DE102018123808A1 - Driver with voltage-controlled distinction between bootstrap capacity recharging and short-circuit failure - Google Patents

Abstract

The invention relates to a driver stage with a first power transistor (M), a second power transistor (M) and a monitoring device (UV). The first power transistor (M) and the second power transistor (M) are connected to form a half bridge (M, M) with a phase output (PH). The monitoring device (UV) detects the amount of the drain-source voltage (U) at the first power transistor (M) and determines a drain-source voltage value (V). The monitoring device (UV) compares the amount of the drain-source voltage (V) with a detection threshold (TH) (the first threshold value SW1). When the detection threshold (TH) (the first threshold value SW1) is exceeded by the drain-source voltage value (V), the monitoring device (UV) initiates a first switch-off of the first power transistor (M) and in this case a switch-on of the second power transistor (M ). Subsequently, the monitoring device (UV) causes the second power transistor (M) to be switched off and the first power transistor (M) to be switched on again. The monitoring device (UV) again detects the drain-source voltage (U) at the first power transistor (M) and determines a further drain-source voltage value (V). The monitoring device (UV) compares the amount of the further drain-source voltage value (V) with a further detection threshold, which can be equal to the detection threshold (TH). The monitoring device (UV) initiates a second switch-off of the first power transistor (M) when the further detection threshold is exceeded by the further drain-source voltage value (V).

Description

Generic term

The invention is directed to a device for reloading bootstrap capacities in a driver circuit.

General introduction

Push-pull stages for driving electrical loads are preferably made up of complementary MOS transistor pairs or IGBT transistor pairs. However, since the mobility of the holes is only about half as large as that of the electrons, P-channel transistors have a larger chip area on the one hand and a higher on-resistance on the other.

For this reason, the P-channel transistors, which are typically the high-side switches in the push-pull stages, are often replaced by N-channel transistors.

However, the problem arises here that, in the event of a faulty activation, such a high-side transistor, which is designed as an N-channel transistor, is opened by a voltage drop of any kind at its control connection, and it thus leads to a cross current in the push -Pull level can come in the event of a fault, which can lead to fire.

Various circuits are known from the prior art which ensure that the high-side transistors cannot be opened unintentionally even if the supply voltage to the gate driver drops.

State of the art

The prior art is explained with the aid of the figures.

1 shows a half bridge bridged by bootstrap. Compared to the prior art, it also shows a monitoring unit ( UV ), which is expressly not state of the art. Nevertheless, the problem can already be determined using 1 are explained. On the left side of the 1 is an integrated circuit ( IC ). To the right of them are all related to the integrated circuit ( IC ) typically external components ( C _VG , D , C _B , M _H , M _L ). The integrated circuit ( IC ) can of course also contain other components which are irrelevant for the discussion of the invention and the prior art and are not shown or mentioned here for the sake of simplicity. The half bridge ( M _H , M _L ) itself consists of a first power transistor ( M _H ) and a second power transistor ( M _L ). The integrated circuit ( IC ) and the half bridge ( M _H , M _L ) are mostly from the same positive supply line ( U _S ) with a positive supply voltage and from the same negative supply line ( GND ) with a negative supply voltage. From the voltage difference between the potential of the positive supply voltage line (Us) and the potential of the negative supply voltage line ( GND ) - typically the reference potential - generates the integrated circuit ( IC ) by means of a voltage supply circuit ( SV ), which is typically part of the integrated circuit ( IC ), preferably a constant voltage ( V _VG ) at their voltage regulator output ( VG ). This constant voltage ( V _VG ) at the voltage regulator output ( VG ) is preferred with the external support capacity ( C _VG ) supported. The constant voltage ( V _G ) at the external support capacity ( C _VG ) serves as supply for the second gate driver ( GT _L ) of the integrated circuit ( IC ), which is the gate of the second power transistor ( M _L ) via the second gate drive output ( GL ) of the integrated circuit ( IC ) controls. Furthermore, this constant voltage ( V _VG ) at the voltage regulator output ( VG ) of the integrated circuit ( IC ) the bootstrap capacity ( C _B ) then always via the diode ( D ) loaded when the phase voltage ( V _PH ) at the phase output ( PH ) against the negative supply voltage line ( GND ) is at reference potential. This is particularly the case if the second power transistor ( M _L ) is switched on and is therefore conductive.

The first gate driver ( GT _H ) of the integrated circuit ( IC ), which is the gate of the first power transistor ( M _H ) via the first gate drive signal ( GH ), in contrast to the second gate driver ( GT _L ) a floating reference potential, the potential ( V _PH ) of the phase output ( PH ) across from the negative supply voltage line ( GND ) and must e.g. the gate potential at the voltage regulator output ( V _GH ) of the first gate driver ( GT _H ) in the case of a switched off first power transistor ( M _H ) always with the voltage potential ( V _PH ) at the phase output ( V _PH ) to the first power transistor ( M _H ) always switch off safely. In the example of the 2nd is the first gate driver ( GT _H ) is drawn in such a way that it V _PH ) of the phase output ( PH ) and the potential of the bootstrap knot ( BST ), i.e. through the bootstrap capacity ( C _BST ), is supplied with electrical energy.

If the first power transistor ( M _H ) the first gate driver ( GT _H ) a potential at the first gate control signal ( GH ) of the first gate driver ( GT _H ) which is a constant amount above that of the phase voltage ( V _PH ) at the phase output ( PH ) against the potential of the negative supply voltage line ( GND ) lies around the gate oxide of the first power transistor ( M _H ) not to be destroyed by overvoltage. The bootstrap tension ( V _BST ) utilized. The bootstrap tension ( V _BST ) is attached to the additional bootstrap node ( BST ) of the integrated circuit ( IC ) preferably by the power supply circuit ( SV ) in the integrated circuit ( IC ) via the diode ( D ) delivered. Since the bootstrap capacity ( C _B ) only when the second power transistor is switched on ( M _L ) from the power supply circuit ( SV ) via their voltage regulator output ( VG ) is loaded, this bootstrap capacity ( C _B ) consequently no longer reloaded as long as the first power transistor ( M _H ) is active because then the second power transistor ( M _L ) is blocked to cross currents in the half-bridge ( M _H , M _L ) to prevent.

Due to the self-consumption of the integrated formwork ( IC ) at the bootstrap node ( BST ) of the integrated circuit ( IC ) and external leakage currents discharge the bootstrap capacity ( C _B ) slowly as long as the first power transistor ( M _H ) is active, i.e. is leading. Consequently, the corresponding switch-on time of the first power transistor ( M _H ) limited and by the capacity value of the bootstrap capacity ( C _B ) and on leakage currents and thus on the temperature, as well as on other variables subject to scattering.

Various methods exist for reloading the bootstrap capacity in the prior art ( C _B ). These are all characterized by the need for additional external and analog components within the integrated circuit ( IC ) and are therefore associated with corresponding costs in the production of the respective integrated circuit ( IC ) connected. They will not be discussed further here.

Monitoring on the driver

U _DS monitoring

So-called U _DS monitoring is now included in the many gate drivers. Here the designation refers U _DS the drain-source voltage at the first power transistor ( M _H ). If the measured value ( V _DS ) the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) a preferably configurable threshold, the corresponding first power transistor ( M _H ) is switched off and an error message is saved, which can be called up by the control processor that is typically present in the system and usually has to be actively reset by it before the relevant first gate driver ( GT _H ) of the relevant first power transistor ( M _H ) can be switched on again. A configurability of the threshold can possibly be solved quite differently in the different integrated circuits.

Such U _DS monitoring typically serves primarily for the detection of short circuits at the phase output ( PH ) and the avoidance of dangerous conditions in the event of such short circuits. However, if the bootstrap voltage becomes too low ( V _BST ) trigger because then the first power transistor ( M _H ) can no longer be fully controlled.

This results in the problem that the second power transistor ( M _L ) must be switched on regularly to increase the bootstrap capacity ( C _B ) in order to avoid faulty triggering of the short-circuit detection. This limits the permissible duty cycle of the PWM signal generated by the driver circuit to a value below 100%, which is not desired and should be improved here.

Bootstrap monitoring

In the prior art there is also monitoring of the bootstrap voltage ( V _BST ). Falls below this bootstrap tension ( V _BST ) a minimum value, the first power transistor ( M _H ) is switched off and an error message is saved, which can be called up again by the typically available processor and which must preferably be actively reset before the first gate driver ( GT _H ) the first power transistor ( M _H ) can turn on.

Bootstrap monitoring is rarely implemented because it causes additional silicon and test costs and ultimately the corresponding error via the previously described U _DS monitoring with intercepted handshaking for automatic reloading of the bootstrap capacity ( C _B ) leads.

Handshaking uses either U _DS monitoring or bootstrap monitoring for detection. On the interface side, between the integrated circuit ( IC ) and the processor the mostly existing interrupt line for signaling an error, as well as the standard control lines for activating the first power transistor ( M _H ) and the second power transistor ( M _L ) utilized.

Only in the digital parts of the integrated circuit ( IC ) and the processor are minor adjustments to implement.

Object of the invention

The invention is therefore based on the object of providing a solution which does not have the above disadvantages of the prior art, is able to distinguish between a recharge of the boost trap capacity and a short circuit, and has further advantages.

This object is achieved by a device according to claim 1.

Solution of the task according to the invention

Modified behavior in the gate driver = reaction to the U _DS monitoring

The typical immediate response of the gate driver, 2nd , remains as it usually is in the respective integrated circuit ( IC ) realized: When the detection threshold is exceeded ( TH ) by the measured value ( V _DS ) the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) becomes the first power transistor ( M _H ) immediately by a monitoring device ( UV ) using the first enable line ( EN _H ) and the interrupt line ( INTN ) of the integrated circuit ( IC ) to the typically available processor to signal the processor the problem.

In contrast to the usual behavior to date, the second gate driver ( GT _L ) however the activation of the second power transistors ( M _L ) the half bridge ( M _H , M _L ) to give the processor the chance to increase the bootstrap capacity ( C _B ) reload. This is precisely not the case in the prior art. Here a drop in the bootstrap voltage ( V _BST ) for short-circuit fault detection, which is used to switch off the first gate driver ( GT _H ) and the second gate driver ( GT _L ) and thus to block the first power transistor ( M _H ) and the second power transistor ( M _L ) leads. It has now been recognized according to the invention that this behavior is counterproductive, since in the event of an insufficient bootstrap voltage ( V _BST ) not the half-bridge has to be switched off, but the bootstrap capacity ( C _B ) must be reloaded. It was also recognized that the half-bridge ( M _H , M _L ), but that another short-term short circuit is possible and can be permitted without danger. The idea according to the invention is therefore first of all from an insufficient loading of the bootstrap capacity ( C _B ) going out and trying to reload the bootstrap capacity ( C _B ) try to fix this error and only if this does not lead to success, the first power transistor ( M _H ) and the second power transistor ( M _L ) to block and only then to signal a short circuit to the processor.

The integrated circuit ( IC ) - more precisely the monitoring device ( UV ) - the half bridge ( M _H , M _L ) switch off permanently (at most) until the error is actively reset, for example by means of a suitable register access of the processor, if immediately after the first power transistor is switched on again ( M _H ) (after the debounce time has elapsed) the U _DS monitoring stops again. In this case, the previous reloading of the bootstrap capacity ( C _B ) not successful. If a short circuit was the cause for the first shutdown, the U _DS monitoring will trigger it again immediately after the next switch on after the attempted recharging. If the device is switched off twice in quick succession, a short circuit can certainly be assumed. In this way it is possible to distinguish between the cause of the fault of a discharged bootstrap capacity ( C _B ) and a short circuit can be safely distinguished. This has the advantage that the bootstrap capacity ( C _B ) has to be reloaded, but only has to be reloaded if necessary. This in turn means that the maximum possible effective duty cycle is closer to the ideal value of a maximum duty cycle for the PWM of 100%, since preventive reloading of the bootstrap capacity ( C _B ) is omitted.

A variant is also conceivable, which in 3rd is shown.

The time in which the first power transistor ( M _H ) is switched on at different intervals ( T _D , T _A , T _EA ) assigned. The time starts with the switch on at the switch on time ( t ₀ ). Switching on ( t ₀ ) and the possibly parameterizable debounce time ( T _D ) remain unchanged until the U _DS monitoring is taken into account. The debounce time ( T _D ) serves to ensure that the first power transistor ( M _H ) can switch on completely before the U _DS monitoring is activated.

• • Schlägt die U_DS-Überwachung innerhalb der der Debounce-Zeit (T_D ) nachfolgenden Aktivzeit (T_A ) an, so wird der betreffende erste Leistungstransistor (M_H ) oder die Halbbrücke (M_H , M_L ) abgeschaltet, ein Fehler über die Interrupt-Leitung (INTN) dem Prozessor signalisiert und der erste Leistungstransistor (M_H ) kann beispielsweise erst wieder eingeschaltet werden, nachdem aktiv ein entsprechendes Fehlerregister in der integrierten Schaltung durch den Prozessor beschrieben wurde.
• • Schlägt die U_DS-Überwachung innerhalb der der Aktivzeit (T_A ) und der Debounce-Zeit (T_D ) nachfolgenden erweiterten Aktivzeit (T_EA ) an, wird der betreffende erste Leistungstransistor (M_H ) abgeschaltet, wie in 2, jedoch kann er, bevorzugt ohne ein Register der integrierten Schaltung (IC) beschreiben zu müssen, wieder vom Prozessor eingeschaltet werden, da in der erweiterten Aktivzeit (T_EA ) typischerweise davon ausgegangen werden kann, dass eine Entladung der Bootstrapkapazität (C_BST ) die Ursache für die U_DS -Detektion war.

The following classification should have the following function:

• • Does the U _DS monitoring fail within the debounce time ( T _D ) subsequent active time ( T _A ), the relevant first power transistor ( M _H ) or the half bridge ( M _H , M _L ) switched off, an error via the interrupt line ( INTN ) signals to the processor and the first power transistor ( M _H ) can only be switched on again, for example, after an appropriate error register in the integrated circuit has been actively written by the processor.
• • Does the U _DS monitoring within the active time ( T _A ) and the debounce time ( T _D ) subsequent extended active time ( T _EA ), the relevant first power transistor ( M _H ) switched off, as in 2nd , but it can, preferably without a register of the integrated circuit ( IC ) must be switched on again by the processor, since the extended active time ( T _EA ) it can typically be assumed that a discharge of the bootstrap capacity ( C _BST ) the cause of the U _DS -Detection was.

The duration of the active time (T) is preferably constant and can preferably be parameterized, for example, by programming. The extended active time ( T _EA ) closes the active time ( T _A ) and always ends with the deactivation of the first power transistor ( M _H ). If the first power transistor ( M _H ) activated only briefly, such as with normal PWM control with duty cycles <100%, the extended active time ( T _EA ) may no longer be reached.

Modification in the processor

The response of the integrated circuit described here ( IC ) could theoretically also in software in the processor to control the integrated circuit ( IC ) are implemented. Due to the disturbing latency of the interrupt routine, however, an implementation in hardware, preferably in the PWM generator ( PWMG ) which is the first PWM signal ( PWMH ) for the first power transistor ( M _H ) and the second PWM signal ( PWML ) for the second power transistor ( M _L ) generated, or in the monitoring device ( UV ) recommended. Monitoring device ( UV ) and PWM generator ( PWMG ) can form a unit and can be designed as a microcomputer, for example.

In 4th is the interrupt signal ( INTN ) for fault messages from the integrated circuit ( IC ) to the processor. The first PWM signal ( PWMH ) is the logic control signal of the processor to the first gate driver ( GT _H ) of the first power transistor ( M _H ) to activate the first power transistor ( M _H ). The second PWM signal ( PWML ) is the logic control signal of the processor to the second gate driver ( GT _L ) of the second power transistor ( M _L ) to activate the second power transistor ( M _L ).

