WO2023232648A1

WO2023232648A1 - Power electronic converters and power and propulsion systems comprising power electronics converters

Info

Publication number: WO2023232648A1
Application number: PCT/EP2023/064118
Authority: WO
Inventors: Stanley Buchert; Uwe Waltrich; Marco BOHLLÄNDER; Claus Müller
Original assignee: Rolls-Royce Deutschland Ltd & Co Kg
Priority date: 2022-05-31
Filing date: 2023-05-25
Publication date: 2023-12-07
Also published as: DE102022205501A1

Abstract

A power electronics converter is disclosed. The power electronics converter comprises a converter commutation cell comprising a power circuit and a gate driver circuit, the power circuit comprising at least one power semiconductor switching element and at least one capacitor. Each power semiconductor switching element has at least three terminals including a gate terminal. The gate driver circuit is electrically connected to and configured to provide switching signals to the gate terminal of each of the at least one power semiconductor switching elements. The power electronics converter may be an AC-DC power electronics converter (i.e., an inverter or a rectifier) or a DC-DC power electronics converter

Description

POWER ELECTRONIC CONVERTERS AND POWER AND PROPULSION SYSTEMS COMPRISING POWER ELECTRONICS CONVERTERS

TECHNICAL FIELD

This disclosure relates to power electronics converters, and particularly but not exclusively to power electronics converters comprising power semiconductor prepackages for use in the power and propulsion systems of aircraft.

BACKGROUND

In aerospace, aircraft and their power and propulsion systems are becoming increasingly electrical in their design. Some proposed platforms are purely electric, relying completely on batteries or fuel cells for all their power and propulsion requirements. Other proposed platforms are of the hybrid-electric type, and others still are ‘more electric’ in that they derive most or all their propulsive power from on-board engines (e.g., gas turbine engines) but have an increased number of electrically powered aircraft and engine systems, sub-systems and accessories.

The electrical power systems that feature in these platforms include power electronics converters. AC-DC converters (inverters and rectifiers) convert between AC and DC, for example to deliver AC to an electrical machine configured as a motor from a DC power source (e.g., a battery or DC power channel), or to deliver DC power from an electrical machine configured as a generator to a DC power channel or rechargeable battery. DC-DC converters may be used to regulate the DC voltage delivered from a battery to a DC power channel, for example. The electrical power systems may also include other power electronics devices, for example protection devices such as solid- state power controllers (SSPCs) and solid-state circuit breakers, some of which may be incorporated into the converters themselves.

So-called power modules, or power electronics modules, are the dominant state of the art technology for power electronics converters. In a converter including a power module, the components of the converter circuit - which include power semiconductor devices including transistors and diodes and smoothing DC-link or input capacitor(s) - are fixed to a carrier substrate and electrically connected to each other, typically using wire bonds. Power module-based converters are widely used in, for example, the automotive industry and are used in existing aerospace systems, where they provide generally acceptable performance with average operating efficiencies as high as about 95%. The performance of converters based on the power module topology is acceptable for most applications, especially when due account is taken of their relatively low cost and high availability. From the point of view of aerospace applications, however, improving the efficiency and power-to-weight ratio of power electronics converters would be advantageous. Compared with ground-based applications including automotive applications, aerospace applications are highly sensitive to weight. For purely electric aircraft applications especially, even a relatively small increase in converter efficiency could provide meaningful improvements in aircraft performance and mission range.

Improvements in the performance of converters based on the existing power module topologies are likely to be limited to what can be achieved through advances in the underlying semiconductor technologies. This is at least in part because of the inherently high parasitic inductance of the commutation cell of a power module, much of which is introduced by the electrical connections between the commutation cell components. Parasitic inductance in the commutation cell is associated with transistor switching losses and voltage overshoot during turn-off, which not only limits efficiency and generates heat which must be removed but also limits other performance characteristics including the transistor switching frequency.

SUMMARY

A power electronics converter is disclosed. The power electronics converter comprises a converter commutation cell comprising a power circuit and a gate driver circuit, the power circuit comprising at least one power semiconductor switching element and at least one capacitor. Each power semiconductor switching element has at least three terminals including a gate terminal. The gate driver circuit is electrically connected to and configured to provide switching signals to the gate terminal of each of the at least one power semiconductor switching elements. The power electronics converter may be an AC-DC power electronics converter (i.e. , an inverter or a rectifier) or a DC-DC power electronics converter.

An electrical power system comprising an electrical machine and an AC-DC power electronics converter is also disclosed. The electrical machine comprises one or more windings. The AC-DC power electronics converter comprises a commutation cell comprising a power circuit and a gate driver circuit. The power circuit comprises a plurality of power semiconductor switching elements and at least one capacitor. An AC- side connection of the power circuit is connected to the one or more windings of the electrical machine to supply current to or receive current from the electrical machine. The electrical machine may be a motor or a generator, and the AC-DC power electronics converter may be an inverter or a rectifier. The electrical machine may be a motorgenerator and the AC-DC converter may be a bi-directional converter operable as an inverter or a rectifier.

The following metric prefixes are used herein to abbreviate numerical values:

Table 1

Any of the following features below may be applied singularly or in combination with each other and with the power electronics converter and the electrical power system set out above.

Each power semiconductor switching element may be a transistor. Each transistor may be a MOSFET having at least the gate terminal, a source terminal and a drain terminal. The MOSFETs may be Silicon Carbide (SiC) MOSFETs. In other examples the MOSFETS are Gallium Nitride (GaN) MOSFETs.

Each power semiconductor switching element may be comprised in a power semiconductor prepackage. Each power semiconductor prepackage comprises one or more power semiconductor switching elements embedded in a solid insulating material. Each power semiconductor prepackage may comprise precisely one power semiconductor switching element.

The power electronics converter may comprise one or more power semiconductor logical switches each comprising one or more parallel-connected power semiconductor switching elements. Each power semiconductor logical switch may comprise one or more power semiconductor prepackages, each power semiconductor prepackage comprising at least one (and optionally precisely one) power semiconductor switching element. The number of power semiconductor switching elements per logical switch may be greater than or equal to three. The number may be in the range of three to twelve.

A peak rated power output of the power electronics converter may be greater than 10 kW (equivalently kVA) and may be greater than 25 kW. The peak rated power may be greater than or equal to 40 kW, or greater than or equal to 50 kW. The peak rated power may be less than or equal to 500 kW. The peak rated power output may be less than or equal to 400 kW, or less than or equal to 300kW. The peak rated power may be in the range 50 kW to 300 kW.

A maximum efficiency of the power electronics converter may be greater than 97%. The maximum efficiency may be greater than 97.5%, greater than 98%, greater than 98.5% or even greater than 99%.

A value of a converter parameter a may be less than or equal to 5 pHm³. a is a product of the smallest cuboidal volume that encloses the commutation cell and a parasitic inductance of the power circuit of the commutation cell, a may be greater than or equal to 0.3 pHm³. a may be is less than or equal to 4 pHm³. a may satisfy 0.4 pHm³

< a < 3.5 pHm³. a may satisfy 0.5 pHm³ < a < 2.5 pHm³.

The value of a divided by the peak rated power output of the power electronics converter may be greater than or equal to 1.5 aHm³/W. The value of a divided by the peak rated power output of the power electronics converter may be less than or equal to 100 aHm³/W. The value of a divided by the peak rated power output of the power electronics may be in the range 2.5 aHm³/Wto 50 aHm³/W.

A product of the parasitic inductance of the power circuit of the commutation cell and the peak rated power output may be in the range 0.05 mHW to 1.5 mHW. The product may be in the range 0.1 mHW to 1.2 mHW, or in the range 0.2 mHW to 1.0 mHW.

A value of a parameter p may be greater than or equal to 0.3 PV/s². p is a product of a maximum switching frequency of the switching signals and a maximum rate of change of a source-drain voltage of the plurality of power semiconductor switching elements during operation. The value of p may be less than or equal to 10 PV/s². The value of p may be greater than or equal to 0.5 PV/s². The value of p may satisfy 0.8 PV/s²

< P < 5 PV/s² or may satisfy 1.0 PV/s² < p < 2.5 PV/s². A value of a converter parameter y may be less than or equal to 150 fFs/W. y is a total rated capacitance of the at least one capacitor of the power circuit divided by a product of the peak rated power output of the power electronics converter and a maximum switching frequency of the switching signals. The value of y may be greater than or equal to 1 .0 fFs/W. The value of y may be less than or equal to 100 fFs/W, less than or equal 75 fFs/W or less than or equal to 50 fFs/W. The value of y may satisfy 2.0 fFs/W < y < 50 fFs/W or may satisfy 4.0 fFs/W < y 25 fFs/W.

A value of a converter parameter δ may be greater than or equal to 0.5 PV/FH. δ is a maximum blocking voltage of the one or more power semiconductor switching elements of the power circuit divided by a product of a parasitic inductance of the power circuit of the commutation cell and a total rated capacitance of the at least one capacitor of the power circuit. The value of δ may be less than or equal to 40 PV/FH. The value of δ may be greater than or equal to 1.5 PV/FH. The value of 5 may be in the range 2.5 PV/FH to 25 PV/FH. The value of δ may be in the range 4.0 PV/FH to 15 PV/FH.

A value of a converter parameter E may be greater than or equal to 10²⁶ V/s⁴. E is equal to:

In this equation, f_max is a maximum switching frequency of the switching signals, |dv/dt|max is a maximum rate of change of a source-drain voltage of the one or more power semiconductor switching elements during operation, L is a parasitic inductance of the power circuit of the commutation cell and C is a total rated capacitance of the at least one capacitor of the power circuit. The value of E may be less than or equal to 10²⁹ V/s⁴. The value of E may be greater than or equal to 5x10²⁶ V/s⁴. The value of E may be in the range 10²⁷ V/s⁴ to 5x10²⁸ V/s⁴ or may be in the range 1.5x10²⁷ V/s⁴ to 3x10²⁸ V/s⁴.

The parasitic inductance of the power circuit of the commutation cell may be less than or equal to 16 nH, less than or equal to 10nH, less than or equal to 8 nH, less than or equal to 6 nH, less than or equal to 4 nH, less than or equal to 3 nH or even less than or equal to 2 nH. The parasitic inductance may be in the range 2 nH to 8nH.

The total rated capacitance of the power circuit of the commutation cell divided by the peak rated power may be less than or equal to 5 nF/W, and may be less than or equal to 3 nF/W. The total rated capacitance of the power circuit of the commutation cell divided by the peak rated power may be in the range 0.1 nF/W to 2.5 nF/W. The smallest cuboidal volume that encloses the commutation cell may be less than or equal to 1 ,000 cm³. The smallest cuboidal volume may be less than or equal to 900 cm³, less than or equal to 800 cm³, less than or equal to 700 cm³ or less than or equal to 600 cm³. The smallest cuboidal volume may be in the range 100 cm³ to 800 cm³, 100 cm³ to 700 cm³, 100 cm³ to 600 cm³, 150 cm³ to 600 cm³, or 200 cm³ to 450 cm³.

A maximum rate of change of a source-drain voltage of the one or more power semiconductor switching elements during operation may be greater than or equal to 10 kV/ps, greater than or equal to 15 kV/ps, or greater than or equal to 20 kV/ps. The maximum rate of change may be less than 150 kV/ps, less than 120 kV/ps, less than 100 kV/ps, less than 90 kV/ps, or less than 80 kV/ps. The maximum rate of change may be in the range 10 kV/ps to 120 kV/ps, in the range 15 kV/ps to 100 kV/ps, in the range 20 kV/ps to 90 kV/ps, or in the range 25 kV/ps to 80 kV/ps. The maximum rate of change may be in the range 10 kV/ps to 60 kV/ps, in the range 15 kV/ps to 50 kV/ps, in the range 20 kV/ps to 50 kV/ps, in the range 25 kV/ps to 50 kV/ps, or in the range 30 kV/ps to 50 kV/ps. The maximum rate of change may be in the range 30 kV/ps to 40 kV/ps.

The maximum switching frequency of the switching signals (f_max) may be greater than or equal to 10 kHz. It may be greater than or equal to 20 kHz, greater than or equal to 30 kHz, greater than or equal to 40 kHz or even greater than or equal to 50 kHz. The maximum switching frequency may be less than 100 kHz. It may be in the range 30 kHz to 70 KHz.

The blocking voltage (i.e., the ‘source-drain blocking voltage’, sometimes called the ‘rated voltage’) of each power semiconductor switching element may be greater than 600 V, greater than 700 V or greater than 800 V. The blocking voltage may be less than 1 ,800 V, or less than 1 ,700V. The blocking voltage may be in the range 800 V to 1 ,600 V, 900 V to 1 ,500 or 1 ,000 V to 1 ,400 V.

The power electronics converter may further comprise a multi-layer planar carrier substrate. The multi-layer planar carrier substrate may define an x-y direction parallel to a planar surface of the substrate and a z-direction perpendicular to the x-y direction. The carrier substrate may comprise a plurality of electrically conductive layers extending in the x-y direction and at least one electrical connection extending in the z-direction. The carrier substrate may comprise an outer conductive layer on one or both of its opposed planar surfaces. The multi-layer planar substrate may be a rigid printed circuit board (PCB). It may be a flexible PCB. It may be a ceramic carrier substrate. The multi-layer planar substrate may be a structural component of the converter.

Each power semiconductor prepackage may further comprise at least one electrical connection extending in the z-direction from at least one terminal of each of the one or more power semiconductor switching elements through the solid insulating material to an electrical connection side of the power semiconductor prepackage. At least one of the terminals of each of the one or more power semiconductor switching elements of the prepackage may be connected to at least one of the conductive layers of the multilayer planar carrier substrate at the electrical connection side of the power semiconductor prepackage.

The electrical connection side of the power semiconductor prepackage may be spaced apart in the z direction from the multi-layer planar carrier substrate so as to define a gap (referred to herein as the prepackage gap) between the multi-layer planar carrier substrate and the electrical connection side of the prepackage. A size in the z-direction of the prepackage gap may be less than or equal to 300pm. The size of the prepackage gap may be less than or equal to 250pm. The size of the prepackage gap may be less than or equal to 200pm. The size of the prepackage gap may be less than or equal to 150pm. The size of the prepackage gap may be greater than or equal to 10pm, or greater than or equal to 20pm, or greater than or equal to 50pm, or greater than or equal to 80pm. The prepackage gap may be in the range 20pm to 250pm, or in the range 50pm to 150pm.

A value of a converter parameter 0 may be less than or equal to 300 pm²/V. 0 is a size in the z-direction of the prepackage gap divided by a maximum electric field strength in the prepackage gap. Accordingly, the converter parameter 0 can be expressed as follows:

In this equation, Gi is the size of the prepackage gap in the z-direction, and Ei is the maximum electric field strength in the prepackage gap. 0 may be greater than or equal to 0.1 pm²/V. 0 may be less than or equal to 250 pm²/V. 0 may be in the range 2.0 pm²/V to 20 pm²/V, or in the range 3.0 pm²/V to 10 pm²/V. 0 may be in the range 0.5 pm²/V to 100 pm²/V, or in the range 2.0 pm²/V to 50 pm²/V.

