Switch and associated electronic device
The switch design addresses the Miller effect in power transistors by integrating a smaller protection transistor and control circuits, effectively reducing switching losses and the risk of short circuits while being straightforward to implement.
Patent Information
- Application Number
- FR2021014066
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-12-21
- Publication Date
- 2025-05-09
- Estimated Expiration
- 2041-12-21
AI Technical Summary
Rapid switching power transistors often experience self-commutation phenomena due to the Miller effect, leading to potential short circuits and increased switching losses, for which existing solutions are either complex to implement or result in significant switching losses.
A switch comprising a transistor to be protected and a Miller effect protection unit, where the protection unit includes a smaller protection transistor connected between the drain and gate of the main transistor, along with a link circuit and control circuit to manage leakage currents and voltage dividers, respectively.
The solution effectively mitigates the Miller effect by controlling the voltage rise, reducing the risk of short circuits and switching losses, while being easy to implement and integrating the protection unit monolithically with the transistor.
Smart Images
Figure 00000011_0000 
Figure 00000012_0000 
Figure 00000013_0000
Abstract
Description
Title of the invention: Switch and associated electronic device
[0001] The present invention relates to a switch. It also relates to an electronic device comprising such a switch.
[0002] Certain types of fast-switching power transistors undergo self-switching phenomena due to the Miller effect. Such phenomena occur when the voltage between the drain and source of the power transistor increases rapidly, inducing a potential rise in the gate voltage that can lead to re-conductivity and a temporary short circuit.
[0003] To remedy this problem, several solutions are known in the prior art.
[0004] Document WO201984899 A1 proposes a control circuit for a switch, the control circuit comprising a voltage sensor for detecting a control voltage and an electrical source for providing a control signal having a control value. The control circuit is designed to adjust the control value based on the control voltage in order to limit the switching current flowing through the switch.
[0005] Such a circuit makes it possible to eliminate grid lift, but at the expense of switching losses.
[0006] There are also techniques proposing transistor manufacturing processes aimed at adjusting the capacitance values between the transistor electrodes to induce better natural immunity.
[0007] However, such a technique is relatively complex to implement.
[0008] There is therefore a need for a switch protected at least in part against the Miller effect with a technique that is easy to implement.
[0009] To this end, the description describes a switch comprising a transistor to be protected, the transistor to be protected comprising a gate, a drain, and a source; the switch comprising a Miller effect protection unit, the protection unit comprising a protection transistor, the protection transistor comprising a gate, a drain, and a source, the drain of the protection transistor being connected to the gate of the transistor to be protected, and the source of the protection transistor being connected to the source of the transistor to be protected. The switch includes a linking circuit, the linking circuit being a high-pass filter for the leakage current of the transistor to be protected, the linking circuit being disposed between the gate of the protection transistor and the drain of the transistor to be protected. The switch includes a control circuit, the control circuit being interposed between the gate of the protection transistor and the source of the transistor to be protected.
[0010] According to particular embodiments, the switch has one or more of the following characteristics, taken individually or in all technically possible combinations:
[0011] - the linking circuit is a capacitor.
[0012] - the control circuit is formed by a Zener diode in parallel with a resistor resistance.
[0013] - the control circuit is formed by a discharge transistor and a unit of discharge transistor control.
[0014] - the control unit comprises a resistor bridge including a midpoint, a first resistance and a second resistance, the control unit also including a capacitor in parallel with the first resistance, the midpoint being connected to the gate of the discharge transistor.
[0015] - the transistor to be protected is made of gallium nitride or silicon carbide.
[0016] - the protection unit and the transistor to be protected are made on the same chip.
[0017] - each transistor has a dimension, the ratio between the dimension of the transistor protection and the dimension of the transistor to be protected being less than or equal to 0.1.
[0018] The description also describes an electronic device comprising a switch as previously described.
[0019] According to a particular embodiment, the electronic device further comprises a voltage source and a bridge arm formed by a transistor and the switch.
