TWI616740B - Computing device - Google Patents

Abstract

A computing device includes a first connector adjacent a first wall. The first connector is in communication with a chip package that is positioned to be spaced apart from the first wall by a first cable. The wafer package includes a wafer supported by a support layer. The wafer can be supported by a substrate and/or a circuit board. A second connector can be positioned adjacent a second wall and can also be in communication with the wafer package via a second cable. If desired, the substrate or circuit board can include a signal board connector configured to terminate at an end of the first cable and an end of the second cable The board connectors are joined.

Description

Computing device

The present invention relates to the field of high frequency signal transmission, and more particularly to a computing system positioned in a base.

Computing devices (such as routers, servers, switches, etc.) need to operate at high data transfer speeds to meet the increasing demand for bandwidth and streaming audio and video delivery in many end-user devices. . These devices include a base that supports a circuit board that in turn supports various circuits and that uses a master chip component (such as an application specific integrated circuit (ASIC)) mounted on the circuit board. A signal transmission line extending between a field programmable gate array (FPGA), a digital signal processor (DSP), and the like, and a connector mounted on the circuit board. These transmission lines are formed with conductive traces on the board or within the board and extend between the main wafer element and an external connector or circuit of the device.

As can be appreciated, integrated circuits (often referred to as wafers) are the heart of these electronic devices. These wafers typically include a processor having a die that is connectable to a substrate (package of the substrate) through conductive solder bumps. The package may include micro-vias or plated vias that extend through the substrate to the solder balls. These welds The ball may include a ball grid array that is mounted to the board through a ball grid array. The circuit board includes a plurality of traces designated to define a transmission line including a ground signal for differential signal pairs, associated with differential signal pairs, and various low speeds for power, clock, and other functions. Transmission line. These traces are routed from the ASIC to the device's I/O connector (external connector is connected to the I/O connector) and other traces are routed from the ASIC to the backplane connector (backplane connector) Allowing the device to be connected to an overall system (such as a network server, etc.) or other traces still arranged to be arranged from the ASIC to the electronic components and circuits on the motherboard or another circuit board of the device .

A typical circuit board is typically formed from an inexpensive material such as the well-known inexpensive FR4. Although FR4 is inexpensive, it is well known that FR4 will cause losses in high speed signal transmission lines that transmit data at rates of about 6 Gbps and higher (e.g., at signal transmission frequencies above 3 GHz). These losses increase with increasing frequency, and thus FR4 materials are unsatisfactory for high speed data transmission applications at signal transmission frequencies of about 10 GHz and above. In order to use FR4 as a board material for high frequency signal transmission lines, designers may have to employ amplifiers and equalizers, which increases the final cost of the device.

The overall length of the signal transmission line in the FR4 board can exceed threshold lengths (about 10 inches) and can include bends and turns that can create signal reflection and noise problems as well as additional losses. As mentioned above, losses can sometimes be corrected by using amplifiers, repeaters, and equalizers, but these components also add to the final cost of manufacturing the board and further complicate the layout of the board. In addition, the signal transmission line is on the board The arrangement in the middle may require multiple turns and/or transitions. These turns and transitions occurring at the termination points along the signal transmission line tend to reduce the signal to noise ratio. In addition, transitions and terminations tend to create impedance discontinuities, and impedance discontinuities cause reflections in the signal, which makes it difficult to overcome the signal to noise issue simply by increasing the power of the transmission. As a result, as data rates increase, the use of a circuit board (especially when using FR4) has become increasingly difficult to transmit signals over an excessive distance even with more expensive materials. As a result, some people will appreciate further improvements.

A bypass cable assembly is used to provide a high speed data transmission line extending between a chip or chip set of a device and a backplane or circuit board. The bypass cable assembly includes a cable containing a signal transmission line that avoids or circumvents the use of a supported circuit board regardless of the material of the structure.

In such an application, an integrated circuit in the form of a wafer, such as an application specific integrated circuit (ASIC) or field programmable gate array (FPGA), is provided as part of an overall chip package. The wafer can be mounted on a package substrate through a conventional solder bump or the like and can be enclosed in the substrate through an encapsulating material and integrated with the substrate, and the sealing material overlies the wafer. And a portion of the substrate. The package substrate can include traces or leads extending from the solder bumps on the bottom of the wafer to one end regions on the substrate. The substrate may also support a connector or contact pad or may alternatively be mounted on a circuit board including connecting a communication point on the substrate to the circuit Contact pads on the board. A first connector can then be mounted on the circuit board for interfacing with a cable assembly. The first connector can interface with a second connector if desired. The cable terminated to the first connector or the second connector extends to an external interface, such as an I/O connector and a backplane connector.