The maximum electric field strength in the prepackage gap may be greater than or equal to 1 kV/mm, or greater than or equal to 3 kV/mm, or greater than or equal to 5 kV/mm, or greater than or equal to 10 kV/mm. The maximum electric field strength in the prepackage gap may be less than or equal to 50 kV/mm, less than or equal to 40 kV/mm, or less than or equal to 25 kV/mm. The maximum electric field strength in the prepackage gap may be in the range 5 kV/mm to 40 kV/mm, or in the range 10 kV/mm to 25 kV/mm. The maximum electric field strength in the prepackage gap may be in the range 3 kV/mm to 25 kV/mm.

The power electronics converter may further comprise a heat sink for removing heat from power semiconductor prepackages. The heat sink may be spaced apart in the z-direction from the multi-layer planar carrier substrate so as to define a gap (referred to herein as the heat sink gap) between the multi-layer planar carrier substrate and the heat sink. The size in the z-direction of the heat sink gap between the multi-layer planar carrier substrate and the heat sink may be less than or equal to 10 mm, or less than or equal to 5 mm, or less than or equal to 3 mm, or less than or equal to 2.5 mm, or less than or equal to 1 mm, or less than or equal to 0.3 mm. The size of the heat sink gap may be greater than or equal to 0.1 mm, or greater than or equal to 0.5 mm, or greater than or equal to 1 mm. The size of the heat sink gap may be within the range 0.5 mm to 3 mm, or in the range 0.5 mm to 2.5 mm, or in the range 0.5 mm to 2 mm, or in the range 1 mm to 2 mm, or in the range 1.3 mm to 1.7 mm.

A value of a converter parameter (p may be less than or equal to 20 nm²/V. (p is a size in the z-direction of the heat sink gap divided by the maximum electric field strength in the heat sink gap. Accordingly, the converter parameter (p can be expressed as follows:

In this equation, G2 is the size of the heat sink gap in the z-direction, and E2 is the maximum electric field strength in the heat sink gap. (p may be less than or equal to 15 nm²/V. (p may be greater than or equal to 0.01 nm²/V. (p may be in the range 0.25 nm²/V to 2.5 nm²/V, or in the range 0.5 nm²/V to 1.5 nm²/V. cp may be in the range 0.02 nm²/V to 10 nm²/V, or in the range 0.05 nm²/V to 5 nm²/V.

The maximum electric field strength in the heat sink gap may be greater than or equal to 0.1 kV/mm, or greater than or equal to 0.2 kV/mm, or greater than or equal to 1 kV/mm. The maximum electric field strength in the heat sink gap may be less than or equal to 20 kV/mm, or less than or equal to 15 kV/mm, or less than or equal to 10 kV/mm, or less than or equal to 5 kV/mm. The maximum electric field strength in the heat sink gap may be in the range 0.2 kV/mm to 10 kV/mm, or in the range 1 kV/mm to 2 kV/mm, or in the range 1.3 kV/mm to 1.7 kV/mm.

Those skilled in the art will appreciate that the maximum electric field strength in the prepackage gap and/or the heat sink gap may be determined by means of a mathematical and/or computational simulation method (e.g., Finite Element Analysis).

The maximum electric field strengths Ei, E2 are maximum homogenous field strength in the respective gaps, i.e., the maximum field strengths may exclude highly localized maxima such as singularities that occur at or near sharp edges. The maximum electric field strengths may be determined at a point or in a region in the respective gap, wherein said point or region is located away in the x-y direction from a singularity, for example an edge or boundary, of the multi-layer planar carrier substrate, power semiconductor prepackage and/or heat sink. The distance in the x-y direction between said point or region and the singularity, for example the edge of the multi-layer planar carrier substrate, the power semiconductor prepackage and/or the heat sink, may be defined by a number of mesh cells of the simulation method, e.g., three mesh cells. By determining the maximum electric field strength in said point or region, the maximum electric field strength is determined in a region of the electric field that is relatively homogenous.

Generally, the maximum electrical field strengths may be determined in a region between two opposed and substantially parallel surface regions.

Where a prepackage gap is present, at least a portion of the prepackage gap may be filled with an electrically insulating material. The electrically insulating material, which may be a resin (e.g., a dielectric resin or polymer resin) or another suitable insulating material, may have a plurality of voids within the region (volume) it occupies.

A converter parameter a may be greater than or equal to 10/mm. a is defined as an insulation fill factor of the electrically insulating material divided by a maximum void size of the plurality of voids. Accordingly, the a can be expressed as follows:

In this equation, F is the insulation fill factor and R_max is the maximum void size of the plurality of voids. The insulation fill factor is defined as a cumulated volume of the plurality of voids (‘void volume’), subtracted from a volume of the electrically insulating material, divided by the volume of the electrically insulating material. This may be expressed as follows:

In this equation, VIM is the volume of the electrically insulating material and Vv is the cumulated volume of the plurality of voids. Hence, the converter parameter a may also be expressed as:

a may be greater than or equal to 15/mm, or greater than or equal to 18/mm, or greater than or equal to 50/mm, or greater than or equal to 80/mm, or greater than or equal to 80/mm. a may be less than or equal to 1000/mm, or less than or equal to 800/mm, or less than or equal to 500/mm, or less than or equal to 200/mm, or less than or equal to 150/mm. a may be in the range 30/mm to 200/mm, or in the range 50/mm to 150/mm.

The insulation fill factor may be greater than or equal to 90%, greater than or equal to 95%, greater than or equal to 99%, or greater than or equal to 99.99%.

The electrically insulating material may be described as an underfill material of the semiconductor prepackage. However, in contrast to some applications of so-called underfill material in the field of electronics, which provide improved mechanical properties (e.g., stiffness), the underfill material according to the present disclosure additionally or alternatively has an electrically insulating function to resist the high electric fields developed in the power dense converters of the present disclosure.

The maximum void size of the plurality of voids may be less than or equal to 100 pm, or less than or equal to 50 pm, or less than or equal to 20 pm, or less than or equal to 10 pm, or less than or equal to 5 pm, or less than or equal to 1 pm.

Those skilled in the art will appreciate that the term “voids” refers to the inclusion of ‘foreign’ material, different to the electrically insulating material, in the region occupied by the electrically insulating material. The voids may be in solid, (e.g., particle-like), liquid or gaseous form. Example void materials may include solder material, solder flux residues, air, washing liquid, and the like. Voids may be unintentionally introduced during the manufacturing process.

The maximum void size may be defined as the diameter of an equivalent spherical body having the same volume as the largest void. The largest void and/or the maximum void size can be determined by methods known to those skilled in the art, for example by statistical methods determining the void size of a representative number of voids. Said representative number of voids is smaller than the total number of the plurality of voids in the electrically insulating material.

A converter parameter τ may be less than or equal to 10,000 V. T is as a product of a dielectric strength of the electrically insulating material and the maximum void size of the plurality of voids. Accordingly, τ can be expressed as: τ = D X R_max

In this equation, D is the dielectric strength of the electrically insulating material and R_max is the maximum void size of the plurality of voids, τ may be less than or equal to 1 ,000 V, or less than or equal to 500 V, or less than or equal to 250 V. τ may be greater than or equal to 1 V, or greater than or equal to 10 V, or greater than or equal to 100 V. T may be greater than or equal to 150 V. τ may be in the range 100 V to 300 V, or in the range 150 V to 250 V.

The dielectric strength of the electrically insulating material, D, may be greater than or equal to 1 kV/mm, or greater than or equal to 10 kV/mm, or greater than or equal to 15 kV/mm. D may be less than or equal to 250 kV/mm, or less than or equal to 200 kV/mm, or less than or equal to 100 kV/mm, or less than or equal to 50 kV/mm. D may be in the range 10 kV/mm to 30 kV/mm, or range 15 kV/mm to 25 kV/mm.

One or more (e.g., each) power semiconductor prepackage may further comprise one or more electrically conductive layers. The one or more electrically conductive layers may extend in the x-y direction. At least one of the one or more electrically conductive layers may be embedded within the solid electrically insulating material of the power semiconductor prepackage. At least one of the one or more electrically conductive layers may be located on an opposite side of the power semiconductor switching element to the electrical connection side of the prepackage and connect to at least one of the terminals of the power semiconductor switching element. At least one electrical connection may extend in the z-direction from an electrically conductive layer of the prepackage to the electrical connection side of the prepackage.

Each power semiconductor prepackage may further comprise an electrical isolation layer. The electrical isolation layer may be located on an opposite side of the power semiconductor switching element to the electrical connection side of the prepackage. The electrical isolation layer may be of a ceramic material. The electrical isolation layer may be embedded within the solid electrically insulating material of the power semiconductor prepackage. The electrical isolation layer may extend in the x-y direction. The at least one capacitor may be connected to the at least one power electronics switching element through at least one of the conductive layers of the multi-layer planar carrier substrate.

The at least one capacitor may be a ceramic capacitor.

The gate driver circuit may be electrically connected to the gate terminal of each power semiconductor switching element by one or more electrical connections extending in the z-direction.

The power electronics converter may further comprise a heat sink for removing heat from power semiconductor prepackages.

The converter may further comprise a thermal interface layer (TIL) between a heat removal side of the prepackage and the heat sink, the heat removal side of the prepackage being opposite the electrical connection side of the prepackage.

The at least one prepackage may be located between the multi-layer carrier substrate and the heat sink. The heat sink may comprise one or more recessed regions defining one or more chambers for receiving the prepackages. Adjacent chambers may be separated by a wall.

The at least one prepackage may be embedded within the multi-layer carrier substrate. The heat sink may be located adjacent a heat removal side of the prepackage opposite the electrical connection side of the prepackage.

A converter parameter q may be greater than or equal to 100 kW/m³K. q is a heat transfer coefficient between the heat removal side of the power semiconductor prepackage and a cooling medium of the heat sink divided by the size in the z-direction of the gap between the heat removal side of the power semiconductor prepackage and the heat sink. Accordingly, q can be expressed as:

In this equation, h is the heat transfer coefficient between the heat removal side of the power semiconductor prepackage and the cooling medium of the heat sink, and G3 is the size in the z-direction of the gap between the heat removal side of the power semiconductor prepackage and the heat sink, q may be greater than or equal to 500 kW/m³K, or greater than or equal to 1 MW/m³K, or greater than or equal to 10 MW/m³K. q may be smaller than or equal to 1000 MW/m³K. q may be less than or equal to 500 MW/m³K, or less than or equal to 150 MW/m³K, or less than or equal to 100 MW/m³K. q may be in the range 1 MW/m³K to 1000 MW/m³K, or in the range 10 MW/m³K to 100 MW/m³K, or in the range 20 MW/m³K to 50 MW/m³K. The converter parameter q may be in the range 125 kW/m³K to 75 MW/m³K, or in the range 30 MW/m³K to 45 MW/m³K, or in the range 35 MW/m³K to 40 MW/m³K.

The heat transfer coefficient between the heat removal side of the power semiconductor prepackage and a cooling medium of the heat sink, h, may be greater than or equal to 0.1 kW/m²K, or greater than or equal to 1 kW/m²K, or greater than or equal to 5 kW/m²K. h may be less than or equal to 50 kW/m²K, or less than or equal to 30 kW/m²K, or less than or equal to 20 kW/m²K. h may be in the range 2.5 kW/m²K to 15 kW/m²K, or in the range 5 kW/m²K to 10 kW/m²K. The size in the z-direction of the gap between the heat removal side and the heat sink, G3, may be less than or equal to 2 mm, or less than or equal to 1 mm, or less than or equal to 0.8 mm, or less than or equal to 0.5 mm. G3 may be greater than or equal to 0.05 mm, or greater than or equal to 0.1 mm, or greater than or equal to 0.2 mm.

The thermal interface layer (TIL) may have a thermal conductivity and a mechanical compressibility.

A converter parameter Q may satisfy 0.1 MNK/Wm < Ω < 1 GNK/Wm. Ω is the mechanical compressibility of the thermal interface layer divided by the thermal conductivity of the thermal interface layer. Accordingly, Ω can be expressed as:

In this equation, M Is the mechanical compressibility of the thermal interface layer and k is the thermal conductivity of the thermal interface layer. Ω may be greater than or equal to 0.2 MNK/Wm, or greater than or equal to 0.4 MNK/Wm, or greater than or equal to 0.6 MNK/Wm. Ω may be less than or equal to 500 MNK/Wm, or less than or equal to 100 MNK/Wm, or less than or equal to 10 MNK/Wm, or less than or equal to 5 MNK/Wm. Ω may be in the range 0.25 MNK/Wm to 2 MNK/Wm, or in the range 0.7 MNK/Wm to 1.5 MNK/Wm, or in the range 0.7 MNK/Wm to 1.5 MNK/Wm, or in the range 0.8 MNK/Wm to 0.9 MNK/Wm.

The mechanical compressibility of the thermal interface layer, M, may be less than or equal to 3000 MN/m² (300 MPa), or may be less than or equal to 100 MN/m² (10 MPa), or may be less than or equal to 10 MN/m² (1 MPa). M may be greater than or equal to 0.5 MN/m² (0.05 MPa). M may be in the range 1 MN/m² (0.1 MPa) to 5 MN/m² (0.5 MPa), or in the range 2.5 MN/m² (0.25 MPa) to 3.5 MN/m² (0.35 MPa). The thermal conductivity of the thermal interface layer, k, may be less than or equal to 100 W/mK, or less than or equal to 25 W/mK, or less than or equal to 10 W/mK, or less than or equal to 5 W/mK. k may be greater than or equal to 0.5 W/mK, or greater than or equal to 1 W/mK. k may be in the range 2 W/mK to 10 W/mK, or in the range 3 W/mK to 4 W/mK.

The thickness (size in the z-direction) of the thermal interface layer may be in the range 0.05 mm to 3 mm, or in the range 0.075 mm to 1.5 mm, or in the range 0.1 mm to 0.75 mm, or in the range 0.15 mm to 0.25 mm.

An electrical isolation layer may be arranged between the one or more power semiconductor switching elements of the prepackage and the heat sink. The electrical isolation may form part of the power prepackage or be between the prepackage and the heat sink. The electrical isolation layer may be arranged between the prepackage and the thermal interface layer. The electrical isolation layer may be arranged between the thermal interface layer and the heat sink.

A converter parameter p may be greater than or equal to 5 MVW/m²K. p is a product of a thermal conductivity of the thermal interface layer and a breakdown electrical field strength of the electrical isolation layer. Accordingly, p can be expressed as: p=k x E_Break

In this equation k is the thermal conductivity of the thermal interface layer and E_{Brea k} is the breakdown electric field strength (which may also be referred to as a dielectric strength) of the electrical isolation layer, p may be greater than or equal to 10 MVW/m²K, or greater than or equal to 25 MVW/m²K, or greater than or equal to 50 MVW/m²K. p may be less than or equal to 25 GVW/m²K, or less than or equal to 5 GVW/m²K, or less than or equal to 500 MVW/m²K, or less than or equal to 250 MVW/m²K. p may be in the range 25 MVW/m²K to 5 GVW/m²K, or in the range 50 MVW/m²K to 250 MVW/m²K.

The breakdown electric field strength (dielectric strength), EBreak, of the electrical isolation layer may be greater than or equal to 5 kV/mm. E_Break may be less than or equal to 250 kV/mm. E_Break may be in the range 10 kV/mm to 50 kV/mm, or in the range 10 kV/mm to 100 kV/mm, or in the range 15 kV/mm to 25 kV/mm.

