Two-stage switch driver for avalanche protection

Abstract

The present disclosure relates to a two-stage switch driver for preventing avalanche breakdown. The provided technique is used to avoid avalanche breakdown voltage across a synchronous rectification (SR) switch located on the secondary side of an isolated switch mode power converter, the converter operating in a low power mode, such as a burst mode, during which the current consumed by the load of the power converter is negligible. This is achieved by using a two-stage switch driver, which is used to control a power switch on the primary side of the power converter. The two-stage switch driver is configured to cause a low current level to flow out to the control terminal (e.g., gate) of the power switch during burst mode operation. The low current reduces the rising edge slope of the voltage pulse on the primary and secondary sides of the power converter, limiting the peak value of the voltage ringing across the SR switch. By limiting the voltage in this way, the SR switch is prevented from entering avalanche breakdown.

Description

Technical Field

[0001] The present application relates to a switch-mode power converter in which a power switch is controlled by a driver having two strength levels, wherein the driver limits the ratio at which the power switch is turned on to prevent the voltage across a synchronous rectification (SR) switch from reaching an avalanche breakdown level. Background Art

[0002] Mobile consumer devices such as smart phones, tablet computers, and notebook computers continue to grow in functionality and popularity. Such devices are typically battery powered and require a charger for charging the device battery from a mains power source. The battery charger is preferably small and inexpensive. Battery chargers must be efficient, in part due to their small size, so that the heat generated inside them does not cause the housing temperature to be too high. Although a single charger consumes relatively little power, the large number of such chargers connected to the power grid can place a significant load on the grid. Therefore, regulators have imposed strict requirements on the power efficiency of battery chargers during active and standby operating modes to limit the energy wasted by such chargers. In order to meet these specifications and efficiency targets, modern battery chargers increasingly use isolated switch-mode power converters. Previous power converters relied on large converters to directly step down the alternating current (AC) main voltage before rectification, especially compared to the high switching speeds used by such converters, which make the transformer inside the converter relatively small and light.

[0003] Many modern battery chargers rely on the flyback converter topology and its variations for converting a high voltage of direct current (DC) that has been rectified from the AC mains voltage to a low voltage that is used to charge the battery. A power switch converts the high DC voltage into a high-frequency AC voltage that is applied to the primary side of a transformer. The AC voltage induced on the secondary side of the transformer must be rectified to provide the DC voltage required by the load of the converter, such as a battery being charged. Such rectification relies on current interrupters, such as diodes and / or synchronous rectification (SR) switches. Compared to diodes, SR switches have low power losses and low associated heating, so in most applications, including typical battery chargers, SR switches are preferred for rectification.

[0004] In addition to providing efficient power conversion when the power converter is in active use, such as battery charging, the switching power converter preferably consumes minimal power when no load is attached to it. For example, consumers often keep the battery charger connected to the main power supply even after the charged device has been removed and / or the battery is fully charged. In order to minimize power loss during such standby periods, power converters often employ burst mode operation. During burst mode operation, switching of the power switch is suspended for relatively long time intervals. During the idle interval, once the output voltage has been reduced to the minimum allowed standby mode voltage level, switching is initiated and the output voltage is raised to the maximum standby mode voltage desired during the burst interval. Note that even during standby mode, leakage current drains charge from the output capacitor, so occasional recharging of the capacitor is required, such as by burst mode operation.

[0005] Burst mode operation potentially creates the problem that the voltage across the secondary side rectifying device (such as the SR switch) reaches a higher level than during normal operation. During standby / burst mode operation, because the load (e.g., battery) consumes no (or negligible) current, the voltage ringing produced by the sharply rising voltage edge of each voltage pulse on the secondary side of the transformer is larger across the SR switch than during normal operating mode. The upper voltage excursion can exceed the maximum allowed voltage of the SR switch. During burst mode operation, this maximum voltage can be repeatedly exceeded and damage the rectifying device.

[0006] Circuits and methods are desired to limit the maximum voltage appearing across a rectifier circuit in a switch-mode power converter during burst mode operation. Such circuits and methods should require minimal additional circuit components and should be energy efficient. Summary of the invention

[0007] According to one aspect of the switch driver, the switch driver is configured to control a power switch in a switch-mode power converter to prevent excessive voltage levels across a synchronous rectification (SR) switch of the power converter. The switch driver includes a pull-down driver switch, two pull-up driver switches, a mode indication input, a switch control input, and a high-side control circuit. The pull-down driver switch injects a current from a control terminal of the power switch. The pull-up driver switch causes the current to flow out to the control terminal. The first of the pull-up switches has a first on-state resistance, and the second of the pull-up switches has a second on-state resistance, wherein the second on-state resistance is substantially higher than the first on-state resistance. The mode indication input indicates whether the power converter is operating in a burst mode. The switch control input controls the conductivity of the following components: the pull-down driver switch, the second (high resistance) pull-up driver switch, and (in conjunction with the mode indication input) the first pull-up driver switch. When burst mode operation is indicated at the mode indication input, the high-side control circuit disables the first (low resistance) pull-up driver switch.

[0008] According to a method performed within a switch-mode power converter, the method is used to control a power switch of the power converter to prevent excessive voltage levels across a synchronous rectification (SR) switch of the power converter. The power converter includes a switch driver configured as described above, and in particular includes a first pull-up driver switch and a second pull-up driver switch. The method includes detecting that a load of the power converter is consuming power below a low power threshold, such as the load is in a standby mode. In response to such detection, a burst mode of operation is entered, and the first (low resistance) pull-up driver switch is disabled.

