This disclosure relates to an internally-cooled flange for mounting electronic devices.
Many of today's electronic devices are in search of greater performance, but in smaller package sizes. This increases the power-density of the devices, which can be challenging for thermal management systems. For example, power amplifiers for defense, radar, and telecom look to increase the power levels of devices while reducing the sizes of packages. Insulated-gate bipolar transistors (IGBTs) for electric vehicles seek to increase the amperage rating of devices, while maintaining reliable device temperatures. Laser diodes for defense, communications, and industrial equipment can increase optical power, but may then experience undesirable wavelength shifts due to temperature variations. In these examples, and others, the thermal management challenges created by increasing power-density require the introduction of higher performance thermal management approaches, and ones that also come in small packages.
Current approaches to cooling these types of devices involve the attachment of the device to a conductive metal flange. This flange typically forms part of the final component's package. Heat is conductively transferred from the device to the flange, where it is spread out and then dissipated to a system-level heat sink. These strategies have historically been effective but are reaching their performance limits for high power-density devices. Further, these solutions require large, heavy, and expensive system-level metal heat sinks to help dissipate the heat from the flange. Increasing the size of these heat sinks is incompatible with the desired miniaturization of modern electronics.
Advanced approaches to cooling high power devices include modification to the actual electronic device, for example by etching microchannels into the chip or substrate. These approaches are very effective and can be built into small package sizes. However, such invasive approaches may not be compatible with existing assembly processes and must typically be considered during initial device design and fabrication.
It would, therefore, be useful to have an electronics flange that includes high performance cooling features to: eliminate the need for additional metal heat sinks; produce better cooling to accommodate higher power devices; remain compatible with existing flange assembly architectures; and require no modification or fluid-contact to the electronic device itself.
All examples and features mentioned below can be combined in any technically possible way.
In one aspect a flange for cooling an electronic component includes a heat transfer portion with an inner surface, and an opposed outer surface that is configured to be thermally coupled to the electronic component, a high-pressure fluid reservoir, a fluid inlet in fluid communication with the high-pressure reservoir, the inlet configured to conduct single-phase cooling fluid into the flange, a low-pressure fluid reservoir that is in fluid communication with the inner surface of the heat transfer portion, a fluid outlet in fluid communication with the low-pressure reservoir, the outlet configured to conduct the fluid out of the flange, and a plurality of fluid nozzles that are each configured to transmit the fluid from the high pressure reservoir to the low pressure reservoir in the form of jets that are configured to strike the inner surface of the heat transfer portion.
Some examples include one of the above and/or below features, or any combination thereof. In an example a perimeter can be drawn around the plurality of fluid nozzles without encompassing the fluid outlet. In an example the fluid nozzles are configured non-uniformly relative to the heat transfer portion, to provide more effective cooling to certain areas for reduction of temperature gradients across the electronic component. In some examples the flange is of unitary structure. In an example the flange is fabricated using additive manufacturing. In some examples the plurality of fluid nozzles form microjet nozzles. In an example the microjet nozzles serve to form jets that are configured to strike substantially perpendicularly to the inner surface of the heat transfer portion, to create fluid flow with substantially high momentum in said perpendicular direction. In an example the flange is configured to serve as an electronics base plate.
Some examples include one of the above and/or below features, or any combination thereof. In some examples the flange is fabricated from at least two distinct members that are joined together. In an example a first member comprises the heat transfer portion that is made from a material with high heat conductivity. In an example a second member is made from a material with lower heat conductivity than that of the first member. In an example the flange further includes at least one hole or slot that is configured to attach the flange to another structure. In an example the heat transfer portion is configured to provide a short, direct path from a primary thermal interface of the electronics component to the inner surface of the heat transfer portion. In an example the fluid nozzles comprise orifices through a thickness of an internal microjet nozzle plate of the flange. In an example the electronic component comprises at least one transistor. In an example the electronic component comprises at least one laser diode.
