GB2417625A

GB2417625A - Digital adaptive control of IGBT or MOS gate charging current in a converter for a railway traction motor

Info

Publication number: GB2417625A
Application number: GB0419257A
Authority: GB
Inventors: Waldemar Belwon
Original assignee: Bombardier Transportation GmbH
Current assignee: Alstom Transportation Germany GmbH
Priority date: 2004-08-31
Filing date: 2004-08-31
Publication date: 2006-03-01
Also published as: GB0419257D0

Abstract

The capacitance of an IGBT gate is determined without turning on the IGBT (figure 7), and the driving current waveform for the IGBT gate when the IGBT is conducting is modified in dependence on the capacitance. In this way, an optimum switching speed may be obtained despite replacement of the IGBT by a newer or different IGBT. The optimum switching speed reduces switching losses, prevents shoot-through current and allows conformance with EMC (electromagnetic compatibility) regulations. The gate current in distinct phases of the charge or discharge cycle can be adapted independently (figures 3 and 4). The switching response of the IGBT in one cycle may be monitored and used to modify the gate drive waveform in subsequent cycles.

Description

24 1 7625 Method and Arrangement for Operating an Electronic Valve

Description

The invention relates to a method of operating an electronic valve, in particular a valve adapted for use in a high power converter such as an Insulated Gate Bipolar Transistor (1GBT), wherein a process of changing a charging state of a control electrode of the valve is repeatedly performed in a similar manner. More particularly, the invention relates to converters comprising such valves, e.g. for delivering electric energy to a driving motor of a railroad traction vehicle. Furthermore, the invention relates to a corresponding arrangement for operating the electronic valve.

Controlling the valves by using pulse signals generated by a control device is well .

known in the art. In particular, a first edge of a pulse signal initiates a process of switching-on of at least one of the valves and a second edge of the pulse signal,., initiates a process of switching-off of the valve or of the valves. The first edge may. A be defined by a step-like increase of a voltage from a low potential to a high potential at a control signal input of a driver circuit, or, in the case of a light pulse, by a transition from dark to light at a control signal input. The second edge may be.. . defined by a step-like decrease of the voltage from a high potential to a low potential at the control signal input or by a transition from light to dark at the control signal input. In particular, the initiation of the switching process causes a current flow to or from the control electrode (e.g. gate) of the valve, whereby the control electrode is charged or discharged and, as a result, a current flow through the valve is enabled or disabled. However, other ways of initiating the switching process are as well possible.

In recent times, collector-emitter voltages of IGBTs have been increased significantly. For example, it has become possible to switch voltages of some thousand volts, for example voltages ranging between 3300 or 6500 volts. As a result, the task of reducing switching losses has gained in significance, particularly where IGBTs are concerned, which are used in converters. The switching speed is an important factor which is to be taken into account for any strategy in reducing switching losses. Unfortunately, the switching speed depends on several quantities, such as the individual properties of the IGBT, the operation temperature, the level of the voltage to be switched, the level of the current flowing through the valve and any inaccuracies of a driving device for driving an electric current to or from the control electrode. If a valve, connected in series to another valve is switched too fast, a short circuit may occur. If the switching speed cannot be controlled precisely, it is therefore necessary to slow down the switching process.

There are different approaches of performing the control process for switching the valve on or off. All these approaches can be realised in connection with the invention, which will be described later. For example, a voltage source with a controllable voltage can be used in combination with a resistor, located in an electrical connection between the voltage source and the control electrode of the valve. According to another approach, a voltage amplifier with pre-programmed .

voltage slopes may be used. Furthermore, a current source amplifier with pre programmed levels of the current to or from the control electrode may be provided. 8, In any case, the process of switching the valve on or switching it off may be divided, r into a plurality of time zones, wherein the voltage of the voltage source, the voltage slope of the voltage amplifier or the current level of the current source amplifier '.

differ from time zone to time zone. The respective time zone is terminated by an.

event, such as the event when the gate-emitter voltage of an IGBT reaches a predefined level.

Power semiconductors, in particular valves, are continuously improved by developers. As a result, there is an ongoing demand for replacing the valves in existing devices, such as in converters. For example, the lifespan of a converter may be 30 years or longer. However, the power semiconductors may be replaced several times during the life-span.

The control devices which control the operation of the valves are adapted to the properties of the valves. Especially in view of the problems and requirements mentioned above and in view of the fact that switching occurs at high switching frequencies, the control devices need to be adapted to the characteristics and properties of the valves. When the valves are replaced, the control device can be replaced as well or - preferably - the control device can be adapted to the new valves (e.g. by updating the hardware and firmware). However, the control device must be removed and re-arranged in either case. Highly qualified staff is required for this task and several tests must be performed afterwards in order to ensure correct operation. However, it might be difficult to find staff with specific knowledge about the control device and about the test equipment when the valves are replaced after 10 or 15 years.

It is an object of the present invention to provide a method and an arrangement of the kind indicated above, which reduce the effort related to replacement of the valve or valves. In particular, the operation of the valve after replacement should comply with safety regulations and regulations concerning electromagnetic compatibility (EMC), ë.; ë;e The following is proposed: A method of operating an electronic valve, in particular a valve adapted for use in a high power converter, wherein the valve comprises a...

control electrode, in particular an insulated gate. The control electrode is repeatedly. , ., charged and discharged during an operation of the valve in order to change a ë e switching state of the valve. A charging process of the control electrode is . .

investigated and a corresponding determination result is obtained. The determination.. : result is used to control charging and/or discharging of the control electrode during the operation of the valve.