The wafer package may include a plurality of contacts in the form of solder bumps disposed on a lower side of the wafer package for providing slave logic circuits, clock circuits, power supply circuits, and low speed signal circuits as well as high speed signals The connection between the circuit to the traces on the motherboard of a device that uses a chip package. If the substrate directly supports a connection interface, the contacts associated with the high speed signal circuit of the wafer are removed from the bottom of the wafer package because the high speed traces are no longer arranged at the bottom of the wafer package. However, if the substrate is mounted on a circuit board, the high speed traces can be arranged on the board and extended a short distance to a suitable connector.

Cables for such assemblies can be designed for differential signaling and preferably biaxial cables that are encased in pairs of signal conductors within the dielectric cladding. To form two wires or a pair of signal wires. The first ends of the pair of wires are typically terminated to respective wafer packages and the second ends of the pairs of wires are directly terminated to the inlet or outlet connectors (such as I/O connectors and backplane connectors) Terminal.

The computing device of the present invention includes a housing having a first wall, a chip package including a wafer electrically coupled to a first signal board connector, and a first connector supported adjacent the first wall, The first connector includes a housing having a pair of interfaces and a base, the base supporting a first pair of terminals, each of the terminals of the first pair of terminals each having a contact portion in the docking interface a first cable having a pair of conductors, the first cable including a first end and a second end The first end of the first cable is connected to the first pair of terminals; a first board connector, the first board connector is terminated to the second end of the first cable and configured Docking the first signal board connector; and an electronic component circuit that supplies power to the wafer.

In some implementations, the wafer is supported by a signal board and the first signal board connector is mounted on the signal board.

In some implementations, the housing includes a second wall, the computing device includes a second connector supported adjacent the second wall, and the second connector is terminated to a second cable The first cable has a second end terminated by a second end of the second board connector, wherein the signal board supports a second signal board connector, and the second signal board is connected The device is electrically coupled to the wafer, and the second board connector is configured to interface with the second signal board connector.

In some implementations, the wafer package includes a substrate that supports the wafer.

In some implementations, the substrate is supported by a signal board that does not extend to the first connector.

In some implementations, the first connector is supported by a tray.

In some implementations, the wafer is electrically connected to the plurality of first signal board connectors, and the plurality of first signal board connectors are each docked to a different first via a corresponding first board connector. Connector.