[0020] Some features and advantages of the invention will become apparent from the following description, given solely by way of non-limiting example, and made with reference to the accompanying drawings, in which:
[0021] - [Fig.1] [Fig.1] is a schematic representation of an example of a switch,
[0022] - [Fig.2] [Fig.2] is a schematic representation of another example of com mutator, and
[0023] - [Fig.3] [Fig.3] is a schematic representation of an electronic device including a switch according to [Fig.1] or 2.
[0024] An example of a switch 10 is shown schematically in [Fig.1].
[0025] A switch 10 is, by definition, an electronic component designed to switch.
[0026] The switch 10 includes a transistor to be protected 12 and a Miller effect protection unit 14.
[0027] According to the example described, the transistor to be protected 12 and the protection unit 14 are made on the same chip.
[0028] In other words, the protection unit 14 is monolithically integrated with the transistor to be protected 12 to form a single component, which is the switch 10 which ultimately corresponds to a controlled Miller effect transistor.
[0029] The transistor to be protected 12 is subject to the Miller effect. This generally means that the transistor has a very high switching speed, i.e. a switching speed greater than or equal to 50kV / ps.
[0030] The transistor to be protected 12 is, for example, a high electron mobility transistor (more often referred to by the acronym HEMT, which refers to the corresponding English name of "High Electron Mobility Transistor").
[0031] Alternatively, the transistor to be protected 12 is an insulated gate bipolar transistor (more often referred to by the acronym IGBT, which refers to the corresponding English name "Insulated Gate Bipolar Transistor").
[0032] The transistor to be protected 12 is, for example, made of Gallium nitride (GaN) or Silicon Carbide (SiC).
[0033] The transistor to be protected 12 has three electrodes, a gate 12G, a drain 12D and a source 12S.
[0034] The gate 12G of the transistor to be protected 12 is supplied by a control voltage 16.
[0035] The protection unit 14 is intended to protect the transistor to be protected 12 against the Miller effect.
[0036] Such protection can be absolute or relative, that is to say that the protection unit 14 can reduce or cancel the Miller effect as the case may be, that is to say reduce at least in part the magnitude of the Miller effect.
[0037] The protection unit 14 could thus be described as a "Miller clamp circuit" as the name is sometimes used.
[0038] The protection unit 14 includes a protection transistor 18, a linking circuit 20 and a control circuit 22.
[0039] The protection transistor 18 also has three electrodes, namely a gate 18G, a drain 18D and a source 18S.
[0040] The protection transistor 18 is such that the protection transistor 18 has a size much smaller than that of the transistor to be protected 12.
[0041] The size of a transistor is, for example, defined as the surface occupied by the transistor on the chip on which the transistor is made in the chosen manufacturing technology.
[0042] Alternatively, the size of a transistor can also be obtained by measuring the on-state value of the transistor in question. Indeed, the on-state value of a transistor is related to the area occupied by the transistor's gate.
[0043] Thus, the ratio between the dimension of the protection transistor 18 and the dimension of the transistor to be protected 12 is less than or equal to 0.1.
[0044] Furthermore, the protection transistor 18 has a conducting state having a value in less than the impedance of the capacitance CGs at the equivalent frequency of the drain-source voltage front (capacitance between the gate 12G and the source 12S) of the transistor to be protected 12.
[0045] The drain 18D of the protection transistor 18 is connected to the gate 12G of the transistor to be protected 12.
[0046] By the expression "connected", it is understood here that the potential of the drain 18D of the protection transistor 18 and the potential of the gate 12G of the transistor to be protected 12 are identical thanks to a conductive element connecting the two electrodes.
[0047] The source 18S of the protection transistor 18 is also connected to the source 12S of the transistor to be protected 12.
[0048] The 12G gate of the transistor to be protected 12 is connected to a first gate clamp 16 and similarly, the 18G gate of the protection transistor 18 is connected to a second gate clamp 24, the two gate clamps 16 and 24 being distinct.