The term "charging process" includes the possibility that the control electrode is charged and/or discharged during at least a part of the process.

The electronic valve may be any valve whose switching process can be controlled by directly or indirectly adjusting the current to or from the control electrode. Examples arc: IGBTs and Metall Oxide Field-Effect Transistors (MOSFETs), or more generally speaking, valves that have an insulated control electrode (gate). However, IGBTs are preferred for highpower applications, such as converters used to provide electric energy to power consumers like driving motors or industrial machines.

The charging process is preferably investigated prior to the operation of the valve and/or at the beginning of the operation of the valve. In particular, the charging process which is investigated takes place before the first charging process that is performed in order to switch on the valve or the investigated charging process may be the first charging process or part of the first charging process that is performed in order to switch on the valve. The term "first" denotes the first process after a period of time (e.g. a period of some seconds to more than several hours) in which the valve is not switched on (non-operation period) or it denotes the first charging process of the valve at all after integrating the valve in an electronic arrangement, such as a converter. Furthermore, the investigation of the charging process may be repeated during the operation of the valve and/or again after a non-operation period. . Furthermore, it is preferred to charge the control electrode until a defined charging state is reached. In particular, the charging state is characterized by a voltage of the...

control electrode (e.g. a gate-emitter voltage) which is smaller than a threshold value, . . . wherein the threshold value is smaller than a voltage that is required for switching on the valve. . The term "investigation of the charging process" includes the possibility of investigating a plurality of charging processes.

It is preferred to perform the investigation when a switching voltage between switching electrodes of the valve is zero and/or is smaller than during the operation of the valve.

The term "switching voltage" denotes a voltage across the valve when the valve is switched off, wherein the switching voltage becomes zero (or approximately zero) by switching on the valve. In particular, the valve comprises a pair of switching electrodes (such as collector and emitter in case of an IGBT) which are electrically connected by the valve when the valve is switched on and which are electrically insulated by the valve when the valve is switched off and wherein the charging process of the control electrode is investigated when a switching voltage between the switching electrodes is zero and/or is smaller than during the operation of the valve.

The determination result may be used to set a set value (e.g. an initial value for the operation of the valve) of a charging quantity, wherein the charging quantity is used to control and/or is used to perform the charging and/or the discharging of the control electrode during the operation of the valve. Examples for the charging quantity are - the length of a time interval for charging or discharging of the control electrode, wherein the charging current or discharging current is constant during the time interval, - the level of the charging current or discharging current during charging or discharging the control electrode, - a time derivative of a voltage between the control electrode and another electrode of the valve (e.g. the voltage between the gate and emitter of an IGBT), wherein the voltage rises or falls according to the time derivative. .

The value of the charging quantity may be adapted during the operation of the valve.

For example, the charging quantity corresponds to the length of a time interval..

during which a charging current used to charge and/or discharge the control electrode...

is kept constant or corresponds to a predetermined point in time at which a level of the constant charging current is changed. :e .

A preferred way of investigating the charging process of the control electrode comprises determining an electrode charge and/or an electrode charge difference which is carried by the control electrode. The term "charge difference" means a difference between an earlier and a later charging state of the electrode. For example, the charge is necessary to produce a specific voltage between the control electrode and another electrode of the valve. Since the necessary charge differs for different types of valves, the charge is a direct measure for the properties of the individual valve and gives useful information how to perform the charging process.

In particular, the charging current and/or the time interval of charging can be adapted to the individual valve. Furthermore, this enables the user and/or an automatically operating arrangement to define the control strategy for operating the valve before the valve is switched on. At least, an initial control strategy and/or an initial value for control can be defined before the valve is switched on for the first time.

However, the electrode charge is not the only measure for the properties of the individual valve. Other ways of investigating the charging process are described later.

In particular, a time interval needed to charge or discharge the control electrode can be measured and the time interval can be used to calculate the electrode charge and/or the electrode charge difference. The time interval may start at a first defined state of the valve and may end at a second defined state of the valve. For example, the first and/or the second defined state are defined by an electric voltage between the control electrode and a second electrode of the valve (e.g. the gateemitter voltage of an IGBT).

According to a preferred embodiment of the invention - the charging process, which is investigated, is repeatedly performed, . - a current used to charge the control electrode is kept constant during each of the charging processes, - the charging time is equal for all of the charging processes and A - the current is increased or decreased from charging process to charging. . . process until a predefined operation state of the valve is reached at the end of the charging process. The predefined operation state may be a voltage level, e.g. a voltage level of the gate-emitter voltage of an IGBT. A corresponding event may be defined and/or the predefined operation state may be detected using a comparator, in particular a voltage comparator. The comparator compares the instantaneous operation state of the valve with the Redefined operation state of the valve. The defined event "happens" when the Redefined operation state is reached and the comparator may indicate this by outputting a corresponding signal. An operatability of the comparator may be checked before the charging process.

In particular, the control electrode may be charged or discharged at a level or at levels of a charging current during the charging process of the control electrode which is investigated. The same level or levels of the charging current may be used to charge and/or discharge the control electrode during the operation of the valve.