• Wird am Interrupt-Eingang (INTN) des Prozessors ein Fehler gemeldet, so schaltet der Prozessor das erste PWM-Ansteuersignal (PWMH) für den ersten Leistungstransistor (M_H ) durch entsprechende Signalisierung an die integrierte Schaltung (IC) sofort ab. Dies geschieht beispielsweise durch geeignete Programmierung der Überwachungsvorrichtung (UV) und/oder des PWM-generators (PWMG). Hierdurch geht der erste Leistungstransistor (M_H ) in einen abgeschalteten, typischerweise hochohmigen Zustand. (Übergang 1 in 4). Nach Ablauf der für den ersten Gate-Treiber (GT_H ) eingestellten Totzeit wird dann das zweite PWM-Ansteuersignal (PWML) für die Ansteuerung des zweiten Leistungstransistors (M_L ) für die Dauer der Ladezeit (T_L ) aktiviert. Dadurch wird das zweite Gate-Ansteuersignal (GL) zur Ansteuerung des zweiten Leistungstransistors (M_L ) aktiviert. Hierdurch wird der Phasenausgang (PH) mit dem Potenzial der negativen Versorgungsspannungsleitung (GND) verbunden und die Bootstrap-Kapazität (C_B ) wird über die Diode (D) aus dem Spannungsreglerausgang (VG) der Spanungsversorgungsschaltung (SV) nachgeladen. Die Ladezeit (T_L ) ist vorzugsweise über eine Registerprogrammierung eines Registers der integrierten Schaltung konfigurierbar, um sie den Anforderungen der jeweiligen Anwendung flexibel anpassen zu können. Nach Aktivierung des zweiten PWM-Ansteuersignals (PWML) sollte der zweite Gate-Treiber (GT_L ) das Fehlersignal über die Interrupt-Leitung (INTN) wieder deaktivieren, um dem Prozessor zu signalisieren, dass der Entsättigungsfehler nun beseitigt ist. Nach Ablauf der Ladezeit (T_L ) wird das zweite PWM-Ansteursignal (PWML) wieder deaktiviert. Es schließt sich eine weitere Totzeit zur Vermeidung von Querströmen an, nach deren Ablauf das erste PWM-Ansteuersignal (PWMH) wieder aktiviert wird. Danach befinden sich Prozessor, integrierte Schaltung (IC), deren Treiber und die Halbbrücke (M_H , M_L ) wieder im ursprünglichen Zustand.

If the PWM generator ( PWMG ) on the control signal pair ( G _H , G _L ) for the half bridge ( M _H , M _L ) a duty cycle of 100%, or one above a configurable threshold> 90% + x, then the following reaction on this control signal pair ( GL , GH ) realized:

• Is at the interrupt input ( INTN ) the processor reports an error, the processor switches the first PWM control signal ( PWMH ) for the first power transistor ( M _H ) by appropriate signaling to the integrated circuit ( IC ) immediately. This is done, for example, by suitable programming of the monitoring device ( UV ) and / or the PWM generator ( PWMG ). The first power transistor ( M _H ) in a switched off, typically high-resistance state. (Crossing 1 in 4th ). After the first gate driver ( GT _H ) set dead time, the second PWM control signal ( PWML ) for the control of the second power transistor ( M _L ) for the duration of the charging time ( T _L ) activated. As a result, the second gate drive signal ( GL ) to control the second power transistor ( M _L ) activated. The phase output ( PH ) with the potential of the negative supply voltage line ( GND ) connected and the bootstrap capacity ( C _B ) is via the diode ( D ) from the voltage regulator output ( VG ) of the power supply circuit ( SV ) reloaded. The charging time ( T _L ) is preferably configurable via register programming of a register of the integrated circuit so that it can be flexibly adapted to the requirements of the respective application. After activation of the second PWM control signal ( PWML ) the second gate driver ( GT _L ) the error signal via the interrupt line ( INTN ) again to signal to the processor that the desaturation error has now been eliminated. After the charging time ( T _L ) the second PWM control signal ( PWML ) deactivated again. This is followed by another dead time to avoid cross currents, after which the first PWM control signal ( PWMH ) is reactivated. Then there are processor, integrated circuit ( IC ), their drivers and the half bridge ( M _H , M _L ) in its original condition.

This signal game can also be implemented in the same way if the second gate driver ( GT _L ) the bootstrap tension ( V _BST ) between bootstrap nodes ( BST ) and phase output ( PH ) monitors itself. In this case, the error signal on the interrupt line ( INTN ) go out as soon as the bootstrap voltage ( V _BST ) between bootstrap nodes ( BST ) and phase output ( PH ) has assumed a sufficient value again. Then, as a further implementation variant, instead of the constant loading time ( T _L ) the charging interval is ended as soon as the error signal ( INTN ) is inactive again or a certain time has passed after the error signal was deactivated.

Certain measurements, such as current measurements by the PWM generator ( PWMG ) which are normally triggered and which fall within the reloading interval should be suspended for the duration of the reloading interval or recognized as invalid by the control algorithm of the application in order to avoid malfunctions in the control due to the reloading.

Description of the features of the invention

The following section repeats the above description in a claim-like form.

The invention relates to a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ), a first gate driver ( GT _H ), a second gate driver ( GT _L ), a power supply circuit ( SV ) with a voltage regulator output ( VG ), a diode ( D ), a bootstrap capacity ( C _B ), a positive supply voltage line (Us), a negative supply voltage line ( GND ) and a monitoring device ( UV ). The first power transistor ( M _H ) and the second power transistor ( M _L ) are preferably MOS transistors or IGBT transistors. Other power semiconductors are also appropriate. The first power transistor ( M _H ) and the second power transistor ( M _L ) are to a half bridge ( M _H , M _L ) with a phase output ( PH ) between the positive supply voltage line ( U _S ) and the negative supply voltage line ( GND ) connected. The first power transistor ( M _H ) with its drain connection with the positive supply voltage line (Us) and with its source connection with the phase output ( PH ) connected. The second power transistor ( M _L ) is with its source connection with the negative supply voltage line ( GND ) and with its drain connection to the phase output ( PH ) connected. The first control connection of the first power transistor ( M _H ) using a first gate driver ( GT _H ) via a first gate driver output ( GH ) controlled. The second control connection of the second power transistor ( M _L ) is activated by means of a second gate driver ( GT _L ) via a second gate driver output ( GL ) controlled. The logic state of the first gate driver output ( GH ) depends on a first PWM control signal ( PWMH ). The logic state of the second gate driver output ( GL ) depends on a second PWM control signal ( PWML ). The first gate driver ( GT _H ) is via a bootstrap node ( BST ) with electrical energy for switching on the first power transistor ( M _H ) at least if the potential at the bootstrap node ( BST ) related to the potential of the negative supply voltage line ( GND ) above the potential of the positive supply voltage line ( U _S ) related to the potential of the negative supply voltage line ( GND ) lies. The diode ( D ) is between the voltage regulator output ( VG ) and the bootstrap knot ( BST ) switched. The bootstrap capacity ( C _B ) is between the bootstrap nodes ( BST ) and the phase output ( PH ) the half bridge ( M _H , M _L ) switched. The monitoring device ( UV ) captures the potential difference between the potential at the bootstrap node ( BST ) and the potential at the phase output ( PH ) the half bridge ( M _H , M _L ) and determines an associated bootstrap potential difference value ( ΔV _BST ). The monitoring device ( UV ) compares the bootstrap potential difference value determined in this way ( ΔV _BST ) in terms of amount with a first threshold. The monitoring device switches the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) on if the bootstrap potential difference value determined in this way ( ΔV _BST ) is below the first threshold. There is only one discharge of the bootstrap capacity ( C _B ), it will now be loaded.

After switching off for the first time because the bootstrap potential difference value ( ΔV _BST ) was below the first threshold in terms of amount, the monitoring device switches the first power transistor in a first modification of the proposal ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) off if the bootstrap potential difference value determined in this way ( ΔV _BST ) then again amounts above a second threshold value, which can be equal to the first threshold value. This now makes it possible to test whether the bootstrap capacity was discharged and is now loaded again or whether there is a short circuit.

It is particularly preferred that the device is switched on again after the first switch-off only after a charging time has elapsed ( T _L ) so that the bootstrap capacity ( C _B ) is then safely loaded.

In a further preferred embodiment, the monitoring device then detects ( UV ) the potential difference between the potential at the bootstrap node ( BST ) and the potential at the phase output ( PH ) the half bridge ( M _H , M _L ) again and determines an associated additional bootstrap potential difference value ( ΔV _BST ). The monitoring device compares ( UV ) the further bootstrap potential difference value determined in this way ( ΔV _BST ) in terms of amount with a further threshold value, which can be equal to the first threshold value. The monitoring device switches the first power transistor ( M _H ) using the first gate driver ( GT _H ) after switching on again and the second power transistor ( M _L ) using the second gate driver ( GT _L ) also from if the further bootstrap potential difference value determined in this way ( ΔV _BST ) is below the further threshold. This happens because a short circuit must then be assumed.

Second variant

Instead of monitoring the bootstrap voltage, it is also necessary to monitor the U _DS voltage when the first power transistor is on ( M _H ) possible.

It is then again a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ), a first gate driver ( GT _H ), a second gate driver ( GT _L ), a power supply circuit ( SV ) with a voltage regulator output ( VG ), a diode ( D ), a bootstrap capacity ( C _B ), a positive supply voltage line (Us), a negative supply voltage line ( GND ) and with a monitoring device ( UV ). The first power transistor ( M _H ) and the second power transistor ( M _L ) are preferably MOS transistors or IGBT transistors. Other power semiconductors are also appropriate. The first power transistor ( M _H ) and the second power transistor ( M _L ) are back to a half bridge ( M _H , M _L ) with a phase output ( PH ) between the positive supply voltage line ( U _S ) and the negative supply voltage line ( GND ) connected. The first power transistor ( M _H ) is again with its drain connection with the positive supply voltage line (Us) and with its source connection with the phase output ( PH ) connected. The second power transistor ( M _L ) and with its source connection with the negative supply voltage line ( GND ) and with its drain connection to the phase output ( PH ) connected. The first control connection of the first power transistor ( M _H ) using a first gate driver ( GT _H ) via a first gate driver output ( GH ) controlled. The second control connection of the second power transistor ( M _L ) is activated by means of a second gate driver ( GT _L ) via a second gate driver output ( GL ) controlled. The logic state of the first gate driver output ( GH ) depends on a first PWM signal ( PWMH ). The logic state of the second gate driver output ( GL ) depends on a second PWM signal ( PWML ). The first gate driver ( GT _H ) becomes a bootstrap knot ( BST ) with electrical energy for switching on the first power transistor ( M _H ) at least if the potential at the bootstrap node ( BST ) related to the potential of the negative supply voltage line ( GND ) above the potential of the positive supply voltage line (Us) in relation to the potential of the negative supply voltage line (Us GND ) lies. The diode ( D ) is between the voltage regulator output ( VG ) and the bootstrap knot ( BST ) switched. The bootstrap capacity ( C _B ) is between the bootstrap nodes ( BST ) and the phase output ( PH ) the half bridge ( M _H , M _L ) switched. The monitoring device ( UV ) detects the potential difference between the potential at the drain connection ( U _S ) of the first power transistor ( M _H ) and the potential at the source ( PH ) of the first power transistor ( M _H ) and determines an associated U _DS potential difference value. The monitoring device ( UV ) compares the U _DS potential difference value determined in this way with a detection threshold ( TH ). The monitoring device switches the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) when the U _DS potential difference value determined in this way is above the detection threshold ( TH ) lies. This enables the bootstrap capacity to be reloaded ( C _B ).

In a first sub-variant, the monitoring device ( UV ) after the previous switch-off because the U _DS potential difference value determined is above the detection threshold ( TH ), the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) out. This now makes it possible to check again whether the bootstrap capacity ( C _B ) acted and whether it is now loaded or whether it is a short circuit. The monitoring device ( UV ) again records the potential difference between the potential of the drain connection ( U _S ) of the first power transistor ( M _H ) and the potential at the source ( PH ) of the first power transistor ( M _H ) and again determines an associated further U _DS potential difference value. The monitoring device ( UV ) again compares the further U _DS potential difference value determined in this way in terms of amount with a further detection threshold which is equal to the detection threshold ( TH ) can be. The monitoring device then switches the first power transistor ( M _H ) using the first gate driver ( GT _H ) again and the second power transistor ( M _L ) using the second gate driver ( GT _L ) now also decreases if the further U _DS potential difference value determined in this way is above the further detection threshold, since a short circuit must then be assumed.

Third variant

The third variant also relates to a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ), a positive supply voltage line (Us), a negative supply voltage line ( GND ) and one Monitoring device ( UV ). The first power transistor ( M _H ) and the second power transistor ( M _L ) are preferably MOS transistors or IGBT transistors. Other power semiconductors are also appropriate. The first power transistor ( M _H ) and the second power transistor ( M _L ) are back to a half bridge ( M _H , M _L ) with a phase output ( PH ) between the positive supply voltage line ( U _S ) and the negative supply voltage line ( GND ) connected. The monitoring device ( UV ) detects the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) in terms of amount and determines an associated drain-source voltage value ( V _DS ). The monitoring device ( UV ) or another control device compares the drain-source voltage value ( V _DS ) with a detection threshold ( TH ). The monitoring device ( UV ) causes the detection threshold to be exceeded ( TH ) by the drain-source voltage value ( V _DS ) a first shutdown of the first power transistor ( M _H ) and switching on the second power transistor ( M _L ). Subsequently, the monitoring device ( UV ) or the other control device in this case, especially after a charging time ( T _L ), switching off the second power transistor ( M _L ) and switching the first power transistor on again ( M _H ). The monitoring device ( UV ) then detects the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) again in terms of amount and thus determines a further drain-source voltage value ( V _DS2 ) (U _DS voltage value). The monitoring device ( UV ) then compares the amount of the further drain-source voltage value ( V _DS2 ) (U _DS voltage value) with a further detection threshold that is equal to the detection threshold ( TH ) can be. The monitoring device ( UV ) causes the further drain-source voltage value to exceed the further detection threshold in terms of amount ( V _DS2 ) (U _DS voltage value) a second shutdown of the first power transistor ( M _H ).

In a first sub-variant, the monitoring device ( UV ) if the additional detection threshold is exceeded by the additional drain-source voltage value ( V _DS2 ) also switching off the second power transistor ( M _L ).

Fourth variant

The fourth variant again provides a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ), a positive supply voltage line (Us), a negative supply voltage line ( GND ) and a monitoring device ( UV ). The first power transistor ( M _H ) and the second power transistor ( M _L ) are preferably MOS transistors or IGBT transistors. Other power semiconductors are also appropriate. The first power transistor ( M _H ) and the second power transistor ( M _L ) are back to a half bridge ( M _H , M _L ) with a phase output ( PH ) between the positive supply voltage line ( U _S ) and the negative supply voltage line ( GND ) connected. The monitoring device ( UV ) records the amount of the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) and determines a drain-source voltage value ( V _DS ). The monitoring device ( UV ) compares the drain-source voltage value ( V _DS ) with a detection threshold ( TH ). The monitoring device ( UV ) causes the detection threshold to be exceeded ( TH ) by the drain-source voltage value ( V _DS ), hereinafter referred to as exceeding for the first time, a first shutdown of the first power transistor ( M _H ). The monitoring device ( UV ) or the other control device in this case causes the second power transistor to be switched on ( M _L ). Subsequently, the monitoring device ( UV ) or the other control device in this case, especially after a charging time ( T _L ), switching off the second power transistor ( M _L ) and allows the first power transistor to be switched on again ( M _H ) too.

In a first sub-variant of this variant, the monitoring device again detects ( UV ) the amount of the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) after switching off for the first time when switching off again and thereby determines a further drain-source voltage _value (V _DS2 DS). The monitoring device ( UV ) compares the further drain-source voltage value ( V _DS2 ) with a further detection threshold that is equal to the detection threshold ( TH ) can be. The monitoring device ( UV ) causes the further drain-source voltage value to exceed the further detection threshold in terms of amount ( V _DS2 ) a second shutdown of the first power transistor ( M _H ).

In a second sub-variant of this variant, the monitoring device signals ( UV ) or another sub-device of the driver only when the detection threshold is exceeded again ( TH ) by the further drain-source voltage value ( V _DS2 ) a short circuit.

