Zero current switching

[0001] The present application benefits from U.S. provisional application 61/039,046 filed on Mar. 24, 2008 by the present inventors.

[0002] 1. Technical Field

[0003] The present invention relates to switching converters and to specifically a method and devices for zero current switching for reducing switching losses in switching converters.

[0004] 2. Description of Related Art

[0005] FIG. 1 shows a typical conventional buck-boost DC-to-DC converter circuit 10. The buck circuit of buck-boost DC-to-DC converter 10 has an input voltage V_in with an input capacitor C₁ connected in parallel across V_in. Two switches are implemented as field effect transistors (FET) with integral diodes: a high side buck switch Q₁ and a low side buck switch Q₂ connected in series by connecting the source of Q₁ to the drain of Q₂. The drain of Q₁ and the source of Q₂ are connected parallel across an input capacitor C₁. A node A is formed between switches Q₁ and Q₂ to which one end of an inductor 106 is connected. The other end of inductor 106 is connected to the boost circuit of buck-boost DC-to-DC converter 10 at a node B. Node B connects two switches: a high side boost switch Q₄ and a low side boost switch Q₃ together in series where the source of Q₄ connects to the drain of Q₃ to form node B. The drain of Q₄ and the source of Q₃ connect across an output capacitor C₂ to produce the output voltage V_out of buck-boost DC-to-DC converter 10.

[0006] FIG. 2a illustrates the buck phase or on-state circuit of DC-to-DC converter circuit 10 shown in FIG. 1, the input voltage source V_in is directly connected to inductor 106 and the load is isolated from V_in because Q₁ is on, Q₂ is off, Q₃ is on and Q₄ is off. These switch positions: Q₁ on, Q₂ off, Q₃ on and Q₄ off; result in accumulating energy in inductor 106 since source V_in is directly connected to inductor 106. In the on-state, output capacitor C₂ supplies energy to the load.

[0007] FIG. 2b illustrates the boost phase or off-state circuit of DC-to-DC converter circuit 10, Inductor 106 is connected in parallel across the load and capacitor C₂ because Q₁ is off, Q₂ is on, Q₃ is off and Q₄ is on. Q₁ being off isolates inductor 106 from the input voltage (V_in) and capacitor (C₁). The stored energy in inductor 106 (as a result of the previous On-state) is transferred from inductor 106 to C₂ and the load.

[0008] Two common methods of operating DC-to-DC converter circuit 10 are in either continuous mode or discontinuous mode. If the current through the inductor 106 never falls to zero during a commutation cycle (i.e. the time period to perform both the on-state and the off-state), DC-to-DC converter circuit 10 is said to operate in continuous mode and typically the on-state operates for a shorter period of time when compared to the off-state. Discontinuous mode of operation for DC to DC converter circuit 10 occurs when the amount of energy required by the load is small enough to be transferred in a time period smaller than the whole commutation cycle. Typically, the current through inductor 106 falls to zero for a short time period after the off-state period and therefore inductor 106 is completely discharged at the end of the commutation cycle. The commutation cycle therefore includes the on-state, the off-state and the short time period during which the inductor current is zero.

[0009] A conventional “resonant” method for achieving virtually zero power loss when switching a switch is to apply a direct current voltage input voltage V_in across a switch (with a diode connected across the switch, the diode is reverse biased with respect to V_in) in series with an inductor L and a capacitor C. The output voltage of the circuit is derived across the capacitor. The output voltage of the circuit could then in principle be connected to the input of a power converter, for example a buck-loaded series tank circuit with load. The resonant frequency of the series inductor L and capacitor C is given by Eq. 1 and the corresponding resonant periodic time T given in Eq. 2.
f₀=½π(LC)^1/2  Eq.1
T=1/f₀  Eq.2

[0010] A pulse response of the circuit means that when the switch turns on, there is both zero current in the inductor and zero voltage across the capacitor (Power=Volts×Current=0×0=zero power loss at turn on). During steady state operation of the circuit, the inductor current and capacitor voltage are sinusoidal and have a 90 degrees phase shift with respect to each other. When the switch turns off (the on period of the switch corresponds to half of the resonant periodic time) there is zero current in the inductor and maximum positive voltage (i.e. V_capacitor=V_in) across the capacitor (Power=Volts×Current=V_in×0=zero power loss at turn off).