[0009] According to an embodiment of an isolated power converter, the isolated power converter includes an input, an output, a transformer, a power stage, an SR switch, and a controller. The input is used to couple to an input power supply, and the output is used to couple to a load of the power converter. The transformer includes a primary winding and a secondary winding. The power stage is configured to switch the coupled power from the input to the primary winding. The power stage includes a power switch and a switch driver for controlling the power switch. The switch driver is configured as described above, in particular including a first pull-up driver switch and a second pull-up driver switch, wherein in response to the power converter entering a burst mode (standby) of operation, the first pull-up driver switch is disabled. The SR switch is configured to rectify the power provided through the secondary winding so that DC power is provided to the output, and the SR switch has an SR voltage across its terminals. The SR switch has an avalanche breakdown voltage level associated with the SR voltage. The controller is configured to control the transfer of power to the output by generating a switch control signal to drive the power stage. As described above, in response to detecting a low power state, the controller also generates a burst mode indication for disabling the first pull-up driver switch. The low power state is detected based on a sensed current and / or a sensed voltage of the load.

[0010] Those skilled in the art will recognize additional features and advantages upon reading the following detailed description, and upon viewing the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The elements of the drawings are not necessarily to scale relative to each other. The same reference numerals identify corresponding similar parts. The features of the various illustrated embodiments can be combined unless they exclude each other. The embodiments are depicted in the drawings and described in detail in the description below.

[0012] Figure 1 A schematic diagram showing an isolated switch mode power converter.

[0013] Figure 2 A schematic diagram of a two-stage switch driver is shown, which is used, for example, in Figure 1 isolated power converter.

[0014] Figure 3 FIG. 1 shows voltage waveforms corresponding to the voltage across the secondary-side synchronous rectification (SR) switches during burst mode operation for different power converter circuits, such as in FIG. Figure 1 SR switch in isolated power converter.

[0015] Figure 4 Shown are waveforms corresponding to energy dissipation during burst mode operation for different power converter circuits.

[0016] Figure 5A method of controlling a primary side power switch is shown such that excessive voltage levels across a secondary side SR switch are prevented by using a two-stage switch driver. DETAILED DESCRIPTION

[0017] The embodiments described herein provide techniques for switching power switches in switch-mode power converters. These techniques are mainly described in the context of isolated power converters, in which the power switch is arranged on the primary side of the power converter and the synchronous rectification (SR) switch is arranged on the secondary side of the power converter. These techniques use a two-stage switch driver so that the power switch can be converted to its conduction mode at different ratios. The switch driver includes: a low resistance driver switch; and a high resistance pull-up driver switch for causing current to flow out to the control terminal (e.g., gate) of the power switch. During the normal operating mode of the power converter, both pull-up drivers are used so that the control terminal of the power switch can be charged as quickly as possible. This causes fast switching conversion and low switching losses because the power switch spends little time in the triode region where its power consumption is the highest. Fast switching causes a steep rising edge of the voltage pulse across the primary winding of the transformer, which is located in the power converter. The synthetic pulse on the secondary winding of the transformer also has a steep rising edge. As also explained in detail below, only the high resistance pull-up driver switch is used during the burst mode of the power converter. This reduces the slope of the rising edge of the voltage pulse, and associated problems on the secondary side of the power converter.

[0018] The alternating current (AC) voltage provided by the secondary winding must be rectified in order to provide a direct current (DC) voltage to the load of the power converter. Such rectification is provided by a current interrupter, such as a diode or a synchronous rectification (SR) switch. An SR switch provides a voltage drop lower than that of a diode and has lower associated power losses. Therefore, in applications where power efficiency is very important, SR switches are preferred. The embodiments herein are primarily described in the context of a power converter using an SR switch for rectification, but for a power converter using a diode for rectification, the described two-stage switch driver has similar advantages.

[0019] The voltage pulse induced across the secondary winding causes a voltage pulse corresponding to the conductance terminal across the SR switch, such as the drain and source of a metal oxide semiconductor field effect transistor (MOSFET). The rising edge of each voltage pulse causes ringing in the voltage applied to the SR switch. During the normal operating mode of the power converter, the current injected by the load limits the extreme case of such voltage ringing. However, during the standby operating mode, the load does not inject any appreciable current, which means that if no mitigation techniques are used, the voltage across the SR switch oscillates to a level higher than during the normal operating mode. The SR switch (or diode) can only safely handle voltage levels below certain thresholds. Once the voltage across the SR switch reaches this level, the SR switch enters the avalanche breakdown region. The voltage across the SR switch is constrained (clamped) by the avalanche breakdown voltage of the switch, and once the SR voltage reaches this breakdown voltage, the current through the SR switch increases significantly.

[0020] SR switches can tolerate a limited number of avalanche breakdowns. However, an isolated power converter operating in burst mode (standby) can cause its SR switch to be repeatedly avalanche-broken. Specifications for SR switches, such as MOSFETs, typically provide an avalanche breakdown threshold and warn against exceeding this threshold. SR switches in avalanche breakdown cause excessive currents and associated excessive temperatures that damage the SR switch. In addition, repeated avalanche breakdowns of the MOSFET can cause charge to be trapped within the MOSFET, thereby changing the internal field distribution and ultimately causing the MOSFET to be unable to block voltage. Regardless of the specific mechanism, avalanche breakdown, especially repeated avalanche breakdown, changes the switching characteristics of the MOSFET, thereby damaging and / or destroying the MOSFET. Therefore, repeated avalanche breakdowns must be avoided.