In a third sub-variant of this variant, the monitoring device again detects ( UV ) the amount of the drain-source voltage ( U _DS ) on the first power transistor ( M _H ) after switching off the first power transistor ( M _H ) when the first power transistor is switched on again ( M _H ) and determines another drain-source voltage value ( V _DS ). The monitoring device ( UV ) compares the further drain-source voltage value ( V _DS2 ) with a further detection threshold that is equal to the detection threshold ( TH ) can be. The monitoring device ( UV ) then behaves when the amount falls below the further detection threshold due to the further drain-source voltage value ( V _DS2 ) in the event of a crossing again as if it were a first crossing.

Fifth variant

• • Schritt S21: Einschalten (S21) des ersten Leistungstransistors (M_H ) und Ausschalten des zweiten Leistungstransistors (M_L ) zu einem Einschaltzeitpunkt (t₀ ).
• • Schritt S22: Erfassen (S22) der Potenzialdifferenz zwischen dem ersten Anschluss und dem zweiten Anschluss der Bootstrap-Kapazität (C_B ) und Ermitteln eines zugehörigen Bootstrap-Potenzialdifferenzwerts (ΔV_BST );
• • Schritt S23: Vergleich (S23) des so ermittelten Bootstrap-Potenzialdifferenzwerts (ΔV_BST ) betragsmäßig mit einem ersten Schwellwert;
• • Schritt S24: Erstes Abschalten (S24) des ersten Leistungstransistors (M_H ) und erstes Einschalten des zweiten Leistungstransistors (M_L ), wenn der Vergleich (S23) ergibt, dass das so ermittelte Bootstrap-Potenzialdifferenzwert (ΔV_BST ) betragsmäßig unterhalb des ersten Schwellwerts (SW1) liegt. In dem Fall wird zunächst davon ausgegangen, dass es sich aufgrund der Erfahrung in der Regel nicht um einen Kurzschluss, sondern um eine entladene Bootstrap-Kapazität handelt. Im Gegensatz zum Stand der Technik, wird also nicht sofort auf einen Kurzschluss geschlossen.

The fifth variant relates to a process (see 7 ) for operating a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ) and a bootstrap capacity ( C _B ) with a first connection and a second connection. The bootstrap capacity ( C _B ) is charged when the second power transistor ( M _L ) is switched on. The process includes the steps:

• • step S21 : Turn on ( S21 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) at a switch-on time ( t ₀ ).
• • step S22 : Capture ( S22 ) the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) and determining an associated bootstrap potential difference value ( ΔV _BST );
• • step S23 : Comparison ( S23 ) of the bootstrap potential difference value determined in this way ( ΔV _BST ) in terms of amount with a first threshold value;
• • step S24 : First shutdown ( S24 ) of the first power transistor ( M _H ) and switching on the second power transistor for the first time ( M _L ) if the comparison ( S23 ) shows that the bootstrap potential difference value determined in this way ( ΔV _BST ) below the first threshold ( SW1 ) lies. In the case, it is initially assumed that, based on experience, it is usually not a short circuit, but rather a discharged bootstrap capacity. In contrast to the prior art, a short circuit is not immediately concluded.

• • Schritt S25: erneutes Einschalten (S25) des ersten Leistungstransistors (M_H ) nach einem ersten Abschalten aufgrund einer betragsmäßigen Unterschreitung des ersten Schwellwerts (SW1) durch den ermittelten Bootstrap-Potenzialdifferenzwert (ΔV_BST ), und erneutes Ausschalten (S25) des zweiten Leistungstransistors (M_L ), nach Vergehen der Ladezeit (T_L ) seit dem Beginn des Nachladevorgangs mit dem Schritt S24. Es wird also nun davon ausgegangen, dass die Bootstrap-Kapazität (C_B ) ausreichend nachgeladen sein sollte. Damit das so ist, ist es vorteilhaft mit dem Beginn dieses Schrittes S25 zu warten, bis eine Ladezeit (T_L ) seit dem Beginn des Ladevorgangs mit dem Schritt S24 vergangen ist;
• • Schritt S26: erneutes Erfassen (S26) der Potenzialdifferenz zwischen dem ersten Anschluss und dem zweiten Anschluss der Bootstrap-Kapazität (C_B ) und Ermitteln eines zugehörigen weiteren Bootstrap-Potenzialdifferenzwerts (ΔV_BST2 );
• • Schritt S27: Vergleich (S27) des so ermittelten weiteren Bootstrap-Potenzialdifferenzwerts (ΔV_BST2 ) betragsmäßig mit einem weiteren Schwellwert, der gleich dem ersten Schwellwert sein kann.
• • Schritt S28: Ausschalten (S28) des ersten Leistungstransistors (M_H ) und Ausschalten (S28) des zweiten Leistungstransistors (M_L ), wenn der so ermittelte weitere Bootstrap-Potenzialdifferenzwert (ΔV_BST2 ) betragsmäßig wieder unterhalb eines zweiten Schwellwerts (SW2) liegt, der gleich dem ersten Schwellwert (SW1) sein kann. In diesem Fall wird also davon ausgegangen, dass ein Defekt vorliegt. In diesem Fall wird nicht zwischen der Art des Defekts unterschieden. Das Verfahren ist aber besonders einfach. Es wird also einmal versucht, die Bootstrap-Kapazität nachzuladen und wenn das nicht gelingt, wird ein Fehlerfall angenommen und ein sicherer Zustand eingenommen;
• • Schritt S29: Einschalten (S29) (= Beibehalten des Zustands aus S25) des ersten Leistungstransistors (M_H ) und Ausschalten (S29) des zweiten Leistungstransistors (M_L ), wenn der so ermittelte weitere Bootstrap-Potenzialdifferenzwert (ΔV_BST2 ) betragsmäßig wieder oberhalb des zweiten Schwellwerts (SW2) liegt.

A first sub-variant of this fifth variant comprises the additional steps:

• • step S25 : switch on again ( S25 ) of the first power transistor ( M _H ) after a first shutdown due to the amount falling below the first threshold value ( SW1 ) by the determined bootstrap potential difference value ( ΔV _BST ), and switching off again ( S25 ) of the second power transistor ( M _L ), after the charging time has elapsed ( T _L ) since the start of the reload process with the step S24 . It is now assumed that the bootstrap capacity ( C _B ) should be sufficiently reloaded. In order for this to be the case, it is advantageous to start this step S25 wait for a load time ( T _L ) since the start of the loading process with the step S24 has passed;
• • step S26 : recapture ( S26 ) the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) and determining an associated additional bootstrap potential difference value ( ΔV _BST2 );
• • step S27 : Comparison ( S27 ) of the further bootstrap potential difference value determined in this way ( ΔV _BST2 ) in terms of amount with a further threshold value, which can be equal to the first threshold value.
• • step S28 : Turn off ( S28 ) of the first power transistor ( M _H ) and switch off ( S28 ) of the second power transistor ( M _L ) if the further bootstrap potential difference value determined in this way ( ΔV _BST2 ) again below a second threshold in terms of amount ( SW2 ) which is equal to the first threshold ( SW1 ) can be. In this case, it is assumed that there is a defect. In this case, no distinction is made between the type of defect. The process is particularly simple. An attempt is made to reload the bootstrap capacity and if this fails, an error case is assumed and a safe state is assumed;
• • step S29 : Turn on ( S29 ) (= Keep status from S25 ) of the first power transistor ( M _H ) and switch off ( S29 ) of the second power transistor ( M _L ) if the further bootstrap potential difference value determined in this way ( ΔV _BST2 ) in terms of amount again above the second threshold ( SW2 ) lies.

It is preferable to switch on again ( S29 ) after the first shutdown ( S24 ) only after a charging time has passed ( T _L ) for the bootstrap capacity ( C _B ).

Sixth variant

• • Schritt S31: Einschalten des ersten Leistungstransistors (M_H ) und Ausschalten des zweiten Leistungstransistors (M_L ) zu einem Einschaltzeitpunkt (t₀ ). Dieses ist nicht notwendig, wenn der erste Leistungstransistor (M_H ) bereits zum Einschaltzeitpunkt (t₀ ) eingeschaltet ist und wenn der zweite Leistungstransistor (M_L ) bereits zum Einschaltzeitpunkt (t₀ ) ausgeschaltet ist;
• • Schritt S32: Erfassen der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen U_DS-Spannungswerts (V_DS );
• • Schritt S33: Vergleich des so ermittelten U_DS-Spannungswerts (V_DS ) betragsmäßig mit einem ersten Schwellwert (SW1);
• • Schritt S34: Erstes Abschalten des ersten Leistungstransistors (M_H ) und erstes Einschalten des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass der so ermittelte U_DS-Spannungswert (V_DS ) betragsmäßig unterhalb des ersten Schwellwerts (SW1) liegt.

The sixth variant relates to a process ( 8th ) for operating a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ) and with a bootstrap capacity ( C _B ) with a first connection and a second connection. The bootstrap capacity ( C _B ) is loaded again when the second power transistor ( M _L ) is switched on. The proposed procedure includes the steps:

• • step S31 : Switching on the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) at a switch-on time ( t ₀ ). This is not necessary if the first power transistor ( M _H ) at the time of switching on ( t ₀ ) is switched on and if the second power transistor ( M _L ) at the time of switching on ( t ₀ ) is switched off;
• • step S32 : Detection of the U _DS voltage at the first power transistor ( M _H ) and determining an associated U _DS voltage value ( V _DS );
• • step S33 : Comparison of the U _DS voltage value determined in this way ( V _DS ) in terms of amount with a first threshold value ( SW1 );
• • step S34 : First shutdown of the first power transistor ( M _H ) and switching on the second power transistor for the first time ( M _L ) if the comparison shows that the U _DS voltage value determined in this way ( V _DS ) below the first threshold ( SW1 ) lies.

• • Schritt S35: Erneutes Einschalten des ersten Leistungstransistors (M_H ) nach dem Schritt S34 und erneutes Ausschalten des zweiten Leistungstransistors (M_L ), wobei dies insbesondere nach einer Ladezeit (T_L ) erfolgt;
• • Schritt S36: Erneutes Erfassen der U_DS-Spannung (U_DS ) am ersten Leistungstransistor (M_H ), wenn der erste Leistungstransistor (M_H ) eingeschaltet ist und erneutes Ermitteln eines zugehörigen weiteren U_DS-Spannungswerts (V_DS2 );
• • Schritt S37: Vergleich des so ermittelten weiteren U_DS-Spannungswerts (V_DS2 ) betragsmäßig mit einem weiteren Schwellwert (SW2), der gleich dem ersten Schwellwert (SW1) sein kann;
• • Schritt S38: Erneutes Abschalten des ersten Leistungstransistors (M_H ) und erneutes Ausschalten des zweiten Leistungstransistors (M_L ) erfolgt, wenn der so ermittelte weitere U_DS-Spannungswert (V_DS2 ) betragsmäßig oberhalb eines zweiten Schwellwerts (SW2) liegt, der gleich dem ersten Schwellwert (SW1) sein kann;
• • Schritt S39: Einschalten (S39) (= Beibehalten des Zustands aus S35) des ersten Leistungstransistors (M_H ) und Ausschalten (S29) des zweiten Leistungstransistors (M_L ), wenn der so ermittelte weitere U_DS-Spannungswert (V_DS2 ) betragsmäßig wieder unterhalb des zweiten Schwellwerts (SW2) liegt.

A first sub-variant of this sixth variant comprises the steps:

• • step S35 : Switch on the first power transistor again ( M _H ) after the step S34 and again switching off the second power transistor ( M _L ), this especially after a charging time ( T _L ) he follows;
• • step S36 : Recapture of the U _DS voltage ( U _DS ) on the first power transistor ( M _H ) when the first power transistor ( M _H ) is switched on and a further determination of an associated further U _DS voltage value ( V _DS2 );
• • step S37 : Comparison of the further U _DS voltage value determined in this way ( V _DS2 ) with another threshold value ( SW2 ), which is equal to the first threshold ( SW1 ) can be;
• • step S38 : Switching off the first power transistor again ( M _H ) and the second power transistor is switched off again ( M _L ) occurs when the further U _DS voltage value determined in this way ( V _DS2 ) above a second threshold ( SW2 ) which is equal to the first threshold ( SW1 ) can be;
• • step S39 : Turn on ( S39 ) (= Keep status from S35 ) of the first power transistor ( M _H ) and switch off ( S29 ) of the second power transistor ( M _L ) if the further U _DS voltage value determined in this way ( V _DS2 ) again below the second threshold in terms of amount ( SW2 ) lies.

It is preferred to switch on again ( S35 ) after the first shutdown ( S34 ) only after a charging time has passed ( T _L ).

Seventh variant

• • Einschalten (S1) des ersten Leistungstransistors (M_H ) und Ausschalten des zweiten Leistungstransistors (M_L ) zu einem Einschaltzeitpunkt (t₀ );
• • Erfassen (S2) der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen U_DS-Spannungswerts (V_DS ) in einer Aktivzeit (T_A ) nach dem Verstreichen einer Debounce-Zeit (T_D ) nach dem Einschaltzeitpunkt (t₀ );
• • Vergleich (S3) des so ermittelten U_DS-Spannungswerts (V_DS ) betragsmäßig mit einem ersten Schwellwert (SW1) in der Aktiv-Zeit (T_A ) und Abschalten des ersten Leistungstransistors (M_H ) und
• • Abschalten (S4) des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass der so innerhalb der Aktivzeit (T_A ) ermittelte U_DS-Spannungswert (V_DS ) betragsmäßig oberhalb des ersten Schwellwerts (SW1) liegt. Dieser Schritt findet hier deswegen statt, da bei einem derartig schnellen Auftreten der Verletzung des ersten Schwellwerts (SW1) davon ausgegangen werden muss, dass ein Kurzschluss vorliegt. Daher wird hier auch der zweite Leistungstransistor (M_L ) abgeschaltet, da dann ein Querstrom für den möglicherweise vorhandenen Fall eines geschädigten ersten Leistungstransistors (M_H ) ausgeschlossen werden soll. Da der Treiber selbst die Notabschaltung durchführt, ist ein schneller Eingriff eines externen Steuerrechners typischerweise nicht erforderlich. Daher wird bevorzugt, die Information, dass durch die Vorrichtung ein Kurzschluss angenommen wird nicht über eine Interrupt-Leitung signalisiert, sondern das Signal der Interrupt-Leitung (INTN) dient nur zur Signalisierung, dass etwas geschehen ist. Die eigentliche Information wird in einem Datenspeicher des Treibers abgelegt, wo sie von dem externen Steuerrechner gelesen werden kann. Der externe Steuerrechner wird dann typischerweise erst versuchen, den ersten Leistungstransistor (M_H ) auszuschalten und den zweiten Leistungstransistor (M_L ) einzuschalten, um die Bootstrap-Kapazität (C_B ) nachzuladen. Erst dann wird er die Treiber-Register über den Datenbus (DB) lesen und den Kurzschluss als solchen erkennen. Damit der Steuerrechner im Falle eines Kurzschlusses den zweiten Leistungstransistor (M_H ) nicht einschalten kann, blockiert beispielsweise die Überwachungsvorrichtung (UV) und/oder der PWM-Generator (G) ein solches Einschalten nach dem Auftreten dieses Fehlerfalles, bis ein spezielles, separates Freigabekommando des externen Rechnersystems dieses Einschalten des zweiten Leistungstransistors (M_L ) explizit wieder zulässt.; Erfassen (S5) der der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen weiteren U_DS-Spannungswerts (V_DS2 ) in einer erweiterten Aktivzeit (T_EA ) nach dem Verstreichen der Aktivzeit (T_A ) und der Debounce-Zeit (T_D ) nach dem Einschaltzeitpunkt (t₀ );
• • Vergleich (S6) des so ermittelten weiteren U_DS-Spannungswerts (V_DS2 ) betragsmäßig mit einem zweiten Schwellwert (SW2) in der erweiterten Aktiv-Zeit (T_EA ) und
• • Abschalten (S7) des ersten Leistungstransistors (M_H ) und Einschalten des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass die so innerhalb der erweiterten Aktivzeit (T_EA ) ermittelte weitere U_DS-Spannungswert (V_DS2 ) betragsmäßig oberhalb des zweiten Schwellwerts (SW2) liegt, wobei der zweite Schwellwert (SW2) gleich dem ersten Schwellwert (SW1) sein kann. Der erste Leistungstransistor (M_H ) ist in diesem Fall also nicht ausreichend eingeschaltet, weist einen zu hohen Leistungsumsatz auf und muss daher abgeschaltet werden. Dieser Schritt findet in diesem Falle deswegen statt, da dann davon ausgegangen wird, dass ein Kurzschluss zu einer schnelleren Verletzung der Schwellwerte (SW1, SW2) geführt hätte und es sich somit, da die Verletzung für einen Kurzschluss nicht schnell genug erfolgte, um eine Entladung des Bootstrap-Kondensators (C_B ) handeln muss. Für die Nachladung der Bootstrap-Kapazität wird daher in diesem Fall der zweite Leistungstransistor (M_L ) eingeschaltet. Der Schritt S7 ist also der Nachladeschritt. Mit dem Abschalten erfolgt typischerweise eine Signalisierung über eine Interrupt-Leitung (INTN). Mit dem Schritt S7 wird somit ein Laden dieser vermutlich entladenen Bootstrap-Kapazität (C_B ) gestartet.