[0011] According to the present invention there is provided a method for providing non-resonant zero-current switching in a switching power converter operating in a continuous current mode. The switching power converter converts power from input power to output power. The switching power converter includes a main switch connected to a main inductor, wherein an auxiliary inductor is connected in parallel with the main inductor. The main current flows from an input to an output. The auxiliary inductor is connected with the main inductor thereby charging the auxiliary inductor so that an auxiliary current flows from the output to the input opposing main current. Upon a total current including a sum of the main current and the auxiliary current. substantially equals or approaches zero, the main switch is turned on. When the auxiliary current is or approaches zero current the auxiliary inductor is disconnected from the main inductor.

[0012] According to the present invention there is provided a switching converter including a buck stage or a boost stage or a buck-boost stage including: a main switch connecting an input voltage terminal to a first node, a main inductor connected at one end to the first node and at the other end operatively connected at a second node to a voltage output; and an auxiliary inductor adapted for connecting in parallel with the main inductor between the first and second nodes. Upon connecting the auxiliary inductor with the main inductor, the auxiliary inductor is charged so that an auxiliary current flows from the second node to the first node opposing the main current flowing between the first and second nodes through the main inductor. The total current includes a sum of the main current and the auxiliary current. When the total current substantially equals or approaches zero. The main switch is switched on. Energy stored within the auxiliary inductor is substantially all available for converting to output power by the switching converter. The current of the auxiliary inductor is naturally discharged to the input and/or the output. When the auxiliary current approaches zero, the auxiliary inductor is disconnected from the main inductor. The switching converter may include a first auxiliary switch adapted for connecting the auxiliary inductor to the first node; a second auxiliary switch adapted for connecting the auxiliary inductor to the second node. A discharge diode for discharging the auxiliary inductor is connected between the auxiliary conductor and a second input voltage terminal or ground in case of reverse current in the auxiliary inductor due to reverse recovery charge of one of the auxiliary switches. The first and second auxiliary switches are typically implemented as field-effect transistors each with parasitic diodes, with the parasitic diodes connected in opposite directions. The main switch is usually any of a: silicon controlled rectifier (SCR), insulated gate bipolar junction transistor (IGBT), bipolar junction transistor (BJT), field effect transistor (FET), junction field effect transistor (JFET), switching diode, electrical relay, reed relay, solid state relay, insulated gate field effect transistor (IGFET), diode for alternating current (DIAC), and/or triode for alternating current TRIAC.

[0013] According to the present invention there is provided a plurality of main switches interconnected in a bidirectional current full bridge topology, the main switches including a first switch, a second switch, a third switch and a fourth switch. A pair of input voltage terminals are attachable at a first node connecting the first and third switches and at a second node connecting the second and fourth switches. A first output voltage terminal is operatively attached at a third node connecting the third and fourth switches and a second output voltage terminal is operatively attached at a fourth node connecting the first and second switches. A first main inductor or a single inductor and some other circuitry (a transformer for example) first node is attachable between the first output voltage terminal and the third node and a second main inductor or a single inductor and some other circuitry second node attachable between the second output voltage terminal and the fourth node. An auxiliary inductor is connectible between the third node and the fourth node. Upon connecting the auxiliary inductor with the first and/or second main inductors or single inductor with circuitry, the auxiliary inductor is charged so that an auxiliary current flows between the third node and the fourth node. The auxiliary current opposes the main current. The total current includes a sum of the main current and the auxiliary current.

[0014] When the total current substantially equals or approaches zero, the main switches are switched (on). The current of the auxiliary inductor is naturally discharged to the input and/or the output. When the auxiliary current approaches zero, the auxiliary inductor is disconnected from the main circuit. Energy stored within the auxiliary inductor is substantially all available for converting to output power by the switching converter. A first auxiliary switch is preferably adapted for connecting the auxiliary inductor to the third node and a second auxiliary switch is adapted for connecting the auxiliary inductor to the fourth node. One or two discharge diodes are preferably connected between said auxiliary conductor to the second node. The discharge diodes serve to protect against reverse recovery current of the auxiliary switches, depending on the main current direction. The first and second auxiliary switches are typically implemented as field-effect transistors each with parasitic diodes; the parasitic diodes are connected on opposite directions. The first inductor and the second inductor are preferably split inductors. The main switches are silicon controlled rectifier (SCR), insulated gate bipolar junction transistor (IGBT), bipolar junction transistor (BJT), field effect transistor (FET), junction field effect transistor (JFET), switching diode, electrical relay, reed relay, solid state relay, insulated gate field effect transistor (IGFET), diode for alternating current (DIAC), and/or triode for alternating current TRIAC.

