Electrical switching module and switching device comprising such a module

Abstract

The present invention relates to an electric switch module (1) comprising a first electrode and a second electrode forming the ends of an electrically conductive path and at least one control electrode for selectively allowing or preventing the passage of an electric current through the electrically conductive path. The electrodes of the switch module (1) are arranged on a main face (1a) of a housing (2) formed of an electrically insulating resin which completely encloses two transistors formed on and / or in at least one chip and arranged back-to-back and in series in the electrically conductive path, and which encloses electrically conductive tracks (P) electrically associating the electrodes and the transistors.

Description

[0001] This invention relates to the field of microelectronic systems with very low power loss, extremely fast switching, and operation at high voltages (preferably above 600V). These electronic systems can be used in grid protection applications to instantly interrupt the flow of AC or DC current, such as in networks powering computer servers or in “microgrid” networks connecting renewable energy sources (such as solar panels). These electronic systems are also used for battery protection. Background Technology

[0002] Such protection systems use integrated circuit-based switching components electrically arranged in the current path of the electrical network to be protected. These integrated circuits are controlled to change state very quickly (e.g., from an on state to an off state) to respond to network faults and prevent damage to connected devices. The desired switching time is on the order of approximately one microsecond, or even less.

[0003] The term "fault" refers to any condition that results in a significant and excessive increase in current in the network ("overcurrent" or short circuit). It can also refer to overvoltage, excessive temperature rise, or any other anomaly that may damage network equipment or the network itself.

[0004] The protection system is associated with or equipped with sensors, such as current sensors and / or voltage sensors and / or temperature sensors, depending on the type of fault being monitored. The control components of the protection system are associated with the sensors and switching components to enable their operation.

[0005] Such protection systems use integrated circuit-based switching components whose losses (switching and / or conduction) are dissipated as heat, which may require a large cooling system depending on the level of loss to be dissipated.

[0006] Using integrated circuits based on so-called “wide bandgap” materials (such as SiC or GaN) or “ultra-wide bandgap” materials (such as diamond, AlN, or gallium oxide) makes it possible to significantly reduce the on-state resistance of these circuits (and thus reduce conduction losses) while withstanding target voltages (above 600V) and providing smaller dimensions compared to more conventional integrated circuits (such as those based on silicon).

[0007] Typically, protection systems use multiple switching components, each comprising an integrated circuit housed in a housing; these switching components are arranged electrically in parallel on a printed circuit board. This parallel arrangement allows for a reduction in the current flowing through each component, which also tends to reduce their on-state resistance, and thus reduces conduction losses and the need for heat dissipation in the system.

[0008] However, implementing such a parallel arrangement is complex because it requires the use of switching components with very similar (ideally identical) electrical characteristics, particularly regarding the threshold voltage (Vth) and on-state resistance (Ron). It also necessitates the design of highly precise printed circuit boards that seek to minimize parasitic capacitance and / or inductance in the electrical signal wiring traces and to position the control components as close as possible to the switching components to minimize the length of the control loop (referred to in the technical terminology of the art as the "drive loop"), and thereby minimize the required switching time of the system.

[0009] The present invention aims to solve these problems. Summary of the Invention

[0010] To achieve one of these objectives, the present invention provides an electrical switching module comprising a plurality of electrodes, particularly: a. The first and second electrodes forming the ends of the conductive path; b. At least one control electrode, which is used to selectively allow current to pass through a conductive path or to prevent current from passing through a conductive path.

[0011] Multiple electrodes of the switching module are arranged on the main surface of the housing. According to the invention, the housing is formed of an electrically insulating resin that completely encapsulates two transistors formed on and / or in at least one chip, arranged back-to-back and in series in a conductive path. The resin also encapsulates conductive traces that electrically associate the multiple electrodes and transistors.

[0012] Other advantageous non-limiting features according to the invention (alone or in any technically feasible combination): - At least one chip includes a surface layer of material having a wide bandgap, and at least one of two transistors is formed on and within the surface layer; - The electrical path between the first electrode and the second electrode consists of multiple parallel paths, each of which includes two transistors arranged back-to-back and in series. - The housing has a secondary surface opposite the main surface, which has at least one thermally conductive pad; - Both transistors have a vertical architecture; Both transistors have a lateral architecture.

[0013] According to another aspect, the present invention provides a switching device comprising a printed circuit board and a switching module as described above mounted on the board.