[0021] One technique to avoid avalanche breakdown of the SR switch is to connect a snubber circuit in parallel with the SR switch, such as a resistor-capacitor-diode (RCD) snubber. The RCD snubber uses a resistor and a capacitor connected in series to dissipate the energy trapped in the leakage inductance of the transformer. By redirecting the current associated with this energy to flow away from the SR switch, the maximum voltage level that occurs across the SR switch when a voltage pulse is applied is reduced, i.e., the extent of voltage ringing is limited. The current charges the capacitor, thereby storing energy associated with the current directed away from the SR switch. The charge stored on the capacitor is returned to the circuit via the current flowing through the diode. In a typical topology, the energy stored on the capacitor is returned to the inductor on the secondary side, which can be the inductance of the secondary winding, so the energy is not lost.

[0022] Even if most of the energy stored by the RCD snubber is returned to the circuit, the RCD snubber still causes some power losses. Especially when the SR voltage is limited (i.e., current is flowing through the resistor and charging the capacitor), the voltage drop across the RCD resistor causes power losses, and when the energy from the RCD capacitor is returned to the circuit, the voltage drop across the RCD diode causes power losses. When the isolated power converter is being operated in its burst mode, these power losses are quite small. However, power losses of the RCD snubber can also be caused during the normal operating mode, and thus the overall power efficiency of the isolated power converter is degraded. For this reason, and because the RCD snubber requires additional circuit components, the RCD snubber does not provide an ideal solution for avoiding avalanche breakdown.

[0023] Rather than using an RCD snubber to limit the voltage across the SR switch, the circuit here reduces the slope of the rising edge of the voltage pulse across the SR switch. Such a reduced slope reduces the high frequency portion of the voltage ringing and results in a maximum voltage across the SR switch that is much lower than the maximum voltage that would otherwise be produced. The reduced slope is only needed during standby (low power) mode when the load does not sink appreciable current, such as when the isolated power converter is operated in burst mode.

[0024] By slowing down the rising edge of the corresponding voltage slope on the primary side, a reduced rising slope for the secondary side voltage pulse is generated. In turn, this is achieved by slowing down the transition of the primary side power switch from its off state to its on state. Taking a MOSFET power switch as an example, its gate is charged more slowly so that the MOSFET spends more time in its triode region, rather than quickly switching from its "off" (non-conducting) state to its saturation (fully on) state. When the power converter is in its burst mode of operation, the switch driver driving the gate of the MOSFET operates in a reduction-current mode to achieve slow charging of the gate. The two-stage switch driver includes two pull-up driver switches. Only the high resistance pull-up driver switch is activated during the burst mode of operation. During the normal operating mode, both pull-up driver switches are activated so that the power switch can be converted from its cut-off state to its saturation state as quickly as possible.

[0025] Embodiments will be described below primarily through specific examples of the following components and methods: an isolated switch mode power converter, a switch driver, and a method for controlling a power switch in an isolated switch mode power converter. These examples have a common feature in that they use a switch driver with two current outflow stages, where the current outflow stages are based on burst mode indications. The switch driver technology can also be applied to non-isolated power converters.

[0026] It should be understood that the examples below are not intended to be limiting. In order to avoid obscuring the unique aspects of the present invention, circuits and techniques known in the art are not described in detail. Unless mutually exclusive, features and aspects from the example embodiments may be combined or rearranged.

[0027] The description provided below uses a flyback converter topology and uses an SR switch on the ground side of the transformer secondary winding. It should be understood that the present invention can be used in variations of the flyback converter topology described herein, as well as in variations of other isolated switch mode power converter topologies. For example, a half-bridge, full-bridge, or LLC power stage can be used on the primary side. On the secondary side, the SR switch can be connected between the other terminals of the secondary winding and the output terminal of the power converter. Other rectification topologies can also be used, including, for example: a center-tapped secondary winding with an SR switch on each outer leg; a current doubler; or a full-wave rectifier bridge. In addition to using an SR switch, a diode-based topology can be used alternatively. Diode-based rectification is usually at the expense of increased power losses, but like SR switch-based rectification, it has the advantage of not requiring switch control. Diodes often have an avalanche breakdown mode, which should be avoided in the same manner as described above for MOSFETs.

[0028] Figure 1 An isolated switch mode power converter 100 is illustrated that utilizes a two-stage switch driver 160 that limits the voltage V across the synchronous rectification (SR) switch Q2 during burst mode operation. DS_Q2 The power converter 100 includes an AC input 102 , a DC input 104 , an output 106 , a rectifier 120 , a power factor correction (PFC) circuit 130 , a transformer 140 , a power stage 150 , and a controller 190 .

[0029] The AC input 102 is used to connect to an AC power source 110, which provides an AC voltage V AC_IN (AC power source 110 is not part of isolated power converter 100). Rectifier 120 converts AC voltage V AC_IN rectified, providing a DC voltage at its output. Optional power factor correction (PFC) circuits and protection circuits can provide a variety of functions including: correcting the power factor of the power converter by boosting the DC voltage provided by rectifier 120; providing protection against voltage spikes or other input anomalies; and filtering the input voltage. In some applications, rectifier 120 and PFC 130 can be switched in order or can be integrated into a combined circuit. PFC 130 will output a DC voltage V suitable for coupling to transformer 140. DC_INThe illustrated power converter 100 provides AC to DC conversion, but a DC to DC variation may be achieved by omitting the rectifier 120 and PFC 130 and coupling the DC input 104 to a DC power source, such as a battery or other power converter that provides DC power.

[0030] The transformer 140 includes a primary winding 142 having N1 turns, a secondary winding 144 having N2 turns, and a core 146. The transformer provides voltage step-down (or voltage step-up) according to a turns ratio N1:N2.