[0025] The foregoing and/or other aspects will become apparent from the following detailed description when considered in conjunction with the accompanying drawing figures.

[0026] Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. The embodiments are described below to explain the present invention by referring to the figures.

[0027] The term switch as used herein refers to any of: silicon controlled rectifier (SCR), insulated gate bipolar junction transistor (IGBT), bipolar junction transistor (BJT), field effect transistor (FET), junction field effect transistor (JFET), switching diode, mechanically operated single pole double pole switch (SPDT), SPDT electrical relay, SPDT reed relay, SPDT solid state relay, insulated gate field effect transistor (IGFET), diode for alternating current (DIAC), and triode for alternating current (TRIAC).

[0028] The term “zero current switching” (or “ZCS”) as used herein is when the current through a switch is reduced to significantly zero amperes prior to when the switch is being turned either on or off.

[0029] The term “power converter” as used herein applies to DC-to-DC converters, AC-to-DC converters, DC-to-AC inverters, buck converters, boost converters, buck-boost converters, full-bridge converters and half-bridge converters or any other type of electrical power conversion/inversion known in the art.

[0030] The term “zero voltage switching” (or “ZVS”) as used herein is that the peak voltage across a switch, is reduced to substantially zero volts when the switch is being turned either on or off.

[0031] The term “cycle” or “commutation cycle” refers to the periodicity of main switch positions in a circuit which performs a process of electrical power conversion or inversion.

[0032] The term “non-resonant” as used herein to exclude resonant and quasi-resonant circuits used in the prior art for zero current switching. Resonant switching implies that the switching frequencies are similar to a resonant frequency of a resonant tank circuit in the switching converter.

[0033] The terms “charging” and “discharging” in the context of the present invention in reference to charging and discharging a capacitor, are used herein interchangeably except that current flow while charging and discharging is usually in the opposite direction.

[0034] Reference is now made to FIG. 3 (including FIGS. 3a-3e) showing a buck-boost DC-to-DC converter 30 according to an embodiment of the present invention. A buck circuit 32 of buck-boost DC-to-DC converter 30 has an input voltage V_in with an input capacitor C₁ connected in parallel across V_in. Two switches Q₁ and Q₂ are connected in series at node A by connecting the source of Q₁ to the drain of Q₂. The drain of Q₁ and the source of Q₂ are placed in parallel across capacitor C₁. The source of Q₂ is connected to ground. The other end of inductor 106 is connected in the present example to a boost circuit 34 of buck boost DC-to-DC converter 30 at node B.

[0035] A zero-voltage switching feature according to embodiments of the present invention is provided using switch module 300. Switch module 300 has a switch Q₅, the drain of Q₅ is connected to node A of buck circuit 32 via link 308. The source of Q₅ connects to one end of an auxiliary inductor 302 to form a node E. A cathode of discharge diode D₁ connects to node E and the anode of discharge D₁ connects to ground. The other end of auxiliary inductor 302 connects to the source of switch Q₆. The drain of Q₆ connects to node B of boost circuit 34 via a link 310.

[0036] Reference is still made to buck-boost DC-to-DC converter 30 shown in FIGS. 3a-3e which illustrate operation of buck-boost DC-to-DC converter 30 and FIG. 4 showing a flow diagram of a method for zero current switching, for zero current switching in boost and/or buck topologies running in continuous mode during the turn on of main switch Q₃ and/or Q₁ according to embodiments of the present invention. Throughout the following illustration using FIGS. 3a-3e, an electrical property of an inductor is relied upon; namely that when a voltage V is applied across an inductor, the initial current through the inductor is zero and after a certain time period inductor current I_L builds up linearly. In FIGS. 3a-3e, current flow in buck circuit 32, boost circuit 34 and switch module 300 is indicated by arrow markings and gray shaded lines.