[0014] Other advantageous non-limiting features of this aspect of the invention (alone or in any technically feasible combination): - The switching device includes at least one sensor, which is electrically connected to the control component; - The switching device includes a current sensor arranged to measure the current flowing through the conductive path; - Current sensors and / or control components are integrated into the electrical switch module. Attached Figure Description

[0015] Other features and advantages will become apparent from the following detailed description of the invention with reference to the accompanying drawings, in which: [ Figure 1 ] [ Figure 1 The first embodiment is described; [ Figure 2 ] [ Figure 2 The second embodiment is described; [ Figure 3 ] [ Figure 3 This illustrates an example of two transistors in a "back-to-back" configuration; [ Figure 4 ] [ Figure 4 A switching device utilizing a switching module according to the present invention is shown; [ Figure 5 ] [ Figure 5 The diagram shows a switch module according to another embodiment. Detailed Implementation

[0016] Two implementation methods are illustrated by reference. Figure 1 and Figure 2 The diagram shows a switch module 1 according to one aspect of the present invention.

[0017] The switch module 1 takes the form of a housing 2 in the shape of a roughly parallelepiped, which has multiple electrodes T1 to T4 that allow it to be electrically connected to other components of the protection system.

[0018] Specifically, on surface 1a, referred to as the "main" surface of the housing, there are a first electrode T1 and a second electrode T2, which form the ends of a conductive path passing through module 1. This path is designed to carry current that supplies power to one or more devices that are to be protected from failure, as explained in the description of this application. For this purpose, the switch module 1 includes a control electrode T3, also on the main surface 1a of the housing 2, designed to receive a control signal that allows the module to be configured to selectively allow or block the flow of current in the conductive path. By arranging the electrodes of the switch module, and particularly the first electrode, the second electrode, and the control electrode, on a single surface of the housing 2, it is easier to directly mount the module onto a printed circuit board using automated manufacturing techniques.

[0019] The housing 2 of the switching module 1 has a secondary surface 1b opposite the primary surface 1a, which has at least one thermally conductive pad Tt1, Tt2. The function of the at least one pad is to dissipate heat generated by the transistors integrated into the module 1, as will be explained in more detail later in this description. These pads can be made of metal, such as copper, which is a good thermal conductor. The thermally conductive pads T1, T2 can be associated with other heat dissipation elements, such as heat sinks mounted on the switching module 1, which are in thermal contact with the pads.

[0020] The switching function is implemented by two transistors integrated into housing 2, which are arranged back-to-back and in series within the conductive path. The first transistor C1 and the second transistor C2 each have sources S1 and S2, drains D1 and D2, and gates G1 and G2. A channel extends between the source and drain of each transistor.

[0021] "Back-to-back" means that the first transistor C1 and the second transistor C2 are electrically connected via their sources S1 and S2 or their drains D1 and D2. Figure 3 An example of such an arrangement is shown in the figure, where the source poles S1 and S2 are directly connected to each other.

[0022] Transistors C1 and C2 are connected in series in the conductive path; that is, the channels of these transistors C1 and C2 extending between the source S1 and S2 and the drain D1 and D2 are arranged in the path between the first electrode T1 and the second electrode T2 of the module.

[0023] This configuration allows current to flow through the path in both directions, and the current can be selectively blocked or allowed to flow by applying appropriate control signals to gates G1 and G2.

[0024] Transistors C1 and C2 can be manufactured on a single semiconductor chip, or on two chips P1 and P2, respectively. Figure 1 and Figure 2The situation is the same as in the implementation described in the text.

[0025] Preferably, the semiconductor chips P1 and P2 of module 1 consist of a relatively thick substrate, at least 30 micrometers thick and typically in the range of 30 to 300 micrometers, made of a material with high thermal conductivity, such as SiC, AlN, or diamond. The substrate material can be single-crystal or polycrystalline. The substrate supports a relatively thin crystalline surface layer, preferably a single-crystal surface layer, with a thickness between 2 and 10 micrometers, made of a wide-bandgap material, such as selected from SiC, GaN, AlN, Ga2O3, or diamond. When the substrate is single-crystal, the surface layer may constitute the surface thickness of the substrate, or the surface layer may be formed by deposition onto the substrate or by transfer. In some cases, the semiconductor chips P1 and P2 may also consist of only a surface layer, without the need for a substrate to support them. For example, this may involve removing or disassembling the substrate after the transistors C1 and C2 have been manufactured and before they have been integrated into the housing 2 of module 1. Depending on the architecture suitable for the application and the materials chosen, at least one of transistors C1 and C2 is formed on and / or within the surface layer of semiconductor chips P1 and P2. These transistors can be configured to be "normally on" or "normally off".