[0049] A grid clamp is used to clamp the voltage between the grid and the source, that is to say to limit it.
[0050] In this case, each gate clamp 16 or 24 includes a transistor.
[0051] Unlike a conventional diode which only allows electric current to pass in one direction, the forward direction, Zener diodes are designed to also allow reverse current to pass, but only if the voltage across its terminals is higher than the avalanche effect threshold.
[0052] According to the example given, the Zener diode has a Zener voltage equal to the control voltage of the protection transistor 18.
[0053] For example, the control voltage of a protection transistor 18 which is a GaN HEMT is 5 Volts (V).
[0054] According to another example, for a protection transistor 18 which is a MOSFET (acronym which refers to the English name of "Metal Oxide Semiconductor Field Effect Transistor" which translates to metal-oxide-semiconductor field effect transistor) in SiC, the control voltage is between 10V and 20V.
[0055] The connecting circuit 20 blocks the flow of direct or low-frequency current. Thus, the connecting circuit 20 allows high-frequency current to flow to the control circuit 22.
[0056] In this sense, the linking circuit 20 is a high-pass filter for the leakage current of the transistor to be protected 12.
[0057] This helps to prevent leakage at the transistor to be protected 12.
[0058] In the case described, the linking circuit 20 is a capacitor 26 disposed between the gate 18G of the protection transistor 18 and the drain 12D of the transistor to be protected 12.
[0059] More specifically, one terminal of capacitor 26 is connected to the potential applied to the gate 18G of the protection transistor 18 and the other terminal of capacitor 26 is connected to the potential applied to the drain 12D of the transistor to be protected 12.
[0060] The capacitance of the capacitor 26 is adapted according to the size of the protection transistor 18.
[0061] By way of example, for a transistor to be protected 12 which has a resistance in a conducting state of 50 mQ (allowing this transistor to withstand 30 A) and a protection transistor 18 having a resistance in the conducting state of 500 mQ for a capacitance between the gate 18G and the drain 18D of 8 pF, the capacitance of the capacitor 26 can be chosen at a value of 9 pF.
[0062] The control circuit 22 is used to discharge (respectively charge) the protection transistor 18 and more specifically to discharge (charge) the gate 18G of the protection transistor 18.
[0063] The control circuit 22 is interposed between the gate 18G of the protection transistor 18 and the source 12S of the transistor to be protected 12.
[0064] For the case of [Fig.1], the control circuit 22 is formed by a Zener diode 28 in parallel with a resistor 30.
[0065] The Zener diode 28 is connected on one side to the potential applied to the gate 18G of the protection transistor 18 and on the other side to the source 12S of the transistor to be protected 12. The Zener diode 28 is connected so as to be in the blocked state in the presence of a positive voltage between the gate 18G of the protection transistor 18 and the source 12S of the transistor to be protected 12.
[0066] Similarly, the resistor 30 is connected by one of its terminals to the potential applied to the gate 18G of the protection transistor 18 and connected by the other of its terminals to the source 12S of the transistor to be protected 12.
[0067] The value of the resistor 30 is between 100 Q and 1000 Q. Such a value allows the resistor 30 to have a good discharge capacity.
[0068] To give an order of magnitude of the set of values of the components that can be used for the switch 10, as a particular example, to protect a transistor having the following electrical characteristics respectively: 650 V (breakdown voltage), 30 A (current rating) and 50 mQ (resistance in the on state), the capacitor 26 has a capacitance of 11 pF, the resistor 30 has a resistance of 1000 Q and the Zener diode 28 has a maximum current of 3A and an effective current of 120 mA at 1 MHz.
[0069] The operation of the protection circuit 14 is now described.
[0070] The Miller effect arises from a variation in voltage between the 12D drain and the source 12S of the transistor to be protected 12 with a significant equivalent frequency.
[0071] The protection transistor 18 is used to maintain the voltage between the gate 12G and the source 12S at zero.
