according to Pig. 3 fed back at a point with the voltage zero via a potentiometer 52, the Tap 53 via a co-determining the time constant fixed resistor 55 is connected to an input of an integrator 54. The potentiometer 52 determines the proportional Negation or a derivative term as described above.

The modulator control amplifier 48 has a fixed low gain. The modulator is powered by an alternating voltage of conveniently 50 Hz or mains frequency fed, and the modulating triacs contained in the modulator are set so that they are at a fixed Switch positive or negative voltage at the modulator input, for example positive or negative at 3 volts. is If the gain factor of the modulator amplifier is set to the value, this corresponds to a voltage of j-1 volt at the output of the error and integration summing amplifier. The two Zener diodes 50, 51 switch on, for example + 5 volts. and act only when the modulator control amplifier is saturated. As in FIG. 2, the two triacs are arranged in the modulator 49 so that one switches on and the positive half-wave of the applied AC voltage conducts when the gate signal is positive, and the other the negative Half-wave conducts when the gate signal is negative. A third triac is used to generate the positive and negative half-wave pulses from the other two triacs to combine to a voltage opposite the zero voltage line. The two Transistors which are used to short-circuit when the output signal from the modulator control amplifier occurs prevent the triacs from igniting except during the desired Perio.de at the beginning of the respective half-wave.

The integrator 54 is fed with two signals, from. those the first, as described above, comes from the two Zener diodes 50, 51, while the second signal comes from the

209885/0973

- 223536?

dulator 49 generated positive or negative half-wave pulses is formed. These output signals are combined and the integrated signals are connected to ' the non-inverted input of the error and integration summing amplifier 47 is fed back. A variable resistor 48 is parallel to integration capacitors 59, 60, which are advantageously two mutually polarized Electrolytic capacitors. Resistor 58 has two functions: It determines the gain factor of the integrator and thus the maximum output voltage, and it determines the capacitor discharge rate. The size of the error signal, against which the integration is effective, as well as the speed with which the integration capacitors in a given errors are thus determined by the resistor 58.

The output signal of the modulator becomes a pulse shaping stage 61 supplied. The input of this pulse forming stage includes pnp and npn transistors, which are respectively the positive or Switch negative half-wave of the modulator output signal. The leading edge of the resulting square wave is differentiated and fed via the contacts of the transfer switching relay to one of two thyristor pulse ignition transformers.

Two transition stages 62, 63 are connected to the above-mentioned pulse control relay 46, the output signal of the The pulse forming stage is controlled via the contacts 64 of the relay 46 and fed to one or the other transition stage 62, 63; the transition stages in turn control triacs 65 and 66, which give a motor 68 a lower or higher voltage feed from the winding of a feed transformer 67.

The arrangement of FIG. 4 is somewhat more detailed in FIG shown. In both figures wear the same parts the same reference numerals. In view of the above description, a complete description of FIG. 5 should be understood superfluous,

209 8 8 5/0973

223536?

In Pig. 5 are the triacs of the modulator at 68x and 69 shown in connection with another triac 70. The pulse shaping stage 61 comprises input transistors 71 and 72 as well an output transistor 73. - '

In the transition stages, the input signal from the pulse forming stage via the contacts 64 and a pulse transformer 75 to a thyristor shown approximately at 76 which has a capacitor 77 with a resistor 78 connected in parallel in its cathode circuit. the The anode of the thyristor 76 is connected via a resistor 79 to a capacitor 80, which is connected via a resistor 81 is fed from a bridge rectifier 82. Another capacitor 83 is parallel to the output of the bridge rectifier. For the output pulses of the pulse forming stage 61, which are of the opposite polarity, a similar circuit is provided which includes a transformer 86, a Thyristor 87, a capacitor 88, a resistor 89 and has a resistor 90. When an input pulse is received, the relevant thyristor 76 or 87 switches on and charges the capacitor 77 or 88 until the flowing current falls below the required holding current, whereby the thyristor then switches off; the supply is limited by the time constant of the circuit 80, 81.