[0026] Using wide bandgap materials (i.e., materials with a bandgap threshold greater than 3 eV) to form transistors here particularly allows for very low on-state resistance, small semiconductor chip size, and withstand voltages exceeding 600V or even 1,200V, as well as very fast switching speeds.

[0027] The source pads, drain pads, and gate pads of chips P1 and P2, which are respectively associated with the source S1 and S2, drain D1 and D2, and gate G1 and G2 of transistors C1 and C2, are preferably formed of aluminum-based metal and have a thickness of at least one micrometer.

[0028] exist[ Figure 1 In the embodiment shown, transistors C1 and C2 have a vertical configuration. In other words, conductive paths run between the front and back sides of at least one semiconductor chip P1, P2 on which transistors are fabricated. Voltage withstand is ensured by applying an electric field between the source and drain pads (associated with the drains D1, D2 and sources S1, S2 of the transistors) arranged on opposite surfaces of the semiconductor chip. Gate pads are arranged on the surface supporting the source pads.

[0029] In this case, transistors C1 and C2 may correspond to transistors formed on a substrate with a JFET, MOSFET, FET, superjunction, or FinFET architecture. The substrate may be bulk and composed of SiC 4H, GaN, AlN, diamond, or even Ga2O3. Alternatively, it may be a composite substrate including a thin layer of SiC 4H transferred onto a SiC substrate (such as a polycrystalline substrate). Alternatively, these transistors C1 and C2 may be formed on and within a surface layer of GaN, AlN, Ga2O3, or diamond transferred onto a single-crystal or polycrystalline SiC substrate or a polycrystalline AlN substrate.

[0030] In the case of such vertical architectures, it can be advantageous for the substrate forming part of the semiconductor chip to have a relatively thin thickness (due to resistance and thermal resistance), less than 150 micrometers, or even less than 100 micrometers or 80 micrometers.

[0031] exist[ Figure 2 In the embodiment shown, transistors C1 and C2 have a lateral architecture. In other words, the conductive path travels along the surface of semiconductor chips P1 and P2 without crossing their thickness. Voltage withstand is ensured by applying an electric field between the source pads and drain pads (associated with the drains D1 and D2 and sources S1 and S2 of the transistors) arranged on the same side of the chip (referred to as the "front side"). The gate pad is arranged on this same side. The back side of the semiconductor chip, opposite the front side, may optionally be electrically biased by connecting it to the source or drain pads to limit certain transient effects, such as the trapping of electrons and / or holes.

[0032] In this case, transistors C1 and C2 may correspond to transistors with a HEMT architecture. Specifically, these transistors may be formed on and within a GaN surface layer disposed on a silicon substrate, or formed on a bulk SiC substrate. Transistors may also be formed on a composite substrate obtained by transferring the surface layer onto a bulk substrate. For example, this could be a transistor with a lateral architecture, formed in a surface layer of AlN, Ga2O3, or diamond transferred onto a single-crystal or polycrystalline SiC substrate or a polycrystalline AlN substrate.

[0033] Return to Figure 1 and Figure 2 In general, conductive traces P, made of copper, are advantageously used for thermal and electrical conductivity reasons to electrically connect multiple electrodes to the pads of transistors C1 and C2.

[0034] Advantageously, one or more chips and conductive traces are integrated into a multilayer or 3D printed circuit board. This integration can be achieved using techniques for integrating chips into printed circuit boards (according to the terminology used in the art, "embedded substrate packaging" or "embedded PCB technology"). According to this method, instead of simply mounting the chips on the surface of a carrier and connecting their bonding pads to electrodes via connecting wires, as is typically the case, the chips are "embedded" or "buried" within the thickness of the printed circuit board itself. Therefore, preferably, the chips P1 and P2 of the switching module and the conductive traces P that electrically associate electrodes T1 to T4 with transistors C1 and C2 are completely encapsulated in an electrically insulating resin.

[0035] To achieve this, and according to certain methods, a stack of resin layers or insulating sheets (such as fiber resin prepreg) can be placed on a chip that has already been mounted on a carrier. A metal layer is then formed in a plane parallel to the carrier, and this metal layer is etched according to a defined pattern to form conductive traces. Vias are also typically created by laser or mechanical drilling, and these vias are then filled with metal (in most cases, copper). These vias are arranged perpendicular to the carrier, allowing electrodes, chip pads, and traces arranged in different planes and parallel to the carrier to be interconnected.

[0036] Therefore, the conductive trace P of the switching module 1 according to the invention is advantageously formed by a portion disposed on a carrier above a carrier on which one or more chips are disposed, and a portion extending perpendicularly to the carrier (i.e., vias), which provide electrical conduction between the two layers.