30‧‧‧Second connector

45‧‧‧ solder bumps

47‧‧‧Substrate

47a‧‧‧Underside

48‧‧‧through holes, microvias or traces

49‧‧‧Contacts

50‧‧‧ Computing equipment

51‧‧‧ solder bumps

52‧‧‧ wafer

52-1‧‧‧ Chip package

52-2‧‧‧ motherboard

52a, 52b, 52c‧‧‧ conductive traces

53‧‧‧Substrate

54‧‧‧ Chip package

54-1‧‧‧ Traces

54-2‧‧‧Contact pads

54-3‧‧‧ Termination area

54-4‧‧‧Edge

54-5‧‧‧ Sealing body

55‧‧‧ side wall

56‧‧‧First Wall

57‧‧‧ second wall

58‧‧‧Power supply

59‧‧‧ Refrigeration components

60‧‧‧ Cable

61‧‧‧Signal conductor

61-1‧‧‧ dielectric coating

61-2‧‧‧Addition line

61-3‧‧‧ outer shield

61-4‧‧‧Outer insulation skin

62‧‧‧ motherboard

63‧‧‧Shell

63'‧‧‧Shell

63a'‧‧‧ partition wall

63b'‧‧‧ tail

64a, 64b‧‧‧ side wall

65‧‧‧ Back wall

66‧‧‧Wire-to-board connector

67‧‧‧ entrance

68‧‧‧ bottom wall

69‧‧‧ feet

70‧‧‧Socket connector

71‧‧‧ radiator

72‧‧‧floor

73‧‧‧ Side connecting wings

74‧‧‧ openings

75‧‧‧ inner wing

75a‧‧‧Contact sheet

76‧‧‧Snap or sheet

77a, 77b‧‧‧bump or flange

78‧‧‧Internal

79‧‧‧ top wall

80, 80'‧‧‧ first connector

Port 80a'‧‧‧

81‧‧‧ access opening

82‧‧‧heat transfer element

84‧‧‧ base

85‧‧‧Bottom contact surface

86‧‧‧ Holder

87‧‧‧ top board

88‧‧‧Contact arm

89‧‧‧EMI gasket

90‧‧‧Back cover

91‧‧‧ recess

92‧‧‧ Herringbone EMI Gasket

94‧‧‧ Groove

95‧‧‧Joint hook

96‧‧‧ highlight sheet

97‧‧‧ Groove

99‧‧‧fasteners

100‧‧‧With screw bosses

102a, 102b‧‧‧ EMI Absorbing Pad

103‧‧‧ ribs

104, 104a, 104b‧‧‧ connection elements

106‧‧‧After opening

106a, 106b‧‧‧ side wall

108‧‧‧Ontology

109‧‧‧ card slot

110‧‧‧Danger prevention channel

110a, 110b‧‧‧ legs

111‧‧‧Terminal receiving cavity

115a, 115b‧‧‧ terminals

116‧‧‧ tail

120‧‧‧ screws

121‧‧‧Wire

121a, 60a‧‧‧Free end

122‧‧‧ openings

123‧‧‧Interior screw

123a‧‧‧Female

124‧‧‧Retaining strip

125‧‧‧ Grounding plate

125a‧‧‧ board

126‧‧‧Contact leg

127‧‧‧ screws

128‧‧‧Contacts

130‧‧‧   nest

132‧‧‧Frame

133‧‧‧ wall

134‧‧‧ bottom wall

135‧‧‧ openings

136‧‧‧ Groove

138‧‧‧ recess

140‧‧‧Cylinder

141‧‧‧ hole

142‧‧‧ Ground plane

144‧‧‧ buckle

146‧‧‧ openings

148‧‧‧ top

150‧‧‧Installation of the nest

152‧‧‧ base

153‧‧‧Extended side wall

154‧‧‧Connection flange

155‧‧‧ openings

156‧‧‧ panel

163‧‧‧ first end

164‧‧‧ second end

200, 201, 202‧‧‧ components

222‧‧‧ housing

224‧‧‧ Back wall

226‧‧‧Internal space

229‧‧‧Internal cabin

230‧‧‧ front opening

232‧‧‧ openings

233‧‧‧ top surface

240‧‧‧heatsink assembly

241‧‧‧Heat Transfer Department

242‧‧‧ base

244‧‧‧ skirt

246‧‧‧ recess

247‧‧‧EMI gasket

248‧‧‧Flexible fingers

250‧‧‧ Thermal contact surface

252‧‧‧ Department of heat dissipation

254‧‧‧ base

256‧‧‧ Heatsink

258‧‧‧Refrigeration channel

260‧‧‧ Holder

262‧‧‧ Rivets

263‧‧‧Vertical column

264‧‧‧ openings

265‧‧‧ base

267‧‧‧Flexible arm

268‧‧‧Free end

269‧‧‧ root

271‧‧‧Socket connector

272‧‧‧ wire

274‧‧‧heat transfer element

275‧‧‧heat pipe

278‧‧‧ trench

279‧‧‧Evaporation zone

280‧‧‧Condensation zone

282‧‧‧ lumen

283‧‧‧ side wall

284‧‧‧ wick

290‧‧‧heat pipe

300‧‧‧ Computing equipment

301‧‧‧Base

305a‧‧‧ first wall

305b‧‧‧ second wall

312‧‧‧ main board

321‧‧‧First connector

322‧‧‧ Cable

330‧‧‧Power connection device

331‧‧‧Second connector

332‧‧‧ Cable

340‧‧‧ Chip package

400‧‧‧ Computing equipment

401‧‧‧Base

405a‧‧‧ first wall

405b‧‧‧ second wall

406‧‧‧ main board

407‧‧‧Electronic component circuit

408‧‧‧Support connection device

410‧‧‧Signal board

411‧‧‧ front board

411a‧‧‧Connector support

421‧‧‧First connector

Port 421a‧‧‧

422‧‧‧ Cable

422a‧‧‧ first end

422b‧‧‧ second end

423‧‧‧ first board connector

426‧‧‧Signal board connector

426a‧‧‧Base

429‧‧‧ terminals

431‧‧‧Second connector

432‧‧‧ Cable

432a‧‧‧ first end

432b‧‧‧second end

440‧‧‧ Chip package

441‧‧‧ Refrigeration seat

441a‧‧‧Refrigeration base

441b‧‧‧Refrigeration heat sink

442‧‧‧heating frame

443‧‧‧ Thermal compound

445‧‧‧ wafer

522‧‧‧ Cable

523‧‧‧ first board connector

523a‧‧‧Base

526‧‧‧Signal board connector

527a‧‧‧Base

527b‧‧‧shell

527c‧‧‧ legs

528a‧‧‧Pressure element

528b‧‧‧Actuating element

614‧‧‧Support layer

626‧‧‧Signal board connector

640‧‧‧chip package

645‧‧‧ wafer

646‧‧‧Substrate

690‧‧‧Flexible circuit

F‧‧‧vertical plane

H1, H2‧‧‧ water level

R‧‧‧ Space

The invention is illustrated by way of example and not limitation, in the accompanying drawings Similar components are shown in the drawings: FIG. 1 is a perspective view of an electronic device (such as an exchanger, router, etc.) in which the top cover of the electronic device is removed and shows the overall layout of the components of the device and A bypass cable assembly is located within the device; Figure 2 is the same view as Figure 1, with the bypass assembly removed from the device for clarity; Figure 2A is a perspective view of the bypass only assembly of Figure 2; 2B is the same view as FIG. 2A, but with the wafer package substrate and/or the sealing body removed for clarity; FIG. 2C is an enlarged detail of the termination area around a wafer for the bypass assembly of FIG. Figure 2D is a perspective view of the end of a twinaxial cable for a bypass assembly; Figure 3A is a schematic cross-sectional view of a known structure conventionally used in an electronic device (such as a router) , a switch, etc.) is connected to a motherboard through a trace arranged through the motherboard or on the motherboard; FIG. 3B is a schematic cross-sectional view, similar to FIG. 3A, but shown in FIG. The structure of the bypass assembly as shown, which is used to The chip package is connected to the connector or other connector of the device of FIG. 1, which employs a cable and thereby eliminates the use of conductive traces on the motherboard as signal transmission lines in the device of FIG. 1 as shown; FIG. Is a perspective view of a cable-direct connector assembly constructed in accordance with the principles; FIG. 4A is a