[0031] The secondary side of the isolated power converter 100 includes an output capacitor C O , SR switch Q2, and output 106. Output capacitor C O The voltage pulses provided by the secondary winding 144 and the SR switch Q2 are filtered, thereby providing a relatively stable voltage V at the output 106 OUT Output 106 is intended to be connected to a load 170 which is not actually part of power converter 100. SR switch Q2 switchably connects one side of secondary winding 144 to ground, thereby providing a voltage V OUT Rectification. Figure 1 The SR switch Q2 shown in FIG. 1 is an enhancement mode metal oxide semiconductor field effect transistor (MOSFET), but other types of switches may be used. For example, a crystalline field effect transistor (JFET), a bipolar junction transistor (BJT), an insulated gate bipolar transistor (IGBT), a high electron mobility transistor (HEMT), a gallium nitride (GaN) MOSFET, or other types of power transistors may be preferred in some applications. It should be noted that most transistor types have a maximum allowable voltage, such as an avalanche breakdown voltage.

[0032] The controller 190 and its components may be implemented using a combination of analog hardware components (such as transistors, amplifiers, diodes, and resistors), and a processor circuit that primarily includes digital components. The processor circuit may include one or more digital signal processors (DSPs), general-purpose processors, and application-specific integrated circuits (ASICs). The controller 190 may also include a memory, such as a non-volatile memory such as flash memory, that includes instructions or data for use by the processor circuit, and one or more timers. The controller 190 inputs a sensor signal, such as a signal corresponding to an output voltage V OUT and the sensed output current I OUT_SENSE signal.

[0033] The controller 190 is responsible for controlling the power converter 100 to provide the necessary power to the load 170. The controller 190 senses the output voltage V OUT and IOUT_SENSE The current is sensed based on the current through the SR switch Q2, such as the drain-to-source voltage V DS_Q2 The rectifier controller in the controller 190 generates a control signal CTL_Q2 based on the sensed secondary side voltage, or the timing from the pulse width modulation (PWM) generator 192. For ease of illustration, such a rectifier controller is not directly shown, and because such rectification techniques are well known in the art, further details about rectification control are not provided herein. The control signal CTL_Q2 is provided to the gate of the SR switch Q2 via the driver 180 to control the conduction of the SR switch Q2.

[0034] The controller 190 includes a PWM generator 192 and a low power detector 194. The PWM generator 192 generally includes a circuit for regulating the output voltage V OUT A linear feedback controller such as a proportional integral derivative (PID) controller is used. During normal operation mode of power converter 100, PWM generator 192 senses output voltage V OUT , and compares this voltage with the reference (target) voltage V TARGET , to determine the control parameters for generating the control signal CTL_Q1 for the power stage 150. For example, the PWM generator 192 may generate the control signal CTL_Q1 with a fixed switching frequency and a variable duty cycle, in which case the determined control parameter is the duty cycle. Alternatively, the PWM generator 192 may generate a control signal with a fixed pulse width and a variable frequency, in which case the control parameter is the switching frequency. These techniques and other techniques are well known in the field of feedback control and are not critical to understanding the unique aspects of the power switch control described herein. Therefore, further details regarding feedback control techniques are not provided.

[0035] Low power detector 194 determines whether power converter 100 should operate in normal mode or burst mode. OUT_SENSE With the low power current threshold, the low power detector 194 can detect that the load 170 is inactive. The output current I OUT_SENSE The current may be sensed using a shunt resistor, a current transformer, the DC resistance (DCR) of an inductor associated with the secondary winding 144, or other current sensing techniques. Because such current sensing techniques are well known in the art, they will not be described in further detail herein.

[0036] In addition to directly sensing the secondary current I OUT_SENSEFor example, the primary side current flowing through the primary winding 142 and the power switch Q1 can be sensed and used to sense that the load 170 is inactive. In some applications, sensing of the primary side current may be preferred, such as when the controller is located on the primary side of the power converter 190 rather than on the secondary side, and / or when current mode control is used for the power switch Q1. The sensing of the primary side current can be performed using similar techniques as described above for measuring the output current I OUT_SENSE For example, measuring the voltage across the shunt resistor connecting the source of the power transistor Q2 to the ground, measuring the source-drain voltage of the power transistor Q2, etc. Figure 1 The output current I is sensed by measuring the primary side current OUT_SENSE .

[0037] Via the burst mode indicator BRST, the low power detector 194 indicates that the power converter 100 should enter a burst mode of operation in response to detecting that the load 170 is inactive, such as in a low power state. Such inactivity may be detected, for example, by detecting the output current I OUT_SENSE The output voltage is not precisely regulated to match the target voltage V TARGET , but the PWM generator 192 generally uses a loose output voltage limit during the burst mode. For example, the PWM generator 192 prohibits the generation of the control signal CTL_Q1 until the output voltage drops to the burst mode lower limit V LOW_STBY Until then, the burst mode lower limit is lower than the normal mode target voltage V TARGET . For the target voltage V TARGET =20V, the standby voltage can be allowed to drop to V LOW_STBY =19.5V. Then, the PWM generator 192 allows the generation of the switch control signal CTL_Q1 until the output voltage V OUT Reach the target voltage V TARGET =20V or upper limit voltage V HIGH_STBY , the upper limit voltage is slightly higher than the target voltage V TARGET =20 V, for example 20.5 V. The net effect is that during burst mode, the PWM generator 192 provides bursts of pulses on the control signal CTL_Q1 separated by longer inactive intervals.

[0038] Although the above description is made in the context of a flyback converter 100 having a power stage 150 with a power switch, it should be understood that isolated power converters having other topologies may have additional power switches and that the PWM generator 192 will generate control signals corresponding to such additional power switches.

[0039] The controller 190 may be located on the primary side or the secondary side of the power converter 100. In order to maintain the integrity of the primary to secondary boundary, signals input or output from the controller 190 may need to be passed through isolators, which are not shown for ease of illustration. For example, a controller 190 located on the secondary side will require output signals CTL_Q1, BRST to pass through isolators, which are provided to the (primary side) power stage 150. In turn, the primary side controller will require other input and output signals to the controller 190 to be passed through isolators.