The thickness (size in the z-direction) of the thermal interface layer may be in the range 0.05 mm to 3 mm, or in the range 0.075 mm to 1.5 mm, or in the range 0.1 mm to 0.75 mm, or in the range 0.15 mm to 0.25 mm. The thickness (size in the z-direction) of the electrical isolation layer may be in the range 0.025 mm to 2 mm, or in the range 0.025 mm to 1 mm, or in the range 0.05 mm to 1 mm, or in the range 0.1 mm to 0.8 mm, or in the range 0.2 mm to 0.3 mm.

The thermal interface layer may have a relatively low electrical conductivity, (i.e. , a high dielectric strength) so as to reduce the dependence of the converter on additional electrical isolation measures such as the inclusion of the dedicated electrical isolation layer described above. Hence, some embodiments may not comprise an electrical isolation layer. In such embodiments, a converter parameter A may be greater than or equal to 1 TW/SK, A being defined as a thermal conductivity of the thermal interface layer divided by an electrical conductivity of the thermal interface layer. The converter parameter A can be expressed as:

In this equation, k is the thermal conductivity of the thermal interface layer and P is the electrical conductivity of the thermal interface layer. A may be less than or equal to 100 PW/SK, or less than or equal to 10 PW/SK, or less than or equal to 1 PW/SK, or less than or equal to 500 TW/SK. A may be greater than or equal to 10 TW/SK, or greater than or equal to 50 TW/SK, greater than or equal to 100 TW/SK, or greater than or equal to 200 TW/SK. A may be in the range 100 TW/SK to 500 TW/SK, or in the range 300 TW/SK to 400 TW/SK.

The electrical conductivity of the thermal interface layer, P, may be less than or equal to 0.1 pS/m (i.e., 1x10^-13 S/m). P may be greater than or equal to 1 fS/m (i.e., 1x10- ¹⁵ S/m). P may be in the range 5 fS/m to 50 fS/m, or in the range 7.5 fS/m to 25 fS/m.

The thermal conductivity of the thermal interface layer, k, may be greater than or equal to 0.1 W/mK. k may be less than or equal to 150 W/mK. k may be in the range 0.5 W/mK to 100 W/mK, or in the range 1 W/mK to 25 W/mK, or in the range 2 W/mK to 10 W/mK, or in the range 2.5 W/mK to 5 W/mK.

The multi-layer planar carrier substrate may comprise at least one outer conductive layer and/or at least one internal conductive layer. The at least one inner conductive layer may be thicker than (e.g., at least twice or at least three times as thick as) the outer conductive layer. The outer conductive layer and the inner conductive layer may be electrically connected by at least one connection (e.g., a plurality of connections) extending in the z-direction through the carrier substrate. Each power semiconductor prepackage may further comprise an electrical connection from at least one terminal of the one or more power semiconductor switching elements to the outer conductive layer of the multi-layer planar carrier substrate, the electrical connection extending in the z- direction through the solid insulating material.

Each power semiconductor prepackage may further comprise an electrical connection from at least one of its terminals to an electrical connection side of the power semiconductor prepackage, the electrical connection extending in the z-direction through the solid insulating material. Electrical connections may extend from each of the terminals to the electrical connection side of the power semiconductor prepackage, each electrical connection extending in the z-direction through the solid insulating material.

At least one of the terminals of each of the at least one power semiconductor switching elements may be connected to at least one of the conductive layers of the multi-layer planar carrier substrate at the electrical connection side of the power semiconductor prepackage. The at least one of the terminals may be a source and/or drain terminal of the power semiconductor switching element.

For each power semiconductor prepackage, the electrical connection side of the prepackage may form a flat surface. The prepackage may be surface mounted at its electrical connection side to a planar surface of the multi-layer planar carrier substrate.

For each power semiconductor prepackage, each electrical connection extending from the at least one of the terminals of the power semiconductor switching element through the solid insulating material may terminate at the flat surface of the prepackage. Each power semiconductor prepackage may be surface mounted to the surface of the multi-layer planar carrier substrate by soldering, sintering or gluing of the terminated electrical connection to an electrical connection of the multi-layer planar carrier substrate.

The flat surface of the electrical connection side may further comprise conductive pads (e.g., solder pads) for connecting the terminals of the power semiconductor switching element to the surface of the multi-layer planar carrier substrate.

The terminated electrical connection may be connected (e.g., soldered, sintered or glued) to an external conductive layer of the multi-layer carrier substrate. These connections space apart each power semiconductor prepackage from the planar surface of the multi-layer planar carrier substrate to define a gap (referred to herein as the prepackage gap) between the multi-layer planar carrier substrate and the electrical connection side.

The power electronics converter may be an AC-DC converter (i.e., an inverter or a rectifier). The AC-DC converter may be a multi-phase AC-DC converter. For each of the plurality of phases, the power circuit may comprise a phase leg. In this case, the parasitic inductance of the power circuit of the commutation cell is the parasitic inductance of one phase leg. Those skilled in the art will understand that, aside from inherent variation due to variation in components and manufacture, the parasitic inductance of each phase leg is the same.

The AC-DC converter may be a two-level converter comprising two logical switches per phase. The number of power semiconductor prepackages per logical switch may be greater than or equal to three. In some examples, a multi-phase (e.g., three- phase or four-phase) AC-DC converter comprises greater than 50 power semiconductor prepackages.

The power electronics converter may be a DC-DC converter. The power circuit of the commutation cell may further comprise an inductor.

The at least one capacitor may be a single capacitor, e.g., a single “DC-link” capacitor or “input capacitor”. In other examples, a more complex DC filter comprising a plurality of capacitors may be used.

The electrical machine may be of any suitable type and configuration. In one specific example, the electrical machine is a transversal flux (i.e., transverse flux) electrical machine. The transversal flux electrical machine may be air-cooled. The transversal flux electrical machine and the power semiconductor prepackages may both be air-cooled by a common cooling system.

There is also provided an electrical propulsion unit (EPU) for an aircraft. The EPU comprises an electric motor and an AC-DC power electronics converter in accordance the AC-DC power electronics converters set out above. The AC-DC power electronics converter is configured as an inverter and arranged to supply current to a winding of the electric motor.

A gas turbine engine is also provided. The gas turbine engine comprises a spool; an electrical machine having a rotor mechanically coupled to the spool; and an AC-DC power electronics converter in accordance the AC-DC power electronics converters set out above. The AC-DC power electronics converter is arranged to supply current to or receive current from a winding of the electric machine.

An electrical power system for an aircraft is also provided. The electrical power system comprises a power electronics converter in accordance the power electronics converters set out above.

An aircraft is also provided. The aircraft comprises the EPU, the gas turbine engine or the electrical power system set out above. In one group of embodiments, the aircraft is an electric vertical take-off and landing (eVTOL) aircraft having a plurality of EPUs as set out above.

The skilled person will appreciate that except where mutually exclusive, a feature described in relation to any one of the above aspects may be applied mutatis mutandis to any other aspect. Furthermore, except where mutually exclusive, any feature described herein may be applied to any aspect and/or combined with any other feature described herein.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments will now be described by way of example only with reference to the accompanying drawings, which are purely schematic and not to scale, and in which:

Figure 1 is circuit diagram of the commutation cell of a one-phase, two-level AC- DC converter;

Figure 2A is a schematic cross-section of the one-phase, two-level AC-DC converter of Figure 1 ;

Figure 2B is a further schematic cross-section of the one-phase, two-level AC- DC converter of Figure 1 illustrating further properties such as electric fields;

Figures 3A and 3B are schematic cross-sectional and plan views of the one- phase, two-level AC-DC converter of Figure 2A, further illustrating the calculation of the commutation cell volume;

Figure 4A is a schematic cross-section of a single power semiconductor prepackage, showing details not visible in Figures 2A and 3A;

Figure 4B is a schematic cross-section of the power semiconductor prepackage of Figure 4A sandwiched between a multi-layer carrier substrate and a heat sink;

Figure 4C is a schematic cross-section of a power semiconductor prepackage sandwiched between a multi-layer carrier substrate and a heat sink and comprising an electrically insulating material in a prepackage gap;

Figure 5A illustrates how low- and high-side power semiconductor switching elements of a half-bridge may be electrically connected through a multi-layer carrier substrate;

Figure 5B illustrates how power semiconductor switching elements may be electrically connected in parallel through a multi-layer carrier substrate;

Figures 6A and 6B illustrate how the one-phase arrangements of Figures 2-5 may be extended to multiple phases;

Figure 7A is a further schematic cross-sectional illustration of the power electronics converter of Figure 2A; Figure 7B is a schematic cross-sectional illustration of an alternative arrangement in which power semiconductor prepackages and a heat sink are arranged on opposite sides of the multi-layer carrier substrate;

Figures 8A and 8B illustrate alternative arrangements in which power semiconductor prepackages are embedded within a multi-layer carrier substrate;

Figure 9A illustrates a two-level, three phase inverter powering a three-phase motor;

Figure 9B illustrates a plurality of H-bridge inverter circuits powering a four-phase motor;

Figure 9C illustrates a two-level, three phase inverter with a more complex DC- side filter;

Figure 10A illustrates a DC-DC converter connected with the terminals of a battery;

Figure 10B illustrates a DC-AC-DC converter connected with the terminals of a battery;

Figure 11 A is a schematic illustration of an electric aircraft power and propulsion system;

Figure 11B is a schematic illustration of a multi-channel electric power and propulsion system for an aircraft including six electrically powered propulsors;

Figure 12A is a perspective view of an electric vertical take-off and landing (eVTOL) aircraft including six electrically powered propulsors in an eVTOL flight configuration;

Figure 12B is a perspective view of the eVTOL aircraft of Figure 2A in a forward flight configuration;

Figure 13 shows a general arrangement of a turbofan engine for an aircraft;

Figure 14A is a schematic illustration of a hybrid electric aircraft propulsion system; and

Figure 14B illustrates an electrically powered propulsor such as may be used in an electric or hybrid electric propulsion system.

DETAILED DESCRIPTION

Figure 1 is a circuit diagram of a power electronics converter 10. For simplicity and ease of explanation, one phase of a two-level AC-DC inverter is illustrated. The present disclosure is not, however, limited to converters of this circuit type, and further examples are disclosed herein. The power electronics converter 10 has a commutation cell having two parts: a power circuit and a gate driver circuit.

The power circuit has two DC inputs (DC-IN_L and DC-IN_H) and a single-phase AC output (AC-OUT). Connected between the DC inputs and the AC output of the power circuit is a half-bridge circuit comprising two power semiconductor switching elements 121 L, 121 H. The letters ‘L’ and ‘H’ designate the low- and high-voltage sides of the halfbridge which are connected to the low- and high-voltage DC inputs. The power circuit also includes an intermediate smoothing capacitor 14, often referred to as the ‘DC-link capacitor’ in the context of an AC-DC converter. The function of the DC-link capacitor will be familiar to those skilled in the art.

Each power semiconductor switching element 121 L, 121 H has three terminals. In the case of a MOSFET the terminals are named the source (S), the drain (D) and the gate (G). Current flowing from the DC inputs to the AC output passes between the source (S) and drain (D), whilst the voltage and current at the gate (G) controls whether or not the path between the source (S) and drain (D) conducts. The power semiconductor switching elements are preferably MOSFETs, in particular Silicon Carbide (SiC) based MOSFETs, though other semiconductor technologies such as Galium Nitride (GaN) could be used. As will be understood by those skilled in the art, the use of MOSFETs may allow for the omission of discrete parallel-connected diodes due to the inherent ‘body diode’ character of a MOSFET.

The gate driver circuit 13L, 13H is electrically connected to and configured to supply switching signals to the gate terminal of the MOSFET 121 L, 121 H to control the conduction of the MOSFETs i.e. , to control whether current can flow between the source and drain terminals or whether the flow of current is blocked. The gate driver circuit effectively acts as an amplifier of signals received from a controller (not shown), e.g., a digital controller which operates at and supplies signals having lower voltages, e.g., 3 V to 5 V. In this example the gate driver circuit is also connected to the drain terminals of the MOSFETs 121 L, 121 H, though those skilled in the art will appreciate this is not necessarily the case and that the gate and drain terminals may be isolated from each other.

The power circuit is further shown to include an inductor Lp. It is to be understood that the inductor LP is not a discrete component of the power circuit, but instead represents the combined parasitic inductance of the power circuit. Parasitic inductance is the inherent inductance of components and the connections between components, and which is not intentionally introduced into the circuit. The gate driver circuits 13L, 13H are also shown to include inductors LG; these too represent the parasitic inductances of the gate driver circuits 13L, 13H and are not discrete components.

Parasitic inductance is a notable problem in power electronics converters because it creates a loss mechanism: switching losses. The higher the parasitic inductance, the higher the switching losses. The magnitude of the switching losses also increases with the operating voltage of the power semiconductor switches 121 L, 121 H and with the switching frequency of the power semiconductor switches 121 L, 121 H. This means that a high parasitic inductance also limits a system designer’s ability to select higher values of the converter operating voltage and maximum switching frequency, as these must be kept lower to keep the switching losses at a tolerable level. These are undesirable limitations. The use of a lower voltage necessitates the use of a higher current to achieve the same power (P = I x V), which increases resistive losses (i.e. , l²R losses) in the power circuit and in e.g., electric machine windings to which the converter is connected. The use of a lower switching frequency limits the quality of the output voltage/current waveform, which leads to undesirable effects such a torque ripple in the rotor of an electrical machine connected to the power electronics.

The present disclosure provides power electronics converters with commutation cells having reduced parasitic inductances. As well as reducing the switching losses, this allows the adoption of increased operating voltages, increased switching frequencies and higher voltage and current ramp rates during switching. Overall, this provides for a significant increase in the operating efficiency of the converters compared with state-of- the-art power electronics converters.

Table 2 provides exemplary values of the parasitic inductance of the power circuit and the maximum operating efficiency of AC-DC converters in the power range 50- 400kW. The powers quoted are peak rated powers, i.e., the highest electrical power that can be controlled by the converter. This differs from the continuous rated power which depends on, e.g., environmental conditions during operation and the capabilities of the converter cooling system.

Table 2

Only the parasitic inductance LP of the power circuit, and not of the whole commutation cell, is quoted. This is because the gate driver circuit is electrically decoupled from the power circuit, and so the parasitic inductances of the two circuits are not combined. However, it should be appreciated that according to the present disclosure, the close integration of the gate driver circuit within the commutation cell results in a reduced value of LG.

As can be seen from Table 2, the value of parasitic inductance of power electronics converters according to the present disclosure may be approximately five times lower than the comparable state-of-the-art example. This, along with other measures disclosed herein, results in operating efficiencies up to and in excess of 99%, which compares with values of 95-96% commonly achieved in state-of-the-art power electronics modules.

It can also be seen from Table 2 that the parasitic inductance of the power circuit generally decreases as the power rating increases. This is because the peak rated power is generally increased through parallelization of the power semiconductors of the power circuit, i.e., at higher power ratings each e.g., low-side MOSFET 121 L of a given phase is implemented by multiple MOSFETs connected in parallel. This parallel connection of the components has the additional effect of reducing the parasitic inductance of the power circuit. Whilst this provides a mechanism for reducing the parasitic inductance of the power circuit to any desired low value, the additional components significantly add to the weight and volume of the converter, reducing the power density. Thus, a power electronics converter may be characterized by a converter inductance-volume parameter a, defined as the product of the parasitic inductance of the power circuit and the volume of the commutation cell: a = Lp x Vol. (1)

LP = Parasitic inductance of power circuit

Vol. = Smallest cuboidal volume enclosing commutation cell

As stated above, the volume is defined as the smallest cuboidal volume which encloses the entire commutation cell, i.e. , the combination of the power circuit and the gate driver circuit. An example illustrating the commutation cell volume is shown and will be described with reference to Figures 3A and 3B.