The seventh variant again relates to a process ( 5 ) for operating a driver stage with a first power transistor ( M _H ), one second power transistor ( M _L ) and with a bootstrap capacity ( C _B ) with a first connection and a second connection. The bootstrap capacity ( C _B ) is recharged when the second power transistor ( M _L ) is switched on. The method of the seventh variant again has the following steps:

• • Turn on ( S1 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) at a switch-on time ( t ₀ );
• • Capture ( S2 ) of the U _DS voltage at the first power transistor ( M _H ) and determining an associated U _DS voltage value ( V _DS ) in an active period ( T _A ) after a debounce time has passed ( T _D ) after the switch-on time ( t ₀ );
• • Comparison ( S3 ) of the U _DS voltage value determined in this way ( V _DS ) in terms of amount with a first threshold value ( SW1 ) in the active time ( T _A ) and switching off the first power transistor ( M _H ) and
• • Switch off ( S4 ) of the second power transistor ( M _L ), if the comparison shows that within the active time ( T _A ) determined U _DS voltage value ( V _DS ) above the first threshold ( SW1 ) lies. This step takes place here because if the violation of the first threshold value occurs so quickly ( SW1 ) it must be assumed that there is a short circuit. Therefore, the second power transistor ( M _L ) switched off because then a cross current for the possibly existing case of a damaged first power transistor ( M _H ) should be excluded. Since the driver performs the emergency shutdown himself, rapid intervention by an external control computer is typically not required. It is therefore preferred that the information that a short circuit is assumed by the device is not signaled via an interrupt line, but rather the signal of the interrupt line ( INTN ) is only used to signal that something has happened. The actual information is stored in a data memory of the driver, where it can be read by the external control computer. The external control computer will then typically first try to connect the first power transistor ( M _H ) turn off and the second power transistor ( M _L ) to turn on bootstrap capacity ( C _B ) reload. Only then will it register the driver via the data bus ( DB ) read and recognize the short circuit as such. So that the control computer, in the event of a short circuit, the second power transistor ( M _H ) cannot switch on, for example the monitoring device blocks ( UV ) and / or the PWM generator (G) such a switching on after the occurrence of this fault, until a special, separate release command of the external computer system switches on the second power transistor ( M _L ) explicitly allows again .; Capture ( S5 ) of the U _DS voltage at the first power transistor ( M _H ) and determining an associated further U _DS voltage value ( V _DS2 ) in an extended active period ( T _EA ) after the active time has passed ( T _A ) and the debounce time ( T _D ) after the switch-on time ( t ₀ );
• • Comparison ( S6 ) of the further U _DS voltage value determined in this way ( V _DS2 ) in terms of amount with a second threshold ( SW2 ) in the extended active time ( T _EA ) and
• • Switch off ( S7 ) of the first power transistor ( M _H ) and switching on the second power transistor ( M _L ), if the comparison shows that within the extended active time ( T _EA ) determined further U _DS voltage value ( V _DS2 ) above the second threshold ( SW2 ), with the second threshold ( SW2 ) equal to the first threshold ( SW1 ) can be. The first power transistor ( M _H ) is not switched on sufficiently in this case, has too high a power conversion and must therefore be switched off. This step takes place in this case because it is then assumed that a short circuit leads to a faster violation of the threshold values ( SW1 , SW2 ) and, as the short-circuit injury did not occur quickly enough, the bootstrap capacitor ( C _B ) must act. In this case, the second power transistor ( M _L ) switched on. The step S7 is the reload step. When switching off, there is typically signaling via an interrupt line ( INTN ). With the step S7 a loading of this presumably discharged bootstrap capacity ( C _B ) started.

• • Einschalten (S8) des ersten Leistungstransistors (M_H ) und Abschalten des zweiten Leistungstransistors (M_L ) nach einer Ladezeit (T_L ), wenn vor dem Verstreichen der Ladezeit (T_L ) ein Vergleich des ermittelten weiteren U_DS-Spannungswerts (V_DS2 ) betragsmäßig mit dem zweiten Schwellwert (SW2) in der erweiterten Aktiv-Zeit (T_EA ) ein Abschalten des ersten Leistungstransistors (M_H ) und ein Einschalten des zweiten Leistungstransistors (M_L ) ausgelöst hatte. Im vorhergehenden Schritt (S7) wurde vermutet, dass es sich um einen entladenen Bootstrap-Kondensator (C_B ) handelt. Nun wird das Laden dieser vermutlich zuvor entladenen Bootstrap-Kapazität (C_B ) in der Annahme beendet, dass die Bootstrap-Kapazität (C_B ) ausreichend geladen ist. Bevorzugt wird dies über die Interrupt-Leitung (INTN) signalisiert.

A first sub-variant of this seventh variant (see also 5 ) concerns the timed reloading of the bootstrap capacity ( C _B ) and has the following additional step compared to the basic process of this seventh variant:

• • Turn on ( S8 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) after a charging time ( T _L ) if before the charging time has passed ( T _L ) a comparison of the determined further U _DS voltage value ( V _DS2 ) in terms of amount with the second threshold ( SW2 ) in the extended active time ( T _EA ) switching off the first power transistor ( M _H ) and switching on the second power transistor ( M _L ) had triggered. In the previous step ( S7 ) was suspected to be a discharged bootstrap capacitor ( C _B ) acts. Now the loading of this presumably previously discharged bootstrap capacity ( C _B ) ended on the assumption that the bootstrap capacity ( C _B ) is sufficiently charged. This is preferred via the interrupt line ( INTN ) signals.

• • Erfassen (S9) der U_DS-Spannung (U_DS ) am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen dritten U_DS-Spannungswerts (V_DS3 ) in einer Ladezeit (T_L ), die mit dem Abschalten des ersten Leistungstransistors (M_H ) und dem Einschalten des zweiten Leistungstransistors (M_L ) im vorhergehenden Schritt (S7) beginnt;
• • Vergleich (S10) des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit einem dritten Schwellwert (SW3), der gleich dem ersten Schwellwert (SW1) und dem zweiten Schwellwert (SW2) sein kann;
• • Einschalten (S8) des ersten Leistungstransistors (M_H ) und Abschalten des zweiten Leistungstransistors (M_L ), wenn vor dem Verstreichen der Ladezeit (T_L ) der Vergleich des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit dem dritten Schwellwert (SW3) ergibt, dass der ermittelte dritte U_DS-Spannungswert (V_DS3 ) betragsmäßig oberhalb des dritten Schwellwerts (SW3) liegt. Der Ladevorgang der Bootstrap-Kapazität (C_B ) wird hier also nicht zeitgesteuert, sondern in Abhängigkeit vom Abschaltzustand des ersten Leistungstransistors (M_H ) durchgeführt. Hierdurch wird bei hohen Duty-Cyclen in der Nähe von 100% der jeweilige Duty-Cycle durch das Nachladen der Bootstrap-Kapazität (C_B ) nur noch in dem unbedingt nötigen Umfang durchgeführt. Bei einer reinen Zeitsteuerung mit einer Ladezeit (T_L ), länger als die maximale Ladezeit, muss ein zeitlicher Vorhalt aus Sicherheitsgründen berücksichtigt werden, um jede Art von Querstrom auszuschließen. Daher ist die Störung im Falle einer reinen Zeitsteuerung massiv größer als bei der hier ebenso vorgeschlagenen Nachlademethode über die U_DS-Steuerung, bei der die Unterscheidung zwischen Kurzschlussfall und Nachladefall über den Zeitpunkt des Auftretens der Schwellwertverletzung nach dem Einschalten erfolgt. Die erweiterte Aktivzeit (T_EA ) kann übrigens so gewählt werden, dass die Summe der Debounche-Zeit (T_D ) plus der Aktivzeit (T_A ) plus der erweiterten Aktivzeit (T_EA ) gleich der PWM-Periode ist, sodass dann bis auf die Debounce-Zeiten (T_D ) stets eine Überwachung stattfindet. Bevorzugt wird die Aktivzeit (T_A ) in jeder PWM-Periode durchlaufen, also auch dann, wenn keine Schaltzustandsänderung der Leistungstransistoren (M_H , M_L ) zwischen zwei PWM-Perioden erfolgt, also die Debounce-Zeit (T_D ) zu 0s gesetzt werden kann.

A second sub-variant of a U _DS control for reloading the bootstrap capacity ( C _B ) of this seventh variant has the following additional steps compared to the basic method and the first sub-variant of this seventh variant:

• • Capture ( S9 ) of the U _DS voltage ( U _DS ) on the first power transistor ( M _H ) and determining an associated third U _DS voltage value ( V _DS3 ) in one loading time ( T _L ), which occurs when the first power transistor ( M _H ) and switching on the second power transistor ( M _L ) in the previous step ( S7 ) begins;
• • Comparison ( S10 ) of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with a third threshold ( SW3 ), which is equal to the first threshold ( SW1 ) and the second threshold ( SW2 ) can be;
• • Turn on ( S8 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) if before the charging time has passed ( T _L ) the comparison of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with the third threshold ( SW3 ) shows that the determined third U _DS voltage value ( V _DS3 ) above the third threshold ( SW3 ) lies. Bootstrap Capacity Loading ( C _B ) is not time-controlled here, but depending on the switch-off state of the first power transistor ( M _H ) carried out. With high duty cycles close to 100%, the respective duty cycle is thereby reloaded by the bootstrap capacity ( C _B ) only carried out to the extent absolutely necessary. With a pure time control with a charging time ( T _L ), longer than the maximum charging time, a time reserve must be taken into account for safety reasons in order to exclude any type of cross current. Therefore, the disorder is in the case of a pure Time control massively larger than in the reloading method also proposed here via the U _DS control, in which the distinction between short-circuit case and reload case is made via the time of occurrence of the threshold violation after switching on. The extended active time ( T _EA ) can be chosen so that the sum of the debounche time ( T _D ) plus the active time ( T _A ) plus the extended active time ( T _EA ) is equal to the PWM period, so that apart from the debounce times ( T _D ) surveillance always takes place. The active time is preferred ( T _A ) run through in each PWM period, even if there is no change in the switching state of the power transistors ( M _H , M _L ) between two PWM periods, i.e. the debounce time ( T _D ) can be set to 0s.

• • Erfassen (S9) der U_DS-Spannung (U_DS ) am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen dritten U_DS-Spannungswerts in einer Ladezeit (T_L ), die mit dem Abschalten des ersten Leistungstransistors (M_H ) und dem Einschalten des zweiten Leistungstransistors (M_L ) beginnt;
• • Vergleich (S10) des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit einem dritten Schwellwert (SW3), der gleich dem ersten Schwellwert (SW1) und dem dritten Schwellwert (SW3) sein kann;
• • Abschalten (S11) des ersten Leistungstransistors (M_H ) und Abschalten des zweiten Leistungstransistors (M_L ), wenn ein Vergleich des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit einem dritten Schwellwert (SW3) ergibt, dass der ermittelte dritte U_DS-Spannungswert (V_DS3 ) betragsmäßig unterhalb des dritten Schwellwerts liegt. Hier war also das Nachladen der Bootstrap-Kapazität (C_B ) wahrscheinlich nicht erfolgreich oder es liegt doch ein Kurzschluss vor. Daher erfolgt wie im Kurzschlussfall in Schritt S4 hier bevorzugt wieder eine Signalisierung über die Interrupt-Leitung (INTN) und ein Register der Integrierten Schaltung (IC) bzw. der Überwachungsvorrichtung (UV) und/oder des PWM-generators (PWMG). Wie in Schritt S4 verhindern bevorzugt der PWM-Generator (PWMG) und/oder die Überwachungsvorrichtung (UV) das Wiedereinschalten des zweiten Leistungstransistors (M_L ) durch einen externen Steuerrechner solange, bis durch einen speziellen Entriegelungsbefehl des externen Steuerrechners über ein spezielles Register ein solches Wiedereinschalten wieder zugelassen wird. Bevorzugt signalisiert die integrierte Schaltung (IC), insbesondere die Überwachungsvorrichtung (UV) und/oder der PWM-Generator (PWMG), in diesem Fehlerfall über ein Register und dem Datenbus (DB) dem externen Steuerrechner einen anderen Fehler-Code als im Falle der Kurzschlusserkennung im Schritt S4.

A third sub-variant (see also 6 ) of this seventh variant has the following additional steps compared to the basic method and the first and second sub-variants of this seventh variant:

• • Capture ( S9 ) of the U _DS voltage ( U _DS ) on the first power transistor ( M _H ) and determining an associated third U _DS voltage value in a charging time ( T _L ), which occurs when the first power transistor ( M _H ) and switching on the second power transistor ( M _L ) begins;
• • Comparison ( S10 ) of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with a third threshold ( SW3 ), which is equal to the first threshold ( SW1 ) and the third threshold ( SW3 ) can be;
• • Switch off ( S11 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) if a comparison of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with a third threshold ( SW3 ) shows that the determined third U _DS voltage value ( V _DS3 ) is below the third threshold. So here was the reload of the bootstrap capacity ( C _B ) probably not successful or there is a short circuit. Therefore, as in the event of a short circuit, step S4 signaling via the interrupt line ( INTN ) and a register of the integrated circuit ( IC ) or the monitoring device ( UV ) and / or the PWM generator ( PWMG ). As in step S4 prevent the PWM generator ( PWMG ) and / or the monitoring device ( UV ) switching on the second power transistor ( M _L ) by an external control computer until such a restart is permitted again by a special unlock command from the external control computer via a special register. The integrated circuit preferably signals ( IC ), especially the monitoring device ( UV ) and / or the PWM generator ( PWMG ), in this case of an error via a register and the data bus ( DB ) the external control computer has a different error code than in the case of short circuit detection in the step S4 .

Advantage of the invention

The main advantage of the invention described above is that the achievable duty cycle at the phase output ( PH ) is closer to 100% than can be with solutions in the prior art and that between a parasitic discharge of the bootstrap capacity ( C _B ) and a short circuit can be distinguished. If the recharge of the bootstrap capacity occurs too often, i.e. if the period between two recharges falls below a minimum recharge period, this indicates a fault in the bootstrap capacitor ( C _B ) close, which may be relevant to security and which may be reported and treated separately if necessary. This distinction is also not possible in the prior art. The advantages are not limited to this.