This embodiment allows to directly use the result of the investigation. For example, a charging current level is determined during investigation, which current level is exactly the level that will cause the valve to reach the predefined operation state at the end of the fixed charging time. Then, the same current level is used to operate the valve under normal operating conditions. Alternatively, the determined current level may be used to calculate the current level, which is used during operation of the valve. Furthermore, different current levels may be used during operation of the valve and at least one of the current levels may be equal to the determined current level and/or may be calculated using the determined current level. Preferably, the determined current level is a level used to pre-eharge the control electrode in a charging process with a plurality of charging time intervals (e.g. the time zones described below), wherein the pre-charging prepares the switching of the valve, but does not switch on the valve.

Furthermore, the invention includes an arrangement for operating an electronic .

valve, in particular a valve adapted for use in a high power converter, wherein the arrangement comprises: ..

- a driver device adapted to drive an electric current to and/or from a control. . . electrode, in particular from an insulated gate, of the valve, - - a determining device adapted to investigate a charging process of the control ë- electrode and adapted to obtain a corresponding determination result, .. :.

- a control unit, wherein the control unit is connected to the determining device or comprises the determining device and wherein the determining device is adapted to produce information which can be used to control charging and/or discharging of the control electrode during the operation of the valve.

The determining device may be connected to at least one current control unit and the current control unit may be adapted to save at least a part of the information.

Furthermore, the current control unit may be adapted to control the driver device and the current control unit may directly be connected to an output unit which is connected to the driver device.

In addition, the invention includes a converter for converting a current, in particular for a high power application such as providing electric energy to a driving motor of a railroad traction vehicle, wherein the converter comprises the arrangement as described in this description, wherein the converter can be operated by repeatedly switching-on and off the valve and wherein the arrangement and the control electrode are connected to each other via a connection for discharging and/or charging the control electrode.

Also, the invention covers a programmable logic device comprising the control unit and the determining device of the arrangement as described in this description, wherein the programmable logic device is adapted to investigate the charging process of the control electrode according to the method as described in this description. For example, modern logic devices, such as CPLDs (Complex Programmable Logic Devices) and/or FPGAs (Field Programmable Gate Arrays) may be used as the logic device. ' : The method and/or the arrangement of the invention may be combined with a method and/or arrangement of operating the valve. In this context, it is proposed to define a specific operating state of the valve, which operating state can be reached by changing a charging state of the control electrode of the valve, and to monitor whether the specific operating state is reached at a predetermined point in time. For '''. -e

example, the predetermined point in time may be predetermined by a time interval of a set length. The starting point of such a time interval may be defined by receipt of a control signal at a control unit for controlling the process of changing the charging state of the control electrode, wherein the process is initiated on receipt of the control signal or a time period of predetermined length after the receipt. Preferably, the length of the time interval may be fixed for all similar processes of changing the charging state of the valve, wherein each of the similar processes is to be performed in one of a plurality of consecutive procedures of switching on or switching off the valve.

In particular, the following is proposed: A method of operating an electronic valve, in particular a valve adapted for use in a high power converter, wherein a process of changing a charging state of a control electrode (in particular of an insulated gate of the valve) is repeatedly performed, it is determined whether a specific operating state is reached at a predetermined point in time and a corresponding determination result is obtained and the process of changing the charging state is adapted using the determination result.

An arrangement for operating an electronic valve may comprise: a driver device adapted to drive an electric current to and/or from a control electrode of the valve, a determining device adapted to determine whether a specific operating state of the valve is reached at a predetermined point in time, and adapted to obtain a corresponding determination result and a control unit, wherein the control unit is connected to the determining device or wherein it comprises the determining device, and wherein the.

control unit is adapted to perform a future execution of a process of. ..

changing a charging state of the control electrode by controlling the driver device in a corresponding manner, taking the determination result .

into account.

The specific operating state is to be understood as a state in which at least one '2; quantity, in particular an electric quantity (such as the collector-emitter voltage, the '^.' gate-emitter voltage, the collector current or a derivative) fulfils a defined condition, e.g. a state in which the quantity is equal to a defined value.

Strategies for optimising the switching process have been discussed in detail in many publications before the invention. It is now possible to optimise the switching procedure in a simple, but highly effective way. The switching speed of the valve can precisely be controlled by monitoring whether the specific operating state is reached at the predetermined point in time or not.

An embodiment of the operation will be described by way of example later. The investigation of the charging process, in particular before the operation, can explore all information which is necessary to adapt the operation to a specific valve. In particular, the charging or discharging current may be determined by the investigation as already mentioned above and the same level of the current may be used during operation.

Preferred embodiments of the invention will be described in more detail, with reference to the accompanying schematic drawing by way of example. However, the following description is not intended to limit the scope of protection of the invention.

In the drawing, the same reference numerals are used for identical parts or units and for parts or units with a similar function. The figures of the drawing show: Fig. 1 an arrangement according to the most preferred embodiment of the invention; Fig. 2 an arrangement with a converter and a load; Fig. 3 a gate current as a function of time during a process of switching on an 1GBT; - ^ Fig. 4 a gate current as a function of time during a process of switching off an IGBT; Fig. 5 a part of the arrangement shown in Fig. 1; Fig. 6 a part of the arrangement shown in Fig. 5; ; Fig. 7 four diagrams illustrating a charging process of the control electrode of a valve according to a first method of investigation; Fig. 8 four diagrams illustrating a first charging process of the control electrode of a valve according to a second method of investigation; and Fig. 9 four diagrams illustrating a second charging process of the control electrode of a valve according to the second method of investigation, wherein the charging current has been amended compared to the first charging process.

Fig. I shows an arrangement 31 (a so-called Gate Drive Unit) with an IGBT 18. The insulated gate of the IGBT 18 is denoted with the reference numeral 17, the collector with the reference numeral 20 and the emitter with the reference numeral 19.