• [0037] A. Reference is now made to FIG. 3a which illustrates the state of DC-to-DC converter 30 prior to buck turn on. Switch Q₂ is on. Switches Q₁, Q₅ and Q₆ are off. Current flows through inductor 106 from node A to node B and switch Q₂. Switch Q₂ turns off and current continues to flow through the parasitic diode of switch Q₂.
• [0038] B. Reference is now made to FIG. 3b which continuous to illustrate the state of DC-to-DC converter 30 prior to buck leg turn on. Switches Q₅ and Q₆ turn on at zero current at I_aux. Current I_aux through auxiliary conductor and switches Q₅ and Q₆ increases linearly as auxiliary inductor 302 is charged by output voltage V_out. The direction of current I_aux opposes current I_p through main inductor 106 and flows from node B to node A. Switches Q₁, and Q₂, are off but current I_p flows through inductor 106 from node A to node B through the parasitic diode of switch Q₂.
• [0039] C. Reference is now made to FIG. 3c which illustrates the state of DC-to-DC converter 30 as buck leg turns on. When current I_aux equals current I_p through inductor 106 plus Q₂ diode recovery current, the voltage V_buck across Q₂ starts to rise. Switch Q₁ then turns on at zero current. Switch Q₂ remains off. Switches Q₅ and Q₆ remain on and I_aux begins to decrease discharging from V_in to V_out.
• [0040] D. Reference is now made to FIG. 3d which continues to illustrate the state of DC-to-DC converter 30 as buck leg turns on; Q₅ turns off when current through I_aux gets low (i.e. substantially zero current switching), the current I_aux flows through Q₅ parasitic diode. The current I_aux reverses because of Q₅ reverse recovery current. This reverse recovery current discharges through the discharge diode D₁ and Q₆.
• [0041] E. Reference is now made to FIG. 3e which illustrates the state of DC-to-DC converter 30 after buck leg turns on. Current I_p flows through switch Q₁ through inductor 106 and through boost circuit 34. Current I_aux falls to zero and switch Q₆ is turned off.

[0042] FIG. 4 shows a simplified flow diagram of a method for zero current switching according to an embodiment of the present invention. Still referring to the embodiment of FIG. 3, in step 41, auxiliary inductor 302 is connected to main inductor 106 in parallel by auxiliary switches Q₅ and Q₆. When the total current is substantially zero in decision box 43, main switch Q₁ is switched on (step 45) at zero current. A preferred method of determining when the total current reached is zero is to use a zero current sensing circuit or to turn Q₁ on at a time according to the timing transients of either auxiliary inductor 302 or inductor 106. When the current in auxiliary inductor 302 is substantially zero in decision box 47, auxiliary switch Q₆ is switched off (step 49) at zero current and inductor 302 is disconnected across inductor 106. A preferred method of determining when the current in auxiliary inductor 302 approaches zero current, is to use a zero current sensing circuit or to turn Q₆ off at a time according to the timing transients of either auxiliary inductor 302 or inductor 106.

[0043] FIG. 5 shows a further embodiment of present invention as applied to a full bridge DC to DC converter 50. Full bridge DC to DC converter 50 has four main switches S_m,1, S_m,2, S_m,3 and S_m,4 connected together in a full bridge configuration. Each of four main switches (S_m,1, S_m,2, S_m,3 and S_m,4) have respective diode shunts connected in parallel thereto. The diodes placed across switches S_m,1 and S_m,2 are in both the same direction similarly the diodes of S_m,3 and S_m,4 are both in the same direction. All diodes connected across switches S_m,1, S_m,2, S_m,3 and S_m,4 are reverse biased with respect to the input voltage V_in. An input voltage (V_in⁻) of full bridge DC-to-DC converter 50 is connected across the node between switches S_m,2 and S_m,4 and an input voltage (V_in⁺) is connected at the node between switches S_m,1 and S_m,3. An output voltage (V_out⁻) of full bridge DC-to-DC converter 50 is connected across the node (M) between switches S_m,1 and S_m,2 connected through a split inductor 500a and output voltage V_out+ is connected at the node (L) between switches S_m,3 and S_m,4 through a split inductor 500b. An auxiliary switch S_a,1 is connected between nodes M and K. A diode is placed across in parallel with S_a,1 with the cathode of the diode connected to node M and the anode of the diode connected to node K. A discharge diode D₁ is connected between node K and V_in⁻ with the anode of D₁ connected to V_in⁻ and the cathode of D₁ connected to node K. One end of an auxiliary inductor 502 connects to node K. The other end of auxiliary inductor 502 connects to node J. A discharge diode D₂ is connected between node J and V_in⁻ with the anode of D₂ connected to V_in⁻ and the cathode of D₂ connected to node J. An auxiliary switch S_a,2 is connected between nodes J and L. A diode is placed across in parallel with S_a,2 with the cathode of the diode connected to node L and the anode of the diode connected to node J.