Figure list

• 1 shows a schematically simplified block diagram of the driver stage according to the invention.
• 2nd shows the course of the drain-source voltage ( V _DS ) on the first power transistor ( M _H ) versus time ( t ) and the course of the interrupt signal ( INTN ).
• 3rd shows the course of the bootstrap tension ( V _BST -V _PH ) and the gate-source voltage ( V _GS ) on the first power transistor ( M _H ) against the time after switching on the first line transistor ( M _H ) at a switch-on time ( T ₀ ).
• 4th shows the waveform of various signals during a reload for the bootstrap capacity ( C _B ).
• 5 shows the procedure for a time and demand-controlled reloading of the bootstrap capacity ( C _B ).
• 6 shows the procedure for a U _DS -controlled and demand-controlled reloading of the bootstrap capacity ( C _B ).
• 7 shows the procedure for a bootstrap voltage ( V _BST ) and the reload time ( T _L ) controlled reloading of the bootstrap capacity ( C _B ) without distinction between a short circuit and a defect in the bootstrap capacity ( C _B ).
• 8th shows the procedure for a drain-source voltage ( V _DS ) and the reload time ( T _L ) controlled reloading of the bootstrap capacity ( C _B ) without distinguishing between a short circuit and a defect in the bootstrap capacity ( C _B ).
• 9 shows the procedure for a bootstrap voltage ( V _BST ) and the reload time ( T _L ) controlled reloading of the bootstrap capacity ( C _B ) without termination condition.

Description of the figures

1
Figure shows a schematically simplified block diagram of the driver stage according to the invention. The core of the driver stage is the half bridge ( M _H , M _L ), consisting of a first power transistor ( M _H ) and a second power transistor ( M _L ). The exit of the half bridge ( M _H , M _L ) forms the phase output ( PH ). The control electrode of the first power transistor ( M _H ) is controlled by the first gate control signal ( GH ) controlled. The first gate control signal ( GH ) is supported by the first gate driver ( GT _H ) depending on the first PWM control signal ( PWMH ) for driving the control electrode of the first power transistor ( M _H ) generated. The first gate driver ( GT _H ) either from the bootstrap capacity ( C _B ) or from the positive supply voltage line ( U _S ) supplied directly or indirectly with electrical energy. Is the potential of the positive supply voltage line (Us) compared to the reference potential ( GND ) too low, the first gate driver is supplied ( GT _H ) from the bootstrap capacity ( C _B ), which is why the bootstrap capacity ( C _B ) must always be sufficiently charged. The monitoring device ( UV ) can turn off the first power transistor ( M _H ) via the first gate control signal ( GH ) by forcing the monitoring device ( UV ) using the first enable signal ( EN _H ) the first gate driver ( GT _H ) signals the first gate control signal ( G _H ) in such a state that the first power transistor ( M _H ) turns off, so locks. The control electrode of the second power transistor ( M _L ) is controlled by the second gate control signal ( G _L ) controlled. The second gate control signal ( G _L ) is controlled by the second gate driver ( GT _L ) depending on the second PWM control signal ( PWML ) for controlling the control electrode ( G _L ) of the second power transistor ( M _L ) generated. The second gate driver ( GT _L ) typically only from the positive supply voltage line ( U _S ) supply directly or indirectly with electrical energy. The monitoring device ( UV ) can turn off the second power transistor ( M _L ) via the second gate control signal ( GL ) by forcing the monitoring device ( UV ) by means of the second enable signal ( EN _L ) the second gate driver ( GT _L ) signals the first gate control signal ( G _L ) to such a state that the second power transistor ( M _L ) turns off, so locks. A power supply circuit ( SV ) preferably generates a constant voltage ( V _VG ) at their voltage regulator output ( VG ) against the reference potential ( GND ). This constant voltage ( V _VG ) is supported by a backup capacitor ( C _VG ) stabilized. The voltage regulator output ( VG ) of the power supply circuit ( SV ) can be via the diode ( D ) the bootstrap capacitor ( C _B ) charge when the second power transistor ( M _L ) is switched on and the first power transistor ( M _H ) is switched off because the other connection of the bootstrap capacity ( C _B ) with the reference potential ( GND ) connected is. The bootstrap capacitor can be the output stage of the first gate driver ( GT _H ) supply with electrical energy especially if the electrical potential on the positive supply voltage line ( U _S ) for whatever reason breaks down at short notice. This creates a sufficiently high potential for the control connection of the first power transistor ( M _H ) ensured if this is to be switched off. The monitoring device ( UV ) the potential at the bootstrap node ( BST ) against the reference potential and on the other hand the voltage drop as U _DS voltage ( U _DS ) via the first power transistor ( M _H ) monitor when switched off. In the event of a fault, a deviation can be determined and the power transistors ( M _H , M _L ) can be switched off. The PWM generator ( PWMG ) preferably provides the first PWM signal ( PWMH ) for driving the first power transistor ( M _H ) and the second PWM signal ( PWML ) for the control of the second power transistor ( M _L ) ready. In the event of an error, the monitoring unit communicates in this example using an interrupt line ( INTN ) with a processor not shown.

2nd shows the course of the drain-source voltage ( V _DS ) on the first power transistor ( M _H ) versus time ( t ) and the course of the interrupt signal ( INTN ) when the first power transistor is switched on ( M _H ). Discharging the bootstrap capacitor ( C _B ) there is an increase in the drain-source voltage ( V _DS ) on the first power transistor ( M _H ), because the transistor can no longer be opened sufficiently. This increases its on-resistance and the drain-source voltage ( V _DS ) increases. The monitoring device ( UV ) detects this rise and activates the interrupt signal ( INTN ) when the detection threshold is exceeded ( TH ). The exemplary interrupt signal is drawn here as low active.

3rd shows the course of the bootstrap tension ( V _BST -V _PH ) above the bootstrap capacity ( C _B ) and the gate-source voltage ( V _GS ) on the first power transistor ( M _H ) against the time ( t ) after switching on the first line transistor ( M _H ) at a switch-on time ( T ₀ ). In the debouncing period ( T _D ) the signals must first stabilize. Only then does the active time begin ( T _A ). If an error occurs during this time, it is most likely a short circuit. However, the bootstrap capacity ( C _B ) act. In this case, the monitoring device ( UV ) therefore first the second power transistor ( M _L ) and the first power transistor ( M _H ) because then the bootstrap capacity ( C _B ) from the power supply circuit ( SV ) via the diode ( D ) can be loaded. After a charging time ( T _L ) or if the voltage is above the bootstrap capacity ( C _B ) is sufficient, the monitoring device switches the first power transistor again ( M _H ) and the second power transistor ( M _L ) out. If the error is still there, it is most likely a short circuit. For example, one of the two power transistors ( M _H , M _L ) be alloyed. Therefore, both power transistors ( M _H , M _L ) switched off. It is also conceivable to fundamentally evaluate such an error in the active time as a short circuit and then immediately switch both power transistors ( M _H , M _L ) switch off. If an error occurs in the extended active time ( T _EA ), it is most likely a discharged bootstrap capacitor ( C _B ). In this case, the second power transistor ( M _L ) may be waived. However, it is recommended to use the two-step procedure.

4th shows the exemplary signal curve of various signals during a reloading process for the bootstrap capacity ( CB ). The monitoring device ( UV ) first activates the interrupt signal ( INTN ). In this example, the monitoring unit ( UV ) by means of this signal ( INTN ) and possibly one or more other control signals ( ST ) the PWM generator ( PWMG ). Because of the chain of action ( 1 ) is then in this example by the PWM generator ( PWMG ) the first PWM signal ( PWMH ) deactivated and the first power transistor ( M _H ) is switched off. Alternatively, this connection can of course also be done via the first enable signal ( EN _H ) and the first gate driver ( GT _H ) respectively. After a first dead time ( T _T1 ), which is typically generated by the PWM generator ( PWMG ) or the monitoring device ( UV ) is ensured, the second power transistor ( M _L ) by means of the second PWM signal ( PWML ) turned on what the interrupt signal ( INTN ) by means of the chain of action ( 2nd ) resets.

5 shows the procedure for a time and demand-controlled reloading of the bootstrap capacity ( C _B ).

• • Einschalten (S1) des ersten Leistungstransistors (M_H ) und Ausschalten des zweiten Leistungstransistors (M_L ) zu einem Einschaltzeitpunkt (t₀ );
• • Erfassen (S2) der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen U_DS-Spannungswerts (V_DS ) in einer Aktivzeit (T_A ) nach dem Verstreichen einer Debounce-Zeit (T_D ) nach dem Einschaltzeitpunkt (t₀ );
• • Vergleich (S3) des so ermittelten U_DS-Spannungswerts (V_DS ) betragsmäßig mit einem ersten Schwellwert (SW1) in der Aktiv-Zeit (T_A ) und Abschalten des ersten Leistungstransistors (M_H ) und
• • Abschalten (S4) des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass der so innerhalb der Aktivzeit (T_A ) ermittelte U_DS-Spannungswert (V_DS ) betragsmäßig oberhalb des ersten Schwellwerts (SW1) liegt. Dieser Schritt findet hier deswegen statt, da bei einem derartig schnellen Auftreten der Verletzung des ersten Schwellwerts (SW1) davon ausgegangen werden muss, dass ein Kurzschluss vorliegt. Daher wird hier auch der zweite Leistungstransistor (M_L ) abgeschaltet, da dann ein Querstrom für den möglicherweise vorhandenen Fall eines geschädigten ersten Leistungstransistors (M_H ) ausgeschlossen werden soll. Da der Treiber selbst die Notabschaltung durchführt, ist ein schneller Eingriff eines externen Steuerrechners typischerweise nicht erforderlich. Daher wird bevorzugt, die Information, dass durch die Vorrichtung ein Kurzschluss angenommen wird nicht über eine Interrupt-Leitung signalisiert, sondern das Signal der Interrupt-Leitung (INTN) dient nur zur Signalisierung, dass etwas geschehen ist. Die eigentliche Information wird in einem Datenspeicher des Treibers abgelegt, wo sie von dem externen Steuerrechner gelesen werden kann. Der externe Steuerrechner wird dann typischerweise erst versuchen, den ersten Leistungstransistor (M_H ) auszuschalten und den zweiten Leistungstransistor (M_L ) einzuschalten, um die Bootstrap-Kapazität (C_B ) nachzuladen. Erst dann wird er die Treiber-Register über den Datenbus (DB) lesen und den Kurzschluss als solchen erkennen. Damit der Steuerrechner im Falle eines Kurzschlusses den zweiten Leistungstransistor (M_H ) nicht einschalten kann, blockiert beispielsweise die Überwachungsvorrichtung (UV) und/oder der PWM-Generator (G) ein solches Einschalten nach dem Auftreten dieses Fehlerfalles, bis ein spezielles, separates Freigabekommando des externen Rechnersystems dieses Einschalten des zweiten Leistungstransistors (M_L ) explizit wieder zulässt.; Erfassen (S5) der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen weiteren U_DS-Spannungswerts (V_DS2 ) in einer erweiterten Aktivzeit (T_EA ) nach dem Verstreichen der Aktivzeit (T_A ) und der Debounce-Zeit (T_D ) nach dem Einschaltzeitpunkt (t₀ );
• • Vergleich (S6) des so ermittelten weiteren U_DS-Spannungswerts (V_DS2 ) betragsmäßig mit einem zweiten Schwellwert (SW2) in der erweiterten Aktiv-Zeit (T_EA ) und
• • Abschalten (S7) des ersten Leistungstransistors (M_H ) und Einschalten des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass der so innerhalb der erweiterten Aktivzeit (T_EA ) ermittelte weitere U_DS-Spannungswert (V_DS2 ) betragsmäßig oberhalb des zweiten Schwellwerts (SW2) liegt, wobei der zweite Schwellwert (SW2) gleich dem ersten Schwellwert (SW1) sein kann. Der erste Leistungstransistor (M_H ) ist in diesem Fall also nicht ausreichend eingeschaltet, weist einen zu hohen Leistungsumsatz auf und muss daher abgeschaltet werden. Dieser Schritt findet in diesem Falle deswegen statt, da dann davon ausgegangen wird, dass ein Kurzschluss zu einer schnelleren Verletzung der Schwellwerte (SW1, SW2) geführt hätte und es sich somit, da die Verletzung für einen Kurzschluss nicht schnell genug erfolgte, um eine Entladung des Bootstrap-Kondensators (C_B ) handeln muss. Für die Nachladung der Bootstrap-Kapazität wird daher in diesem Fall der zweite Leistungstransistor (M_L ) eingeschaltet. Der Schritt S7 ist also der Nachladeschritt. Mit dem Abschalten erfolgt typischerweise eine Signalisierung über eine Interrupt-Leitung (INTN).
• • Einschalten (S8) des ersten Leistungstransistors (M_H ) und Abschalten des zweiten Leistungstransistors (M_L ) nach einer Ladezeit (T_L ), wenn vor dem Verstreichen der Ladezeit (T_L ) ein Vergleich des ermittelten weiteren U_DS-Spannungswerts (V_DS2 ) betragsmäßig mit dem zweiten Schwellwert (SW2) in der erweiterten Aktiv-Zeit (T_EA ) ein Abschalten des ersten Leistungstransistors (M_H ) und ein Einschalten des zweiten Leistungstransistors (M_L ) ausgelöst hatte. Im vorhergehenden Schritt (S7) wurde vermutet, dass es sich um einen entladenen Bootstrap-Kondensator (C_B ) handelt. Nun wird ein Laden dieser vermutlich entladenen Bootstrap-Kapazität (C_B ) gestartet.

It essentially corresponds to the seventh method variant for operating a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ) and with a bootstrap capacity ( C _B ) with a first connection and a second connection. The bootstrap capacity ( C _B ) is recharged when the second power transistor ( M _L ) is switched on. The process of the seventh variant again has the following steps:

• • Turn on ( S1 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) at a switch-on time ( t ₀ );
• • Capture ( S2 ) of the U _DS voltage at the first power transistor ( M _H ) and determining an associated U _DS voltage value ( V _DS ) in an active period ( T _A ) after a debounce time has passed ( T _D ) after the switch-on time ( t ₀ );
• • Comparison ( S3 ) of the U _DS voltage value determined in this way ( V _DS ) in terms of amount with a first threshold value ( SW1 ) in the active time ( T _A ) and switching off the first power transistor ( M _H ) and
• • Switch off ( S4 ) of the second power transistor ( M _L ), if the comparison shows that within the active time ( T _A ) determined U _DS voltage value ( V _DS ) above the first threshold ( SW1 ) lies. This step takes place here because if the violation of the first threshold value occurs so quickly ( SW1 ) it must be assumed that there is a short circuit. Therefore, the second power transistor ( M _L ) switched off because then a cross current for the possibly existing case of a damaged first power transistor ( M _H ) should be excluded. Since the driver performs the emergency shutdown himself, rapid intervention by an external control computer is typically not required. It is therefore preferred that the information that a short circuit is assumed by the device is not signaled via an interrupt line, but rather the signal of the interrupt line ( INTN ) is only used to signal that something has happened. The actual information is stored in a data memory of the driver, where it can be read by the external control computer. The external control computer will then typically first try to connect the first power transistor ( M _H ) turn off and the second power transistor ( M _L ) to turn on bootstrap capacity ( C _B ) reload. Only then will it register the driver via the data bus ( DB ) read and recognize the short circuit as such. So that the control computer, in the event of a short circuit, the second power transistor ( M _H ) cannot switch on, for example the monitoring device blocks ( UV ) and / or the PWM generator (G) such a switching on after the occurrence of this fault, until a special, separate release command of the external computer system switches on the second power transistor ( M _L ) explicitly allows again .; Capture ( S5 ) of the U _DS voltage at the first power transistor ( M _H ) and determining an associated further U _DS voltage value ( V _DS2 ) in an extended active period ( T _EA ) after the active time has passed ( T _A ) and the debounce time ( T _D ) after the switch-on time ( t ₀ );
• • Comparison ( S6 ) of the further U _DS voltage value determined in this way ( V _DS2 ) in terms of amount with a second threshold ( SW2 ) in the extended active time ( T _EA ) and
• • Switch off ( S7 ) of the first power transistor ( M _H ) and switching on the second power transistor ( M _L ), if the comparison shows that within the extended active time ( T _EA ) determined further U _DS voltage value ( V _DS2 ) above the second threshold ( SW2 ), with the second threshold ( SW2 ) equal to the first threshold ( SW1 ) can be. The first power transistor ( M _H ) is not switched on sufficiently in this case, has too high a power conversion and must therefore be switched off. This step takes place in this case because it is then assumed that a short circuit leads to a faster violation of the threshold values ( SW1 , SW2 ) and, as the short-circuit injury did not occur quickly enough, the bootstrap capacitor ( C _B ) must act. In this case, the second power transistor ( M _L ) switched on. The step S7 is the reload step. When switching off, there is typically signaling via an interrupt line ( INTN ).
• • Turn on ( S8 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) after a charging time ( T _L ) if before the charging time has passed ( T _L ) a comparison of the determined further U _DS voltage value ( V _DS2 ) in terms of amount with the second threshold ( SW2 ) in the extended active time ( T _EA ) switching off the first power transistor ( M _H ) and switching on the second power transistor ( M _L ) had triggered. In the previous step ( S7 ) was suspected to be a discharged bootstrap capacitor ( C _B ) acts. Now a loading of this presumably discharged bootstrap capacity ( C _B ) started.