[0040] The power stage 150 includes a power switch Q1 and a two-stage switch driver 160. The power switch Q1 switchably connects one terminal of the primary winding 142, corresponding to the illustrated node VSW, to ground so as to generate a voltage pulse across the primary winding 142. The illustrated power switch Q1 is a MOSFET having gate, drain and source terminals. In some applications, other switch types may preferably be crystalline field effect transistors (JFETs), bipolar junction transistors (BJTs), insulated gate bipolar transistors (IGBTs), gallium nitride (GaN) MOSFETs, or high electron mobility transistors (HEMTs). According to the burst mode indication BRST input to the power stage 150, the gate of the power switch Q1 is charged at different rates by the two-stage switch driver 160. When the burst mode operation is indicated, the switch driver 160 disables at least some of its source currents so as to charge the gate of the power switch Q1 more slowly.

[0041] Figure 2A specific embodiment for the power stage 250 is illustrated in further detail, including a two-stage switch driver 250. The switch driver 260 provides a gate signal GS_Q1 to the gate (G) of the power switch Q1 by causing current to flow out or sink into the gate via a push-pull arrangement of the driver switch. The switch driver 260 includes pull-up driver switches MPa, MPb and a pull-down driver switch MN1. The pre-drivers 264, 266, 268 provide gate signals GHa, GHb, GL for each driver switch MPa, MPb, MN1. The pull-down driver switch MN1 is an n-channel MOSFET, meaning that the switch is set to turn on when a high voltage is provided to its gate. The pull-up driver switches MPa, MPb are p-channel MOSFETs, meaning that they are set to turn on when a low voltage (e.g., 0V) is provided to their respective gates. The first pull-up driver switch MPa has a low on-state resistance R DS_ON , while the second pull-up driver switch MPb has a high on-state resistance.

[0042] The first pull-up driver switch MPa is controlled by a gate signal GHa output from the high-side control circuit 262. When a burst mode operation is indicated, the high-side control circuit 262 effectively disables the first pull-up driver switch MPa. The high-side control circuit 262 includes an OR gate 263 and a first high-side pre-driver 264. During normal operation, the burst mode indication BRST is low, and the OR gate 263 passes the control signal CTL_Q2 to the first high-side pre-driver 264. In this state, according to the control signal CTL_Q2, the first pull-up driver switch and the second pull-up driver switch are switched simultaneously. More specifically, the high voltage (VDD) for the control signal CTL_Q2 is passed to the gate terminals of the pull-up driver switches MPa and MPb so that their gate-to-source voltage V GS_MPa , V GS_MPb is zero, and the switch is turned off. Conversely, the low voltage (0V) for the control signal CTL_Q2 is delivered to the gate terminals of the pull-up driver switches MPa and MPb so that their gate-source voltages V GS_MPa , V GS_MPb is (-VDD), and the switch is turned on.

[0043] During the burst mode operation, the burst mode indicator BRST is high, and the output of the OR gate 263 always remains high regardless of the state of the control signal CTL_Q2. The first pull-up driver switch MPa remains off because its gate-source voltage V GS_MPa =0 V. The second pull-up driver switch MPb is controlled by the control signal CTL_Q2 and is independent of the burst mode indication BRST.

[0044] The low-side driver switch MN1 is turned on when the control signal CTL_Q2 is high (logic "1"), and turned off when the signal is low (0 V). When turned on, the low-side driver switch MN1 sinks current from the gate of the power switch Q1, turning the power switch Q1 off.

[0045] The first pull-up driver switch MPa has a low on-state resistance, such as R MPa_DS_ON =5Ω); however, the second pull-up driver switch MPb has a relatively high on-state resistance, such as R MPa_DS_ON =600Ω. The current source capability of the pull-up driver stage is determined by its on-state resistance. During burst mode, only the second (high resistance) pull-up driver switch is enabled, meaning that the current source capability is greatly reduced relative to normal operation. With a lower source current supplied to the gate of the power switch Q1, the power switch Q1 will turn on more slowly and the voltage pulse generated across the primary winding 142 has a low slew rate.

[0046] The on-state resistance of the switches including the pull-up driver switches MPa and MPb is generally determined by their dimensions, such as channel length and width. In one embodiment, the channel widths of the first and second pull-up driver switches MPa and MPb can be set to achieve a desired ratio of on-state resistances, or similarly a ratio of the absolute desired on-state resistance of the second pull-up driver switch MPb, which results in a voltage across the secondary side SR switch, such as Figure 1The desired voltage slew rate of Q2 in the pull-up driver switch MPa and MPb. In another embodiment, the pull-up driver switches MPa and MPb can be parts of the same device, but they have isolated control terminals. For example, a MOSFET can have multiple source and drain regions connected in parallel, where gate fingers are provided to control each source / drain region. Multiple gate fingers can be tied together to effectively generate a large MOSFET with a lower on-state resistance than a single source / drain portion of the MOSFET. The gate fingers can be divided into a group assigned to the first pull-up driver switch MPa and another group assigned to the second pull-up driver switch MPb. For example, a MOSFET with 120 gate fingers is characterized in that each drain / source pair can have an effective on-state resistance of 600Ω. One such finger can be assigned to control a drain / source pair, which constitutes the second (high resistance) pull-up driver switch MPb. The remaining 119 fingers can be assigned to control the remaining drain / source pairs, which constitute the first (low resistance) pull-up driver switch MPa. The allocation of gate fingers between the first and second pull-up driver switches MPa and Mpb can be achieved during manufacturing via a metal mask, or can otherwise be programmable, for example using a digital code to allocate the MSOFET segments between the first and second pull-up driver switches.