The voltage that builds up across the capacitor 77 or 88 is applied to the gate of a triac via a diac 92 or 93 94 or 95 »with pulses being generated until the capacitor 77 or 88 is discharged. It should be noted that it is with a Low-inertia motor with low inductance, such as that formed by one with a printed armature circuit, is possible to use the back EMF generated by the armature as an indication of the armature movement, in order to further feed the armature to interrupt if, for example, a driven element reaches a positive stop. With a conventional one The winding inductance of the motor is too large to produce a rapid Allow drop in back EMF and the present result

209 8 8 5/0973

to be significantly outbid. By using, for example, a thyristor and a resistor for the negative or successful Half-cycle of the applied voltage thus generates a measuring voltage that can be used for control.

Fig. 6 shows a circuit detail of an arrangement which enables the generation of a train of pulses of high energy in Dependence on a single pulse of low energy is permitted and can be used to trigger thyristors. An input pulse is applied to the primary winding 100 of one Pulse transformer 101. The secondary winding of the transformer is connected between gate and cathode of a thyristor 102, its anode is connected to the connection point of the cathodes of diodes 104 and 105 via a resistor 103 is. The anodes of the diodes 104 and 105 are connected to windings 106 and 107 of a transformer 108 with a center tap, so that the voltages on the two anodes have the same amplitude and are 180 ° out of phase. The smock tap between windings 106 and 107 is taken off at a point of zero potential. The cathode of thyristor 102 is connected to zero potential via a capacitor 109, to which a resistor 110 'is connected in parallel. Between between the cathode of the thyristor 102 and a resistor 113, a diac 111 is connected in series with a limiting resistor 112. The other end of the resistor 113 is connected to the zero potential.

An input pulse on winding 100 switches the thyristor 102 and the capacitor 109 charges at a rate determined by the resistor 103. The voltage across the capacitor 109 rises to a value at which the diac 111 switches on, whereupon the capacitor 109 overcomes the diac discharges to resistor-113. The tension on this Resistance is, for example, fed to the gate of a triac 114, which turns on and continues to discharge capacitor 109.

209885/0973

223536?

The voltage across resistor 109 drops below the voltage which is necessary to maintain the flow of current through the diac; however, the thyristor 102 remains by the action of resistor 110 and capacitor 109 in the conductive state, until the current falls below the holding value. The voltage on capacitor 109 then rises again until the Diac 111 switches on again. This cycle repeats itself over part of the half cycle of the voltage from the transformer 108, at the end of this period the thyristor 102 switches off; this becomes dependent on the input pulse generated a series of pulses.

As described here, circuits can be designed that provide a highly satisfactory level of control, The output pulse frequency can be used as an integral effect in relation to the input signal by adjusting a previously set Control knob can be changed. With the minimum setting, the output pulse has the highest repetition frequency, i.e. the Besugs supply frequency, when the input signal of the controller just exceeds the selected electrical dead band. at the maximum setting reaches the 'output pulse frequency the maximum frequency as long as the input signal has the preselected maximum value. The relationship between the input signal level and the output signal frequency can be linear can be made, and the range of available integral times is permitted due to the flexibility of the design use as a fast position servo mechanism or as a valve timing and control mechanism in operations, where delays of minutes can occur.

If the supply is used as the reference frequency, then a bidirectional thyristor used to supply the negative or the positive half-cycles to the output load to switch. The number of half-periods switched per unit period is a function of the input signal level and the setting of the integral action. The printed engine

209885 / Ü973

¹⁵ "223536?

is particularly suitable as output load for the control ", and "when used in conjunction with a triac, high performance "switch cheaply. -With higher reference frequencies alternately switching semiconductors are used. The maximum switching frequency is determined by the motor losses, the utilization factor and the permissible motor efficiency.