At the end of the PWM period ( T _PWM ) the cycle preferably starts all over again.

• • Einschalten (S1) des ersten Leistungstransistors (M_H ) und Ausschalten des zweiten Leistungstransistors (M_L ) zu einem Einschaltzeitpunkt (t₀ );
• • Erfassen (S2) der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen U_DS-Spannungswerts (V_DS ) in einer Aktivzeit (T_A ) nach dem Verstreichen einer Debounce-Zeit (T_D ) nach dem Einschaltzeitpunkt (t₀ );
• • Vergleich (S3) des so ermittelten U_DS-Spannungswerts (V_DS ) betragsmäßig mit einem ersten Schwellwert (SW1) in der Aktiv-Zeit (T_A ) und Abschalten des ersten Leistungstransistors (M_H ) und
• • Abschalten (S4) des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass der so innerhalb der Aktivzeit (T_A ) ermittelte U_DS-Spannungswert (V_DS ) betragsmäßig oberhalb des ersten Schwellwerts (SW1) liegt. Dieser Schritt findet hier deswegen statt, da bei einem derartig schnellen Auftreten der Verletzung des ersten Schwellwerts (SW1) davon ausgegangen werden muss, dass ein Kurzschluss vorliegt. Daher wird hier auch der zweite Leistungstransistor (M_L ) abgeschaltet, da dann ein Querstrom für den möglicherweise vorhandenen Fall eines geschädigten ersten Leistungstransistors (M_H ) ausgeschlossen werden soll. Da der Treiber selbst die Notabschaltung durchführt, ist ein schneller Eingriff eines externen Steuerrechners typischerweise nicht erforderlich. Daher wird bevorzugt, die Information, dass durch die Vorrichtung ein Kurzschluss angenommen wird nicht über eine Interrupt-Leitung signalisiert, sondern das Signal der Interrupt-Leitung (INTN) dient nur zur Signalisierung, dass etwas geschehen ist. Die eigentliche Information wird in einem Datenspeicher des Treibers abgelegt, wo sie von dem externen Steuerrechner gelesen werden kann. Der externe Steuerrechner wird dann typischerweise erst versuchen, den ersten Leistungstransistor (M_H ) auszuschalten und den zweiten Leistungstransistor (M_L ) einzuschalten, um die Bootstrap-Kapazität (C_B ) nachzuladen. Erst dann wird er die Treiber-Register über den Datenbus (DB) lesen und den Kurzschluss als solchen erkennen. Damit der Steuerrechner im Falle eines Kurzschlusses den zweiten Leistungstransistor (M_H ) nicht einschalten kann, blockiert beispielsweise die Überwachungsvorrichtung (UV) und/oder der PWM-Generator (G) ein solches Einschalten nach dem Auftreten dieses Fehlerfalles, bis ein spezielles, separates Freigabekommando des externen Rechnersystems dieses Einschalten des zweiten Leistungstransistors (M_L ) explizit wieder zulässt. Erfassen (S5) der U_DS-Spannung am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen weiteren U_DS-Spannungswerts (V_DS2 ) in einer erweiterten Aktivzeit (T_EA ) nach dem Verstreichen der Aktivzeit (T_A ) und der Debounce-Zeit (T_D ) nach dem Einschaltzeitpunkt (t₀ );
• • Vergleich (S6) des so ermittelten weiteren U_DS-Spannungswerts (V_DS2 ) betragsmäßig mit einem zweiten Schwellwert (SW2) in der erweiterten Aktiv-Zeit (T_EA ) und
• • Abschalten (S7) des ersten Leistungstransistors (M_H ) und Einschalten des zweiten Leistungstransistors (M_L ), wenn der Vergleich ergibt, dass der so innerhalb der erweiterten Aktivzeit (T_EA ) ermittelte weitere U_DS-Spannungswert (V_DS2 ) betragsmäßig oberhalb des zweiten Schwellwerts (SW2) liegt, wobei der zweite Schwellwert (SW2) gleich dem ersten Schwellwert (SW1) sein kann. Der erste Leistungstransistor (M_H ) ist in diesem Fall also nicht ausreichend eingeschaltet, weist einen zu hohen Leistungsumsatz auf und muss daher abgeschaltet werden. Dieser Schritt findet in diesem Falle deswegen statt, da dann davon ausgegangen wird, dass ein Kurzschluss zu einer schnelleren Verletzung der Schwellwerte (SW1, SW2) geführt hätte und es sich somit, da die Verletzung für einen Kurzschluss nicht schnell genug erfolgte, um eine Entladung des Bootstrap-Kondensators (C_B ) handeln muss. Für die Nachladung der Bootstrap-Kapazität wird daher in diesem Fall der zweite Leistungstransistor (M_L ) eingeschaltet. Der Schritt S7 ist also der Nachladeschritt. Mit dem Abschalten erfolgt typischerweise eine Signalisierung über eine Interrupt-Leitung (INTN).
• • Erfassen (S9) der der U_DS-Spannung (U_DS ) am ersten Leistungstransistor (M_H ) und Ermitteln eines zugehörigen dritten U_DS-Spannungswerts (V_DS3 ) in einer Ladezeit (T_L ), die mit dem Abschalten des ersten Leistungstransistors (M_H ) und dem Einschalten des zweiten Leistungstransistors (M_L ) im vorhergehenden Schritt (S7) beginnt;
• • Vergleich (S10) des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit einem dritten Schwellwert (SW3), der gleich dem ersten Schwellwert (SW1) und dem zweiten Schwellwert (SW2) sein kann;
• • Einschalten (S8) des ersten Leistungstransistors (M_H ) und Abschalten des zweiten Leistungstransistors (M_L ), wenn vor dem Verstreichen der Ladezeit (T_L ) der Vergleich des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit dem dritten Schwellwert (SW3) ergibt, dass der ermittelte dritte U_DS-Spannungswert (V_DS3 ) betragsmäßig oberhalb des dritten Schwellwerts (SW3) liegt. Der Ladevorgang der Bootstrap-Kapazität (C_B ) wird hier also nicht zeitgesteuert, sondern in Abhängigkeit vom Abschaltzustand des ersten Leistungstransistors (M_H ) durchgeführt. Hierdurch wird bei hohen Duty-Cyclen in der Nähe von 100% der jeweilige Duty-Cycle durch das Nachladen der Bootstrap-Kapazität (C_B ) nur noch in dem unbedingt nötigen Umfang durchgeführt. Bei einer reinen Zeitsteuerung mit einer Ladezeit (T_L ) länger als die maximale Ladezeit muss ein zeitlicher Vorhalt aus Sicherheitsgründen berücksichtigt werden, um jede Art von Querstrom auszuschließen. Daher ist die Störung im Falle einer reinen Zeitsteuerung massiv größer als bei der hier ebenso vorgeschlagenen Nachlademethode über die U_DS-Steuerung, bei der die Unterscheidung zwischen Kurzschlussfall und Nachladefall über den Zeitpunkt des Auftretens der Schwellwertverletzung nach dem Einschalten erfolgt. Die erweiterte Aktivzeit (T_EA ) kann übrigens so gewählt werden, dass die Summe der Debounche-Zeit (T_D ) plus der Aktivzeit (T_A ) plus der erweiterten Aktivzeit (T_EA ) gleich der PWM-Periode ist, sodass dann bis auf die Debounce-Zeiten (T_D ) stets eine Überwachung stattfindet. Bevorzugt wird die Aktivzeit (T_A ) in jeder PWM-Periode durchlaufen, also auch dann, wenn keine Schaltzustandsänderung der Leistungstransistoren (M_H , M_L ) zwischen zwei PWM-Perioden erfolgt, also die Debounce-Zeit (T_D ) zu 0s gesetzt werden kann.
• • Abschalten (S11) des ersten Leistungstransistors (M_H ) und Abschalten des zweiten Leistungstransistors (M_L ), wenn ein Vergleich des ermittelten dritten U_DS-Spannungswerts (V_DS3 ) betragsmäßig mit einem dritten Schwellwert (SW3), der gleich dem ersten Schwellwert (SW1) und dem dritten Schwellwert (SW3) sein kann, ergibt, dass der ermittelte dritte U_DS-Spannungswert (V_DS3 ) betragsmäßig unterhalb des dritten Schwellwerts liegt. Hier war also das Nachladen der Bootstrap-Kapazität (C_B ) nicht erfolgreich oder es liegt doch ein Kurzschluss vor. Daher erfolgt wie im Kurzschlussfall in Schritt S4 hier bevorzugt wieder eine Signalisierung über die Interrupt-Leitung (INTN) und ein Register der Integrierten Schaltung (IC) bzw. der Überwachungsvorrichtung (UV) und/oder des PWM-Generators (PWMG). Wie in Schritt S4 verhindern der PWM-Generator (PWMG) und/oder die Überwachungsvorrichtung (UV) das Wiedereinschalten des zweiten Leistungstransistors (M_L ) durch einen externen Steuerrechner solange, bis durch einen speziellen Entriegelungsbefehl des externen Steuerrechners über ein spezielles Register ein solches Wiedereinschalten wieder zugelassen wird. Bevorzugt signalisiert die Integrierte Schaltung (IC), insbesondere die Überwachungsvorrichtung (UV) und/oder der PWM-Generator (PWMG) in diesem Fehlerfall über ein Register und dem Datenbus (DB) dem externen Steuerrechner einen anderen Fehler-Code als im Falle der Kurzschlusserkennung im Schritt S4.

6 shows the procedure for a U _DS -controlled and demand-controlled reloading of the bootstrap capacity ( C _B ). It essentially corresponds to the sequence of the ninth method variant for operating a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ) and with a bootstrap capacity ( C _B ) with a first connection and a second connection. The bootstrap capacity ( C _B ) is recharged when the second power transistor ( M _L ) is switched on. The method of the seventh variant has the following steps:

• • Turn on ( S1 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) at a switch-on time ( t ₀ );
• • Capture ( S2 ) of the U _DS voltage at the first power transistor ( M _H ) and determining an associated U _DS voltage value ( V _DS ) in an active period ( T _A ) after a debounce time has passed ( T _D ) after the switch-on time ( t ₀ );
• • Comparison ( S3 ) of the U _DS voltage value determined in this way ( V _DS ) in terms of amount with a first threshold value ( SW1 ) in the active time ( T _A ) and switching off the first power transistor ( M _H ) and
• • Switch off ( S4 ) of the second power transistor ( M _L ), if the comparison shows that within the active time ( T _A ) determined U _DS voltage value ( V _DS ) above the first threshold ( SW1 ) lies. This step takes place here because if the violation of the first threshold value occurs so quickly ( SW1 ) it must be assumed that there is a short circuit. Therefore, the second power transistor ( M _L ) switched off because then a cross current for the possibly existing case of a damaged first power transistor ( M _H ) should be excluded. Since the driver performs the emergency shutdown himself, rapid intervention by an external control computer is typically not required. It is therefore preferred that the information that a short circuit is assumed by the device is not signaled via an interrupt line, but rather the signal of the interrupt line ( INTN ) is only used to signal that something has happened. The real information is in stored in a data memory of the driver, where it can be read by the external control computer. The external control computer will then typically first try to connect the first power transistor ( M _H ) turn off and the second power transistor ( M _L ) to turn on bootstrap capacity ( C _B ) reload. Only then will it register the driver via the data bus ( DB ) read and recognize the short circuit as such. So that the control computer, in the event of a short circuit, the second power transistor ( M _H ) cannot switch on, for example the monitoring device blocks ( UV ) and / or the PWM generator (G) such a switching on after the occurrence of this fault, until a special, separate release command of the external computer system switches on the second power transistor ( M _L ) explicitly allows again. Capture ( S5 ) of the U _DS voltage at the first power transistor ( M _H ) and determining an associated further U _DS voltage value ( V _DS2 ) in an extended active period ( T _EA ) after the active time has passed ( T _A ) and the debounce time ( T _D ) after the switch-on time ( t ₀ );
• • Comparison ( S6 ) of the further U _DS voltage value determined in this way ( V _DS2 ) in terms of amount with a second threshold ( SW2 ) in the extended active time ( T _EA ) and
• • Switch off ( S7 ) of the first power transistor ( M _H ) and switching on the second power transistor ( M _L ), if the comparison shows that within the extended active time ( T _EA ) determined further U _DS voltage value ( V _DS2 ) above the second threshold ( SW2 ), with the second threshold ( SW2 ) equal to the first threshold ( SW1 ) can be. The first power transistor ( M _H ) is not switched on sufficiently in this case, has too high a power conversion and must therefore be switched off. This step takes place in this case because it is then assumed that a short circuit leads to a faster violation of the threshold values ( SW1 , SW2 ) and, as the short-circuit injury did not occur quickly enough, the bootstrap capacitor ( C _B ) must act. In this case, the second power transistor ( M _L ) switched on. The step S7 is the reload step. When switching off, there is typically signaling via an interrupt line ( INTN ).
• • Capture ( S9 ) of the U _DS voltage ( U _DS ) on the first power transistor ( M _H ) and determining an associated third U _DS voltage value ( V _DS3 ) in one loading time ( T _L ), which occurs when the first power transistor ( M _H ) and switching on the second power transistor ( M _L ) in the previous step ( S7 ) begins;
• • Comparison ( S10 ) of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with a third threshold ( SW3 ), which is equal to the first threshold ( SW1 ) and the second threshold ( SW2 ) can be;
• • Turn on ( S8 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) if before the charging time has passed ( T _L ) the comparison of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with the third threshold ( SW3 ) shows that the determined third U _DS voltage value ( V _DS3 ) above the third threshold ( SW3 ) lies. Bootstrap Capacity Loading ( C _B ) is not time-controlled here, but depending on the switch-off state of the first power transistor ( M _H ) carried out. With high duty cycles close to 100%, the respective duty cycle is thereby reloaded by the bootstrap capacity ( C _B ) only carried out to the extent absolutely necessary. With a pure time control with a charging time ( T _L ) For safety reasons, a time reserve longer than the maximum charging time must be taken into account in order to exclude any type of cross current. Therefore, the disturbance in the case of a purely time control is massively greater than in the reloading method also proposed here via the U _DS control, in which the distinction between a short-circuit case and a reload case is based on the time when the threshold value violation occurred after switching on. The extended active time ( T _EA ) can be chosen so that the sum of the debounche time ( T _D ) plus the active time ( T _A ) plus the extended active time ( T _EA ) is equal to the PWM period, so that apart from the debounce times ( T _D ) surveillance always takes place. The active time is preferred ( T _A ) run through in each PWM period, even if there is no change in the switching state of the power transistors ( M _H , M _L ) between two PWM periods, i.e. the debounce time ( T _D ) can be set to 0s.
• • Switch off ( S11 ) of the first power transistor ( M _H ) and switching off the second power transistor ( M _L ) if a comparison of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with a third threshold ( SW3 ), which is equal to the first threshold ( SW1 ) and the third threshold ( SW3 ) results in the third U _DS voltage value ( V _DS3 ) is below the third threshold. So here was the reload of the bootstrap capacity ( C _B ) Not successful or there is a short circuit. Therefore, as in the event of a short circuit, step S4 signaling via the interrupt line ( INTN ) and a register of the integrated circuit ( IC ) or the monitoring device ( UV ) and / or the PWM generator ( PWMG ). As in step S4 prevent the PWM generator ( PWMG ) and / or the monitoring device ( UV ) switching on the second power transistor ( M _L ) by an external control computer until such a restart is permitted again by a special unlock command from the external control computer via a special register. The integrated circuit preferably signals ( IC ), especially the monitoring device ( UV ) and / or the PWM generator ( PWMG ) in this error case via a register and the data bus ( DB ) the external control computer has a different error code than in the case of short circuit detection in the step S4 .