A control unit 33 for controlling processes of switching on and switching off the IGBT 18 (and optionally at least one further 1GBT) is provided. A driver 34 (in particular a gate current amplif er) for driving a charging current to the gate 17, a driver 36 (in particular a gate current amplifier) for driving a discharging current from the gate 17 and signal outputs of voltage signal generators 35, 37, 39 are connected to the control unit 33. The first voltage signal generator 35 is connected to the gate 17 for generating a signal corresponding to a voltage of the gate 17, in particular the voltage V(-}E between the gate 17 and the emitter 19. The second voltage signal generator 37 is connected to the collector 20 for generating a signal corresponding to a voltage of the collector 20, in particular the voltage VCE between the collector 20 and the emitter 19. The third voltage signal generator 39 is connected to an emitter power terminal 38 of an IGBT module. The third voltage signal generator 39 generates a signal corresponding to a voltage Via i. which is caused due to an inductance of the conductor between the emitter 19 and the emitter.

power terminal 38. In particular, the voltage VEE may be the voltage between the emitter power terminal 38 and a reference potential, wherein the emitter may directly be connected to the reference potential (denoted with "O V" in Fig. 1). The voltage .

VEE IS caused by dlC/dt (the derivative of the collector-emitter current of the IGBT)". . . . ë A further signal generator 32 is connected with the control unit 33. The signal generator 32 comprises a first input for a positive electric potential (denoted with .

"+V") and comprises a second input for a negative potential (denoted with "-V"). A first supply supervision signal is transferred from the signal generator 32 via line SS to the control unit 33 when an arrangement (such as an intermediate circuit of a power converter) is in a normal operation state. The arrangement is connected to the valve 18 or comprises the valve 18. For example, the positive potential is greater than a threshold value and the negative potential is smaller than a threshold value, if the arrangement is normally operated. A second supply supervision signal is transferred from the signal generator 32 via line SS to the control unit 33 when the arrangement is out of normal operation (e.g. switched off) and, therefore, the valve is not operated normally. Thus, the supply supervision signals can trigger and/or enable processes which are to be performed either during normal operation of the valve 18 or when the valve 18 is out of normal operation.

A signal line for transferring control signals from a control device 23 (not shown in Fig. I) to the control unit 33 and from the control unit 33 to the control device is denoted with letters "CC". The control unit 33 may be a programmable logic device, such as a CPLD (Complex Programmable Logic Devices) or a FPGA (Field Programmable Gate Array). The IGBT 18 and further electronic valves - may be part of a converter such as the converter 21 shown in Fig. 2, wherein the converter 21 is connected to an energy supply circuit (e. g. a direct current intermediate circuit) or to an energy supply network 27 (e.g. a single-phase alternating current network of a railway system). Furthermore, the converter 21 is connected to a load 25 (e.g. a driving motor of a railroad traction vehicle) and its operation is controlled by the controlling device 23, which controls the operation using PWM (Pulse Width Modulation) signals, for example.

The operation of the arrangement shown in Fig. 1 will be described in the following with reference to Fig. 3 and Fig. 4. When a corresponding control signal is received by the control unit 33 it initiates switching on the IGBT 18. The procedure is divided into a plurality of time zones (first zone: from time It to time t2, second zone: from .

time t2 to time t3, third zone: from time t3 to time t4, fourth zone: from time t4 to time t5), wherein the control unit 33 causes the driver 34 to drive a gate current IG of....

a given, constant height in each of the time zones. The height or level of the gate. . current IG in at least some of the time zones is dependent on a correction procedure described in the following.

On the other hand, the lengths of the time zones do not depend on the correction procedure. Rather, the lengths are fixed for a given operation state of the arrangement shown in Fig. 2. For example, the lengths may be adapted, if the load 25 is operated in a different manner. In the particular example of Fig. 3, the length of the first time zone is 1 Us plus/minus 20 ns, the length of the second time zone is 0.5 As plus/minus 20 ns, the length of the third time zone is 0.3 Us plus/minus 20 ns and the length of the fourth time zone is 1 As plus/minus 20 ns.

In an alternative embodiment, the length of at least one of the time zones can be adjusted, for example in order to adapt the operation to modified operating conditions and/or in order to increase the safety and/or the accuracy of the switching process thus decreasing switching losses.

In order to perform the correction procedure and in order to decide whether the level of the gate current IG IS to be adjusted, the time of a predefined event occurring is monitored and/or registered. For example, a corresponding signal is transferred from an event detecting device 35, 37 and/or 39 to the control unit 33 when the event occurs. The event happens, for example, when the voltage V(,E between the gate 17 and the emitter 19 reaches a predefined voltage level. The control unit 33 determines whether the event occurred before, at or after a predetermined or defined point in time. If the event did not occur at the point in time (or, optionally, within a tolerance period of defined length, including the point in time), the control unit 33 either adjusts the level of the gate current IG for future use in the same time zone of the next following process of switching the valve, or it uses the information obtained to .

...DTD: decide later whether such a correction is to be performed. .... :.

For example, the correction may be performed if the event repeatedlyoccurs too - early for the corresponding time zones of consecutive switching processes. For example, the correction may be performed only, if the event occurs too late in a. . predefined number (e. g. five) of consecutive switching processes. This may eliminate the influence of noise and/or any interference.