In many applications for servo mechanisms that do not have require continuous control, the described control with a printed motor, for example, has the advantage of that quickly reacting position servo devices can be built without having to resort to tacho generators, because the increment method allows flexible approach speeds. In doing so, overshoot can be eliminated, and mechanical play can be achieved by slowly approaching the Switch off point of deviation from only one direction. Is will always develops maximum torque. The problems of inertia and delay known in DC systems can be solved by setting the integral action control to optimal approach speeds. Of the Operation with triacs allows high performance to be controlled with minimal cost or space. Control units encapsulated and with externally accessible devices for controlling the proportional band and the integral time be provided.

The resolving power of the system depends on the duration of an output signal increment or step and the total Movement time of the exit device is determined. For example would have an output device with a full movement time of 2 seconds and a step pulse length of 10 m / sec a maximum release capacity of about 1: 200. A significant improvement in performance can be achieved automatic reduction of the step size within the achieve equal range of the system.

2098 85/097 3

This is achieved in that the two output triacs and a feed transformer with tap are operated at a previously set point determined by the dynamic properties of the system, the logic circuit returning the gate pulses to the second triac connected to a low supply voltage. The above-mentioned operating point can be preset using a potentiometer that is mounted on the front panel of the device.

In order to overcome the limitations of the relatively low-frequency steps of the system operating with a reference frequency of 50 Hz an internally generated reference frequency can be used. This "frequency" can have any value; for example, a 1 KHz system has proven practical.

Pig. 7 shows a circuit diagram of an arrangement which is used for generating positive and negative values relative to a zero value Signals suitable and at high frequency instead of referring to Pig. 2 or 5 described triacs can be used can.

According to FIG. 7, two triacs 130, 131 are triggered by input pulses via diodes of opposite polarity 132, 133 are supplied. The triacs are connected via diodes 134 and 135 as well as resistors 136 and 137 are fed. The voltage at the resistors is at the base electrodes of transistors 138 and 139, which are of opposite conductivity types and at their collector load resistors I40 and 139, respectively. 141 output signals are generated at 142 and 143, respectively.

Pig. 8 shows a circuit diagram for the power stage of a control circuit suitable for use at high frequency. The circuit shown includes terminals 150, 151 for connection to a DC power source, such as a rectified AC power source. This source is used to feed a motor 152.

20988b / 0973

^'17 "" 223536?

The motor is powered by a number of controllable Silicon rectifiers or thyristors or their equivalents Semiconductor switching devices in conjunction with a capacitor 153 controlled by corresponding Schalttä'tigkeit the thyristors are charged and discharged via the motor winding ".

¹ The trigger potentials "for the thyristors are supplied by a trigger control unit 154 which operates as described above and generates trigger pulses, but at a relatively high frequency; therefore, the current supplied to the motor is also in the form of rapidly occurring current pulses . ·. ·

In the circuit eight thyristors are shown, which the Bear reference numerals 161 to 168. There are the thyristors

161 and 162 and the thyristors 163 and 164 each in series switched on between terminals 150 and 151. The motor lies between the connection point A of the thyristors 161 and

162 and the connection point B of the thyristors 163 and I64. The thyristors 165 and 166 are between the connection point A and a terminal of the capacitor 153 with opposite polarity connected in parallel; the other terminal of the capacitor is connected to terminal I5I of the supply. The thyristors 167 and 168 are also polarized opposite and parallel between the connection point B and the mentioned one terminal of the capacitor 153. Die'mit a to d designated trigger electrodes of all thyristors are on the correspondingly designated outputs a to d of the control unit 154 are connected.

First the thyristors 161 and 168 are switched on, and a current flows from the terminal 150, via the thyristor 161, through the motor 152, the thyristor 168 and the capacitor 153, so that it charges and the charge current charges the motor excited. The charging current decreases exponentially until the thyroid

209885/0973

223536?

Switch off transistors 161 and 168 due to the lack of holding current. Then the thyristors 166 and 164 are turned on, so that the capacitor is discharged, the discharge current passing through the motor in the same direction and thus maintains engine excitation. This cycle is then repeated.

■ The motor can be driven in the opposite direction by first turning the thyristors 163 and 165 to Charging the capacitor and then the thyristors 167 and 162 are switched on to discharge the capacitor.

