Reducing voltage transients across a MOSFET in a synchronous rectifier in a DC/DC converter

[0002] The invention relates generally to DC/DC converters, and more specifically, to a method and an arrangement for controlling synchronous rectifiers in DC/DC converters.

[0003] To rectify high frequency AC voltages in switched DC/DC converters, synchronous rectifiers in the form of MOSFETs are often used instead of diodes. Since the voltage drop across a MOSFET is lower than across a diode, the efficiency of converters with MOSFETs will be higher. However, since a MOSFET cannot be turned off instantaneously, its turn-off will not occur at exactly the same moment as the turn-on of a primary switch but a little later. This will cause reverse currents through the MOSFET, causing voltage transients to appear across the MOSFET. To eliminate these voltage transients, unique control signals are needed for the MOSFET to function properly. It is known to provide these control signals from the primary side of the converter, e.g. via an additional transformer.

[0004] An object of the invention is to control a MOSFET on the secondary side without any need of transferring control signals from the primary side of the converter. Voltage transients across the MOSFET caused by reverse currents when it is turned off are detected. At least a next on-period of the MOSFET is shortened in response to the detection of such transients. As a result, the MOSFET will be controlled more accurately, and transients are eliminated.

[0005] The invention will be described more in detail below with reference to the appended drawing on which FIG. 1 shows a fly-back DC/DC converter with an example embodiment of a control arrangement in accordance with the invention, and FIGS. 2A-2D are diagrams illustrating different signals in the example embodiment in FIG. 1.

[0006] FIG. 1 shows a fly-back DC/DC converter with an example embodiment of a control arrangement in accordance with the invention. The converter comprises a transformer TR having a primary winding connected in series with a primary switch T1 to output terminals of a schematically illustrated DC voltage source V1, and a secondary winding connected in series with a MOSFET T2 to an output capacitor C1 for generating an output DC voltage. The MOSFET T2 is used as a synchronous rectifier and is controlled in accordance with the invention to improve the efficiency of the converter.

[0007] In a manner known per se, the MOSFET T2 comprises a source S, a drain D, and a gate G as well as a body diode D1 connected with its anode to the source S and with its cathode to the drain D of the MOSFET T2. In the example embodiment shown in FIG. 1, the drain D of the MOSFET T2 is connected to the source S of the MOSFET T2 via a resistor R1 in series with a diode D2 and a capacitor C2. Also, the drain D is connected to the source S via a diode D3 in series with a capacitor C3.

[0008] The interconnection point between the diode D2 and the capacitor C2 is connected to the collector of a transistor T3 whose emitter is connected to the gate G of the MOSFET T2 and to the emitter of a transistor T4 whose collector is connected to the source S of the MOSFET T2. The bases of the transistors T3 and T4 are interconnected and connected via a resistor R2 to the collector of the transistor T3. The interconnected bases of the transistors T3 and T4 are also connected to the collector of a transistor T5 whose emitter is connected to the source S of the MOSFET T2.

[0009] The base of the transistor T5 is connected via resistor R3 in series with a capacitor C4 to the drain D of the MOSFET T2. The interconnection point between the resistor R3 and the capacitor C4 is connected via a resistor R4 to the collector of a transistor T6 whose emitter is connected to the interconnection point between the diode D3 and the capacitor C3 via a resistor R5 and whose base is connected to the interconnection point between the diode D2 and the capacitor C2.

[0010] As mentioned above, the switch T1 is the so-called primary switch of the fly-back converter. When the switch T1 is on, magnetic energy is stored in the transformer TR. The voltage U across the secondary winding of the transformer TR is negative. When the primary switch T1 goes off, the voltage U will be positive and energy will be transferred via the MOSFET T2 to the output capacitor C1.

[0011] The transistors T3 and T4 are two emitter-followers that quickly can charge/discharge the gate G of the MOSFET T2. The transistor T5 turns the MOSFET T2 on and off via the emitter-followers T3 and T4. The transistor T5 is turned on and off by the capacitor C4. When the primary switch T1 turns on, the MOSFET T2 is still conducting, but the current through the MOSFET T2 reverses its direction when the voltage U starts to fall. The transistor T5 senses this change via the capacitor C4 and turns off the MOSFET T2. By charging the capacitor C4 via the resistor R4, the time when the transistor T5 turns on, i.e., when the MOSFET T2 turns off, can be varied.

[0012] The transistor T6 senses the voltage difference between the voltage across the capacitor C3 and the voltage across the capacitor C2. The capacitor C3 is charged to the peak voltage of the voltage U including any transient emanating from a non-desired turn-off current from the MOSFET T2, while the capacitor C2 is charged to the voltage U excluding any transients since the resistor R1 filters out any transients.