Clock signals, which are repeatedly transferred from a high precision clock (not shown) to the control unit 33, are used by the control unit 33 in order to precisely determine the defined point in time. For example, the control signal which initiates the process of switching the valve 18 triggers the start of a counting procedure for counting clock signals, wherein the predetermined point in time corresponds to a specific number of counted signals. When the number of clock signals counted is equal to the specific number, the defined point in time has arrived and the control unit 33 evaluates whether the defined event occurred before or after the end of the time zone (optionally plus or minus a tolerance time) . If the event has not occurred yet (optionally including the tolerance time period), it is decided that the event is too late (not in time).

If the control unit 33 decides that the process of charging or discharging the gate 17 is to be adapted in at least one of the time zones, it outputs a corresponding control signal to the driver 34 or 36, wherein the control signal causes the driver 34 or 36 to drive a gate current at the corresponding amended level during the respective time zone.

Furthermore, it is preferred to correct the height of the gate current IG by small steps only, e.g. by steps of smaller or equal to 1 % (preferably smaller or equal to 0.05 %) of a default value or maximum value of the current. The default value may be used as an initial value at the beginning of an operation of the valve, when the correction process has not been performed yet. After each correction step, it is checked whether the current is still too high (or too low) before a further correction step is performed.

The description of the correction procedure applies to all time zones of the switching. . . process for which an event is defined. In the first time zone of a process of switching :.

on the IGBT 18, the gate 17 is pre-charged. The corresponding defined event is the event that the voltage VGE between the gate 17 and the emitter 19 reaches a defined voltage level, e.g. a voltage level of 3 Volt. In the second zone the actual start of a current flow (collectoremitter current) through the IGBT 18 is prepared. The. . defined event is the event that an increase of the collector-emitter current occurs for the first time after the first time zone (or an increase is detected). For the third time zone no event is defined. In the third time zone a reverse recovery of a freewheeling diode (not shown in Fig. 1) is performed. The freewheeling diode is connected in parallel to a second IGBT, wherein the second IGBT and the IGBT 18 are connected in series to each other. E.g., the second IGBT and the IGBT 18 may be part of the converter 21 and may be used to control one phase of a three-phase alternating current.

In the fourth time zone the voltage VCE between the collector 20 and the emitter 19 is reduced. The defined event is the event that the voltage VCE reaches a defined voltage level.

Fig. 3 shows three time-dependent signals Sl, S2, S3, which are transferred from the event detecting device 35, 37 or 39 to the control unit 33 and which indicate the occurrence of a corresponding event.

The step-like increase of the signal S 1 indicates the occurrence of the event defined for the first time zone. As shown in Fig. 3, the event occurs after time t2 which suggests increasing the level (the plateau) of the gate current. The decision whether the level is actually increased or not may depend on the signal S I obtained for further executions of the charging process of the first time zone. A corresponding routine which is capable of making the decision is implemented in the control unit 33.

The step-like increase of the signal S2 indicates the occurrence of the event, defined for the second time zone. As shown in Fig. 3, the event occurs earlier than the time t3. This does not necessarily suggest the adjustment (in this case: the decrease) of the.

level of the gate current during the second time zone, since the event defined for the. ..

first time zone did not occur in time. There are different ways of handling this situation. According to a first approach, the occurrence time of the event is .

compared with time t3, the end time of the second time zone. According to a second approach, the occurrence time of the event is compared with the occurrence time of....

the event defined for the first time zone (when signal S 1 steps-up to a higher signal level), plus the duration of the second time zone. It is preferred to choose the start values for the gate currents in all time zones in such a manner that all events occur too late at the beginning of the valve operation.

The step-like decrease of the signal S3 indicates the occurrence of the event defined for the fourth time zone. As shown in Fig. 3, the event occurs slightly after the time t5.

Fig. 4 shows a corresponding procedure of switching off the IGBT 18. As in the case of Fig. 3, there are four time zones with constant gate current level, wherein an event is defined and the level of the gate current IG can be adjusted (as indicated by short vertical lines pointing up and down) in a corresponding correction procedure for three of the time zones. Three time-dependent signals S4, S5 and S6, which are transferred from the event detecting device 35, 37 and/or 39 to the control unit 33, indicate the occurrences of the corresponding events.

Before the operation of the valve as described above, a charging process of the valve is investigated in order to determine initial values of the gate currents IG used in the different time zones. A preferred embodiment of the control unit 33, which performs the investigation, is described in the following with reference to Fig. 5 and Fig. 6.

The control device 40 shown in Fig. 5 is part of the control unit 33 of Fig. I. The control device 40 comprises a first input for receiving the supply supervision signals via the line SS. Furthermore, it comprises a second input for receiving the control signals from the control device 23 via the signal line CC. The first input is connected with a first delay unit 41 and with a second delay unit 42. An output of the first delay unit 41 is connected with a non-inverted input 44 of a first logic combination.

device 43. An output of the second delay unit 42 is connected with an inverted input.

of the first logic combination device 43 and with an input of a second logic combination device 47. .

This part of the control device 40 (which can be combined with other elements as....

shown in Fig. 5) can operate as follows. When the supply supervision signal SS .

indicates that the supply voltage - and thus the Gate Drive Unit 31 - is in a range for operation, the delay unit 41 generates an output signal with a shorter delay time than the second delay 42. For example, the delay time of the first delay unit 41 may be to 400 ms and the delay time of the second delay unit 42 may be 2 to 4 s.