Table 3 shows exemplary values of a for power electronics converters according to the present disclosure. The values of a are characteristically lower than the prior art and associated with a combination of high efficiency and high power density. Values of a are quoted in pHm³ (i.e., pico-Hm³, or x10^-12 Hm³).

Table

Example values of the parasitic inductance of the power circuit are quoted in Table 2 and are preferably less than or equal to 16 nH for converters having peak power ratings in the range 25-500 kW. The commutation cell volume will generally increase with power rating, and values of less than 1 ,000 cm³ are preferred for converters having powers up to 500 kW. Commutation cell volumes will generally be greater than or equal to 100 cm³, with volumes of 150 to 600 cm³ striking a good balance between power density and relative ease of heat removal.

For reference, Table 4 includes values of a_p, which is the power-normalized value of a, and values of the product of the power and parasitic inductance of the power circuit.

Table 4

Returning to Figure 1 , although a circuit diagram, the components of the commutation cell are shown to be located on a carrier substrate 11 . The carrier substrate 11 is a multi-layer carrier substrate 11 , for example a rigid printed circuit board (PCB), comprising alternating insulating and conductive layers which extend in the x-y plane. The components of the commutation cell, including both the power circuit and gate driver circuit, are mounted on the multi-layer planar carrier substrate 11 and are electrically connected through the multi-layer carrier substrate 11. The electrical connections are made through a combination of the conductive layers of the multi-layer carrier substrate 11 and connections extending through the carrier substrate 11 in a z-direction defined perpendicular to the x-y direction of the planar substrate and its conductive layers. The connections extending the z-direction may, for example, be conductive vias or filled holes (e.g., laser micro-vias having diameters of the order of about 100 pm), and may have an x- and/or y- component as well as z-component (i.e. , the connections may be parallel to or form an angle with respect to the z-direction).

Figure 1 also schematically illustrates that the power semiconductor switching elements 121 L, 121 H of the power circuit are each comprised in a power semiconductor prepackage 12L, 12H. Those skilled in the art will understand this to mean that the power semiconductors (e.g., MOSFETs) are embedded in a solid insulating material which electrically isolates the power semiconductors and their terminals from their surroundings. The prepackages and their connections to each other and to other components of the commutation cell will be described in more detail below.

Figure 2A is a schematic cross-section of a power electronics converter 10 in accordance with the present disclosure. Again, for clarity and ease of explanation, one phase of a two-level AC-DC converter is illustrated. The x-direction and z-direction are indicated. The y-direction extends into the plane of the page. The converter 10 includes a multi-layer carrier substrate 11 , power semiconductor prepackages 12, a gate driver circuit 13, an intermediate (DC-link) capacitor 14, DC inputs (DC-IN) and an AC output (AC-OUT). Figure 2A further illustrates an integrated heat sink 15 which interfaces with a cooling side (also referred to herein as the heat removal side) of the prepackages 12 via a thermal interface layer 16.

The multi-layer carrier substrate 11 has opposed first and second planar surfaces 111a and 111b which define an x-y direction and a z-direction perpendicular to the x-y direction. The multi-layer substrate 11 includes alternating layers of insulating and electrically conductive material. The electrically conductive layers 112 are usually formed of copper, though they may be formed of any suitable conductive material such as silver, gold or aluminium. The insulating layers may be formed of the base material of the carrier substrate 11.

The multi-layer carrier substrate 11 may most preferably be a rigid PCB, in which case the base material and conductive layers may be a glass woven fabric impregnated with resin, as is known in the art of PCB manufacture. It should be appreciated, however, that the multi-layer carrier substrate 11 may take a different form, e.g., a ceramic-based carrier substrate or a flexible PCB having flexible polymer film base. The use of a rigid material is generally preferred, partly so that the carrier substrate 11 can effectively act as a structural component of the converter 10.

The number of layers in the multi-layer carrier substrate 11 can vary between applications, partly depending on the specifics of the power circuit, e.g., the number of phases in an AC-DC converter and the number of power semiconductors connected in parallel in each logical switch. In one specific example there are sixteen layers, including eight insulating layers and eight conductive layers 112.

Briefly referring to Figure 4B, in some examples at least a portion of one of the planar surfaces 111b of the carrier substrate 11 carries a metal layer 1121 that defines one or more electrical connection regions on the surface 111b of the substrate 11. Referring again to Figure 2A, additionally or alternatively to the outer layer, electrical connections in the form of e.g., vias or filled holes vias 114 extend in the z-direction through the carrier substrate 11 and terminate at the planar surface 111 b. The points at which the electrical connections 114 terminate define electrical connection points at the planar surface 111 b of the carrier substrate 11 .

Each prepackage 12 comprises a power semiconductor switching element 121 embedded in a solid insulating material 122. Embedding the power semiconductor switching elements 121 in solid insulating material means there are no air gaps surrounding the semiconductor chips and the terminals, reducing the risk of electrical breakdown even where high converter voltages are used. This allows higher voltages to be used and/or for the power semiconductors 121 and other components to be spaced closer together, increasing the power density of the converter 10.

Table 5 provides exemplary values of the maximum blocking voltage (i.e. , the source-drain blocking voltage or ‘rated voltage’) of the power semiconductor switching elements 121 in accordance with the present disclosure.

Table 5

As can be seen, the source-drain blocking voltages utilized in accordance with the present disclosure are high. Generally, the blocking voltage may be in the range of about 600 V to 1 ,800 V, with values greater than or equal to 800 V preferred to limit the peak current and reduce conduction losses. State of the art power electronics converters typically have much lower blocking voltages, with voltages of even 600 V being rare. It can also be seen that the blocking voltage does not increase with the peak power of the converter. This is because most or all the increase in peak rated power is achieved through parallelization in the power circuit. In other examples, a somewhat higher blocking voltage may be used for converters with higher power ratings (e.g., higher than 200 kW).

Generally, each power semiconductor switching element 121 has at least three terminals, including a gate terminal (G) for switching the conduction state of the switching element 121. In some embodiments the power semiconductor switching element 121 may have more than three terminals, for example if one or more terminals are provided for measurement (e.g., a Kelvin terminal) or if additional shorted terminals are present. In preferred embodiments the power semiconductor switching elements 121 are MOSFETs, in which case the terminals are designated the source (S), gate (G) and drain (D). In principle, however, other materials semiconductor switching devices (e.g., IGBTs) may be used in place of MOSFETs. The solid insulating material 122 can be any suitable insulating material, for example FR4.

As will be described in more detail below with reference to Figures 4A-4B, electrical connections extend from the terminals of the power semiconductor switching elements 121 through the solid insulating material 122 and terminate at an electrical connection side/surface 123a of the prepackage 12. Thus, the electrical connection side 123a of the prepackage 12 forms a substantially flat surface with exposed electrical connection points.

The electrical connection side of the prepackages 123a faces one of the planar surfaces 111 b of the carrier substrate 11. The prepackages 12 are surface mounted at their flat electrical connection sides 123a to the planar surface 111 b, for example by soldering, sintering or gluing (e.g., sinter gluing using a mix of glue and sinter paste) the electrical connection points of the prepackages 12 to the electrical connection points or region(s) of the planar surface of the carrier substrate 11. Figure 2A schematically illustrates soldered/sintered/glued connections 113.

The thickness, in particular the size in the z-direction, of the connections 113, which define a prepackage gap between the opposed surfaces 123a, 111b of the prepackages 12 and the carrier substrate 11 , is small. For example, the thickness and gap may be, when measured parallel to the z-direction, less than 500 pm, for example between 20 pm and 250 pm. In a specific embodiment the gap is about 100 pm.

Terminating the electrical connections from the chip terminals at a flat surface 123a of the prepackage 12 and using surface mounting to form the onward electrical connections through the PCB reduces the overall size of the converter 10 in the z- direction, which reduces the size and weight of the converter 10. Moreover, the surface mounting of the prepackages 12 reduces the impact of ‘open loop’ effects in the electrical connections between the power semiconductors 121 , the gate driver circuit 13 and the DC-link capacitor 14. This may substantially reduce the parasitic inductance of the commutation cell, which reduces switching losses and allows the use of e.g., an increased switching frequency.

The gate driver circuit 13 is electrically connected to and configured to supply switching signals to the gate terminals of the power semiconductor switching elements 121. In the illustrated embodiment the gate driver circuit 13 is mounted to the first planar surface 111a of the carrier substrate 11 , opposite to the second planar surface 111b which faces the power semiconductor prepackages 12. In other embodiments, the prepackages 12 and the gate driver circuit 13 could be mounted at the same side of the carrier substrate 11 , e.g., the second side 111 b. In the illustrated embodiment, the electrical connection 114 between the gate driver circuit 13 and the gate terminals of the power semiconductor switching elements 121 extends in the z-direction through the carrier substrate to the surface 111 b of the carrier substrate 11. The onward path then passes through a solder connection 113 and then through the electrical connection that passes in the z-direction through the solid insulating material 122 of the prepackages 12 to the gate terminal. In other embodiments, the connection between the gate driver circuit 13 and the gate terminals may be made through one or more conductive layers 112 of the substrate 11.

Table 6 provides exemplary values of the switching frequency of the power semiconductor switching elements of a power electronics converter in accordance with the present disclosure. Table 6 also includes maximum absolute values of the rate of change of the source-drain voltage (measured in units of kilo-Volts per micro-second) of the power semiconductor switching elements 121 during a switching cycle. The quoted values are maximum values of the switching frequency to occur during operation.

Table 6

In many converter arrangements the switching frequency of each power semiconductor switching element 121 is the same and the same voltage value is used, such that the maximum values quoted above are the same for each individual power semiconductor switching element 121. However, it is known that some converter architectures, for example multi-level converter architectures (e.g., modular multi-level converter architectures) which interface with multiple network voltage levels, use different voltages and/or switching frequencies for different power semiconductor switching elements 121. In these cases, the quoted values correspond to the maximum value of any individual power semiconductor switching element 121 in the converter. Converters according to the present disclosure use maximum frequencies greater than 10 kHZ, though switching frequencies in excess of 30 kHz are generally preferred, which as discussed below may facilitate a reduction in the required capacitance in the power circuit. Unlike many state-of-the-art systems in which the parasitic inductance of the converter commutation cell limits the maximum usable frequency, the low parasitic inductance of the converter commutation cell may mean other system constraints limit the maximum frequency. For example, a maximum desirable switching frequency may be imposed by the capabilities of the insulation of the windings of an electrical machine that interfaces with an AC-DC converter. Generally, a switching frequency of less than or equal to 100 kHz is used.

The use of a high switching frequency and high source-drain block voltage results in a notably high value of the maximum the rate of change of the source-drain voltage during operation. Rapid switching between the on- and off-states of the power semiconductor switching elements results in clean switching and improved output waveforms, which limits the harmonic content of the waveforms and improves e.g., converter efficiency, and reduces electric machine torque ripple. In the above examples, the maximum rate of change of the source-drain voltage during operation is in the range 30 to 45 kV/ps. However, lower values (e.g., 10 to 20 kV/ps) or higher values (e.g., greater than 50 kV/ps or as high or higher than 100 kV/ps) may be used, for example, where a relatively low or high switching frequency is utilized.

Thus, a power electronics converter may be characterized by a converter switching parameter p, defined as the product the maximum switching frequency of the switching signals and the maximum rate of change of the source-drain voltage of the power semiconductor switching elements during operation. β = f_max X |dv/dt| max (2) f_max = Maximum frequency of switching signals

\dV/dt\_max = Maximum rate of change of source-drain voltage during operation

Table 7 shows exemplary values of p for power electronics converters according to the present disclosure. The values of p are characteristically higher than the prior art and associated with a combination of high efficiency, high power density and high-quality (e.g., highly sinusoidal AC) output waveforms. Values of p are quoted in PV/s² (i.e. , Peta- V/s², or x10¹⁵ V/s²).

Generally, converters in accordance with the present disclosure have a value of P that is greater than or equal to 0.3 PV/s². Values of are preferably less than about 10 PV/s² to mitigate against problems such as e.g., insulation breakdown. Values in the range 0.8 PV/s² < p < 5 PV/s², and particularly, 1.0 PV/s² < p < 2.5 PV/s², may strike a good balance between the competing effects.

Returning to Figure 2A, the intermediate capacitor 14, which forms part of the power circuit of the commutation cell, is electrically connected to the power semiconductor switching elements 121 through one or more conductive layers 112 of the multi-layer carrier substrate 11. In the illustrated embodiment the capacitor 14 is mounted to the first planar surface 111a of the carrier substrate 11 , opposite to the second planar surface 111b which faces the power semiconductor prepackages 12. In other embodiments, the prepackages 12 and the capacitor 14 could be mounted at the same side of the carrier substrate 11 , e.g., the second side 111b. Electrical connections between the capacitor 14 and the conductive layers 112 of the multi-layer carrier substrate 11 may be formed in a number of ways. In one example, the capacitor 14 is surface mounted to the substrate 11 in a similar manner as the prepackages 12, i.e. , by soldering, sintering or gluing electrical contacts of the capacitor 14 to an electrical connection region of the surface 111a of the substrate 11. The electrical connection region may be an outer conductive layer on the surface 111a of the substrate (similar to the layer 1121 shown in Figure 4B) or an exposed end of one or more vias or filled holes 114 which extends in the z-direction from the surface 111a of the substrate 11 to an internal conductive layer 112 of the substrate 11. The connection between the capacitor 12 and the multi-layer substrate 11 may be made through THT (through hole technology) as an alternative to SMT (surface mount technology), both of which will be familiar to those skilled in the art. In state-of-the-art converters, the capacitor(s) - which are an essential component of most AC-DC and DC-DC converter architectures - are a significant contributor to both size and weight. This is a particular problem in the context of aerospace applications, which are both sensitive to size and weight and require relatively high powers (compared to, for example, electric vehicles and domestic appliances) which necessitate a higher total rated power circuit capacitance. According to the present disclosure, however, the reduction in parasitic inductance allows for a lower capacitance per unit rated power. This is partly because the low parasitic inductance allows for a high switching frequency. Increasing the switching frequency may decrease the required capacitance of the power circuit.

Table 8 provides exemplary values of the total rated capacitance of the power circuit. Values of the total rated capacitance normalized by peak rated power of the converter are also included. It will be understood from the above that the values of the total rated capacitance and the normalized capacitance are low compared with state-of- the-art power electronics converters.

Generally, values of the total rated capacitance normalized by the peak rated power will be less than 5 nF/W, with values of less than 1 nF/W preferred to achieve a low size and weight for the converter. The use of low values of the capacitance may also allow for the use of low-weight capacitor technologies, in particular ceramic capacitors, allowing for a further weight reduction. Capacitance is known to somewhat vary with operating conditions and therefore the literature usually quotes nominal values of capacitance. In Table 8 the quoted capacitances are those measured at nominal conditions of 25 °C (298K) and 1 ,000 V DC, which is typical for capacitance measurements. The term “total rated capacitance of the power circuit” will be understood to mean the total capacitance of all the capacitors in the power circuit. In the simplest cases, there may be a single capacitor. For example, the one-phase two-level AC-DC converter circuit of Figure 1 includes a single DC link capacitor, as does the three-phase, two-level AC- DC converter circuit of Figure 9A. In other examples, there may be multiple capacitors: there may be a separate DC-link capacitor for each phase, as is the case in the H-bridge based arrangement of Figure 9B and the three-phase, two-level circuit of Figure 9C. Where there is parallelization of the power semiconductors in the power circuit, each power semiconductor switching element may be associated with its own capacitor, or multiple parallel-connected power semiconductor switching elements 121 may be connected with a single larger capacitor.