[0047] The pre-drivers 264 and 266 of the second pull-up driver switches MPa and MPb must be sized according to the first and second pull-up driver switches MPa and MPb, such as the number of allocated segments. In addition, the overall pull-up driver strength is characterized by the combination of the on-resistances of the parallel pull-up driver switches MPa, MPb, and the size of the pull-up driver strength must be set to provide the source current required by the power switch Q1 during normal operation of the power converter 100.

[0048] Figure 3 The source-drain voltage V corresponding to the SR switch Q2 is shown for different power converter circuits. DS_Q2 The waveform 300 is shown. By using a primary side power switch, such as Figure 1 These waveforms 300 are initiated by a voltage pulse applied across the primary winding by Q1 in the embodiment of the present invention, which induces a secondary side voltage pulse, resulting in the shown source-drain voltage waveforms 300. These waveforms are generated during a low power (standby) mode, during which there is negligible current consumed by the corresponding power converter.

[0049] The first waveform 310 corresponds to a first power converter where the steep rising edge of the voltage pulse is not mitigated. The voltage attempts to ring to a high voltage, but the SR switch clamps 312 the voltage at an avalanche breakdown voltage of approximately 165 V. During the avalanche breakdown period, the SR switch conducts excessive current levels, and the SR switch is potentially damaged, especially when avalanche breakdown occurs repeatedly.

[0050] The second waveform 320 corresponds to a second power converter in which a snubber circuit, such as the RCD snubber described above, limits the maximum voltage across the SR switch. The peak of the second waveform only reaches about 158V, which means that avalanche breakdown of the SR switch is avoided.

[0051] The third waveform 330 corresponds to a third power converter, including as in Figure 2 . The source-drain voltage of the SR switch rises more slowly than either the first waveform 310 or the second waveform 320. Avalanche breakdown of the SR switch is avoided without the need for an RCD circuit and its components. In addition, the maximum voltage of about 125V across the SR switch is significantly lower than a power converter without protection or with an RCD snubber.

[0052] Figure 4 A waveform 400 is shown corresponding to a Figure 3 4 and 5. The energy consumed by the power converter circuit in the depicted embodiment is obtained by examining the energy consumed by the power converter circuit from the input power supply. The first energy waveform 410 corresponds to the energy consumed by the first power converter, where the rising edge of the voltage pulse is not mitigated. The second energy waveform 420 corresponds to the energy consumed by the second power converter, where a snubber circuit is used to avoid avalanche breakdown in the SR switch. Note that due to the energy consumed by the snubber, more energy is consumed for this power converter. The third waveform 430 corresponds to the energy consumed for the third power converter, where a two-stage switch driver, such as Figure 2 The switch driver 260 is used to switch the primary side power stage. It can be seen that the energy consumption generated by the third power converter using the two-stage switch driver is much lower than that of the other illustrated alternatives.

[0053] Figure 5 A method 500 is illustrated for controlling a primary side power switch in an isolated power converter to prevent excessive voltage across the secondary side of the SR switch. The method 500 may be used in an isolated power converter 500, such as in Figure 1 The method 500 begins by sensing 510 the power consumed by the load of the power converter. LOAD , and the power P LOAD With low power threshold P LP_THRESHA comparison is made 520. (In substantially the same technique, the load current, e.g. Figure 1 I OUT_SENSE can be sensed, and with the low power current threshold I LP_THRESH Once the power consumption P LOAD Drop to the low power threshold P LP_THRESH Following this, the power converter enters 530 a burst mode of operation and disables the first (low resistance) pull-up driver switch. LOAD As indicated, the burst mode operation continues until the power consumed by the load P LOAD Exceeding the low power threshold P LP_THRESH Once the load starts to consume power above the threshold P LP_THRESH Above 560, burst mode operation is exited 570, and normal operation continues. The first (low resistance) pull-up driver switch is enabled 580.

[0054] Two-stage switch drivers and control of power switches are described in the context of isolated switch-mode power converters described above. These techniques and circuits can also be applied to power converter topologies that do not provide isolation. For example, a non-isolated buck-boost power converter uses an inductor to temporarily store energy, rather than using a transformer to transfer energy across an isolation barrier. In a typical buck-boost converter, a power switch switchably connects an input power source to an energy storage inductor, thereby generating an AC voltage across the inductor. The AC voltage is rectified, such as using an SR switch, and the resulting DC voltage is provided to a load of the power converter. As in the isolated power converter described above, during periods when the load is inactive and current consumption is not significant, the SR switches in such buck-boost converters can withstand excessive voltages, including avalanche breakdown voltage limits. The foregoing description and Figure 2 The two-stage switch driver illustrated in FIG. 1 may similarly be used in such a non-isolated buck-boost power converter. More particularly, the two-stage switch driver may drive a power switch that couples an input power source to an energy storage inductor. As previously described, during periods when the load is consuming low power, and particularly during periods when such conditions induce burst mode operation, the two-stage driver is positioned in a state where only its high resistance pull-up driver is enabled. This may be accomplished by providing a burst mode indication, such as a Figure 2 By using only high resistance pull-up drivers during Burst Mode operation, the voltage pulses supplied to the energy storage inductor have lower rising edges and therefore the extent of any voltage ringing across the SR switch is reduced.

[0055] According to an embodiment of a switch driver, the switch driver is configured to control a power switch located in a switch-mode power converter to prevent excessive voltage levels across a synchronous rectification (SR) switch of the power converter. The switch driver includes a pull-down driver switch, two pull-up driver switches, a mode indication input, a switch control input, and a high-side control circuit. The pull-down driver switch injects a current from a control terminal of the power switch. The pull-up driver switch causes the current to flow out to the control terminal. The first of the pull-up switches has a first on-state resistance, and the second of the pull-up switches has a second on-state resistance, wherein the second on-state resistance is substantially higher than the first on-state resistance. The mode indication input indicates whether the power converter is operating in a burst mode. The switch control input controls the conductivity of the following components: the pull-down driver switch, the second (high resistance) pull-up driver switch, and the first pull-up driver switch in combination with the mode indication input. When burst mode operation is indicated at the mode indication input, the high-side control circuit disables the first (low resistance) pull-up driver switch.