In another aspect a flange that is configured to serve as a base plate for and to cool an electronic component includes a heat transfer portion with an inner surface, and an opposed outer surface that is configured to be thermally coupled to the electronic component, wherein the heat transfer portion is configured to provide a short, direct path from a primary thermal interface of the electronics component to the inner surface of the heat transfer portion, a high-pressure fluid reservoir, a fluid inlet in fluid communication with the high-pressure reservoir, the inlet configured to conduct single-phase cooling fluid into the flange, a low-pressure fluid reservoir that is in fluid communication with the inner surface of the heat transfer portion, a fluid outlet in fluid communication with the low-pressure reservoir, the outlet configured to conduct the fluid out of the flange, and a plurality of fluid microjet nozzles that are each configured to transmit the fluid from the high pressure reservoir to the low pressure reservoir in the form of jets that are configured to strike the inner surface of the heat transfer portion.
Some examples include one of the above and/or below features, or any combination thereof. In some examples the flange is fabricated from at least two distinct members that are bonded together, wherein a first member comprises the heat transfer portion that is made from a material with high heat conductivity. In an example a second member is made from a material with lower heat conductivity than that of the first member. In an example the fluid microjet nozzles comprises orifices through a thickness of an internal microjet nozzle plate of the flange.
For a better understanding of the present disclosure, reference is made to the accompanying drawings, in which:
This disclosure describes the use of a flange, for use in electronics packaging, that contains fluid microjets within its interior to produce effective cooling for the electronic device to which it is attached. The disclosure further describes several possible embodiments of the microjet-cooled flange as all, or part, of a package for devices like power amplifiers, IGBTs, or laser diodes. The disclosure adds new functionality (e.g., advanced microjet cooling) that is integrated within a common packaging component (e.g., the flange) without requiring fluid contact with the electronic device.
Many electronic components involve a semiconductor device that is packaged into an assembly for use in larger, system-level assemblies. In the packaging process, many of these semiconductor devices are affixed onto a small plate, or flange. This flange offers many benefits, including mechanical support, conductive cooling, and mounting features. These features are critical for later system integrators.
These flanges are commonly made from small metal plates, with defined dimensions and interface specifications. Metal plates offer stiff support to avoid stress on the semiconductor, conductive paths for thermal cooling of the semiconductor, coefficient of thermal expansion matching with the electronic die, and easily recognizable mounting features in the form of leads, screw holes, or slots.
However, these metal flanges and their conductive cooling paths are becoming insufficient for cooling of advanced electronic devices with high power density. As a result, many systems must attach supplemental, large, metal heat sinks to the device's existing flange. This approach leads to very large implementations, and ultimately limits the power, and performance, of new electronic devices that are produced, packaged, and sold with industry-standard flange architectures.
Microjet cooling is a technique for cooling high-power devices that is characterized by fluid moving through a nozzle to form a small jet of fluid with substantially greater momentum in one direction than another. When this high-momentum fluid impacts a surface, it greatly compresses the thermal boundary layer at that surface, producing very high heat transfer. Microjet cooling technology has been demonstrated to produce heat transfer coefficients in excess of 200,000 W/m2K, more than ten times that of competing approaches (e.g., microchannels 20,000 W/m2K). This allows the fluid to extract heat from power-dense devices, without the need for additional metal heat spreaders. Microjet cooling is further described in US Patent Application Publication 2019/0013258 and International Patent Application Publication WO 2019/018597, the disclosures of which are incorporated herein by reference in their entireties for all purposes.
This disclosure describes a microjet-cooled flange for electronics packaging. The flange produces high-performance cooling by using fluid microjets that are contained within the flange. The flange includes one or more fluid inlet, inlet reservoir, microjet nozzle plate, fluid microjet, internal heat transfer surface, exit reservoir, and fluid exit.
The microjets (105) formed by the nozzles (104) in the microjet nozzle plate (103) are issued with substantial fluid velocity and strike the heat transfer surface (106). Notably, in all embodiments this heat transfer surface (106) is interior to the flange. That is, the microjet fluid does not strike the actual semiconductor surface or any other surface outside the extents of the flange. This decouples the electronic device from the flange, leading to greater interchangeability and compatibility with various device types. The semiconductor or electronic device to be packaged (and cooled) may then be attached to the external mating surface (109).