A power electronics converter may be characterized by a converter frequencycapacitance parameter y, defined as the total rated capacitance of the power circuit divided by the product of the peak rated power output of the power electronics converter and the maximum switching frequency of the gate switching signals:

C = Total rated capacitance of power circuit of commutation cell P = Peak rated power of power electronics converter f_max = Maximum frequency of switching signals

Table 9 shows exemplary values of y for power electronics converters according to the present disclosure. The values of y are characteristically lower than the prior art and associated with a combination of high efficiency and high power density. Values of y are quoted in fFs/W (i.e., femto-Fs/W, or x10^-15 Fs/W).

Generally, power electronics converters in accordance with the present disclosure have values of y less than or equal to 150 fFs/W. y may be greater than about 1.0 fFs/W, with the provision of a lower bound limiting problems associated with, e.g., insulation breakdown at high switching frequencies. Values of y may most preferably be in the range 4.0 fFs/W < y < 25 fFs/W, which may strike a good balance between high power density and reliable operation.

A power electronics converter may also be characterized by a converter frequency-capacitance parameter δ, defined as the maximum source-drain blocking voltage of the power semiconductor switching elements of the power circuit divided by the product of the parasitic inductance of the power circuit and the total rated capacitance of the power circuit:

V_block = Maximum source-drain blocking voltage of power semiconductors L = Parasitic inductance of power circuit

C = Total rated capacitance of power circuit of commutation cell

T able 10 shows exemplary values of δ for power electronics converters according to the present disclosure. The values of δ are characteristically higher than the prior art and associated with a combination of high efficiency and high power density. Values of 5 are quoted in PV/FH (i.e., Peta-V/FH, or x10’⁵ V/FH).

Table 10

Generally, power electronics converters in accordance with the present disclosure have values of 5 greater than 0.5 PV/FH. The value of 5 may be less than or equal to 40 PV/s². The value of 5 may be greater than or equal to 1.5 PV/s². The value of Y may be in the range 2.5 PV/s² to 25 PV/s². It may be in the range 4.0 PV/s² to 15 PV/s²

A power electronics converter may also be characterized by a converter frequency-capacitance parameter E, defined as:

f_max = Maximum frequency of switching signals \dV/dt\_max = Maximum rate of change of source-drain voltage during operation L = Parasitic inductance of power circuit

C = Total rated capacitance of power circuit of commutation cell

Table 11 shows exemplary values of E for power electronics converters according to the present disclosure. The values of E are characteristically lower than the prior art and associated with a combination of high efficiency, high power density and goodquality output waveforms. Values of E are quoted in units of x10²⁷ V/s⁴.

Table 11

Generally, power electronics converters in accordance with the present disclosure have values of E less than or equal 10²⁹ V/s⁴. The value of E is generally greater than 5x10²⁶ V/s⁴, as lower values may be associated with e.g., insulation failure, though this will depend to some extent on the application requirements (e.g., if high quality insulation can be provided in an electrical machine). Values of E may preferably be in the range 1.5x10²⁷ V/s⁴ to 3x10²⁸ PV/s², as this strikes a good balance between power density, efficiency and reliability. Returning to Figure 2A, components of the converter 10, particularly the power semiconductor switching elements 121 , generate heat which must be removed. Heat removal may be a particularly notable consideration in power electronics converters in accordance with the present disclosure because of their compactness and high power density. To this end, the converter 10 comprises an integral heat sink 15 which is in close thermal contact with power semiconductor prepackages 12. To ensure efficient heat removal, a solid path is provided between a heat removal side 123b of the prepackages (i.e. , the underside of the prepackage 12 as illustrated in Figure 2, opposite the electrical connection side 123a) so that the heat sink 15 removes heat from the prepackages 12 by thermal conduction. In the illustrated example, a plurality of the prepackages 12 of the converter 10, and optionally all of the prepackages 12 of the converter, share a common heat sink 15.

The heat sink 15 itself may be of any suitable design. It may, for example, be formed of aluminum or another thermally conductive material, cooled by a cooling flow of a coolant which may be a gas (e.g., air) or a liquid (e.g., water or an oil). In some embodiments a surface of the heat sink 15 opposite the prepackages 12 may be subject to an impinging flow of coolant to increase the coefficient of heat transfer between the prepackages 12 and the cooling medium of the heat sink 15.

To ensure efficient heat removal by conduction, it is necessary that there is a good quality and consistent thermal interface between the heat removal side 123b of each of the prepackages 12 and the heat sink 15. This may be challenging because, although the heat removal side 123b of the prepackages 12 can be designed to be flat, there does exist some manufacturing and assembly tolerance. For example, the thickness of the electrical contacts 113 may vary slightly between and within prepackages 12, which may result in prepackage tilting and/or inconsistent distances between the heat removal side 123b and the heat sink 15. As another example, the multilayer carrier substrate 11 may bend or locally deform. In view of this, the converter 10 is also shown to include a thermal interface layer (TIL) 16 between the heat removal side 123b of the prepackages 12 and the heat sink 15. The TIL 16 is included to ensure a good quality thermal interface between the prepackages 12 and the heat sink 15. The TIL accommodates tolerance issues whilst also providing a thermally conductive path in the z-direction between the heat removal side 123b of each of the prepackages 12 and the heat sink 15. In some examples the TIL 16 may also have a high thermal conductivity on the x-y plane to spread heat across the surface of the heat sink 15. This may be of particular use where a single TIL 16 serves multiple prepackages 12. The TIL 16 may take one of a number of different forms, including solids (e.g., a solder layer, a foil or a film), semi-solids (e.g., a paste) or a liquid. In one group of examples, the TIL 16 is a layer of solder (e.g., indium-tin solder). In this case, to ensure a good quality solder connection, each prepackage 12 may have its own TIL 16 rather than a single TIL. In another group of examples, the TIL is a foil, for example indium-tin or graphene foil, which is both thermally conductive and flexible. Utilizing a TIL with some compressibility may be advantageous for both accommodating manufacturing tolerances and preventing separation of the heat sink 15 from the prepackages 12 during e.g., vibration. The TIL 16 preferably has a thermal conductivity of at least 1 W/mK, and more preferably at least 2.5 W/mK. The thickness of TIL 16 will generally be less than a few mm, is preferably less than 1 mm, and in one group of examples has a thickness of between 100 pm and 500 pm.

A power electronics converter may be characterized by a converter heat transfer parameter q, defined as:

h = heat transfer coefficient from prepackage to cooling medium

G3 = distance in z-direction between heat removal side of prepackage and heat sink

The gap G3 is labelled in Figure 2A. Table 12 shows exemplary values of q for power electronics converters according to the present disclosure. The values of q are characteristically high and associated with a combination of a compactness, high efficiency and high power density. Values of q are quoted in units of MW/m³K.

Table 12 Generally, power electronics converters in accordance with the present disclosure have values of q greater than or equal 100 kW/m³K (0.1 MW/m³K). However. Values of q greater than or equal 10 MW/m³K are generally preferred.

Table 13 shows exemplary values for the heat transfer coefficient h and the size of the gap, G3, between the heat removal side of the power semiconductor prepackage and the heat sink.

Table 13

A power electronics converter may also be characterized by a thermal interface parameter, Q, defined as:

M = mechanical compressibility of thermal interface layer k = thermal conductivity of thermal interface layer

Table 14 shows exemplary values for the thermal interface parameter Q. Values of Q are quoted in units of MNK/Wm (i.e. , Mega-NK/Wm, equal to 10⁶ NK/Wm).

Table 14

Converters of the present disclosure preferably have values of Q that satisfy 0.1 MNK/Wm < Q < 1 GNK/Wm, with the range 0.25 MNK/Wm < Q < 2 MNK/Wm being preferred. Thermal interface parameters in this range may provide a good combination and heat transfer and mechanical properties.

Table 15 shows exemplary values for the mechanical compressibility, M, of the thermal interface layer as well as for the thermal conductivity k of the thermal interface layer.

Table 15 Figure 2B is a further illustration of the converter 10 of Figure 2A and illustrates the electric fields developed between components of the converter 10.

The prepackage gap, labelled G1 in Figure 2B, is visible between the second surface 111b of the multi-layer carrier substrate 11 and the electrical connection side/surface 123a of the prepackages 12. The prepackage gap G1 has a size which may be measured in the z-direction. A first electric field 50 is established in the prepackage gap G1 resulting from the potential difference (voltage) between the second surface 111b of the multi-layer carrier substrate 11 and the electrical connection surface 123a of the prepackages 12.

A power electronics converter may be characterized by a converter parameter 0, defined as a size in the z-direction of the prepackage gap divided by a maximum electric field strength in the prepackage gap (referred to herein as the first maximum electric field strength):

Gi = Size of prepackage gap in z-direction

Ei = Maximum electric field strength in prepackage gap

Table 16 shows exemplary values for the converter parameter 0 for power electronics converters according to the present disclosure. The values of 0 are characteristically lower than the prior art and associated with a high power density. Values of 0 are quoted in units of pm²/V (pico-m²/V, or x10^-12 m²/V).

Table 16

Generally, power electronics converters in accordance with the present disclosure have values of 0 less than or equal to 300 pm²/V. The value of 0 is generally greater or equal to 0.1 pm²/V, as lower values may be associated with e.g., greater risk of electrical breakdown. Values of 0 may preferably be in the range 2.0 pm²/V to 50 pm²/V, as this may strike a good balance between power density and reliability.

Those skilled in the art will appreciate that the first maximum electric field strength is a maximum homogenous electric field strength. In other words, the first maximum electric field strength is the maximum field strength determined in a location sufficiently spaced away from sharp edges and/or obstructions in the gap which may result in electric field singularities or other highly localized maxima. For example, Figure 2B shows a first region 52 in the prepackage gap in which the first maximum electric field strength may be determined, as the first region 52 is spaced from any edge or boundary of the multilayer planar carrier substrate or of the power semiconductor prepackage in the x-y direction. Here, it is shown that the first region 52 is spaced in the x-direction by a first distance 54 from a prepackage edge 123e. In a similar manner, the first region 52 is spaced apart from another prepackage edge 123e in the y-direction (not visible here as the y-direction is perpendicular to the drawing plane). Such offset of the first region 52 in the x-y direction towards the inside of the prepackage gap ensures the maximum field strength is determined over a suitably homogenous region of the electric field. Where the maximum electric field strength is determined by modelling (e.g., finite element analysis, FEA), the maximum electric field strength may be determined at a location at least three mesh cells from any singularity in the model (e.g., spikes, sharp edges, triple points).

Table 17 provides exemplary values for the size Gi of the prepackage gap in the z-direction, as well as the maximum electric field strength Ei in the prepackage gap.

The second surface 111b of the multi-layer carrier substrate 11 and a heat sink 15 are spaced apart in the z direction in a substantially parallel manner so as to form a heat sink gap, which is labelled G2 in Figure 2B. A second electric field 56 is established in the heat sink gap G2, resulting from the potential difference (voltage) between the second surface 111 b of the multi-layer carrier substrate 11 and the heat sink.

A power electronics converter may be characterized by a converter parameter cp, defined as a size in the z-direction of the heat sink gap divided by a maximum electric field strength in the heat sink gap (referred to herein as the second maximum electric field strength):

G2 = Size of heat sink gap in z-direction

E2 = Maximum electric field strength in heat sink gap

Table 18 shows exemplary values for the converter parameter cp. The values of (p are characteristically lower than the prior art and associated with a high power density. Values of cp are quoted in units of nm²/V (nano-m²/V, or x10^-9 m²/V).

Table 18

Generally, power electronics converters in accordance with the present disclosure have values of cp less than or equal to 20 nm²/V. The value of cp is generally greater or equal to 0.01 nm²/V, as lower values may be associated with e.g., greater risk of electrical breakdown. Values of cp may preferably be in the range 0.05 nm²/V to 5 nm²/V, as this may strike a good balance between power density and reliability.

As with the first maximum electric field strength E1 , the second maximum electric field strength E2 is a maximum homogenous electric field strength and thus singularities and other highly localized maxima are excluded. By way of example, in Figure 2B a second region 58 is illustrated for determining a second maximum electric field strength of the second electric field 56. The second region 58 is located inside the heat sink gap G2 and spaced away from any obstructions and edges inside the heat sink gap G2, such as those associated with the prepackages 12. Where the maximum electric field strength is determined by modelling (e.g., finite element analysis, FEA), the maximum electric field strength may be determined at a location at least three mesh cells from any singularity in the model. Table 19 provides exemplary values for the size G2 of the heat sink gap between the multi-layer planar carrier substrate 11 and the heat sink 15 in the z-direction and the maximum electric field strength E2 in the heat sink gap.

Table 19

Figures 3A and 3B illustrate the smallest cuboidal volume that encloses the commutation cell of a power electronics converter. Once again, the example of a one- phase, two-level converter is chosen for ease of explanation. Reference is made to Equation 1 , which defines the inductance-volume parameter a in terms of the smallest cuboidal volume that encloses the commutation cell.

Figure 3A is the cross-sectional view also shown in Figure 2A. The x- and z- directions are indicated. The box with dashed lines, labelled A_x.z, is a cross-section in the x-z plane through the smallest volume that encloses the commutation cell of the converter 10. The cross-sectional area A_x.z encloses both the power circuit (i.e., the power semiconductors 121 which are formed in the prepackages 12, the capacitor 14, the DC inputs and the AC outputs and the connections therebetween) and the gate driver circuit 13 and its connections with the power semiconductors 121. The area A_x.z can be calculated as the product of the size of the box measured in the x-direction (L_x) and the size of the box measured in the z-direction (L_z).

Figure 3B is a plan view of converter of Figure 3A. The x- and z-directions indicated. The box with dashed lines, labelled A_x.y is a cross-section in the x-y plane through the smallest cuboidal volume that encloses the commutation cell of the converter 10. The cross-sectional area A_x.y encloses both the power circuit (i.e., the power semiconductors 121 , the capacitor, the DC inputs and the AC outputs and the connections therebetween) and the gate driver circuit 13 and its connections with the power semiconductors. The area A_x.z can be calculated as the product of the size of the box measured in the x-direction (L_x) and the size of the box measured in the y-direction (L_y).

It will be readily appreciated that the smallest cuboidal volume that encloses the commutation cell can be calculated as the product of the three dimensions, L_x, L_x and L_z.

It is noted that all components of the commutation cell are visible in Figure 3B even though some of the components illustrated in the cross-sectional view of Figure 3A are located on the underside of carrier substrate 11. It must be appreciated that this is solely for the purpose of illustration and clear explanation and that, in practice, not all components of the converter 10 of Figures 3A-3B would be visible in a single plan view. It should also be appreciated that the x-y position of the gate driver circuit 13 has been moved in Figure 3B for ease of illustration (i.e. , so as not overlap with the x-y locations of the prepackages 12 as it does in Figure 3A). However, flexibility in the x-y location of commutation cell components is an advantageous feature of converters according to the present disclosure; this is in part because changes in the x-y locations do not significantly alter the parasitic inductance of the connections between components.