[0013] When the voltage difference between the capacitors C3 and C2 exceeds the base-emitter voltage of the transistor T6, the transistor T6 will conduct and via the resistor R4 set the off-time of the transistor T5, i.e. the on-time of the MOSFET T2, so that the voltage difference between the capacitors C3 and C2 will be smaller than a few volts after a number of switch cycles

[0014] With reference to FIGS. 2A-2D, the operation of the converter illustrated in FIG. 1 will now be described in more detail.

[0015] FIG. 2A illustrates a couple of cycles of the voltage UT1 across the primary switch T1. The primary switch T1 is supposed to be on from the beginning and is supposed to be turned off at times t1 and t1′ and turned on at times t2 and t2′. When the primary switch T1 is on, the capacitor C2 is charged via the resistor R1 in series with the diode D2, and the capacitor C3 is charged via the diode D3. At time t1 when the primary switch T1 turns off, the body diode D1 of the MOSFET T2 begins to conduct to charge the output capacitor C1.

[0016] If FIG. 2B, the source-drain voltage US-DT2 of the MOSFET T2 is illustrated. Base current will be supplied to the transistor T3 via the resistor R2, and the gate G of the MOSFET T2 will be charged causing the MOSFET T2 to become saturated.

[0017] FIG. 2C illustrates the gate-source voltage UG-ST2 of the MOSFET T2. At time t2, the primary switch T1 turns on again. When the voltage across the MOSFET T2 changes polarity at time t2 as illustrated in FIG. 2B, the transistor T5 becomes saturated via the capacitor C4 and the resistor R3, and the MOSFET T2 goes off. However, due to the fact that it takes some time to discharge the gate of the MOSFET T2 after that the transistor T5 is turned on, the MOSFET T2 will turn off a little late causing a reverse current to flow from the output capacitor C1 back into the transformer TR. This reverse current causes a voltage transient across the MOSFET T2 when the MOSFET T2 is off.

[0018] That transient, which is illustrated in FIG. 2B at time t2, charges the capacitor C3 via the diode D3 to a voltage that is higher than the voltage across the capacitor C2, causing the transistor T6 to start conducting. The collector current from the transistor T6 then sets the time when the transistor T5 starts to draw gate charge from the gate G of the MOSFET T2. Towards the end of the next off-period of the primary switch T1, i.e. towards time t2′, almost all gate charge has been drawn from the gate G of the MOSFET T2 before the primary switch T1 goes on again at time t2′ as illustrated in FIG. 2C.

[0019] When the primary switch T1 turns on at time t2′, the MOSFET T2 is prepared, i.e., most of the gate charge has been drawn off and the MOSFET T2 is not fully conducting. Thus, no reverse current spike will appear across the MOSFET T2 at time t2′ as apparent from FIG. 2B.

[0020] FIG. 2D illustrates the base-emitter voltage UB-ET5 of the transistor T5. At time t1, the voltage change across the secondary winding of the transformer TR turns off the transistor T5 via the capacitor C4 and the resistor R3. Hereby, the gate G of the MOSFET T2 is charged via the resistor R2 and the transistor T3. Between times t1 and t2, the base-emitter voltage UB-ET5 of the transistor T5 is reversed and, consequently, the transistor T5 is not conducting.

[0021] When the primary switch T1 is turned on at time t2, the reverse voltage across the MOSFET T2 turns on the transistor T5 and turns off the MOSFET T2. Between times t2 and t1′, the base of the transistor T5 is forward biased. Therefore, the gate G of the MOSFET T2 is low, and the MOSFET T2 is not conducting as is apparent from FIG. 2C.

[0022] For the next period, i.e., the time between t1′ and t2′, the voltage difference between the capacitors C2 and C3, caused by the transient appearing at time t2, turns on the transistor T6 causing the capacitor C4 to be discharged via the resistor R4. As a result, the transistor T5 will start conducting at a time tx before time t2′ as indicated in FIG. 2D.

[0023] Also at time tx, via the transistor T4, the transistor T5 will begin to draw gate charge out of the gate G of the MOSFET T2 to turn off the MOSFET T2 earlier than time t2′ as illustrated in FIG. 2C.

[0024] Thus, due to the earlier turn-off of the MOSFET T2, the transient or current spike that appeared at time t2 will not appear at time t2′ as illustrated in FIG. 2B. However, should a current spike be present also at time t2′, the MOSFET T2 will be turned off earlier towards the end of its next on-period. As should be apparent from the above, no control signals have to be transferred from the primary side of the converter to control the MOSFET on the secondary side.