When the microjets (105) strike the heat transfer surface (106), they transfer heat from the electronic device to the coolant fluid. This fluid then occupies the exit reservoir (107) and eventually exits the flange through one or more exit ports (108).
The above description is not meant to limit the versatility of the flange. Other variations may be realized. For example, there may be only a single microjet or an array of microjets. One or more ports may also be located on the same side of the electronic device. More than one electronic device may also be mounted to a single flange. Other variations, including multiple layers between the microjets and device, for example, may also be realized to similar effect.
The microjet-cooled flange may also be manufactured in different ways, which may influence the design. For example,
The fluid cooling and routing is contained completely within the microjet-cooled flange. That is, the coolant fluid does not contact the heat generating device directly. Fluid (and cooling) remains within the flange, except for the inlet and exit ports.
The microjet-cooled flange may also include features to allow it to be compatible with many existing packaged devices. For example, the microjet-cooled flange may include holes or slots (306) for attaching the packaged device to a circuit board or other next-level assembly. Also possible are orientation or alignment fiducials (307) for ease of part identification.
On this microjet-cooled flange may be disposed a power amplifier circuit (501). The power amplifier circuit may be a semiconductor device made, for example, from silicon, gallium nitride, silicon carbide, gallium arsenide, diamond, or another material. The power amplifier (501) may be disposed on a substrate (502) for mechanical support and thermal conductivity. The substrate may be, for example, silicon carbide, copper, copper-molybdenum, aluminum, diamond, or another material. Covering the top side of the power amplifier circuit (501) may be a cap (505) that protects the circuit. The cap (505) may be one material or a combination of materials including, for example, plastic, epoxy, solder, or other materials. The power amplifier (and supporting items) may then be attached to the microjet-cooled flange by, for example, solder, epoxy, adhesive, or other thermal interface material (503).
The power amplifier circuit, substrate, package, and interface material are typically designed by the circuit manufacturer to minimize the thermal resistance (and, therefore, temperature rise) of the encapsulated device. This means that there usually exists a strong thermal path from the power amplifier (501) to the lower surface of the interface material (503). With a low-performance thermal management solution for the power amplifier to mount to, however, the temperatures rises can still be significant. The microjet-cooled flange provides a short, direct path from the circuit's primary thermal interface (503) to the heat transfer surface (511) and the coolant fluid. This greatly minimizes the temperature rise of the device, and it occurs within a compact form factor that is already in use in many packaging facilities, and can easily affix to new or existing circuit boards with standard mounting definitions and fasteners (506).
In operation, a single phase fluid is administered. This fluid may be any suitable coolant, including air, water, ethylene glycol, propylene glycol, ethanol, R134A, ammonia, or any other fluid. A combination of two or more of these fluids may also be used. Fluid-tight seals of any type may be formed between layers containing fluid, which may include the use of epoxies, gaskets, soldering, bonding, or any other suitable method. Of course, the construction may not use different layers and may also be a single monolithic part.
The inlet (507) and outlet (513) may be of any type. They may also, for example, provide transmission to/from other components that require cooling. Such other components may be in neighboring parts of the assembly and may require transmission via tubing to other parts of the system. The inlet and outlet may interface with a fitting (e.g. barb, quick disconnect, compression), tubing, a manifold, or any other suitable method of interfacing fluids.
While
In all embodiments, the electronic device to be cooled (here, for example, the power amplifier) is mounted to the surface opposite the internal heat transfer surface (710). The material between the mounting surface and the internal heat transfer surface may be any material, but is typically a thermally conductive metal like copper, copper-molybdenum, or aluminum, or another conductive material such as silicon carbide or diamond.
Coolant fluid enters the microjet-cooled flange through one or more inlets (707) and proceeds to flow through the inlet reservoir (713). A microjet nozzle plate (708) separates the inlet reservoir (713) from the exit reservoir (711). The fluid is forced through one or more orifice (709) in the nozzle plate (708), forming one microjet per orifice. It is these microjets that then strike the internal heat transfer surface (710) causing heat to be transferred from the surface to the fluid. Due to the close proximity of the device (705) and the transfer of heat to the fluid (at surface 710), and owing to the high heat transfer coefficients produced by microjets, the heat is effectively dissipated from the component to maintain proper operating temperatures, even with very large power dissipation.