It is also noted that whilst only two power semiconductor prepackages 12L and 12H are visible in Figure 3A, there are six prepackages 12LI-3 and I2H1-3 in Figure 3B. This is to illustrate parallelization of the power semiconductor switching elements 121. Specifically, it can be seen from Figure 3B that the low side of the half-bridge comprises three power semiconductor prepackages, 12Li, 12L2 and I2L3, comprising a corresponding three power semiconductor switching elements connected together in parallel. Likewise, the high side of the half-bridge comprises three power semiconductor prepackages, 12Hi, 12H2 and I2H3, comprising a corresponding three power semiconductor switching elements connected in parallel. In Figure 3A, where the y- direction is obscured, only one of three prepackages 12L, 12H of each of the low and high side are visible.

Figure 4A illustrates a single power semiconductor prepackage 12 and shows detail not visible in Figures 2A and 3A.

The power semiconductor switching element 121 , which in this example is a MOSFET in the form of a semiconductor die or chip, is embedded in solid insulating material 122, e.g., FR4. Electrical connections 124, 125i, which may be filled holes, vias or similar, extend in the z-direction from the terminals of the semiconductor switching element 121 to the electrical connection side 123a of the prepackage 12, where they terminate to form a flat surface 123a. Although vertically extending connections 124, 125i are illustrated, it should be appreciated the connections could have a component in the x-y plane too.

The MOSFET 121 has at least three terminals, namely the source, drain and gate terminals. A first electrical connection 124 extends from the source terminal to the flat electrical connection surface 123a. In this example the drain and gate terminals are electrically connected by an electrically conductive metallization layer 125ii on the underside of the MOSFET die 121. A second electrical connection 125i extends from the conductive layer 125ii to the flat electrical connection surface 123a. In other examples, the gate and drain terminals are not connected and e.g., the connections 125i, 125ii correspond only to the drain terminal, with the gate terminal served by a separate connection from the gate terminal to the flat electrical connection surface 123a.

The illustrated prepackage 12 further includes an optional electrical isolation layer (EIL) 126. The purpose of the EIL 126, which in this example is a layer of ceramic material, is to electrically isolate the MOSFET 121 and its terminals from the heat sink

15 which is arranged on the underside of the prepackage 12 (see Figure 4B). In other embodiments the EIL 126 may be omitted from the prepackage 12, for example if a separate EIL is provided between the underside of the prepackage 12 and the heat sink 15, or if the TIL 16 is able to provide sufficient electrical isolation. Where present, the EIL 126 preferably has a thickness of less than 5mm, with thicknesses of less than 1mm preferred to keep the converter compact whilst still achieving the isolation function. Lower thicknesses, e.g., less than 0.1 mm may be used where, for example, an organic substrate such as IMS is used.

The illustrated prepackage 12 further includes an optional metal layer 127 on the underside of the EIL 126. The metal layer 127 improves thermal conduction between the prepackage 12 and TIL 16. The metal layer 127 may also provide a suitable material interface between the underside of the prepackage and TIL 16. For example, if the TIL

16 is a solder layer, this may necessitate that the underside 123b of the prepackage 12 carries a material suitable for a solder connection. The metal layer 127 may be omitted, for example if a TIL other than solder is used.

Figure 4B shows the prepackage 12 of Figure 4A sandwiched between a multilayer carrier substrate 11 and a heat sink 15. It further illustrates how the electrical connections 124, 125ii which extend from the terminals of the power semiconductor 121 to the electrical connection surface 123a may be connected to the conductive elements of the multi-layer carrier substrate 11 and thereby other components of the commutation cell. In the illustrated example, a planar surface 111b of the multi-layer carrier substrate 11 comprises one or more regions carrying an outer conductive layer 1121. These regions 1121 allow for soldered or sintered connections 113 to be formed to connect the substrate 11 to the distal ends of the electrical connections 124, 125i of the prepackage 12. In the illustrated example, metallization regions 1241 (e.g., conductive contact pads such as solder pads) are also provided adjacent to the distal ends of the electrical connections 124, 125i to improve the ease with which soldered, sintered or glued connections can be formed. In other examples, these may be omitted and the solder connections 113 may be made directly on the exposed distal ends of the connections 124, 125ii.

With the terminals of the power semiconductor switching elements 121 now connected to the conductive regions 1121 of the carrier substrate 11 , connections to the other components of the commutation cell are made through connection to the conductive regions 1121. These connections may be formed by a combination of conductive layers 112 of the carrier substrate 11 (see Figure 2A) and electrical connections 114 which extend in the z-direction from the conductive regions 1121 through the carrier substrate. For example, referring to Figures 2A and 4B, the layer 1121 which is soldered to the connection 124 which connects to the source terminal of the MOSFET 121 may be associated with one or more (e.g., many) connections 114a which extend in the z-direction to an internal conductive layer 112. The internal conductive layer 112 may then connect to a terminal of the DC link capacitor 14. As another example, the layer 1121 which is soldered to the connection 125i which connects to the gate terminal of the MOSFET 121 may be associated with one or more connections 114b which extend in the z-direction all the way through the substrate 11 to terminals of the gate driver circuit 13.

Power electronics converters described herein may be characterized by a converter parameter p, defined as follows: p = k x E_Break (10) k = thermal conductivity of thermal isolation layer (TIL)

EBreak = breakdown electric field strength of electrical isolation layer (EIL)

Table 20 shows exemplary values for the converter parameter p. Converters described herein may have characteristically high values of p, which may be associated with a combination of good heat removal from the prepackages and good resistance to electrical breakdown. Values of p are quoted in units of MVW/m²K (Mega-VW/m2K, or x10⁶ VW/m²K).

Table 20

Generally, power electronics converters in accordance with the present disclosure have values of p greater than or equal to 5 MVW/m²K, though values greater than 20 MVW/m²K are preferred.

Table 21 shows exemplary values for the breakdown electric field strength (which may also be referred to as the dielectric strength in some literature), EBreak, of the electrical isolation layer (EIL). For exemplary values for thermal conductivity k of the thermal interface layer, see Table 15.

Table 21

Higher values for EBreak, for instance in Example 4, are for EILs comprising organic materials, whereas the lower values, for instance Examples 1 , 2 and 3 are EILs comprising inorganic materials. As noted above, the EIL 126 is optional. In alternative embodiments, the EIL 126 is omitted and a TIL 16 with suitably electrically isolating properties is provided. Thus, the TIL 16 may provide electrical isolation between the prepackages and the heat sink, as well as a good heat path between the prepackages and heat sink.

Such embodiments may be characterized by a TIL parameter A, defined as the thermal conductivity of the TIL divided by the electrical conductivity of the TIL:

k = thermal conductivity of thermal isolation layer (TIL) P = electrical conductivity of the TIL

Table 21 shows exemplary values of the parameter A. The values of A are characteristically high. Values of A are quoted in units of TW/SK (Tera-W/SK, or x10¹² W/SK).

Table 22

Generally, power electronics converters in accordance with the present disclosure have values of A greater than or equal to 1 TW/SK, though values greater than 100 TW/SK are preferred. Table 23 shows exemplary values for the electrical conductivity P of the thermal interface layer.

Table 23

Figure 4C illustrates a further arrangement of a power electronics converter 10 in which at least a portion of the prepackage gap G1 is filled with electrically insulating material 60, for example a resin. The electrically insulating material 60 is arranged in and, optionally, around the prepackage gap G1. “Around” in this context means that the electrically insulating material 60 extends beyond the semiconductor prepackage 12 in the x-y direction. For example, the material 60 may additionally cover the metal layers 1121 extending from connections 114a to the connections 113.

By means of the electrically insulating material 60, a creepage distance 62 as well as an air gap distance 64 can be reduced and kept physically small. Thus, utilizing the electrically insulating material 60, small distances between the components of the power electronics converter 10, for example between the metal layers 1121 and the connections 114a, can be achieved without a significant risk of adverse electrical effects such as sparking or creeping currents. This is particularly advantageous given the high voltages utilized in converters of the present disclosure, which can result in high potential differences between, e.g., the metal layers 1121 and the prepackage surfaces 12 and heat sink 15.

Generally, the applied electrically insulating material 60 will comprise voids (e.g., air) within its volume due to imperfections in the manufacturing process. A power electronics converter utilizing electrically insulating material in a prepackage gap may be characterized by a converter parameter o, defined as an insulation fill factor divided by a maximum void size of the voids:

F = Insulation fill factor

R_max = Maximum void size in insulation

In this equation, F is the insulation fill factor and R_max is the maximum void size of the plurality of voids. The insulation fill factor is defined as a cumulated volume of the plurality of voids (the ‘void volume’), subtracted from a volume of the electrically insulating material, divided by the volume of the electrically insulating material. This may be expressed as follows:

VIM = Volume of electrically insulating material (including voids)

Vv = Cumulated volume of voids

Hence, the converter parameter σ may also be expressed as:

Table 24 shows exemplary values for the converter parameter o, expressed in units of 1/mm.

Table 24

Converters according to the present disclosure preferably have values of o greater than or equal to 10/mm to provide good electrical insulation properties. However, values greater than or equal to 50/mm may be preferred, especially at higher operating voltages.

Table 25 shows exemplary values for the insulation fill factor F and the maximum void size R_max.

Table 25

Values of R_max can be determined through an equivalent-sphere method in which measurements of the void size are made for a representative sample of the electrically insulating material, and a maximum void size is statistically estimated under the assumption the voids are spherical and the measured sizes are diameters of spheres.

A power electronics converter utilizing electrically insulating material in a prepackage gap may also be characterized by a converter parameter τ, defined as the product of the dielectric strength of the electrically insulating material and the maximum void size: τ = D x R_max (15)

D = Dielectric strength of electrically insulating material

R_max = Maximum void size in insulation

Table 26 shows exemplary values for the converter parameter τ, expressed in units of Volts.

Table 26

Converters according to the present disclosure preferably have values of T less than or equal to 1 ,000 V to provide good electrical insulation properties. However, values less than or equal to 100 V may be preferred, especially at higher operating voltages. Table 27 shows exemplary values for the dielectric strength D of the electrically insulating material.

Table 27

Figure 5A illustrates how low-side and high-side power semiconductor switching elements 121 L, 121 H of the prepackages 12L, 12H may be electrically connected using the multi-layer carrier substrate 11. The x- and z-directions are indicated.

As in Figure 4B, a planar surface 111b of the substrate 11 has regions 1121 i, 1121H, 1121iii carrying an outer conductive layer. These regions 1121i-iii facilitate connection to the terminals of the power semiconductors 121 L, 121 H by soldering or sintering. The multi-layer substrate 11 is further shown to include an internal conductive layer 112a, which electrically connects to one of the outer layer regions 1121 ii through a set of connections 114a which extends in the z-direction through the substrate 11. The source terminal (S) of the high-side power semiconductor switching element 121 H is electrically connected to the high-side DC input (DC+) through a connection (e.g., a soldered, sintered or glued connection) to the third outer layer region 1121 iii. The drain terminal (D) of the low-side power semiconductor switching element 121 L is electrically connected to the low-side DC input (DC-) through a connection to the first outer layer region 1121i. The drain terminal (D) of the high-side power semiconductor switching element 121 H and the source terminal (S) of the low-side power semiconductor switching element 121 L are electrically connected to each other and to the inner conductive layer 112 of the substrate by connections to the second outer layer region 1121ii.

The internal layer 112 is shown to be thicker in the z-direction than the outer layer regions 1121i-1121 iii. This increased thickness reduces the resistance and thus increases the current carrying capability of the inner conductive layer 112. This reflects the fact that, in this example, the inner conductive layer 112 carries a high current whereas the outer layer regions 1121 i-iii are used as electrical contact and not paths for carrying current between components. The thin outer layer regions 1121i-1121 iii may have a thickness of less than 100 pm (e.g., 50 pm), whereas the thicker inner layer 114 may have a thickness of greater than 100 pm (e.g., 100-400 pm). A plurality (e.g., five, ten or more) of vias may be used to connect a thin outer contact region and the thick inner layer. By increasing the number of vias for one electrical path, the current carrying capability can be increased accordingly.

Figure 5B illustrates how power semiconductor switching elements P1 and P2 may be connected in parallel. It will be understood that P1 and P2 may, for example, be two low-side power semiconductors of one phase of a two-level AC-DC converter. The y- and z-directions are indicated for comparison with Figure 5A.

As in Figure 5A, a planar surface 111b of the substrate 11 has regions 1121 iv, 1121v, 1121vi carrying an outer conductive layer. These regions 1121 iv, 1121 v, 1121 vi facilitate connection to the terminals of the power semiconductors P1 and P2 by, e.g., soldering, sintering or gluing. The multi-layer substrate 11 is further shown to include two internal conductive layers 112b, 112c. Each of the internal layers 112b, 112c electrically connects to one of the outer layer regions 1121 v, 1121 vi through a corresponding sets of connections 114b, 114c which extend in the z-direction through the substrate 11. As in Figure 5A, the internal layers 112b, 112c are thicker than the outer layer regions 1121 iv-vi. The source terminal (S) of the first power semiconductor switching element P1 is electrically connected to the AC side of the converter 10 through a connection to the first outer layer region 1121 iv. The source terminal (S) of the second power semiconductor switching element P2 is electrically connected to the AC side of the converter 10 through a connection to the third outer layer region 1121 vi, which connects on to the ticker inner conductive layer 112c through the connection 114c. The drain terminal (D) of the first power semiconductor switching element P1 and the drain terminal (D) of the second power semiconductor switching element P2 are electrically connected to each other and to the inner conductive layer 112b of the substrate 11 by soldered, sintered or glued connections to the second outer layer region 1121 v.

It is to be understood that the connection arrangements of Figures 5A-5B are merely one example and that connections may be made in various different ways, with different combinations of outer layers, inner layers and z-direction connections. It should further be appreciated that:

- The various internal layers 112a-c are shown to be of the same thickness and at the same depth through the substrate 11. This need not be the case and generally will not be the case. Internal layers can have different thicknesses and be staggered in the z-direction and/or x-y direction.

- The outer layers 1121 i-vi and inner layers 112a-c can be of the same or different thicknesses. Suitable thicknesses will depend to some extent on the application requirements (e.g., power). It is generally easier to fabricate thick internal layers 112a-c through e.g., known PCB manufacturing techniques than thick outer layers which may be provided by e.g., deposition.

In each of the examples described above, the AC-DC converter 10 has only a single phase. It should, however, be appreciated this is only for ease of illustration and explanation and that AC-DC converters according to the present disclosure can and generally will have multiple phases. To this end, Figures 6A-6B illustrate how the concepts described above may be extended to multiple phases. Figures 6A-6B illustrate a two-level, three-phase AC-DC converter in which the low-side and high-side of each phase comprises multiple (in this case eight) parallel-connected power semiconductor switching elements.

Figure 6A is a schematic cross-section of the converter 10. The x-direction and z-direction are indicated. The converter 10 has three phases, designated II, V and W. Each phase has its own set of power semiconductor prepackages: the first phase II has prepackages 12LI-L and 12LI-H; the second phase V has prepackages 12V-L and 12V- H; and the third phase W has prepackages 12W-L and 12W-H. The prepackages 12 of each phase are surface mounted and electrically connected to a common multi-layer planar carrier substrate 11 , which as before may be a PCB. The heat removal sides of the prepackages 12 face a heat sink 15, with thermal interface layers 16U, 16V, 16W between the undersides of the prepackages 12LI-W and the heat sink 15.