cross-sectional view of the connector assembly of FIG. 4 taken along line AA thereof; FIG. 5 is a housing and the line of FIG. An exploded view of the cable-direct connector assembly; Figure 5A is the same view as Figure 5, but with the receptacle connector located within the housing and the bottom secured to the side wall of the housing; Figure 6 is a perspective view from below the housing of Figure 5 with the bottom wall removed And the connector assembly is removed from the housing; Figure 6A is the same view as Figure 6, but with the connector assembly located within the housing; Figure 6B is the same view as Figure 6A, but with the bottom wall positioned to direct the cable The connector assembly is sealed within the housing; FIG. 7 is a partially exploded view of the connector assembly of FIG. 4; FIG. 8 is a more complete exploded view of the connector assembly of FIG. 7, and FIG. 8A is for the connector of FIG. An exploded view of one terminal array of components, its associated ground plane, and a corresponding set of cables and wires; Figure 8B is the same view as Figure 8A, but showing its components in an assembled state to form a basic Figure 8C is a perspective view of two basic connecting elements assembled to form a socket connector terminal array; Figure 8D is an enlarged detail view of two of the plurality of cables in a connecting element Showing the elevated use with it ( Elevated ground plate structure; Figure 9 is a front view of the housing showing a portion of the edge card in contact with the contact portion of the terminal of the connector assembly; Figure 10 is an enlarged detail view of the bottom of the rear end of the housing, Where the connector assembly is in place; Figure 11 is a perspective view of a housing for a bypass assembly; Figure 11A is a cross-sectional view of the connector of Figure 11 taken along line AA thereof; Figure 11B is a side view of Figure 11A; Figure 12 Is a perspective view of the connector of FIG. 11 as seen from the rear of the opposite side; FIG. 13 is a bottom view of the connector of FIG. 6A with the bottom removed for clarity; FIG. 13A is a view of an empty housing Cutaway view in which the internal connector and heat transfer element are not in place; Figure 14 is a top view of the housing with the top wall removed from the housing and a portion of an edge card engaging the internal connector; 14A is the same view as FIG. 14, but is horizontally cut at a height below the rear cover to show the manner in which the inner connector and the inner connector engage the body of the housing; FIG. 14B is the approach of the housing A vertical cut-away view of the front of the inner connector, wherein, for the sake of clarity, a portion of the inner casing is removed to show the hollow interior of the casing and the ribs therein, the ribs Contacting the connecting element and holding an EMI absorbing pad in the housing; Figure 15 is a pair of housings a perspective view in which a heat transfer element and a light indicator are vertically stacked on a circuit board; FIG. 15A is an exploded view of FIG. 15; and FIG. 16 is a three-shell arrangement vertically arranged in a horizontal direction. A perspective view of three face plates of a device; Figure 16A is an exploded view of a vertical housing and panel mounting assembly: Figure 17 is a perspective view of a housing in which a modified heat sink assembly constructed in accordance with the principles is mounted to the housing; Figure 18 is an exploded view of a portion of the housing-heat sink assembly of Figure 17, for clarity The heat sink assembly member is removed from its engagement with the top of the housing; Figure 18A is a side elevational view of the exploded view of Figure 18, wherein the components shown herein are shown along line CC of Figure 17; Figure 19 19 is a front view of the housing-heat sink assembly of FIG. 17 along its line 3-3; FIG. 19A is a side view of the housing-heat sink assembly of FIG. 17 as seen along its right side; FIG. 19B is a view 17 is a top view of the housing-heat sink assembly of FIG. 17; FIG. 19C is a longitudinal cross-sectional view of the housing-heat sink assembly of FIG. 17 along its line CC; FIG. 19D is the heat sink assembly of FIG. 17 along line DD of FIG. The heat transfer portion is taken from a transverse cross-sectional view of the housing-heat sink assembly; FIG. 19E is the same view as FIG. 19D, but for the sake of clarity, the heat pipe is removed from the heat sink assembly and one of the alternative structures A pair of heat pipes are shown in dashed lines; FIG. 19F is a heat sink set in FIG. 17 along line FF of FIG. The heat dissipating portion of the member is taken from the rear of the housing-heat sink assembly as seen from the rear; FIG. 19G is taken from the housing along the line GG of FIG. 17 in the heat dissipating portion of the heat sink assembly of FIG. A transverse cross-sectional view of the heat sink assembly as seen from the front, showing the gap between the heat sink and the wires of the connector; FIG. 20 is a perspective view of an alternative housing structure; Figure 21 is a perspective view of an embodiment of a computing device; Figure 22 is a perspective view of another embodiment of a computing device; Figure 23 is an enlarged perspective view of the embodiment of Figure 22; Figure 24 is a bypass system BRIEF DESCRIPTION OF THE DRAWINGS FIG. 25 is a simplified partial exploded view of an embodiment of a bypass system; FIG. 26 is a fragmentary side view of a chip package mounted on a signal circuit board; Is a simplified perspective view of an embodiment of a signal circuit board structure; FIG. 28A is a perspective view of an embodiment of a connector system suitable for connection to a signal circuit board; FIG. 28B is an embodiment of the embodiment of FIG. 29A Another perspective view; and FIG. 29 is a schematic illustration of an alternate embodiment of a wafer package.