After transferring the heat from the surface to the fluid, the fluid traverses the outlet reservoir (711) and exits the microjet-cooled flange through one or more exit ports (712). These ports may be open ports, or may have other methods of attachment including, for example, tubes, barbed fittings, quick disconnects, push-to-connect fittings, or a separate manifold for fluid conveyance.
The microjet-cooled flange can offer near equally high performance even when non-mounting portions (703, 704) are made from any number of other materials, including low conductivity ones such as plastic. These layers do not substantially participate in the heat transfer.
On this microjet-cooled flange may be disposed an IGBT device. The IGBT may be comprised of a semiconductor device made, for example, from silicon, gallium nitride, silicon carbide, gallium arsenide, diamond, or another material. The IGBT chip or chips (901) may be disposed on a substrate (902) for mechanical support and thermal conductivity. The substrate may be, for example, polyimide, fiberglass, ceramic, alumina, direct bonded copper, or another material. Attachment (903) between the device (901) and the substrate (902) may be done, for example, by solder, epoxy, direct bonded copper, or another method. Individual chip or chips (901) within the IGBT package may be electrically connected to each other or to the substrate by, for example, wire bonds (904).
The chips, substrate, and other components forming the IGBT are attached to the microjet-cooled flange (906). This attachment (905) may be done by a variety of methods, including epoxy, solder, adhesive, or other method. In this embodiment, the microjet-cooled flange serves as the primary thermal path and a structural support for the IGBT package. The IGBT circuitry is typically protected by a cover (907). The cover (907) may be one material or a combination of materials, for example, plastic, epoxy, solder, or other materials. Covers typically include cutout features for electrical pin (908) egress. Any volume (909) between the components and the cover may be left empty or may be filled with a gap filler or epoxy, for example.
The IGBT circuit, substrate, package, and interface material are typically designed by the circuit manufacturer to minimize the thermal resistance (and, therefore, temperature rise) of the encapsulated device. This means that there usually exists a strong thermal path from the IGBT chips (901) to the lower surface of the interface material (905). With a low-performance thermal management solution for the IGBT to mount to, however, temperature rises can be significant. The microjet-cooled flange provides a short, direct path from the circuit's primary thermal interface (905) to the heat transfer surface (914) and the coolant fluid. This greatly minimizes the temperature rise of the device, and it occurs within a compact form factor that is already in use in many packaging facilities and can easily affix to new or existing assemblies.
In operation, a single phase fluid is administered. This fluid may be any suitable coolant, including air, water, ethylene glycol, propylene glycol, ethanol, R134A, ammonia, or any other fluid. A combination of two or more of these fluids may also be used. Fluid-tight seals of any type may be formed between layers containing fluid, which may include the use of epoxies, gaskets, soldering, bonding, or any other suitable method. Of course, the construction may not use different layers and may also be a single monolithic part.
The inlet (910) and outlet (916) may be of any type. They may also, for example, provide transmission to/from other components that require cooling. Such other components may be in neighboring parts of the assembly and may require transmission via tubing to other parts of the system. The inlet and outlet may interface with a fitting (e.g. barb, quick disconnect, compression), tubing, a manifold, or any other suitable method of interfacing fluids.
Heat generated by the IGBT is conducted through any package layers, including the substrate (1102) and attachment (1103) to reach the top exterior surface of the microjet-cooled flange. This heat is then conducted through a short distance (e.g., the thickness may be about 500 μm. Other thicknesses are possible) to the internal heat transfer surface (1109) of the microjet-cooled flange.
Coolant fluid enters the microjet-cooled flange through one or more inlet ports (not shown) and into the higher-pressure inlet reservoir (1107). The coolant is then driven through one or more orifices (1108) in the microjet nozzle plate that separates the higher-pressure reservoir (1107) and the lower-pressure reservoir (1110). As it passes through each nozzle (1108), the fluid forms a microjet that is directed at, and later strikes, the internal heat transfer surface (1109). This leads to effective transfer of the heat between the internal heat transfer surface and the fluid, ultimately cooling the IGBT device.