For ease of illustration the other components of the commutation cell, namely the gate driver circuit 13, the capacitor(s) 14 and the electrical connections between the components are not shown. It will however be appreciated that these components and their connections will be substantially as described above with reference to Figures 1-5.

In this example there is a common heat sink 15 which serves the entire converter

10, but it will be appreciated there could instead be a separate heat sink for each phase

11, V, W, similar to the arrangement shown in Figure 2 for a single phase. Further, in this example the prepackages sit in a recess 151 in the heat sink 15. The use of a recess may advantageously reduce the thickness of the converter in the z-direction and can also provide a secondary thermal conduction path between the prepackages 12 and the heat sink 15 which passes through the substrate 11. Furthermore, provided the recess is suitably sealed, a liquid or gaseous cooling medium may be flowed through the recess to directly cool the prepackages 12.

In the illustrated example the heat sink 15 and substrate 11 are secured and pressed together by fasteners 17. This is not essential, but the use of fasteners to press the substrate 11 and prepackages 12 to the heat sink 15 may be preferable to ensure a better thermal interface to the heat sink 15.

The heat sink is shown to define barrier walls 152, which separate the recess 151 into three chambers (one for each phase) to provide isolation between the phases. This may be useful for fault mitigation, but in other examples may be omitted. The barrier walls 152 may also not be integral with the heat sink 15, though integral barrier walls may improve the quality of the secondary thermal conduction path between the prepackages 12 and heat sink 15. Where the barrier walls 152 are omitted, a TIL 16 spanning the prepackages of multiple (e.g., all) of the phases II, V, W of the converter 10 may be used.

Figure 6B is a schematic plan view of the converter 10 of Figure 6A. Only the prepackages 12 are illustrated and, as with Figure 3B, it will be understood the prepackages 12 are shown despite being located on the underside of the substrate and being obscured by the heat sink 15. With the y-direction no longer obscured, it can be seen that the low-side and high-side of each phase II, V, W comprises eight power semiconductor switching elements each provided in a power semiconductor prepackage. For example, the eight prepackages 12U-Li to 1211-Ls are labelled.

Those skilled in the art will appreciate that the example of Figures 6A-B may be extended to any number of phases and to any number of parallel-connected power semiconductor switching elements. It will also be appreciated that it can be extended to DC-DC converter circuits, which may only comprise a single power semiconductor switching element or one set of parallel-connected power semiconductor switching elements.

Generally speaking, any power electronics converter can be understood to be formed of one or more ‘logical switches’ each comprising one or more parallel-connected power semiconductor switching elements. In the case of a two-level AC-DC converter, there are two logical switches per phase (one low-side and one high-side logical switch). In the case of a DC-DC converter there may be as few as one logical switch (see e.g., Figure 10A), though other DC-DC converters may comprise multiple logical switches (see e.g., Figure 10B which has four logical switches; two per side of the transformer 250’). Converters according to the present disclosure may comprise any number of power semiconductors (and thus prepackages) per logical switch, with 3 to 10 prepackages per logical switch being typical.

It is worth considering how a change to the number of phases affects the definition of the volume of the commutation cell and the parasitic inductance of the power circuit. Each phase forms part of the commutation cell. Thus, the smallest cuboidal volume that enclose the commutation cell will enclose every phase of the converter. However, each phase circuit is essentially independent from the other phase circuits, with its switching and the conduction between its DC and AC sides being independent of the other phase circuits. Thus, a multi-phase power circuit can, from the perspective of parasitic inductance, be considered equivalent to multiple independent one-phase power circuits, and the parasitic inductance of the power circuit is therefore equal to the parasitic inductance of one of the phases. The parasitic inductance of each phase will be the same (except for small unavoidable variation due to e.g., component manufacturing tolerance and electrical contact quality), so it does not matter which phase is selected. In principle it is possible to intentionally design a converter in which each phase has a different parasitic inductance, but this would be undesirable. By way of specific examples, Table 28 includes specifications of two example converters in accordance with the present disclosure. Both are two-level, three-phase AC-DC converters, but it will be understood this is not intended to be limiting.

Table 28

Figures 7A-7B and Figures 8A-8B illustrate several alternative ways in which the power semiconductor prepackages 12 can be arranged and connected with respect to the multi-layer carrier substrate 11. In each case the gate driver circuit 13 and capacitor(s) 14 are omitted, but it will be understood these could be incorporated into the alternative arrangements substantially as described above.

Figure 7A shows the general converter arrangement 10 described above with reference to Figures 2-6 and in Table 28. Each power semiconductor prepackage 12 is surface mounted on the underside of the multi-layer carrier substrate 11 , with electrical connections between the two taking the form of soldered, sintered or glued connections 113. The other side of the prepackages 12 faces a common heat sink 15, with a TIL 16 providing the mechanical and thermal interface between the prepackage 12 and the heat sink 15.

Figure 7B shows an alternative converter arrangement 10’ in which the prepackages 12 and the heat sink 15 are mounted on opposite sides of the carrier substrate 11. As before, the prepackages 12 are surface mounted to the substrate 11 using soldered or sintered connections 113. The heat sink 15 mechanically and thermally interfaces with the substrate 11 via a TIL 16. The length and thermal resistance of the path of thermal conduction between the prepackages 12 and the heat sink 15 is increased by mounting the prepackages 12 and heat sink 15 on opposite sides of the substrate 11. For this reason, the arrangement of Figure 7A may be preferred to that of Figure 7B. However, the addition of integrated thermally conductive elements 115 into the carrier substrate may provide a useful and sufficient reduction in the thermal resistance. The thermally conductive elements could be, for example, copper vias.

Figures 8A and 8B show further alternative converter arrangements 10”, 10”’ in which the prepackages 12 are embedded within the multi-layer carrier substrate 11. Embedding the prepackages 12 may offer several advantages. Firstly, it eliminates some air gaps from the converter which may otherwise be vulnerable to electrical breakdown due to the high voltages and electric fields which may be used in converters of the present disclosure. Secondly, it is possible to make the converter somewhat more compact in the z-direction because of the elimination of some gaps. Thirdly, there is a reduced need to form, e.g., soldered or sintered connections between the prepackages and the conductive elements of the substrates 11 , which provides some reduction in manufacturing complexity and eliminates a source of failure. However, removing heat from the arrangements 10”, 10’” of Figures 8A and 8B is more challenging.

Figure 8B differs from Figure 8A in that the prepackages 12 are fully embedded within the substrate 11 and thus entirely surrounded by insulating material. In Figure 8A an underside of the prepackage 12 is exposed and flush with one of the planar surface of the substrate 11. Fully embedding the prepackages 12 in the substrate 11 results in a further improvement in the electrical isolation of the power semiconductors 121 , but also further increases the thermal resistance between the prepackages 12 and heat sink 15. In Figure 8B thermally conductive elements 115 are integrated into the substrate 11 to reduce the thermal resistance of the path of the thermal conduction.

In each of the examples described above, the power electronics converters 10, 10’, 10”, 10”’ have taken the form of a two-level AC-DC converter. Those skilled in the art will appreciate that the concepts described can be equally applied to different types of power electronics converters, including alternative AC-DC converter topologies (including multi-level converter topologies) and DC-DC converters. Figures 9-10 illustrate various AC-DC and DC-DC power electronics converter circuits which may be used in accordance with the embodiments described above.

Figure 9A shows a three-phase electrical machine 140 connected to a two-level, three-phase AC-DC converter 100. The electrical machine 140 is configured as a motor and the converter 100 is configured as an inverter, but it will be understood the arrangement of a generator and rectifier would be very similar.

In this example, one end of each phase winding 140u-w of the motor 140 is connected at a common point (the ‘star’ or ‘Y’ point), though an e.g., Delta connection arrangement of the windings 140u-w could also be used. The other end of each phase winding is connected to a corresponding phase leg 110u-w of the inverter 100 at a phase connection point. Each phase leg 110u-w is further connected to high and low DC inputs DC-H, DC-L which may, for example, connect to a DC bus such as the DC bus 330 of Figure 12A. Each phase leg comprises a high- and low-side transistors 112H, 112L and associated parallel diodes 112H-d, 112L-d connected between the high- and low-side DC inputs DC-H, DC-L.

The inverter 100 further includes a smoothing DC-link capacitor 114, which is connected between the high and low DC inputs DC-H, DC-L and a gate driver circuit 113, which is connected to and configured to supply switching signals to the gate terminals of the transistors 112H, 112L. The gate driver circuit 113 generally receives low-power control signals from a controller (not illustrated) and amplifies the low-power signals to supply the gate terminals with switching signals suitable for controlling the on/off state of the transistors 112H, 112L of the phase legs 110u-w.

In this example the transistors 112L, 112H are MOSFETs (e.g., Silicon Carbide (SiC) MOSFETs). Thus, as will be appreciated by those skilled in the art, the parallel diodes 112H-d, 112L-d associated with the MOSFETs 133, 134 may not be discrete components but rather the so-called ‘body diodes’ of the MOSFETs, i.e. , the inherent diode characters of the MOSFETs. In other examples the diodes 112H-d,112L-d may be discrete components separate from the transistors 112H, 112L.

In use, the inverter 100 receives DC electrical power via the DC connections DC- H, DC-L. The gate terminals of the transistors 112H, 112L receive switching signals from the gate driver circuit 113. As will be understood by those skilled in the art, the switching signals switch the transistors 112L, 112H of each phase leg 110u-w between conductive and non-conductive (‘on’ and ‘off’) states, commutating current from the upper and lower branches of each phase leg 110u-w to the respective phase winding 140u-w of the motor 140. Timings and durations of the switching are controlled so that AC electrical power is supplied to the phase windings 140u-w of the motor 140 via the AC phase connection points.

Figure 9B illustrates another inverter circuit 100’. In this example, a four-phase electrical motor 140’ is supplied with electrical power from a two-level DC supply via an AC-DC power electronics converter 100’ comprising four independent H-bridge circuits. For clarity, Figure 10B only shows two of the four H-bridge circuits 110u’, 110v’, connected with a corresponding two of the four phases 140u’, 140v’ of the motor 140’.

Each H-bridge circuit (e.g., H-bridge circuit 110u’) comprises four transistors 112L-1’, 112H-1’, 112L-2’, 112H-2’ and associated parallel diodes connected in an H- bridge configuration between the high and low DC connections DC-H, DC-L and one of the phase windings 140u’ of the motor 140’. A DC-link capacitor 113 is also connected between the DC connections DC-H, DC-L. During operation, the DC connections supply DC electrical power to the H-bridge circuit 110u’ and the gate driver circuit 113 supplies switching signals to the gate terminals of the transistors 112L-1’, 112H-1’, 112L-2’, 112H- 2’. The switching of the transistors between their conductive and non-conductive states affects inversion of the DC power to AC power for supply to the phase windings 140u’ of the motor 140’.

Figure 9C shows an inverter 100” circuit with a DC-side filter that is more complex than the single DC-link capacitor of the converters 100, 100’ of Figures 9A-B. The inverter 100” is, like the inverter 100 of Figure 9A, a two-level three phase inverter. However, it includes a DC filter having three capacitors 114-1 , 114-2, 114-3 and four inductors 119-1 , 119-2, 119-3, 119-4 connected between the DC connections DC-L, DC- H and the converter phase legs 110u-w. Any suitable DC-side filter circuit may be used in accordance with the embodiments described herein.

Figure 10A illustrates a DC-DC converter 200 of the boost type. Buck type and buck-boost type DC-DC converters are also known. The DC-DC converter 200 is connected to a battery 220 at one of its sides and on the other of its sides is connected to, e.g., a DC power channel such as the DC power channel 330 of Figure 11A or the DC power channel 530 of Figure 14A.

The DC-DC converter 200 includes a transistor 212, a gate terminal (g) of which is connected to a gate driver circuit 213. As in the previous examples, the transistor 212 is a MOSFET and the parallel diode associated with the transistor 212 may be an additional discrete diode of the inherent body diode of the MOSFET. The converter circuit 200 further includes a diode 218, a smoothing capacitor 214 (which may be referred to as the input capacitor in the context of a DC-DC converter) and an inductor 219.

In use, the DC-DC converter circuit 200 receives DC power, either from the terminals of the battery 220 or from the DC connections DC-L, DC-H. The gate driver circuit 213 supplies the gate terminal (g) of the transistor with switching signals to control the on/off state of the transistor 212 to affect the desired voltage increase or decrease.

Figure 10B illustrates another another type of DC-DC converter 200’. The DC- DC converter 200’ is of the DC-AC-DC type and includes back-to-back inverter 210’ and rectifier 230’ stages with an intermediate transformer 250’. In this example each of the inverter 210’ and the rectifier 230’ are of the H-bridge type, though it should be appreciated other inverter and rectifier circuits could be used. Each have associated DC- side filters 214’, 234’.

In use, the DC-DC converter 200’ receives DC current from a DC current source, for example the energy storage system 230 or via the DC connections DC-H, DC-L. The gate terminals of the transistors of the H-bridge circuits 210’, 230’ receive switching signals from the gate driver circuit 213, which affects inversion of DC and rectification of AC to supply current to or receive current from the windings of the transformer 250’. The DC output by one of the first and second H-bridge circuits 210’, 230’, is supplied to either the battery 220 or a DC network via the DC connections DC-L, DC-H.

Figures 11-14 illustrate exemplary aircraft power and propulsion systems 300, 400, 500 comprising power electronics converters. The power electronics converters may advantageously be of the efficient, power-dense types described herein.

Figure 11A is a schematic illustration of a purely electric aircraft propulsion system 300. The system 300 includes an electrical energy storage unit 320, in this case a high-voltage battery pack, which supplies a DC power channel 330 with DC electrical power. The DC power channel 330 supplies electrical power to electrical loads including an electric motor 340 which drives rotation of a propulsor 350 (e.g., a propeller or ducted fan). An inverter 310 converts the DC power from the DC power channel 330 to AC power for supply to the windings of the motor 340. In some embodiments, the inverter 310 may be integrated with (e.g., share a common housing structure with) the motor 340 and/or the motor 340 may be integrated with the propulsor 350.

Although not illustrated, the electrical propulsion system 300 may optionally include a DC-DC converter connected between the terminals of the battery pack 320 and the DC power channel 330 to regulate the voltage on the DC power channel. For example, the terminal voltage of the battery pack 320 will tend to drop, e.g., by a factor of up to about two, as the battery pack 320 discharges from a maximum charge level to a lower charge level. By way of an example, the voltage may drop from a maximum voltage level of 900 V to about 450 V over the course of its discharge. A DC-DC converter may therefore be used to boost the terminal voltage to maintain a constant voltage on the DC power channel 330. Other arrangements may omit the DC-DC converter and compensate for the voltage drop and the associated power drop by increasing the current delivered to the loads (e.g., the motor 340).

The electrical power and propulsion system 300 illustrated in Figure 11A comprises only a single power channel (or power ‘lane’) and a single propulsor. In practice, a power and propulsion system may comprise multiple channels and be of a more complex configuration. To illustrate, Figure 12A shows an exemplary electrical vertical take-off and landing (eVTOL) aircraft 360 which has a distributed propulsion system, which in this case comprises six propulsors 350a-f. In this arrangement, each propulsor 350a-f may be associated with its own power channel, e.g., six power channels of the type shown in Figure 11A. In other example, one or more of the propulsors 350a- f may share a power channel so that there are fewer than six power channels.