Generally, one reason for a delay time of the first delay unit 41 is that the conditions for investigating the charging process of the IGBT (or the valve) are more stable after elapse of the delay time.

Thus, after elapse of the delay time of the first delay unit 41, but before elapse of the delay time of the second delay unit 42, the first logic combination device 43 generates an output signal that enables the pulse generator 48 to output a pulse signal or train of pulses. Furthermore, since no corresponding signal is transferred from the second delay unit 42 to the second logic combination device 47, the normal operation of the valve 18 is disabled, even if there is a corresponding control signal via line CC.

However, after elapse of the delay time of the second delay unit 42, the first logic combination device 43 generates an output signal (logic '0' in the example) so that the pulse generator 48 is disabled (resulting in logic '0' at its output in the example).

Furthermore, since a corresponding signal is transferred from the second delay unit 42 to the second logic combination device 47, the normal operation of the valve 18 is enabled: Any control signal transferred via the line CC can now pass the second logic combination device 47.

Consequently, the functionality described before can be used to perform the investigation of the charging process before each normal operation of the valve. -

Further elements of the control device 40 are shown in Fig. 5. However, only these.

elements are described in the following which are useful to understand the functionality in more detail. . .

The output of the pulse generator 48 is connected to a first logic device 49 for....

switching on the valve 18. Furthermore, the output of the pulse generator 48 is . . connected (via further devices, such as a third logic combination device 51) to a .

second logic device 50 for switching off the valve 18.

The second logic device 50 is indirectly controlled by the pulse generator 48. For example as shown in Fig. 5, the output of the pulse generator 48 is connected to the third logic combination device 51. Furthermore, another input of the third logic combination device 51 may be connected to the output of the second logic combination device 47 and an output of the third logic combination device 51 may be connected to an input of a logic signal inverter 46 (which may be integrated to the second logic device 50), wherein the output of the logic signal inverter 46 is connected to the second logic device 50. If one of the inputs of the third logic combination device 51 receives a defined logic signal (e. g. logic "I " when the output signal of the pulse generator is at a high voltage level), it outputs a logic ON-signal to the first logic device 49, which causes the logic combination device 51 to switch on the valve 18. If none of the inputs of the third logic combination device 51 receives such a logic signal (e.g. when the output signal of the pulse generator is at a low voltage level during investigation of the charging process prior to the normal operation of the valve), the third logic combination device 51 outputs a logic OFF- signal to the first logic device 49. However, the logic signal inverter 46 receives the OFF-signal, inverts this signal and outputs a logic ON- signal to the second logic device 50. This causes the second logic device 50 to switch off the valve 18.

When the pulse generator 48 outputs pulse signals, the investigation of the charging process is enabled: The first logic device 49 and the second logic device 50 can control charging and discharging operations during the time intervals (defined by the higher voltage level) of the pulses. The first logic device 49 controls the driver 34 via signal line PC so that the gate 17 is charged. The second logic device 50 controls the driver 36 via signal line NC so that the gate 17 is discharged. Furthermore, the.. . - first logic device 49 and the second logic device 50 control the level of the charging or discharging current (as will be described by way of example for the first logic device 49 later). .

::-:.

According to a first embodiment, the control electrode is charged using a constant....

charging current and the time interval is determined which is needed to charge the . . control electrode up to a predefined gate-emitter voltage VC;E. The time interval may .

be determined by the first logic device 49. For this purpose, the first logic device 49 is connected to the voltage signal generator 35. Either, the voltage signal generator may output a signal which corresponds to the instantaneous voltage value or the voltage signal generator 35 (it may be a comparator, for example) may output a corresponding signal, if the instantaneous voltage value matches or exceeds the predefined gateemitter voltage VGE.

Fig. 7 shows four diagrams. The time scales (horizontal axes) of all four diagrams are identical. The diagram at the top illustrates a first logic signal L (e.g. the pulse signal on the line PG in Fig. 5 which is output by the pulse generator 48). The first logic signal L has a lower voltage level and a higher voltage level. The step-like increase of the first logic signal L from the lower level to the higher level triggers the commencement of a charging process. The step-like decrease of the first logic signal L from the higher level to the lower level marks the end of the time interval which is available for the investigation of the charging process.

The second diagram from the top schematically illustrates the electric current 1; to and from the control electrode as a function of time t, i.e. the charging and discharging current. The current IG IS constant during the charging process and is constant during the discharging process. Correspondingly, the diagram shows a step function.

The third diagram from the top schematically illustrates the gate-emitter voltage VGE as a function of time t. The dashed horizontal line indicates a voltage level which is the minimum voltage level necessary for an onstate of the valve, i.e. for an operation state in which the valve is conductive for electric currents from the collector to the emitter. The voltage VC,E is negative before charging and rises linearly from the. . beginning of the charging process until it reaches a threshold value (see below). . . Afterwards, the gate-emitter voltage VGF falls linearly until it reaches the same voltage level as before the charging process.

The diagram at the bottom illustrates a second logic signal VL. For example, the ' second logic signal VL is the output signal of the voltage signal generator 35, . . . wherein the voltage signal generator 35 is a comparator which compares the gate- ; , emitter voltage VGE with the threshold value (of O Volt in the example shown). If the c gate-emitter voltage VGE is smaller than the threshold value, the value of the second logic signal VL is at a lower voltage level. If the gate-emitter voltage V(jE IS equal to or greater than the threshold value, the value of the second logic signal VL is at a higher voltage level. The second logic signal VL may be received from the investigation control unit 53 (see below and Fig. 6) and the unit 53 may determine the length of the time interval tp (see third diagram from the top in Fig. 7) from the beginning of the charging process until the gate-emitter voltage V(-;E reaches the threshold value. Using the length of the time interval tp, it may calculate the charge carried by the control electrode at the end of the time interval tp or it may calculate any other quantity which characterizes the charging process.