On this microjet-cooled flange may be disposed a laser diode package. The laser diode may be comprised of some combination of semiconductor laser diode chips (1201), base plate (1202), and other components which may include, for example, optical components and fiber couplings. The laser diode components (1201) may be disposed on a base plate (1202) for mechanical support and thermal conductivity. The substrate may be, for example, aluminum, copper, ceramic, or another material. Attachment between the devices (1201) and the base plate (1202) may be done, for example, by solder, epoxy, or another method.
The chips, substrate, and other components forming the laser diode are attached to the microjet-cooled flange (1204). This attachment (1203) may be done by a variety of methods, including epoxy, solder, adhesive, or other material. In this embodiment, the microjet-cooled flange serves as the primary thermal path and a structural package for the laser diode. The package is typically protected by a lid (1207). The lid (1207) may be, for example, plastic, aluminum, or another material.
In some embodiments, other portions of the package including, for example, the side walls (1205) may be integrated with the microjet-cooled flange (1204). These side walls may form a continuous perimeter around the laser diode components or may have cutouts for inlet/egress from the package, for example cutouts for electrical leads, optical fiber, or an optical window (1206).
The laser diode, base plate, package, and interface material are typically designed by the system integrator to minimize the thermal resistance (and, therefore, temperature rise) of the packaged device. This means that there usually exists a strong thermal path from the diode chips (1201) to the lower surface of the interface material (1203). With a low-performance thermal management solution for the diode to mount to, however, the temperatures rise can be significant which may cause reduced performance, such as undesirable wavelength shifts. The microjet-cooled flange provides a short, direct path from the circuit's primary thermal interface (1203) to the heat transfer surface (1213) and the coolant fluid. This minimizes the temperature rise of the device. Such a microjet-cooled flange for laser diodes can be designed for interchangeability at the system level by consideration of fastener compatibility (1208) with current mounting techniques.
In operation, a single phase fluid is administered. This fluid may be any suitable coolant, including air, water, ethylene glycol, propylene glycol, ethanol, R134A, ammonia, or any other fluid. A combination of two or more of these fluids may also be used. Fluid-tight seals of any type may be formed between layers containing fluid, which may include the use of epoxies, gaskets, soldering, bonding, or any other suitable method. Of course, the construction may not use different layers and may also be a single monolithic part.
The inlet (1209) and outlet (1215) may be of any type. They may also, for example, provide transmission to/from other components that require cooling. Such other components may be in neighboring parts of the assembly and may require transmission via tubing to other parts of the system. The inlet and outlet may interface with a fitting (e.g. barb, quick disconnect, compression), tubing, a manifold, or any other suitable method of interfacing fluids.
The microjet-cooled flange (1303) may be comprised of a single layer, or as distinct layers (1405, 1406, 1407) that are then attached via any leak-free method. The microjet-cooled flange may also be integrated with perimeter walls (1304) that surround the sides of the electrical and optical components, forming much of the laser diode package. These walls (1304) may also include cutouts which may, for example, be used for optical windows or fiber couplings (1305).
Coolant fluid enters the microjet-cooled flange through one or more inlet ports (1410) and occupies the higher-pressure inlet reservoir (1411). This higher-pressure inlet reservoir (1411) is separated from the lower-pressure outlet reservoir (1415) by a microjet nozzle plate (1412). Disposed within this nozzle plate are one or more orifices (1413). As fluid passes through each orifice, a microjet is formed. The microjets are directed toward, and strike, the internal heat transfer surface (1414). At this surface, the heat is transferred from the conductive package to the coolant fluid. The fluid then traverses the outlet reservoir (1415) and exits out one or more exit ports (1416).
In yet another embodiment, a microjet flange can be configured to be compatible with the native flange commonly sold in electronics packages to form a microjet-cooled flange.
The microjets (1605) formed by the nozzles (1604) in the microjet nozzle plate (1603) are issued with substantial fluid velocity and strike the heat transfer surface (1606). Notably, as in all embodiments, this heat transfer surface (1606) is interior to the flange. That is, the microjet fluid does not strike the actual semiconductor surface or any other surface outside the extents of the flange (1630). The semiconductor or electronic device to be packaged (and cooled) (1610) may then be attached to the external mating surface (1609).