[0072] The origin of the effect is exploited to drive the protection transistor 18 with a voltage divider consisting basically of a capacitor 26 and a resistor 30.
[0073] The capacitor 26 does not allow direct or low-frequency current to pass through. This prevents leakage at the transistor to be protected 12.
[0074] The capacitor 26 also has a low impedance at the equivalent frequency of the voltage front between the drain 12D and the source 12S of the transistor to be protected 12.
[0075] Consequently, during a rapid transition of this voltage between the drain 12D and the source 12S of the transistor to be protected 12, the protection transistor 18 will be driven by the voltage divider effect, that is to say by distribution of the voltage between the drain 12D and the source 12S of the transistor to be protected 12 across the assembly of the capacitor 26 and the resistor 30.
[0076] In other words, the capacitor 26 allows part of the current induced by the voltage rise (dV / dt) across the terminals of the transistor to be protected 12 to charge the gate 18G of the protection transistor 18.
[0077] The protection transistor 18 then switches to the conducting state.
[0078] Once the protection transistor 18 is in this state, its on-state resistance is in parallel with the capacitance CGs of transistor 12 (between the gate 12G and the source 12S). As a result, the impedance between the gate 12G and the source 12S is reduced. Consequently, the voltage between the gate 12G and the source 12S decreases due to a voltage divider effect. Indeed, the impedance ratio Z_gate-drain / Z_gate-source increases (this ratio being respectively the ratio between the impedance between the gate 12G and the drain 12D and that between the gate 12G and the source 12S).
[0079] Thus, a rapid voltage rise charges a protection transistor 18 through the capacitor 26. Once conducting, this protection transistor 18 presents a minimum level of parasitic element and allows the discharge of charges which are temporarily stored in the gate 12G - source 12S capacitance of the transistor to be protected 12. Re-conductivity is thus avoided despite the rapid voltage rise.
[0080] The protection unit 14 therefore uses the rate of voltage rise which is the origin of the Miller effect phenomenon to trigger itself autonomously and prevent the gate 12G-source 12S voltage from rising significantly.
[0081] In addition, the protection unit 14 is monolithically integrated with the transistor to be protected 12.
[0082] This allows for greater precision as the action is carried out synchronously with the appearance of the Miller effect and is more reactive because very few parasitic elements are present between the protection unit 14 and the transistor to be protected 12.
[0083] The attenuation of the Miller effect makes the transistor to be protected 12 more reliable and also allows it to switch faster (larger dV / dt) and therefore potentially generate less switching loss.
[0084] In addition, the protection unit 14 is passive and does not involve any dedicated control circuit, which makes its implementation easy.
[0085] Another example of a switch 10 is shown schematically in [Fig.2].
[0086] The elements common to the embodiment of [Fig.1] are not repeated here, only the differences are highlighted in what follows.
[0087] The control circuit 22 is different here.
[0088] The control circuit 22 is formed by a control transistor 32 and a unit control transistor 32.
[0089] The control transistor 32 also includes three electrodes, a gate 32G, a drain 32D and a source 32S.
[0090] The control transistor 32 protects the gate 18G of the protection transistor 18.
[0091] The control transistor 32 is controlled by the driver unit 34. The drain 32D discharge transistor 32 is connected to gate 18G of protection transistor 18, gate 32G of discharge transistor 32 is connected to driver unit 34 and source 32S of discharge transistor 32 is connected to source of other transistors 12 and 18.
[0092] In the case of [Fig.2], the control unit 34 includes a resistor bridge 36 and a capacitor 38.
[0093] The resistance bridge 36 is a set of two resistors 40 and 42 in series.
[0094] The first resistor 40 is connected on one side to the capacitor 26 of the circuit of connection 20 and on the other hand to the midpoint 44 of the resistance bridge 36 while the second resistance 42 is connected on the one hand to the midpoint 44 of the resistance bridge 26 and on the other hand to the source 32S of the discharge transistor 32.
[0095] The midpoint 44 is connected to the gate 32G of the discharge transistor 32.