[0044] The operation of full bridge circuit 50, according to a feature of the present invention is illustrated with reference also to FIGS. 5a-5e and FIG. 6 which shows a timing diagram of selected voltages and currents for steps V to Z is as follows:

[0045] FIGS. 5a-5e illustrate current flow is indicated by arrow markings and gray shaded lines.

• [0046] V) Referring now specifically to FIG. 5a: Main switches S_m,1 and S_m,4 are turned on, all other switches are off. Current I_p flows from V_out⁻ to V_in⁺ through inductor 500a, and through main switch S_m,1. Current flows from V_in⁻ to V_out⁺ through main switch S_m,4 and through inductor 500b.
• [0047] W) Referring now specifically to FIG. 5b; Switches S_m,1 and S_m,4 are turned off. Switches S_a,1 and S_a,2 are turned on. Current I_aux begins at zero current and increases linearly, current I_aux flowing between nodes M and L. Because the current I_aux is initially zero, the switching on of switches S_a,1 and S_a,2, occurs with zero current. Current flows from V_out⁻ to V_in⁺ through inductor 500a, and through the diode of main switch S_m,1 Current flows from V_in⁻ to V_out⁺ through the diode of main switch S_m,4 and through inductor 500b.
• [0048] X) Referring now specifically to FIG. 5c; Once the auxiliary current I_aux reaches a peak value, equaling to I_p, main switches S_m,2 and S_m,3 are turned on with zero current since Kirchhoff's current equation at node M shows current through main switch S_m,2: IS_m,2=I_P−I_aux=0 and Kirchhoff's current equation at node L shows IS_m,3=I_P−I_aux=0. Current flows from V_in⁺ to V_out⁺ through main switch S_m,3, and through inductor 500b. Current now flows from V_out⁻ to V_in⁻ through inductor 500a and through main switch S_m,2.
• [0049] Y) Referring now specifically to FIG. 5d; Auxiliary switch S_a,2 is now turned off as current I_aux has reduced to substantially zero in the direction of node M to node L. Diode D₂ takes any reverse recovery current from switch S_a,2. Current flows into V_out+ from V_in⁺ through inductor 500b, and through main switch S_m,3 Current now flows into V_in⁻ from V_out⁻ through main switch S_m,2 and through inductor 500a.
• [0050] Z) Referring now specifically to FIG. 5e; Auxiliary switch S_a,1 is now turned off with zero current I_aux. Current flows into V_out+ from V_in⁺ through inductor 500b, and through main switch S_m,3 Current now flows into V_in⁻ from V_out⁻ to through main switch S_m,2 and through inductor 500a.

[0051] Similar switching steps occur when the current at the main inductor 500 is reverse in polarity. Switch pairs (S_m,1, S_m,2), (S_m,4, S_m,3), (S_a,1, S_a,2) and diodes (D₂, D₁) are swapped at the above description to accomplish this symmetrical case.

[0052] The definite articles “a”, “an” is used herein, such as “a converter”, “a switch” have the meaning of “one or more” that is “one or more converters” or “one or more switches”.

[0053] Although selected embodiments of the present invention have been shown and described, it is to be understood the present invention is not limited to the described embodiments. Instead, it is to be appreciated that changes may be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and the equivalents thereof.