7 shows the procedure for a bootstrap voltage ( V _BST ) and the reload time ( T _L ) controlled reloading of the bootstrap capacity ( C _B ) without distinguishing between a short circuit and a defect in the bootstrap capacity ( C _B ).

The method is shown as an example, as it is from a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ), a first gate driver ( GT _H ), a second gate driver ( GT _L ), with a power supply circuit ( SV ) with a voltage regulator output ( VG ), a diode ( D ), a bootstrap capacity ( C _B ), a positive supply voltage line ( U _S ), a negative supply voltage line ( GND ) and a monitoring device ( UV ) can be executed. The first power transistor ( M _H ) and the second power transistor ( M _L ) are supposed to go back to a half bridge ( M _H , M _L ) with a phase output ( PH ) between the positive supply voltage line (Us) and the negative supply voltage line ( GND ) be connected. Reference is made here to the previous explanations.

In a first step ( S21 ) becomes the first power transistor ( M _H ) turned on and the second power transistor ( M _L ) switched off. The monitoring device ( UV ) detected ( S22 ) then the potential difference between the potential at the bootstrap node ( BST ) and the potential at the phase output ( PH ) the half bridge ( M _H , M _L ) and determined ( S22 ) an associated bootstrap potential difference value ( ΔV _BST ). The monitoring device ( UV ) compares in a further step ( S23 ) then the bootstrap potential difference value determined in this way ( ΔV _BST ) in terms of amount with a first threshold value ( SW1 ). The monitoring device ( UV ) switches in a conditionally executed further step ( S24 ) then the first power transistor ( M _H ) using the first gate driver ( GT _H ) from ( S24 ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) a ( S24 ) if the bootstrap potential difference value determined in this way ( ΔV _BST ) in terms of amount in the comparison step ( S23 ) below the first threshold ( SW1 ) lies.

Some time after a first shutdown (in step S24 ), which occurred because the bootstrap potential difference value ( ΔV _BST ) during the comparison step ( S23 ) below the first threshold ( SW1 ), the monitoring device switches ( UV ) in a further step ( S25 ) the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) out. It is preferably switched on again (in step S25 ) after the first shutdown (step S24 ) only after a charging time has passed ( T _L ). A security margin should be observed here. The monitoring device ( UV ) recorded in a subsequent step ( S26 ) the potential difference between the potential at the bootstrap node ( BST ) and the potential at the phase output ( PH ) the half bridge ( M _H , M _L ) again and determines an associated additional bootstrap potential difference value ( ΔV _BST ). In a further step ( S27 ) compares the monitoring device ( UV ) the further bootstrap potential difference value determined in this way ( ΔV _BST2 ) with another threshold value ( SW2 ), which is equal to the first threshold ( SW1 ) can be. The monitoring device switches in a further following step ( S28 ) the first power transistor ( M _H ) using the first gate driver ( GT _H ) after switching on again ( S25 ) again and the second power transistor ( M _L ) using the second gate driver ( GT _L ) also if the comparison step ( S27 ) determined further bootstrap potential difference value ( ΔV _BST2 ) in terms of amount below the further threshold ( SW2 ) lies.

8th shows the procedure for a drain-source voltage ( V _DS ) and the reload time ( T _L ) controlled reloading of the bootstrap capacity ( C _B ) without distinguishing between a short circuit and a defect in the bootstrap capacity ( C _B ).

The method is shown as an example, as it is from a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ), a first gate driver ( GT _H ), a second gate driver ( GT _L ), with a power supply circuit ( SV ) with a voltage regulator output ( VG ), a diode ( D ), a bootstrap capacity ( C _B ), a positive one Supply voltage line ( U _S ), a negative supply voltage line ( GND ) and a monitoring device ( UV ) can be executed. The first power transistor ( M _H ) and the second power transistor ( M _L ) are supposed to go back to a half bridge ( M _H , M _L ) with a phase output ( PH ) between the positive supply voltage line ( U _S ) and the negative supply voltage line ( GND ) be connected. Reference is made here to the previous explanations.

In a first step ( S31 ) at a switch-on time ( t ₀ ) becomes the first power transistor ( M _H ) turned on and the second power transistor ( M _L ) switched off. The time t is on this switch-on time ( t ₀ ) based. First, the debounce time ( T _D ) waited so that the transient processes can be completed. In a subsequent step ( S32 ) detects the monitoring device ( UV ) the potential difference between the potential at the drain connection ( U _S ) of the first power transistor ( M _H ) and the potential at the source ( PH ) of the first power transistor ( M _H ) and at least temporarily determines an associated drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) when the first power transistor ( M _H ) is switched on, which is in step S31 happened. The monitoring device ( UV ) compares in the following step ( S33 ) the drain-source voltage value determined in this way ( V _DS ) the drain-source voltage ( U _DS ) with a detection threshold ( TH ) (a first threshold SW1 ). The monitoring device switches in a conditionally following step ( S34 ) the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) when the determined drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) above the detection threshold ( TH ) (the first threshold SW1 ) lies. This starts a reload attempt for the bootstrap capacity ( C _B ). After switching off (in step S34 ), because the comparison step ( S33 ) the determined U _DS potential difference value above the detection threshold ( TH ) (the first threshold SW1 ) lag (S33), switches the monitoring device after a while ( UV ) in a subsequent step ( S35 ) the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) out. The monitoring device ( UV ) recorded in a subsequent measuring step ( S36 ) the potential difference between the potential of the drain connection ( U _S ) of the first power transistor ( M _H ) and the potential at the source ( PH ) of the first power transistor ( M _H ) again and determines an associated further drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ). The monitoring device ( UV ) compares in a subsequent comparison step ( S37 ) the further drain-source voltage value determined in this way ( V _DS2 ) the drain-source voltage ( U _DS ) in terms of amount with a further detection threshold (a second threshold value SW2 ), which is equal to the detection threshold ( TH ) (the first threshold SW1 ) can be. The monitoring device ( UV ) switches in a further conditional step ( S38 ) the first power transistor ( M _H ) using the first gate driver ( GT _H ) again from (S38) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) now also from (S38) if the further drain-source voltage value determined in this way ( V _DS2 ) the drain-source voltage ( U _DS ) in the comparison step ( S37 ) above the further detection threshold (the second threshold value SW2 ) lies. There is then a short circuit or a defect in the bootstrap capacity ( C _B ) in front of the interrupt line ( INTN ) and / or the data bus ( DB ) in cooperation with specially set register information within the integrated circuit ( IC ) can then be signaled to a higher-level computer system and communicated.

If the further drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ) in the comparison step ( S37 ) in terms of amount below the further detection threshold (the second threshold value SW2 ) is the bootstrap capacitor ( C _B ) sufficiently charged and there is then no short circuit. In a subsequent step ( S29 ) the first power transistor ( M _H ) then remain safely switched on and the second line transistor ( M _L ) remains switched off. It is then preferred until the end of the PWM period ( T _PWM ) until the next measurement ( S32 ) waited. In the event that at step S31 no change in the switching states of the first power transistor ( M _H ) and the second power transistor ( M _L ), the debounce time ( T _D ) should preferably be chosen to be 0s, since no transient processes are necessary.

• Schritt S42: Erfassen der Potenzialdifferenz zwischen dem ersten Anschluss und dem zweiten Anschluss der Bootstrap-Kapazität (C_B ) und Ermitteln (S22, S42) eines zugehörigen Bootstrap-Potenzialdifferenzwerts (ΔV_BST );
• Schritt S43: Vergleich des so ermittelten Bootstrap-Potenzialdifferenzwerts (ΔV_BST ) betragsmäßig mit einem ersten Schwellwert (SW1);
• Schritt S44: Abschalten des ersten Leistungstransistors (M_H ) und erstes Einschalten (S24, S44) des zweiten Leistungstransistors (M_L ), wenn der Vergleichsschritt (S43) ergab, dass der so ermittelte Bootstrap-Potenzialdifferenzwert (ΔV_BST ) betragsmäßig unterhalb des ersten Schwellwerts (SW1) lag. Hierdurch wird das Nachladen der Bootstrap-Kapazität (C_B ) gestartet.
• Schritt S45: Wiedereinschalten des ersten Leistungstransistors (M_H ) nach dem, ersten Abschalten in Schritt S44 aufgrund einer betragsmäßigen Unterschreitung des ersten Schwellwerts (SW1) durch den ermittelten Bootstrap-Potenzialdifferenzwert (ΔV_BST ), und erneutes Ausschalten des zweiten Leistungstransistors (M_L ), insbesondere nach einer Ladezeit (T_L );
• Schritt S46: Erneutes Erfassen (S46) der Potenzialdifferenz zwischen dem ersten Anschluss und dem zweiten Anschluss der Bootstrap-Kapazität (C_B ) und Ermitteln (S46) eines zugehörigen weiteren Bootstrap-Potenzialdifferenzwerts (ΔV_BST2 );
• Schritt S47: Vergleich (S47) des so ermittelten weiteren Bootstrap-Potenzialdifferenzwerts (ΔV_BST2 ) betragsmäßig mit einem weiteren Schwellwert (SW2), der gleich dem ersten Schwellwert (SW1) sein kann. Im Gegensatz zu 7 folgt nun aber mit Schritt S44 ein erneutes Ausschalten (S44) des ersten Leistungstransistors (M_H ) und ein erneutes Einschalten (S44) des zweiten Leistungstransistors (M_L ), wenn der so ermittelte weitere Bootstrap-Potenzialdifferenzwert (ΔV_BST ) betragsmäßig wieder unterhalb eines zweiten Schwellwerts (SW2) liegt, der gleich dem ersten Schwellwert (SW1) sein kann. Es wird hier also so lange Nachgeladen, bis die Bootstrap-Spannung (V_BST ) dem zweiten Schwellwert (SW2) entspricht. Dies ist dann sinnvoll, wenn Querströme in der Vorrichtung anders verhindert werden können. Wie zuvor auch, ist es besonders vorteilhaft, wenn das erneute Einschalten in Schritt S45 nach dem ersten Abschalten in Schritt S44 erst nach dem Vergehen einer Ladezeit (T_L ) erfolgt.

9 shows the procedure for a bootstrap voltage ( V _BST ) and the reload time ( T _L ) controlled reloading of the bootstrap capacity ( C _B ) without termination condition. The figure illustrates an exemplary method for operating a driver stage with a first power transistor ( M _H ), a second power transistor ( M _L ) and a bootstrap capacity ( C _B ) with a first connection and a second connection. The bootstrap capacity ( C _B ) is charged when the second power transistor ( M _L ) is switched on. The process includes the steps:

• step S42 : Detecting the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) and determine ( S22 , S42 ) of an associated bootstrap potential difference value ( ΔV _BST );
• step S43 : Comparison of the bootstrap potential difference value determined in this way ( ΔV _BST ) in terms of amount with a first threshold value ( SW1 );
• step S44 : Switching off the first power transistor ( M _H ) and first switch on ( S24 , S44 ) of the second power transistor ( M _L ) if the comparison step ( S43 ) showed that the bootstrap potential difference value ( ΔV _BST ) below the first threshold ( SW1 ) was. This will reload the bootstrap capacity ( C _B ) started.
• step S45 : Switching on the first power transistor ( M _H ) after the first shutdown in step S44 due to the amount falling below the first threshold ( SW1 ) by the determined bootstrap potential difference value ( ΔV _BST ), and the second power transistor is switched off again ( M _L ), especially after a charging time ( T _L );
• step S46 : Recapture ( S46 ) the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) and determine ( S46 ) of an associated additional bootstrap potential difference value ( ΔV _BST2 );
• step S47 : Comparison ( S47 ) of the further bootstrap potential difference value determined in this way ( ΔV _BST2 ) with another threshold value ( SW2 ), which is equal to the first threshold ( SW1 ) can be. In contrast to 7 but now follows with step S44 switching off again ( S44 ) of the first power transistor ( M _H ) and switching on again ( S44 ) of the second power transistor ( M _L ) if the further bootstrap potential difference value determined in this way ( ΔV _BST ) again below a second threshold in terms of amount ( SW2 ) which is equal to the first threshold ( SW1 ) can be. So it is reloaded until the bootstrap voltage ( V _BST ) the second threshold ( SW2 ) corresponds. This is useful if cross currents in the device can be prevented in a different way. As before, it is particularly advantageous if switching on again in step S45 after the first shutdown in step S44 only after a charging time has passed ( T _L ) he follows.

Reference list

BST

Bootstrap entrance;

C _VG

external support capacity to stabilize the constant voltage ( V _VG ) at the voltage regulator output ( VG ) of the power supply circuit ( SV );

D

Diode;

DB

Data bus. An external computer can access the internal register of the integrated circuit (e.g. IC ) and / or the monitoring device ( UV ) and / or the PWM generator ( PWMG ) access. The data bus interface can be used, for example, to signal a short circuit and / or to discharge or sufficiently charge the bootstrap capacity ( C _B ) be used;

ΔV _BST

Bootstrap potential difference value. The bootstrap potential difference value is determined by detecting the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) determined. As a rule, it represents the voltage value between these connections. It is shown in the steps in the examples of the figures S22 and S42 determined;

ΔV _BST2

Another bootstrap potential difference value. The further bootstrap potential difference value is determined by detecting the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) determined. As a rule, it represents the voltage value between these connections. It is shown in the steps in the examples of the figures S26 and S46 determined;

EN _H

first enable signal. The first enable signal is sent by the monitoring device ( UV ) used to connect the first gate driver ( GT _H ) to signal that the first gate control signal ( GH ) is placed in such a state that the first power transistor ( M _H ) turns off, so locks;

EN _L

second enable signal. The second enable signal is used by the monitoring device ( UV ) used to connect the second gate driver ( GT _L ) to signal that the second gate control signal ( GL ) is to be placed in such a state that the second power transistor ( M _L ) turns off, so locks;

GND

negative supply voltage line. The potential of the negative supply voltage line is in the examples listed here the reference potential, unless stated otherwise;

GH

first gate control signal. The first gate control signal is generated by the first gate driver ( GT _H ) depending on the first PWM control signal ( PWMH ) for driving the control electrode of the first power transistor ( M _H ) generated. The first gate driver ( GT _H ) either from the bootstrap capacity ( C _B ) or directly or indirectly supplied with electrical energy from the positive supply voltage line (Us). Is the potential of the positive supply voltage line ( U _S ) compared to the reference potential ( GND ) too low, the first gate driver is supplied ( GT _H ) from the bootstrap capacity ( C _B ), which is why the bootstrap capacity must always be sufficiently charged. The monitoring device ( UV ) can turn off the first power transistor ( M _H ) via the first gate control signal ( GH ) by forcing the monitoring device ( UV ) using the first enable signal ( EN _H ) the first gate driver ( GT _H ) signals to put the first gate control signal into such a state that the first power transistor ( M _H ) turns off, so locks;

GL

second gate control signal. The second gate control signal is generated by the second gate driver ( GT _L ) depending on the second PWM control signal ( PWML ) for driving the control electrode of the second power transistor ( M _L ) generated. The second gate driver ( GT _L ) typically only directly or indirectly supplied with electrical energy from the positive supply voltage line (Us). The monitoring device ( UV ) can turn off the second power transistor ( M _L ) via the second gate control signal ( GL ) by forcing the

GND

Monitoring device ( UV ) by means of the second enable signal ( EN _L ) the second gate driver ( GT _L ) signals to put the first gate control signal into such a state that the second power transistor ( M _L ) turns off, so locks; negative supply voltage line, the potential of which is typically the reference potential;

GT _H

first gate driver. The first gate driver generates the first gate control signal ( GH ) for driving the control electrode of the first power transistor ( M _H );

GT _L

second gate driver. The second gate driver generates the second gate control signal ( GL ) for driving the control electrode of the second power transistor ( M _L ). The second gate driver is preferably used by the voltage supply circuit ( SV ) supplied with electrical energy;

IC

integrated circuit;

INTN

Interrupt line of the integrated circuit ( IC ) to the typically available processor. For example, the interrupt signal of the interrupt line can be monitored by the monitoring device ( UV ) be generated. A is preferred INTN Interrupt line signal as insufficient charge of bootstrap capacity ( C _B ) interpreted, since in this case there is little time for countermeasures. In the event of a short circuit, in addition to the INTN signal, the interrupt line is preferably used via the data bus ( DB ) signals the corresponding error, which is typically not a bootstrap capacity loading error;

M _H

first power transistor;

M _H , M _L

Half bridge. The half bridge is through the first power transistor ( M _H ) and the second power transistor ( M _L ) educated;

M _L

second power transistor;

PH

Phase output;

PWMH

first PWM control signal. The first PWM control signal from the PWM generator is preferred ( PWMG ) with a PWM period;

PWML

second PWM control signal. The second PWM control signal is preferably from the PWM generator ( PWMG ) with a PWM period;

S1

first process step ( 5 and 6 ). In the first step, the first power transistor ( M _H ) at a switch-on time ( t ₀ ) turned on and the second power transistor ( M _L ) at this switch-on time ( t ₀ ) switched off. The switch-on time is preferably equal to the beginning of a PWM period. The switch-on time is therefore preferably repeated with the PWM period. To avoid cross currents, a dead time can be inserted between switching off ML and switching on MH.