[0096] The capacitor 38 is connected in parallel with the first resistor 40.
[0097] More specifically, the capacitor 38 is connected on one side to the capacitor 26 of the linking circuit 20 and on the other side to the midpoint 44 of the resistance bridge 36.
[0098] The operation is relatively similar to that explained for the case of [Fig.1]
[0099] In addition, the two resistors 40 and 42 serve to discharge the 18G gate of the transistor to protect 18 and form a voltage divider to drive the control transistor 32 if the voltage between the gate 18G and the source 18S of the transistor to be protected becomes too high.
[0100] The capacitor 38 allows the ratio of the high-frequency voltage divider to be changed and allows the control transistor 32 to trigger faster.
[0101] The switch 10 of [Fig.2] uses components that are easier to produce today than the components of the control circuit 22 of [Fig.1].
[0102] This results in a significant gain in compactness. Typically, compared to implementing a Zener diode 28 using a plurality of diodes, a 50% reduction in the The surface area occupied on the chip can be obtained.
[0103] Furthermore, by using a discharge transistor 32 instead of the Zener diode 28 and driving it with a resistive divider bridge (resistor bridge 36) and a capacitor 38, the high-frequency dynamics of the switch 10 are improved.
[0104] In each of the aforementioned embodiments, the switch 10 can be advantageously used in an electronic device, and in particular a power electronic device.
[0105] An example of an electronic device 50 is shown in [Fig.3].
[0106] The electronic device 50 is here a converter used conventionally in power electronics.
[0107] In the example, the electronic device 50 includes a voltage source 52 supplying a bridge arm 54.
[0108] The bridge arm 54 is formed by a transistor 56 and the switch 10 in series.
[0109] Alternatively, it is possible that the arm at point 54 first includes the com mutator 10 then transistor 56.
Claims
Claims
1. Switch (10) comprising: - a transistor to be protected (12), the transistor to be protected (12) comprising a gate (12G), a drain (12D) and a source (12S), and - a protection unit (14) against the Miller effect, the protection unit (14) comprising: - a protection transistor (18), the protection transistor (18) comprising a gate (18G), a drain (18D) and a source (18S), the drain (18D) of the protection transistor (18) being connected to the gate (12G) of the transistor to be protected (12), the source (18S) of the protection transistor (18) being connected to the source (12S) of the transistor to be protected (12), - a connection circuit (20), the connection circuit (20) being a high-pass filter for the leakage current of the transistor to be protected (12), the connection circuit (20) being arranged between the gate (18G) of the transistor protection (18) and the drain (12D) of the transistor to be protected (12), and - a control circuit (22),the control circuit (22) being interposed between the gate (18G) of the protection transistor (18) and the source (12S) of the transistor to be protected (12). the control circuit (22) being formed by a discharge transistor (32) and a control unit (34) for the discharge transistor (32), the control unit (34) comprising a resistance bridge (36) comprising a midpoint (44), a first resistor (40) and a second resistor (42), the control unit (34) also comprising a capacitor (38) in parallel with the first resistor (40), the midpoint (44) being connected to the gate (32G) of the discharge transistor (32).,
2. A switch according to claim 1, wherein the connecting circuit (20) is a capacitor (26).
3. Switch according to claim 1 or 2, in which the transistor to be protected (12) is made of gallium nitride or silicon carbide.
4. Switch according to any one of claims 1 to 3, in which the protection unit (14) and the transistor to be protected (12) are produced on the same chip.
5. Switch according to any one of claims 1 to 4, in which each transistor (12, 18) has a dimension, the ratio between the dimension of the protection transistor (18) and the dimension of the transistor to be protected (12) being less than or equal to 0.
1.
6. Electronic device (50) comprising a switch (10) according to
7. any one of claims 1 to 5. The electronic device of claim 6, wherein the electronic device (50) further comprises a voltage source (52) and a bridge arm (54) formed by a transistor (56) and the switch (10).