[0056] According to any embodiment of the switch driver, the pull-down driver switch is an n-channel metal oxide semiconductor field effect transistor (MOSFET), and the first and second pull-up driver switches are p-channel MOSFETs.

[0057] According to any embodiment of the switch driver, the high-side control circuit includes a high-side pre-driver and an inclusive OR gate. The high-side pre-driver is coupled to the first pull-up driver switch and controls its conductivity. The inclusive OR gate inputs a mode indication and a switch control signal, and outputs a high-side switch control signal to the high-side pre-driver.

[0058] According to any embodiment of the switch driver, the second pull-up driver switch is sized to source a current level that limits the voltage across the SR switch to below an upper voltage limit of the SR switch corresponding to an avalanche breakdown voltage of the SR switch.

[0059] According to any embodiment of the switch driver, the first pull up driver switch, when turned on, draws a current level that is at least 100 times greater than a current level drawn by the second pull up driver switch when the second pull up driver switch is turned on.

[0060] According to any of the embodiments of the switch driver, a size ratio between the first pull-up driver switch and the second pull-up driver switch may be programmable by a configurable digital code.

[0061] According to any embodiment of the switch driver, a size ratio between the first pull-up driver switch and the second pull-up driver switch is set by metal selection during fabrication of the pull-up driver switches.

[0062] According to a method performed in a switch mode power converter, the method is used to control a power switch in a switch mode power converter to prevent excessive voltage levels across a synchronous rectification (SR) switch of the power converter. The power converter includes a switch driver configured as described above, particularly including a first pull-up driver switch and a second pull-up driver switch. The method includes detecting that a load of the power converter is consuming power below a low power threshold, such as when the load is in a standby mode. In response to such detection, a burst mode of operation is entered and the first pull-up driver switch is disabled.

[0063] In accordance with any embodiments of the method, the SR switch has an associated avalanche breakdown voltage corresponding to a voltage limit across a first terminal and a second terminal of the SR switch, and the on-state resistance of the second pull-up driver switch is configured to support a current level through the second pull-up driver switch such that the ratio at which the power switch is turned on ensures that the voltage across the first terminal and the second terminal of the SR switch remains below the avalanche breakdown voltage of the SR switch.

[0064] According to any embodiment of the method, the method further comprises, in response to detecting that a load of the power converter is consuming more power than a low power level, exiting the burst mode of operation and activating the first pull-up driver switch.

[0065] According to an embodiment of an isolated power converter, the isolated power converter includes an input, an output, a transformer, a power stage, an SR switch, and a controller. The input is used to couple to an input power supply, and the output is used to couple to a load of the power converter. The transformer includes a primary winding and a secondary winding. The power stage is configured to switch power from the input to the primary winding. The power stage includes a power switch and a switch driver for controlling the power switch. The switch driver is configured as described above, and in particular includes a first pull-up driver switch and a second pull-up driver switch, wherein in response to the power converter entering a burst mode (standby) of operation, the first pull-up driver switch is disabled. The SR switch is configured to rectify the power provided by the secondary winding, and the SR switch has an SR voltage across its terminals. The SR switch has an avalanche breakdown voltage level associated with the SR voltage. The controller is configured to control the power delivered to the output to drive the power stage by generating a switch control signal. In response to detecting a low power state, the controller also generates a burst mode indication, which is used to disable the first pull-up driver switch. The low power state is detected based on a sensed current or a sensed voltage of the load.

[0066] In accordance with any of the embodiments of the isolated power converter, the pull-down driver switch is an n-channel metal oxide semiconductor field effect transistor (MOSFET), and the first and second pull-up driver switches are p-channel MOSFETs.

[0067] According to any embodiment of the isolated power converter, the high-side control circuit includes a high-side pre-driver and an OR gate. The high-side pre-driver is coupled to the first pull-up driver switch and controls its conductivity. The OR gate inputs a mode indication and a switch control signal, and outputs a high-side switch control signal to the high-side pre-driver.

[0068] In accordance with any embodiment of the isolated power converter, the second pull-up driver switch is sized to source a current level that limits the voltage across the SR switch on the secondary side to below an upper voltage limit of the SR switch corresponding to an avalanche breakdown voltage of the SR switch.

[0069] According to any embodiment of the isolated power converter, the first pull-up driver switch, when turned on, sources a current level that is at least 100 times greater than a current level that is sources by the second pull-up driver switch when the second pull-up driver switch is turned on.

[0070] According to any of the embodiments of the isolated power converter, a size ratio between the first pull-up driver switch and the second pull-up driver switch may be programmable by a configurable digital code.

[0071] According to any embodiment of the isolated power converter, a size ratio between the first pull-up driver switch and the second pull-up driver switch is set by metal selection during fabrication of the pull-up driver switches.

[0072] According to any embodiment of the isolated power converter, the isolated power converter has a flyback converter topology.

[0073] As used herein, the terms "having", "comprising", "including", etc. are open terms that indicate the presence of stated elements or features, but do not exclude additional elements or features. Unless the context clearly indicates otherwise, the articles "a", "an", and "the" are intended to include the plural as well as the singular.

[0074] It should be understood that the features of the various embodiments described herein may be combined with each other, unless specifically stated otherwise.