S2

second process step ( 5 and 6 ). In the second step, after the debounce time has elapsed ( T _D ) a drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) determined.

S3

third step. In the third method step, the determined drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) with a first threshold ( SW1 ) compared.

S4

fourth step. The fourth process step is only carried out if the comparison in the third process step showed that the determined drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) greater than the first threshold ( SW1 ) is. In that case the first power transistor ( M _H ) not fully switched on and the voltage drop across the first power transistor ( M _H ) is too big. In this case, a short circuit is assumed because the error is too fast = in the active time ( T _A ) (see also 3rd ) occurred. For this reason, both power transistors ( M _H , M _L ) switched off to safely interrupt a supply current.

S5

fifth step ( 5 and 6 ). In the fifth step, after the debounce time has elapsed ( T _D ) and the active time ( T _A ) another or second drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ) determined. If this value is below a second threshold value during the extended active time ( SW2 ) is the bootstrap capacity ( C _B ) sufficiently charged.

S6

sixth step. In the sixth method step, the determined further or second drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ) with a second threshold ( SW2 ) compared.

S7

seventh step. The seventh method step is only carried out if the comparison in the sixth method step showed that the determined further or second drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ) greater than the second threshold ( SW2 ) is. In that case the first power transistor ( M _H ) in the extended active time ( T _EA ) not fully switched on and the voltage drop across the first power transistor ( M _H ) is too big. In this case, however, a short circuit is not assumed because the error is too slow = not in the active time ( T _A ) (see also 3rd and / or step S4 ) occurred. For this reason, the first power transistor ( M _H ) is switched off and the second line transistor ( M _L ) turned on to increase the bootstrap capacity ( C _B ) reload.

S8

eighth step ( 5 ). In the eighth step, the first power transistor ( M _H ) after the charging time has passed ( T _L ) turned on and the second power transistor ( M _L ) after the charging time has passed ( T _L ) switched off. This will reload the bootstrap capacity ( C _B ) completed.

S9

ninth step ( 6 ). In the ninth step of the process, reloading is also carried out (see also S7 ) a third drain-source voltage value ( V _DS3 ) the drain-source voltage ( U _DS ) determined.

S10

tenth procedural step. In the tenth process step, the third drain-source voltage value ( V _DS3 ) the drain-source voltage ( U _DS ) with a third threshold ( SW3 ) compared.

S11

eleventh procedural step. In the eleventh process step, more than the reload time ( T _L ) the bootstrap capacity ( C _B ) has passed and yet the drain-source voltage value ( V _DS3 ) the drain-source voltage ( U _DS ) the third threshold ( SW3 ) not exceeded. This is then interpreted as an error. For example, this can be a latent short circuit and / or the bootstrap capacity could not be recharged for whatever reason. Therefore, this fault case is preferably treated like a short circuit, but a different signaling to an external computer is preferably carried out, so that this case differs from the case in the method step S4 can be differentiated with certainty. Analogous to the short circuit in the step S4 the first power transistor is thus switched off here ( M _H ) and at the same time switching off the second power transistor ( M _L ), because the comparison of the determined third U _DS voltage value ( V _DS3 ) in terms of amount with a third threshold ( SW3 ), which is equal to the first threshold ( SW1 ) and the third threshold ( SW3 ), the third U _DS voltage value determined ( V _DS3 ) is below the third threshold. As already mentioned, step occurs as in the event of a short circuit S4 signaling via the interrupt line ( INTN ) and an integrated circuit register ( IC ) or the monitoring device ( UV ) and / or the PWM generator ( PWMG ). As in step S4 prevent the PWM generator ( PWMG ) and / or the monitoring device ( UV ) switching on the second power transistor ( M _L ) by an external control computer until such a restart is permitted again by a special unlock command from the external control computer via a special register. The integrated circuit preferably signals ( IC ), especially the monitoring device ( UV ) and / or the PWM generator ( PWMG ) in this error case via a register and the data bus ( DB ) the external control computer has a different error code than in the case of short circuit detection in the step S4 ;

S21

step S21 : In the process step S21 the first power transistor ( M _H ) at a switch-on time ( t ₀ ) turned on and the second power transistor ( M _L ) at this switch-on time ( t ₀ ) switched off. The switch-on time is preferably equal to the beginning of a PWM period. The switch-on time is therefore preferably repeated with the PWM period. However, this step can also be carried out within a PWM period after the charging time ( T _L ) occur.

S22

step S22 : The monitoring device ( UV ) recorded in step S22 after step S21 the potential difference between the potential at the bootstrap node ( BST ) and the potential at the phase output ( PH ) the half bridge ( M _H , M _L ) and determines an associated bootstrap potential difference value ( ΔV _BST );

S23

step S23 : The monitoring device ( UV ) compares in step S23 the one in step S22 determined bootstrap potential difference value ( ΔV _BST ) in terms of amount with a first threshold value ( SW1 ):

S24

step 24th : The monitoring device ( UV ) switches in step S24 the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) a. step S24 is executed when the in step S22 determined bootstrap potential difference value ( ΔV _BST ) in terms of amount in the comparison step S23 below the first threshold ( SW1 ) lies;

S25

step S24 : Some time after a first switch off in step S24 , which was done because of the step 22 determined bootstrap potential difference value ( ΔV _BST ) during the comparison step S23 in terms of amount below the first threshold ( SW1 ), switches the Monitoring device ( UV ) in a further step S25 the first power transistor ( M _H ) using the first gate driver ( GT _H ) and the second power transistor ( M _L ) using the second gate driver ( GT _L ) out. Switching on again preferably takes place in step S25 after the first shutdown in step S24 only after a charging time has expired ( T _L ). A security margin should be observed here.

S26

The monitoring device ( UV ) recorded in one step S25 following step S26 the potential difference between the potential at the bootstrap node ( BST ) and the potential at the phase output ( PH ) the half bridge ( M _H , M _L ) and determines an associated additional bootstrap potential difference value ( ΔV _BST ).

S27

In step ( S27 ) compares the monitoring device ( UV ) the further bootstrap potential difference value determined in this way ( ΔV _BST2 ) with another threshold value ( SW2 ), which is equal to the first threshold ( SW1 ) can be.

S28

The monitoring device ( UV ) switches in step S28 the first power transistor ( M _H ) using the first gate driver ( GT _H ) after switching on again in step S25 again and the second power transistor ( M _L ) using the second gate driver ( GT _L ) also off if the in the previous comparison step S27 the in step S26 determined further bootstrap potential difference value ( ΔV _BST2 ) in terms of amount below the further threshold ( SW2 ) lies.

S31

step S31 : In the process step S31 the first power transistor ( M _H ) at a switch-on time ( t ₀ ) turned on and the second power transistor ( M _L ) at this switch-on time ( t ₀ ) switched off. The switch-on time is preferably equal to the beginning of a PWM period. The switch-on time is therefore preferably repeated with the PWM period.

S32

step S32 : Detection of the U _DS voltage at the first power transistor ( M _H ) when the first power transistor ( M _H ) switched on (see step S31 ) and determining an associated drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS );

S33

step S33 : Comparison of the drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) in terms of amount with a first threshold value ( SW1 );

S34

step S34 : First shutdown of the first power transistor ( M _H ) and switching on the second power transistor for the first time ( M _L ) if the comparison step S33 revealed that in step S32 determined drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ) above the first threshold ( SW1 ) lies.

S35

step S35 : Switching on the first power transistor ( M _H ) after a first shutdown in step S34 due to a previous exceeding of the first threshold value ( SW1 ) by the in step S32 determined drain-source voltage value ( V _DS ) the drain-source voltage ( U _DS ), and switching off the second power transistor again ( M _L ), this especially after a charging time ( T _L ) he follows;

S36

step S36 : Detection of the U _DS voltage at the first power transistor ( M _H ) and again determining an associated further drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS );

S37

step S37 : Comparison of the in step S36 determined further drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ) with another threshold value ( SW2 ), which is equal to the first threshold ( SW1 ) can be;

S38

step S38 : Switching off the first power transistor ( M _H ) and a further switching off of the second power transistor ( M _L ) occurs when the step 36 determined further drain-source voltage value ( V _DS2 ) the drain-source voltage ( U _DS ) according to the comparison step S37 above a second threshold ( SW2 ) which is equal to the first threshold ( SW1 ) can be.

S41

step S41 : In the process step S31 the first power transistor ( M _H ) at a switch-on time ( t ₀ ) turned on and the second power transistor ( M _L ) at this switch-on time ( t ₀ ) switched off. The switch-on time is preferably equal to the beginning of a PWM period. The switch-on time is therefore preferably repeated with the PWM period.

S42

step S42 : Detecting the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) and determining an associated bootstrap potential difference value ( ΔV _BST );

S43

step S43 : Comparison of the in step 42 determined bootstrap potential difference value ( ΔV _BST ) in terms of amount with a first threshold value ( SW1 );

S44

step S44 : Switching off the first power transistor ( M _H ) and switching on the second power transistor for the first time ( M _L ) if the comparison step S43 revealed that in step S42 determined bootstrap potential difference value ( ΔV _BST ) below the first threshold ( SW1 ) was. This will reload the bootstrap capacity ( C _B ) started.

S45:

step S45 : Switching on the first power transistor ( M _H ) after the first shutdown in step S44 due to the amount falling below the first threshold ( SW1 ) by the in step S42 determined bootstrap potential difference value ( ΔV _BST ), and the second power transistor is switched off again ( M _L ), especially after a charging time ( T _L ). It is particularly advantageous if switching on again in step S45 after the first shutdown in step S44 only after a charging time has passed ( T _L ) he follows.

S46

step S46 : Detecting the potential difference between the first connection and the second connection of the bootstrap capacity ( C _B ) and determining an associated additional bootstrap potential difference value ( ΔV _BST2 );

S47

step S47 : Comparison of the in step S46 determined further bootstrap potential difference value ( ΔV _BST2 ) with another threshold value ( SW2 ), which is equal to the first threshold ( SW1 ) can be. In contrast to step S37 but now follows with step S44 the first power transistor is switched off again ( M _H ) and the second power transistor is switched on again ( M _L ) if the in step S45 determined further bootstrap potential difference value ( ΔV _BST ) in terms of amount in

SV

Comparison step S46 again below a second threshold ( SW2 ) which is equal to the first threshold ( SW1 ) can be. So it is reloaded until the bootstrap voltage ( V _BST ) the second threshold ( SW2 ) corresponds. This is useful if cross currents in the device can be prevented in a different way. Power supply circuit. The power supply circuit ( SV ) preferably generates a constant voltage ( V _VG ) at their voltage regulator output ( VG ) against the reference potential ( GND );

SW1

first threshold;

SW2

second threshold;

SW3

third threshold;

ST

further control signal from the monitoring device to the PWM generator ( PWMG )

t

Time related to the switch-on time ( t ₀ ) within a PWM period ( T _PWM ).

t '

Time related to the time when the bootstrap capacity is reloaded ( C _B ) was started.

t ₀

Switch-on time. As a rule, the switch-on time ( t ₀ ) equal to the start of each PWM period of the PWM generator ( PWMG );

T _A

Active time;

T _D

Debounce time;

T _EA

extended active time;

TH

Detection threshold;

T _L

Charging time;

T _PWM

PWM period;

T _T1

first dead time;

T _T2

second dead time;

U _DS

Drain-source voltage at the first power transistor ( M _H ).

U _S

positive supply line ( U _S ) with a positive supply voltage compared to the potential of the negative supply voltage line ( GND );

UV

Monitoring device;

V _DS

Drain-source voltage value of the drain-source voltage ( U _DS ) on the first power transistor ( M _H );

V _DS2

further or second drain-source voltage value of the drain-source voltage ( U _DS ) on the first power transistor ( M _H ). The second drain-source voltage value is typically recorded in time after the detection of the drain-source voltage value ( V _DS ) detected;

V _DS3

third drain-source voltage value of the drain-source voltage ( U _DS ) on the first power transistor ( M _H ). The third drain-source voltage value is typically recorded in time after the drain-source voltage value ( V _DS ) and after the detection of the second drain-source voltage value ( V _DS3 ) detected;

VG

Voltage regulator output of the voltage supply circuit ( SV );

V _BST

Bootstrap tension;

V _GH

Voltage between the first gate drive signal ( VG ) and the negative supply voltage line ( GND );

V _PH

Phase voltage ( V _PH ) at the phase output ( PH ) against the reference potential ( GND );

V _VG

Constant voltage at the voltage regulator output ( VG ) of the power supply circuit ( SV ) against the reference potential ( GND );

Claims (2)

Driver stage - with a first power transistor (M _H ) and - with a second power transistor (M _L ) and - with a positive supply voltage line (U _S ) and - with a negative supply voltage line (GND) and - with a monitoring device (UV) and - whereby in particular the first power transistor (M _H ) can be a MOS transistor or an IGBT transistor or another corresponding electronic power switch and - in particular the second power transistor (M _L ) can be a MOS transistor or an IGBT transistor or another corresponding electronic one Circuit breaker can be and - the first power transistor (M _H ) and the second power transistor (M _L ) to a half bridge (M _H , M _L ) with a phase output (PH) between the positive supply voltage line (U _S ) and the negative supply voltage line ( GND) are connected, and - wherein the monitoring device (UV) the drain-source voltage (U _DS ) at the first power transistor (M _H ) in terms of amount and a drain-source voltage value (V _DS ) determined (S32) and - the monitoring device (UV) the drain-source voltage value (V _DS ) with a detection threshold (TH) ( the first threshold value SW1) in terms of amount (S33) and - the monitoring device (UV) initiating a first shutdown of the first power transistor (M _H ) when the absolute value of the detection threshold (TH) (of the first threshold value SW1) is exceeded by the drain-source voltage value (V _DS ), and wherein the monitoring device (UV) or the other control device in this case causes the second power transistor (M _L ) to be switched on and - after which the monitoring device (UV) or the other control device in this case, in particular after a charging time (T _L ), switches off the second power transistor ( M _L ) and the first power transistor (M _H ) is switched on again - the monitoring device (UV) again detecting the amount of the drain-source voltage (U _DS ) at the first power transistor (M _H ) and a further drain-source Voltage value (V _DS2 ) determined and - the monitoring device (UV) the further drain source e-voltage value (V _DS2 ) with a further detection threshold, which can be equal to the detection threshold (TH), compared in terms of amount and - the monitoring device (UV) when the amount of the further detection threshold is exceeded again by the further drain-source voltage value (V _DS2 ) causes a second shutdown of the first power transistor (M _H ). Driver stage according to the preceding claim - wherein the monitoring device (UV) also causes the second power transistor (M _L ) to be switched off when the amount of the further detection threshold is exceeded again by the further drain-source voltage value (V _DS2 ).

Images