According to a second embodiment (see Fig. 8 and 9), the control electrode is repeatedly charged using a constant charging current It; during each charging process. It is determined whether a predefined threshold value (of O Volt in the example shown in Fig. 8 and Fig. 9) is reached by the gate-emitter voltage VGE at the end of the charging time interval tp. The length of the charging time interval tp is predefined and is constant for all charging processes. In particular, the first of the charging processes is performed with a predefined charging current, which is not sufficiently high to reach the predefined gate-emitter voltage VGE within the time interval. If the predefined threshold value has not been reached by the end of the time interval in any one of the charging processes, the charging current is increased in the following charging process by a predefined amount. If the threshold value has been reached by the end of the time interval, the investigation is stopped and the level of the charging current I, of the last charging process can be used for operation of the valve.

The predefined amount is preferably chosen, so that the predefined threshold value cannot be reached in the following charging process before the end of the charging time interval with any type of valves which may be used to replace the valve.

Therein, it is decided that the predefined threshold value has not been reached before , the end of the time interval, if it has not been reached before the end of the charging. ,, time interval minus a defined tolerance time. .... A.

Fig. 8 shows four diagrams (similarly to Fig. 7). The diagrams correspond to a ' '.

charging process which is neither the first nor the last charging process during investigation. The four diagrams shown in Fig. 9 correspond to the last charging process during investigation. Compared to the charging current IG shown in Fig. 8, the charging current IG shown in Fig. 9 has been increased several times. Thus, the gate-emitter voltage VGE can reach the predefined threshold value. In contrast to the first embodiment, the pulse length of the first logic signal L of the second embodiment triggers the commencement and the end of the charging process. The discharging process, which is to be performed after each charging process, is performed immediately after the pulse of the first logic signal L. Preferably, the charging processes of the first and/or of the second embodiment are started when the control electrode is charged to a defined nominal negative value (for example -15V). However, the charging processes may start from any defined state of the control electrode (e.g. when the gate-emitter voltage V(,E has a predefined value).

During operation of the valve 18, the timing of the charging/discharging is controlled by the control signal which is transferred via line CC. However, the first logic device 49 and the second logic device 50 are preferably used as well during operation of the valve.

A preferred embodiment of the first logic device 49 will be described in the following with reference to Fig. 6. A process control unit 52 is connected to an input of the first logic device 49 for receiving the pulse signals from the pulse generator 48 and is connected to an input for receiving an ON signal which enables the first logic device 49 to charge the control electrode. The first logic device 49 performs the investigation of the charging process only, if it receives a pulse signal from the pulse.. . ë generator 48 and an ON signal. If it solely receives an ON signal, the control. . electrode is charged during normal operation of the valve. In the embodiment shown, three current control units 56, 57, 58 are arranged in parallel to each other :.

and only one of the current control units (namely the first unit 56) is used to charge ' the control electrode during investigation. On the other hand, all three of the current.

control units 56, 57, 58 are used to charge the control electrode during operation of. ,, the valve, wherein each current control unit 56, 57, 58 is used in one of the time zones (see Fig. 3 and the corresponding description). For example, the current control unit 56 is used to control the charging process during the first time zone (pre charging), the current control unit 57 is used to control the charging process during the second time zone and the current control unit 58 is used to control the charging process during the fourth time zone.

Outputs of the current control units 56, 57, 58 are connected to an output unit 59, which is connected to signal line PC. Via a separate control line, the process control unit 52 is connected to the output unit 59 in order to control whether an output signal via line PC is enabled or disabled.

Each of the current control units 56, 57, 58 is connected to and combined with one of a plurality of further control units 53, 54, 55, wherein each of the further control units 53, 54, 55 is connected to the process control unit 52. The further control units 53, 54, 55 are adapted to control the corresponding current control unit 56, 57, 58 for charging of the control electrode. In particular, the further control units 53, 54, 55 output signals (during operation of the valve) to their corresponding current control unit 56, 57, 58 in order to increase or decrease the current level, wherein the current level is then saved in the current control unit 56, 57, 58. For example, the current control units 56, 57, 58 may be counters with adaptable counter state which represents the level of a charging current. According to the events described above with reference to Fig. 3, the further control unit 53 is connected to the voltage signal generator 35 (for detecting a signal in connection with the gate-emitter voltage VGl.), the further control unit 54 is connected to the voltage signal generator 39 (for detecting a signal in connection with the emitter voltage VEE) and the further control unit 55 is connected to the voltage signal generator 37 (for detecting a signal in connection with the collector-emitter voltage Vc.).

The further control unit 53 (which is referred to as the investigation control unit in; . the following) is adapted to perform additional tasks: it controls the investigation of the charging process before operation of the valve and sets initial values for all.

current control units 56, 57, 58. The investigation control unit 53 is connected to.. .

each of the current control units 56, 57, 58 for this purpose. For example, the initial .e current level for the current control units 56 (pre-charging) is determined from the .

investigation of the charging process. Then, then current levels for the current. control units 57, 58 are calculated from the determined value or values using a fixed relation and/or using a calculation model of the valve which relation and/or model is valid for different types of valves or might be universal for all types of valves.