[0037] This architecture allows for the formation of very compact modules, minimizing the length of the conductive trace P, which in turn minimizes parasitic factors (resistance, inductance, capacitance). For example, the parasitic resistance of the switching module described herein can be less than 1 milliohm, and its parasitic inductance can be approximately a few nanohenries or less. Furthermore, eliminating the need for connecting wires limits the risk of mechanical failure and also ensures greater reliability.

[0038] In the configuration shown, the first electrode T1 and the second electrode T2 are connected via conductive trace P to two drain pads associated with the drains D1 and D2 of transistors C1 and C2, respectively. The control electrode T3 is then electrically connected to two gate pads associated with the gates G1 and G2 of transistors C1 and C2.

[0039] To facilitate electrical connections between the conductive traces P, which are preferably made of copper, and the pads associated with transistors C1 and C2, copper layers can be formed on the pads before they are integrated into the housing 2. This metallization (typically in the range of 5 μm to 10 μm) can be achieved by electrolytic deposition on the front and back sides of one or more semiconductor chips P1 and P2.

[0040] Advantageously, to facilitate heat dissipation, the conductive trace P also integrates one or more semiconductor chips P1, P2 integrated into the housing 2 with their back surfaces electrically connected to heat-conducting electrodes Tt1, Tt2. [Regarding...] Figure 1 In the vertical architecture shown, due to the back side of the chip where the drains of transistors T1 and T2 are formed and exhibit a significant voltage difference during operation, the secondary surface 1b of the housing 2 is provided with two thermally conductive electrodes Tt1 and Tt2, which are separated from each other and electrically isolated. Separate electrically isolated conductive traces P also connect the two drain pads to the two thermally conductive electrodes Tt1 and Tt2 respectively. Figure 2 In the horizontal architecture shown, a single thermally conductive electrode Tt1 arranged on the secondary surface 1b of the housing 2 is connected to the back sides of two semiconductor chips via conductive trace P, which are biased to the midpoint voltage (typically zero voltage) of two transistors C1 and C2 connected in series.

[0041] Therefore, it should be understood that the conductive traces P serve a dual function of providing electrical connection and dissipating heat. Consequently, they are designed to achieve this dual function as efficiently as possible.

[0042] In an alternative configuration, the switch module 1 may have two control electrodes, each connected via a separate trace to the gate pads of two transistors C1 and C2. In this case, it is necessary to generate two control signals, each applied independently to one of the two control electrodes, to control the blocking or turning operation of the switch module 1. Figure 1 and Figure 2 Similar to the case in [the previous example], switch module 1 may also have additional electrodes, such as a midpoint electrode T4, which is electrically connected to the series junction of the two transistors C1 and C2; in this case, their sources. The midpoint electrode T4 allows a specific voltage, such as zero voltage, to be applied to the midpoint. It may also be at least one Kelvin-type electrode, connected via a Kelvin connection to the sources S1 and S2 or the drains D1 and D2 of transistors C1 and C2; this type of electrode may be needed when the module operates at high current.

[0043] As is well known to those skilled in the art, the Kelvin connection of housing 2 is designed for small signals typically originating from the source. Unlike a direct connection to the source carrying the main current, this method utilizes a small signal to attenuate induced voltage fluctuations in the control signal during current switching. Because the device switches rapidly, this induced voltage drop is sufficient to turn it on or off.

[0044] One or more electrodes may be provided to allow measurement of the junction temperature of one or more transistors C1 and C2.

[0045] In any case, regardless of the number of electrodes in the switch module 1, these are arranged on the main surface 1a of the module housing 2.

[0046] To complete the description of switch module 1, it is therefore presented in the form of a housing 2 made of insulating resin that completely encapsulates the semiconductor chips P1 and P2 that make up transistors C1 and C2, as well as the conductive traces, as described above. More generally, the insulating resin forming housing 2 encapsulates all components integrated into module 1.

[0047] According to a particularly advantageous variant that reduces conduction losses, the electrical path between the first electrode T1 and the second electrode T2 of the switching module 1 comprises multiple paths connected in parallel. As described above, each parallel path comprises two transistors arranged back-to-back and in series. These transistors can be formed on a single semiconductor chip or multiple such chips, as in the embodiment just described.

[0048] In all cases, to meet the requirements of the target application, the aim is to obtain a switching module 1 with an on-state resistance of less than 5 milliohms (typically in the range of 2 to 3 milliohms and ideally around 1 milliohm). It can withstand voltages greater than 600V and up to 2,000V.