The operating frequency can be high and a frequency up to 4 KHz has been found practical. The capacity of capacitor 153 is selected with regard to the operating frequency and resistance of the motor windings. The supply voltage at terminals 150 and 151 is increased the rated voltage of the motor. In practice, the voltage can be in the range from 20 to 230 V, for example; for a motor with a resistance of 1 ohm, the capacitor C had a capacitance of 50 µF.

209885/0973

Claims (16)

Claims 1J engine control system, characterized by means responsive to an input signal and means responsive to a deviation in the input signal responds beyond a dead band and supplies the motor with current pulses in one direction, which corresponds to the deviation of the input signal beyond the upper or lower limit. 2. Motor control system, characterized by a first puncture amplifier (10; 48) with a feedback loop, which comprises a second functional amplifier (11; 54) operating as an integrator, and also one power control stage controlled by the output signal of the first functional amplifier, an input signal being supplied to the input of the first functional amplifier whose polarity and size is a measure of the deviation for measuring a control variable and where the output signal of the first functional amplifier is pulse-shaped and is fed to the power control stage for generating current pulses, the direction of which is determined by the direction of the Deviation is controlled. 3. System according to claim 2, characterized in that that the feedback loop for the first functional amplifier (10) voltage devices (16, 17) for determining an upper and a lower limit level of the through the Feedback loop conducted signal and thus includes for determining a dead band. 2 0 9 Ü B b / Ü 9 7 3 4. System according to claim 2 or 3, characterized in that the power stage has a device (62, -63, 65, 66) for generating current pulses at different levels for feeding the motor (68). , 5. System according to claim 4, characterized in e i c hnet, that the means (62, 63) is responsive to the level of the input signal. 6. System according to one of claims 2 to 5 »marked by means (18; 58) for changing the integration speed of the integrator (11-, 54) according to the output signal of the first puncture amplifier (10; 58) to form a derivative control. 7. System according to claim 6, characterized by two oppositely polarized zener diodes (32, 33; 50, 51) connected in series with a resistor (34; 52), which are connected to the output of the first puncture amplifier (10; 48), the voltage across the resistor at the input of the integrator (11; 54). lies. 8. System according to one of claims 2 "to 7, characterized by a modulator (49) to which the output signal of the first puncture amplifier (48) and a AC voltage and the half-cycle of the applied voltage according to the output signal of the puncture amplifier selects. 9. System according to one of claims 1 to 8, characterized in that the circuit gives the motor current pulses high frequency feeds. 10. System according to claim 8 and 9, characterized in that the modulator has a first circuit a first triac (19; 130) in series with a first diode 20988 5/09 7 3 2235387 (25; 134), a second circuit with a second triac (24; 131) in series with a second diode (23; 135) comprises that the two circuits are parallel to an alternating current source are connected, that the two diodes are polarized in opposite directions, and that an input signal to the trigger electrodes the triacs is supplied via two further diodes (22, 23; 132, 133) in order to control the triacs alternately and the output signals from the first and second circuits alternately to derive. ' 11. System according to claim 10, characterized by a resistor (21) connected in the first and the second circuit, from which an output signal is derived will. 12. System according to claim 10, geke'nnzeic h · η e t in each case switched on in the first and in the second circuit Resistor (136, 137), the output signals of two transistors (138, 139) from opposite Conductivity types are derived and the voltages at the resistors between the emitter and base of the Transistors lie. 13. System according to claim 10, characterized . by a switching device (26, 27) for short-circuiting the trigger inputs of the triacs outside the trigger periods. 14. System according to any one of the preceding claims, characterized in that the power level is one Capacitor (153), means for charging the capacitor and means for discharging the capacitor as well as a switching device (154, 161 ... 168), which carries the charge and discharge currents through the motor (152). 2098öb / Ü9 7 223536? 15. System according to claim 10, characterized net that the switching device is controllable silicon rectifier (161 ... 168), which are triggered by switching pulses being controlled. · 16. System according to one of the preceding claims, characterized by a modulator for . Generation of a large number of output pulses depending on from an input pulse. 209btf5 / 0973 Blank page