[0075] Although specific embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that various alternatives and / or equivalent embodiments may replace the specific embodiments shown and described without departing from the scope of the invention. The present application is intended to cover any modifications or variations of the specific embodiments discussed herein. Therefore, the present invention is intended to be limited only by the claims and their equivalents.

Claims

1. A switch driver for controlling a power switch in a switch-mode power converter to prevent excessive voltage levels across a synchronous rectification (SR) switch of the power converter, the switch driver comprising: a pull-down driver switch for injecting current from a control terminal of the power switch; a first pull-up driver switch for causing current to flow out to the control terminal, the first pull-up driver switch having a first on-state resistance; a second pull-up driver switch for causing current to flow out to the control terminal, the second pull-up driver switch having a second on-state resistance substantially higher than the first on-state resistance; Mode indication input; a switch control input to control the pull-down driver switch, the first pull-up driver switch, and the second pull-up driver switch; as well as A high-side control circuit is configured to disable the first pull-up driver switch in response to a burst mode indication at the mode indication input.

2. The switch driver of claim 1, wherein the pull-down driver switch is an n-channel metal oxide semiconductor field effect transistor (MOSFET), and the first pull-up driver switch and the second pull-up driver switch are p-channel MOSFETs.

3. The switch driver according to claim 2, wherein the high-side control circuit comprises: a high-side pre-driver for controlling the first pull-up driver switch; An inclusive OR gate has the mode indication and the switch control as inputs thereof and has an output coupled to an input of the high-side pre-driver.

4. The switch driver of claim 1 , wherein the second pull-up driver switch is sized to source a current level that limits the voltage level across the SR switch to below an upper voltage limit of the SR switch, the upper voltage limit corresponding to an avalanche breakdown voltage of the SR switch.

5. The switch driver of claim 1, wherein the first pull-up driver switch, when turned on, draws a current level that is at least 100 times greater than a current level drawn by the second pull-up driver switch when the second pull-up driver switch is turned on.

6. The switch driver of claim 1, wherein a size ratio between the first pull-up driver switch and the second pull-up driver switch is programmable by a configurable digital code. 7 . The switch driver of claim 1 , wherein a size ratio between the first pull-up driver switch and the second pull-up driver switch is set by metal selection during fabrication of the pull-up driver switches.

8. A method for controlling a power switch in a switch-mode power converter to prevent excessive voltage levels across synchronous rectification (SR) switches of the power converter, the power converter comprising a switch driver, the switch driver comprising a first pull-up driver switch and a second pull-up driver switch, wherein the second pull-up driver switch has an on-state resistance that is substantially higher than the on-state resistance of the first pull-up driver switch, the method comprising: detecting that a load of the power converter is consuming a low power level; In response to the detecting, a burst mode of operation is entered and the first pull-up driver switch is disabled.

9. The method of claim 8 , wherein the SR switch has an associated avalanche breakdown voltage corresponding to a voltage limit across a first terminal and a second terminal of the SR switch, and wherein the on-state resistance of the second pull-up driver switch is configured to support a current level through the second pull-up driver switch such that the ratio at which the power switch is turned on ensures that the voltage across the first terminal and the second terminal of the SR switch remains below the avalanche breakdown voltage of the SR switch.

10. The method according to claim 8, further comprising: In response to detecting that the load of the power converter is consuming more power than the low power level, the burst mode of operation is exited and the first pull-up driver switch is enabled.

11. An isolated power converter, comprising: an input for coupling to an input power supply; Output for coupling to a load; A transformer, including a primary winding and a secondary winding; a power stage configured to switch power from the input to the primary winding, the power stage comprising a power switch and a switch driver for controlling the power switch; a synchronous rectification SR switch configured to rectify power provided by the secondary winding, the SR switch having an SR voltage across a first terminal and a second terminal of the SR switch and having an avalanche breakdown voltage level associated with the SR voltage; a controller configured to control power delivered to the output by generating a switch control signal and configured to generate a burst mode indication based on detecting that the load is in a low power state; The switch driver comprises: a pull-down driver switch for injecting current from a control terminal of the power switch; a first pull-up driver switch and a second pull-up driver switch for causing current to flow to the control terminal, wherein the second pull-up driver switch has an on-state resistance that is substantially higher than an on-state resistance of the first pull-up driver switch; a switch control input for coupling to the switch control signal; and A high-side control circuit is configured to disable the first pull-up driver switch in response to the burst mode indication.

12. The isolated power converter of claim 11, wherein the pull-down driver switch is an n-channel metal oxide semiconductor field effect transistor (MOSFET), and the first and second pull-up driver switches are p-channel MOSFETs.

13. The isolated power converter of claim 12, wherein the high-side control circuit comprises: a first high-side pre-driver for controlling the first pull-up driver switch; An inclusive OR gate has the mode indication and the switch control as inputs thereof and has an output coupled to an input of the first high-side pre-driver.

14. The isolated power converter of claim 11 , wherein the second pull-up driver switch is sized to source a current level that limits a voltage level across the first terminal and the second terminal of the SR switch to below an upper voltage limit of the SR switch, the upper voltage limit corresponding to an avalanche breakdown voltage of the SR switch.

15. The isolated power converter of claim 11, wherein the first pull-up driver switch, when turned on, draws a current level that is at least 100 times greater than a current level drawn by the second pull-up driver switch when the second pull-up driver switch is turned on.

16. The isolated power converter of claim 11, wherein a size ratio between the first pull-up driver switch and the second pull-up driver switch is programmable by a configurable digital code. 17 . The isolated power converter of claim 11 , wherein a size ratio between the first pull-up driver switch and the second pull-up driver switch is set by metal selection during fabrication of the pull-up driver switches.

18. The isolated power converter of claim 11, wherein the isolated power converter has a flyback converter topology.

Images