[0049] [ Figure 4The diagram illustrates a switching device D utilizing the module 1 just described. Device D includes a printed circuit board 4, which has wiring traces in a very conventional manner for electrically and functionally interconnecting various components. A switching module 1, through which a conductive path I passes and including two transistors C1, C2, is positioned on the board 4. A current sensor K, associated with a control unit 3, is also arranged on the board 4, intersecting with the path I. The current sensor and control unit enable the detection of excessive current exceeding a predetermined threshold. The current sensor K generates a detection signal for the current flowing through path I; this detection signal is transmitted to the control unit 3. The control unit is configured to generate a control signal at the control electrode T3 of the switching module 1; during normal operation, this signal corresponds to the on state of the module. However, when the detection signal generated by the current sensor K and transmitted to the control unit 3 indicates excessive current in the current path I, the control unit 3 is then configured to generate a control signal at the control electrode T3 of the switching module 1, switching it to an off state.

[0050] exist[ Figure 4 The diagram shows an initialization signal (INIT) that can be connected, for example, to a push button or monitoring unit connected to control component 3 to reset the component and, for example, allow switch module 1 to switch back to the ON state and restore normal operation of device D.

[0051] In this diagram, a temperature sensor T is also provided; it is connected to the control unit and operates similarly to a current sensor K, causing the switching module 1 to switch when the temperature exceeds a predetermined threshold. Alternatively, the temperature sensor may not be mounted on the printed circuit board 4, but rather integrated into the device being monitored.

[0052] For clarity, [ Figure 4 The illustration in the figure omits auxiliary components of device D, such as those used to supply power to control unit 3 and sensors K and T, or those used to connect board 4 to the rest of the system.

[0053] Of course, it is conceivable that at least some of the components mounted on the printed circuit board 4 can be directly integrated into the switch module 1. To illustrate this method, [ Figure 5 This illustrates a switching module 1, which, in addition to transistors C1 and C2, also incorporates a current sensor K and a control circuit 3 within a housing, thereby replicating […]. Figure 4 The function of control component 3 is shown in the figure.

[0054] Of course, the present invention is not limited to the described embodiments, and alternative embodiments may be provided without departing from the scope of the invention as defined by the claims.

[0055] In particular, the concepts stated in this description can be extended to a simple series arrangement of two transistors, without having to arrange them back to back. In this way, a half-bridge circuit can be formed, and the transistors can then be controlled by independent control signals applied to their respective gate electrodes.

Claims

1. An electrical switch module (1), the electrical switch module comprising a plurality of electrodes (T1 to T4), particularly: a. The first electrode (T1) and the second electrode (T2) at the ends of the conductive path; b. At least one control electrode (T3), said at least one control electrode being used to selectively allow current to pass through the conductive path or to prevent current from passing through the conductive path; The plurality of electrodes (T1 to T4) of the switching module (1) are arranged on the main surface (1a) of the housing (2), which is formed of an electrically insulating resin that completely covers two transistors (C1, C2) formed on and / or in at least one chip (P1, P2) and arranged back-to-back and in series in the conductive path, and the electrically insulating resin covers the conductive traces (P) that electrically associate the plurality of electrodes (T1 to T4) and the transistors (C1, C2).

2. The electrical switch module (1) according to the preceding claim, wherein the at least one chip (P1, P2) includes a surface layer of a material having a wide bandgap, and at least one of the two transistors (C1, C2) is formed on and within the surface layer.

3. The electrical switch module (1) according to any one of the preceding claims, wherein the electrical path between the first electrode (T1) and the second electrode (T2) consists of a plurality of parallel paths, each parallel path comprising two transistors arranged back-to-back and in series.

4. The electrical switch module (1) according to any one of the preceding claims, wherein the housing (2) has a secondary surface (1b) opposite to the main surface (1a), the secondary surface having at least one thermal conductive pad (Tt1, Tt2).

5. The electrical switch module (1) according to any one of the preceding claims, wherein the two transistors (C1, C2) have a vertical structure.

6. The electrical switch module (1) according to any one of claims 1 to 4, wherein the two transistors (C1, C2) have a lateral architecture.

7. A switching device (D) comprising a printed circuit board (4) and a switching module (1) according to any one of the preceding claims mounted on the board.

8. The switching device (D) according to the preceding claim, the switching device comprising at least one sensor (K, T) electrically connected to the control component (3).

9. The switching device (D) according to the preceding claim, the switching device comprising a current sensor arranged to measure the current flowing through the conductive path.

10. The switching device (D) according to the preceding claim, wherein the current sensor and / or the control component (3) is integrated into the electrical switching module (1).

Images