When the microjets (1605) strike the heat transfer surface (1606), they transfer heat from the electronic device to the coolant fluid. This fluid then occupies the exit reservoir (1607) and eventually exits the flange through one or more exit ports (1608).
In all embodiments, the nozzles may be disposed in arrays so as to provide cooling for electronic devices of a range of different sizes. Such devices may contain length scales that range from 5-50 mm, for example. Therefore, the size, location, and distribution of nozzles are carefully chosen to provide adequate cooling of the entire device. The nozzles may be disposed in linear arrays, circular arrays, or any other pattern that serves to help cover the surface of the heat-generating devices. The nozzles may be far apart or close together, details of which are carefully chosen in balancing thermofluidic considerations such as, for example, heat transfer and pressure drop.
In certain electronic devices, the heat may not be generated uniformly across the device surface to be cooled. Such sections of the surface where more heat is being generated are therefore more prone to increases in temperature, sometimes referred to as “hot spots”. In these cases, it may be advantageous to concentrate nozzles nearer to the hot spots of higher heat generation, while having more sparsity in the array where there is lower or no heat generation. This allows for improved cooling efficiency, as better cooling occurs using the same amount of fluid flow, compared to a case where nozzles are uniformly disposed on the nozzle plate.
In addition to the distribution of nozzles, the size and shape of each individual nozzle may vary across the array to balance tradeoffs of, for example, pressure, flow rate, and heat transfer, with the heat generating character of the electronic device. For example, a set of jets with lower heat transfer capability may be administered around areas of low heat generation, while jets with high heat transfer capability may be administered near hot spots.
As part of this disclosure, electronics flanges (as used on many electronics packages including power amplifiers, IGBTs, and laser diodes) are constructed with internal fluid routing, an internal microjet nozzle plate, and an internal heat transfer surface to bring high-performance cooling within the package. This approach brings active microjet cooling within the device flange, but without direct contact between the coolant fluid and the device. This approach greatly reduces the thermal path for heat generated by the device, in a sealed subassembly with defined inlet/exit ports. The microjet-cooled flanges are designed to take the place of commonly available solid metal flanges.
In one embodiment, the present disclosure discloses a method of cooling a semiconductor device by using the device flange. First, the device or its carrier is disposed onto the mating surface of the microjet-cooled flange. Within this microjet-cooled flange, coolant fluid enters and fills a high-pressure reservoir. An internal microjet nozzle plate separates this high-pressure reservoir from a low-pressure reservoir. The coolant fluid is then driven through one or more orifice in this nozzle plate, forming one microjet per nozzle. The microjets issue into the low-pressure reservoir, striking the internal heat transfer surface of the flange. The device (or carrier) is mounted to the opposite (exterior) surface of the internal heat transfer surface. Heat is transferred from the device, through a short flange thickness, and into the coolant fluid. The fluid then exits the microjet-cooled flange, taking the waste heat with it.
The present disclosure is not to be limited in scope by the specific embodiments described herein. Indeed, other various embodiments of and modifications to the present disclosure, in addition to those described herein, will be apparent to those of ordinary skill in the art from the foregoing description and accompanying drawings. Thus, such other embodiments and modifications are intended to fall within the scope of the present disclosure. Furthermore, although the present disclosure has been described herein in the context of a particular implementation in a particular environment for a particular purpose, those of ordinary skill in the art will recognize that its usefulness is not limited thereto and that the present disclosure may be beneficially implemented in any number of environments for any number of purposes. Accordingly, the claims set forth below should be construed in view of the full breadth and spirit of the present disclosure as described herein.
This application claims priority of PCT/US20/27599, filed on Apr. 10, 2020, which claimed priority of Provisional Patent Application 62/831,888, filed on Apr. 10, 2019. The entire disclosures of both applications are incorporated herein by reference for all purposes.
Number | Date | Country | |
---|---|---|---|
62831888 | Apr 2019 | US |
Number | Date | Country | |
---|---|---|---|
Parent | PCT/US20/27599 | Apr 2020 | US |
Child | 15931284 | US |