Still referring to Figure 12A, four of the propulsors 350a-d are coupled to the wings 361 of the aircraft 360, whilst the remaining two propulsors 350e-f are coupled to rear flight surfaces 362. The wings 361 and rear flight surfaces 362 tilt between a VTOL configuration (shown in Figure 12A), in which the propulsors 350a-f provide thrust for lift, and a forward flight configuration, (shown in Figure 12B) in which at least some of the propulsors (the rear propulsors 350e-f in Figure 12B) provide forward thrust. It will be understood that other numbers of propulsors (e.g., four or eight) are possible, as are other eVTOL configurations (e.g., multicopter designs and variations of the illustrated tilt rotor design are also known).

Figure 11B illustrates how six electrically powered propulsors (e.g., the six propulsors 350a-f of Figures 12A-B) may be arranged within an electrical power system 300. The power system 300 includes three independent power system channels 300a, 300b, 300c, each of which is associated with two of the six propulsors 350a-f. In this specific example, each of the first two power channels 300a, 300b is associated with one front propulsor 350b, 350c and one rear propulsor 350e, 350f whilst the third channel 300c is associated with one front-left propulsor 350a and one front-right propulsor 350d. Each propulsor 350a-f is associated with a respective motor 340a-f and a respective inverter 340a-f. Other power system configurations are possible and will occur to those skilled in the art. For example, the system 300 could feature connections between power channels 300a-c and the use of multi-redundant motor winding arrangements to allow power sharing between some or all the power channels 300a-c and to increase faulttolerance and power availability.

The use of a distributed propulsion system such as that illustrated in Figure 11 B and Figures 12A-B can be highly desirable in terms of improved flight characteristics, reduced aerodynamic noise and reduced drag. However, it also generally increases the number of power electronics converters present in the electrical power system. For instance, applying the system of Figure 11 B, the platform could include nine independent power electronics converters (six inverters and three DC-DC converters), which could add considerable weight to the platform and add losses to the power system. Thus, for platforms of this type, the improvements in the efficiency and power-density of the converters disclosed herein may be particularly advantageous.

Figure 13 shows a general arrangement of an engine 400 for an aircraft. The engine 400 is of turbofan configuration, and thus comprises a ducted fan 401 that receives intake air A and generates two pressurised airflows: a bypass flow B which passes axially through a bypass duct 402 and a core flow C which enters a core gas turbine.

The core gas turbine comprises, in axial flow series, a low-pressure compressor 403, a high-pressure compressor 404, a combustor 405, a high-pressure turbine 406, and a low-pressure turbine 407.

In operation, the core flow C is compressed by the low-pressure compressor 403 and is then directed into the high-pressure compressor 404 where further compression takes place. The compressed air exhausted from the high-pressure compressor 404 is directed into the combustor 405 where it is mixed with fuel and the mixture is combusted. The resultant hot combustion products then expand through, and thereby drive, the high- pressure turbine 406 and in turn the low-pressure turbine 407 before being exhausted to provide a small proportion of the overall thrust. The high-pressure turbine 406 drives the high-pressure compressor 404 via an interconnecting shaft. The low-pressure turbine 407 drives the low-pressure compressor 403 via another interconnecting shaft. Together, the high-pressure compressor 404, high-pressure turbine 406, and associated interconnecting shaft form part of a high- pressure spool of the engine 400. Similarly, the low-pressure compressor 403, low- pressure turbine 407, and associated interconnecting shaft form part of a low-pressure spool of the engine 400. Such nomenclature will be familiar to those skilled in the art. Those skilled in the art will also appreciate that whilst the illustrated engine has two spools, other gas turbine engines have a different number of spools, e.g., three spools.

The fan 401 is driven by the low-pressure turbine 407 via a reduction gearbox in the form of a planetary-configuration epicyclic gearbox 408. Thus in this configuration, the low-pressure turbine 407 is connected with a sun gear of the gearbox 408. The sun gear is meshed with a plurality of planet gears located in a rotating carrier, which planet gears are in turn meshed with a static ring gear. The rotating carrier drives the fan 401 via a fan shaft 410. It will be appreciated that in alternative embodiments a starconfiguration epicyclic gearbox (in which the planet carrier is static and the ring gear rotates and provides the output) may be used instead, and indeed that the gearbox 408 may be omitted entirely so that the fan 401 is driven directly by the low-pressure turbine 407.

It is increasingly desirable to facilitate a greater degree of electrical functionality on the airframe and on the engine. To this end, the engine 400 of Figure 13 comprises one or more rotary electrical machines, generally capable of operating both as a motor and as a generator. The number and arrangement of the rotary electrical machines will depend to some extent on the desired functionality. Some embodiments of the engine 400 include a single rotary electrical machine 420 driven by the high-pressure spool, for example by a core-mounted accessory drive 421 of conventional configuration. Such a configuration facilitates the generation of electrical power for the engine and the aircraft and the driving of the high-pressure spool to facilitate starting of the engine in place of an air turbine starter. Other embodiments, including the one shown in Figure 13, comprise both a first rotary electrical machine 420 coupled with the high-pressure spool and a second rotary electrical machine 430 coupled with the low-pressure spool. In addition to generating electrical power and the starting the engine 400, having both first and second rotary machines 420, 430, connected by power electronics, can facilitate the transfer of mechanical power between the high and lower pressure spools to improve operability, fuel consumption etc. As mentioned above, in Figure 13 the first rotary electrical machine 420 is driven by the high-pressure spool by a core-mounted accessory drive 421 of conventional configuration. In alternative embodiments, the first electrical machine 420 may be mounted coaxially with the turbomachinery in the engine 400. For example, the first electrical machine 420 may be mounted axially in line with the duct between the low- and high-pressure compressors 403 and 403. In Figure 13, the second electrical machine 430 is mounted in the tail cone 409 of the engine 400 coaxially with the turbomachinery and is coupled to the low-pressure turbine 407. In alternative embodiments, the second rotary electrical machine 430 may be located axially in line with low-pressure compressor 403, which may adopt a bladed disc or bladed drum configuration to provide space for the second rotary electrical machine 430. It will of course be appreciated by those skilled in the art that any other suitable location for the first and (if present) second electrical machines may be adopted.

The first and second electrical machines 420, 430 are connected with power electronics. Extraction of power from or application of power to the electrical machines is performed by power electronics converters 440. In the present embodiment, the power electronics converters 440 are mounted on the fan case 411 of the engine 400, but it will be appreciated that it may be mounted elsewhere such as on the core of the gas turbine, or in the vehicle to which the engine 400 is attached, for example.

Control of the power electronics converters 440 and of the first and second electrical machines 420 and 430 is in the present example performed by an engine electronic controller (EEC) 450. In the present embodiment the EEC 450 is a fullauthority digital engine controller (FADEC), the configuration of which will be known and understood by those skilled in the art. It therefore controls all aspects of the engine 400, i.e. , both of the core gas turbine and the first and second electrical machines 420 and 430. In this way, the EEC 450 may holistically respond to both thrust demand and electrical power demand.

The one or more rotary electrical machines 420, 430 and the power electronics converters 440 may be configured to output to or receive electric power from one, two or more DC busses or power channels. The DC power channels allow for the distribution of electrical power to other engine electrical loads and to electrical loads on the airframe.

Those skilled in the art will appreciate that the gas turbine engine 400 described above may be regarded as a ‘more electric’ gas turbine engine because of the increased role of the electrical machines 420, 430 compared with those of conventional gas turbines. Turning now to Figure 14A, this illustrates an exemplary power and propulsion system 500 of a hybrid electric aircraft. The system 500 includes a generator set 501 comprising an engine 560 and electrical generator 540a, and a battery pack 520. Both the generator set 501 and the battery pack 520 are used as energy sources to power a motor-driven propulsor 502, an example of which is shown in Figure 14B.

The illustrated propulsion system 500 further comprises a rectifier 510a, a DC distribution bus 530, an inverter 510b and a DC-DC converter 510c. It will be appreciated that whilst one generator set 501 and one propulsor 502 are illustrated in this example, a propulsion system 500 may include more than one generator set 501 and/or one or more propulsor 502.

A shaft or spool of the engine 560 is coupled to and drives the rotation of a shaft of the generator 540a which thereby produces alternating current. The rectifier 510a, which faces the generator 540a, converts the alternating current into direct current which is fed to various electrical systems and loads via the DC distribution bus 530. These electrical systems include non-propulsive loads (not shown in Figure 14A) and the motor 540b which drives the fan 550 of the propulsor 502 via the inverter 510b.

The battery pack 520, which may be made up of a number of battery modules connected in series and/or parallel, is connected to the DC distribution bus 530 via the DC-DC converter 510c. The DC-DC converter 510c converts between a terminal voltage of the battery pack 520 and a voltage of the DC distribution bus 530. In this way, the battery pack 520 can replace or supplement the power provided by the generator set 501 (by discharging and thereby feeding the DC distribution bus 530) or can be charged using the power from the generator set 501 (by being fed by the DC distribution bus 530).

Referring to Figure 14B, in this example the propulsor 502 takes the form of a ducted fan incorporating an electrical machine 540b. The fan 550 is enclosed within a fan duct 551 defined within a nacelle 552 and is mounted to a core nacelle 553. The fan 550 is driven by the electrical machine 540b via a drive shaft 554, both of which may also be thought of as components of the propulsor 502. In this embodiment a gearbox 555 is provided between the electrical machine 540b and the drive shaft 554.

The electrical machine 540b is supplied with electric power from a power source, for example the generator set 501 and/or the battery 520 via the DC bus 530. The electrical machine 540b of the propulsor 502, and indeed the electrical machine 540a of the generator set 501 , may be of any suitable type, for example of the permanent magnet synchronous type. The inverter 510b may be integrated with (e.g., share a common housing structure with) the electrical machine 540b and thus form a part of the propulsor 502. Likewise, the rectifier 510a may be integrated with (e.g., share a common housing structure with) the electrical machine 540a. The DC-DC converter 510c could itself be integrated with the energy storage system 520.

Those skilled in the art will recognise the propulsion system 500 of Figures 14A- B to be of the series hybrid type. Other hybrid electric propulsion systems are of the parallel type, while still others are of the turboelectric type or have features of more than one type. The configuration of the more electric engine 400 of Figure 13 may be considered similar to a parallel hybrid system, with the main distinction being the roles of the electrical machines. For example, the electrical machines of a more electric engine are generally only used in motor mode to start the engine and to improve engine operability, whereas the electric machines of a parallel hybrid propulsion system are used to motor the spools to meaningfully add to the amount of propulsive thrust produced by the turbomachinery.

Those skilled in the art will also appreciate that the hybrid architecture illustrated in Figure 14A is only one example, and other architectures are known and will occur to those skilled in the art.

Various examples have been described, each of which feature various combinations of features. It will be appreciated by those skilled in the art that, except where clearly mutually exclusive, any of the features may be employed separately or in combination with any other features and the invention extends to and includes all combinations and sub-combinations of one or more features described herein.

It should also be noted that whilst the embodiments have been described with reference to an aircraft, and to turbofan engines, it will be understood that the principles of the described electrical systems may be applied to other installations, for example to aircraft with turboprop and open rotor engines, to marine environments such as on a naval vessel and to other transport applications including trains.

Claims

1. A power electronics converter comprising: a converter commutation cell comprising a power circuit and a gate driver circuit, the power circuit comprising at least one power semiconductor switching element and at least one capacitor; and a multi-layer planar carrier substrate defining an x-y direction parallel to a planar surface of the substrate and a z-direction perpendicular to the x-y direction, the multi-layer carrier substrate comprising a plurality of electrically conductive layers extending in the x-y direction and at least one electrical connection extending in the z-direction, wherein: each power semiconductor switching element is comprised in a power semiconductor prepackage, each power semiconductor prepackage comprising one or more power semiconductor switching elements embedded in a solid insulating material; the plurality of electrically conductive layers comprising an internal conductive layer and an outer conductive layer connected by at least one connection extending in the z-direction through the carrier substrate; each power semiconductor prepackage further comprises an electrical connection from at least one terminal of the one or more power semiconductor switching elements to the outer conductive layer of the multi-layer planar carrier substrate, the electrical connection extending in the z-direction through the solid insulating material; and the internal conductive layer of the multi-layer planar carrier substrate is thicker than the outer conductive layer of the multi-layer planar carrier substrate.

2. The power electronics converter of claim 1 , wherein the internal conductive layer is at least twice as thick as the outer conductive layer.

3. The power electronics converter of claim 2, wherein the internal conductive layer is at least three times as thick as the outer conductive layer.

4. The power electronics converter of any one of the preceding claims, wherein the outer conductive layer has a thickness of less than 100 pm.

5. The power electronics converter of any one of the preceding claims, wherein the electrical connection from the at least one terminal of the one or more power semiconductor switching elements to the outer conductive layer of the multi-layer planar carrier substrate comprises: an electrical connection extending in the z-direction through the solid insulating material to an electrical connection side of the power semiconductor prepackage, the electrical connection side of the prepackage forming a flat surface; and a soldered, sintered or glued electrical connection between the electrical connection side of the power semiconductor prepackage and the outer conductive layer.

6. The power electronics converter of claim 5, wherein the soldered, sintered or glued connection spaces apart the power semiconductor prepackage from the planar surface of the multi-layer planar carrier substrate to define a gap, a size of the gap measured in the z-direction being less than or equal to 300 pm.

7. The power electronics converter of claim 6, wherein the size of the gap measured in the z-direction is less than or equal to 200 pm.

8. The power electronics converter of claim 7, wherein the size of the gap measured in the z-direction is in the range 50 pm to 150 pm.

9. The power electronics converter of any one of the preceding claims, wherein the multi-layer planar substrate is: a rigid printed circuit board (PCB); a flexible PCB; or a ceramic carrier substrate.

10. The power electronics converter of any one of the preceding claims, wherein a peak rated power output of the power electronics converter is greater than 10 kW.

11. The power electronics converter of any one of the preceding claims, wherein a total rated capacitance of the power circuit of the commutation cell divided by a peak rated power output of the power electronics converter is less than or equal to 5 nF/W.

12. The power electronics converter of any one of the preceding claims, wherein a blocking voltage of each power semiconductor switching element is greater than 600 V.

13. The power electronics converter of any one of the preceding claims, wherein a product of a parasitic inductance of the power circuit of the commutation cell and a peak rated power output of the power electronics converter is in the range 0.05 mHW to 1.5 mHW.

14. The power electronics converter of any one of the preceding claims, being an AC-DC converter.

15. The power electronics converter of any one of claims 1 to 13, being a DC- DC converter.

16. An electrical propulsion unit (EPU) for an aircraft, the EPU comprising an electric motor and the AC-DC power electronics converter of claim 14, the AC-DC power electronics converter configured as an inverter and arranged to supply current to a winding of the electric motor.

17. A gas turbine engine comprising a spool; an electrical machine having a rotor mechanically coupled to the spool; and the AC-DC power electronics converter of claim 14, wherein the AC-DC power electronics converter is arranged to supply current to or receive current from a winding of the electric machine.

18. An electrical power system for an aircraft comprising a power electronics converter according to any one of claims 1 to 15.

19. An aircraft comprising the EPU of claim 16, the gas turbine engine of claim 17 or the electrical power system of claim 18.