Furthermore, corresponding current levels for discharging of the valve might be calculated in the same manner. Preferably, the initial values are chosen so that the first charging process during operation of the valve conforms to regulations concerning safety regulations and/or regulations concerning electromagnetic compatibility (EMC). The term "safety regulations" includes a strategy and/or recommendations (e.g. of the manufacturer of the valve) in order to ensure a safe operation of the valve, whereby the valve is protected from destruction or damage.

In other embodiments, more than three current control units may be used and/or the current control units may be operated in a different manner. e . . . ... en.

Claims

Claims I. A method of operating an electronic valve (18), in particular a

valve adapted for use in a high power converter, wherein the valve (18) comprises a control electrode (17), in particular an insulated gate, the control electrode ( 17) is repeatedly charged and discharged during an operation of the valve (18) in order to change a switching state of the valve (18) a charging process of the control electrode (17) is investigated and a corresponding determination result is obtained, the determination result is used to control charging and/or discharging of the control electrode ( 17) during the operation of the valve ( 18).

.
2. The method of the preceding claim, wherein the charging process of the.
control electrode ( 17) is investigated prior to the operation of the valve (18) and/or at the beginning of the operation of the valve (18). ... - ë

The method of one of the preceding claims, wherein the valve (18) comprises ..- a pair of switching electrodes which are electrically connected by the valve (18) when the valve is switched on and which are electrically insulated by the valve when the valve is switched off and wherein the charging process of the control electrode (17) is investigated when a switching voltage between the switching electrodes is zero and/or is smaller than during the operation of the valve.
4. The method of one of the preceding claims, wherein the determination result is used to set a set value of a charging quantity, wherein the charging quantity is used to control or to perform the charging and/or the discharging of the control electrode (17) during the operation of the valve (18).
5. The method of the preceding claim, wherein the set value is an initial value for the operation of the valve (18).
6. The method of the preceding claim, wherein a value of the charging quantity is adapted during the operation ofthe valve (18).
7. The method of one of the three preceding claims, wherein the charging quantity corresponds to the length of a time interval during which a charging current used to charge and/or discharge the control electrode (17) is kept constant or corresponds to a predetermined point in time at which a level of the constant charging current is changed.
8. The method of one of the preceding claims, wherein investigating the charging process of the control electrode (17) comprises determining an electrode charge and/or an electrode charge difference which is carried by the control electrode (17).
9. The method of the preceding claim, wherein a time interval needed to charge.

or discharge the control electrode (17) is measured and the time interval is used to calculate the electrode charge and/or the electrode charge difference. ..
10. The method of the preceding claim, wherein the time interval starts at a first defined state of the valve (18) and ends at a second defined state of the valve (18) .. :.
I I. The method of the preceding claim, wherein the first and/or the second defined state is defined by an electric voltage between the control electrode (17) and a second electrode of the valve (18).
12. The method of one of claims 1 to 6, wherein the charging process, which is investigated, is repeatedly performed, wherein a current used to charge the control electrode (17) is kept constant during each of the charging processes, wherein the charging time is equal for all of the charging processes and wherein the current is increased or decreased from charging process to charging process until a predefined operation state of the valve (18) is reached at the end of the charging process.
13. The method of the preceding claim, wherein a comparator for comparing the instantaneous operation state of the valve (18) with the predefined operation state of the valve (18) is used and wherein the comparator outputs a signal when the predefined operation state of the valve (18) is reached.
14. The method of the preceding claim, wherein an operatability of the comparator is checked before the charging process.
15. The method of one of the preceding claims, wherein the control electrode (17) is charged or discharged at a level or at levels of a charging current during the charging process ofthe control electrode (17) which is investigated, wherein the same level or levels of the charging current is/are used to charge and/or discharge the control electrode (17) during the operation of the valve.
16. An arrangement for operating an electronic valve (18), in particular a valve adapted for use in a high power converter, wherein the arrangement..

comprises: . a . - a driver device (34, 36) adapted to drive an electric current to and/or from a control electrode (17), in particular from an insulated gate, of the valve (18), .

- a determining device (53) adapted to investigate a charging process of the control electrode (17) and adapted to obtain a corresponding determination result, - a control unit (33), wherein the control unit (33) is connected to the determining device or comprises the determining device (53) and wherein the determining device (53) is adapted to produce information which can be used to control charging and/or discharging of the control electrode (17) during the operation ofthe valve (18).
17. The arrangement of the preceding claim, wherein the determining device (53) is connected to at least one current control unit (56, 57, 58) and wherein the current control unit (56, 57, 58) is adapted to save at least a part of the information.
18. The arrangement of the preceding claim, wherein the current control unit (56, 57, 58) is adapted to control the driver device (34) and wherein the current control unit (56, 57, 58) is directly connected to an output unit (59) which is connected to the driver device (34).
19. A converter (21) for converting a current, in particular for a high power application such as providing electric energy to a driving motor (25) of a railroad traction vehicle, wherein the converter (21) comprises the arrangement of one of the preceding claims, wherein the converter (21) can be operated by repeatedly switching-on and off the valve (18) and wherein the arrangement and the control electrode (17) are connected to each other via a connection for discharging and/or charging the control electrode. . e . A....
20. A programmable logic device comprising the control unit (33) and the determining device (53) of one of claims 16 to 18, wherein the programmable. . logic device is adapted to investigate the charging process of the control. . . electrode (17) according to the method of one of claims 1 to 15.

- ..