The detailed description below describes exemplary embodiments and is not intended to be limited to such specifically disclosed combinations. Thus, the features disclosed herein may be combined together to form additional combinations that are not given for the purpose of the invention. Thus, reference to a feature or aspect is intended to illustrate a feature or aspect of an embodiment, and does not imply that each embodiment must have the features or aspects illustrated. In addition, it should be noted that the specification lists a number of features. Although certain features have been combined to illustrate possible system designs, those features are also applicable to other combinations that are not explicitly disclosed. Therefore, the combinations illustrated are not intended to be limiting unless otherwise stated.

In the embodiments illustrated in the figures, the directional representations (such as up, down, left, right, front and back) used to explain the structure and motion of the various components of the present invention are not absolute but relative. These representations are appropriate when the components are in the position shown in the figures. However, if the description of the location of the component changes, then these representations will change accordingly.

As can be appreciated, the following description relates to signal transmission. Depending on the data rate, those skilled in the art often refer to signals at low or high speeds (low data rates are referred to as low speed signals and high data rates are referred to as high speed signals). Although these terms are not technically the most accurate way to refer to these things, in order to be consistent with typical usage, low speed/high speed conventions are also used herein.

The embodiments described herein are suitable for use with high speed data signal transmission line systems that support high data rates from wafers or processors and the like to backplanes, motherboards or other boards with low loss. An assembly is disclosed that connects a device's chip package to an inlet connector (an inlet connector can be used to provide signals into or out of the device) without the need to significantly use traces on a circuit board, thereby reducing losses . If desired, an improved connector for use as an inlet connector is not connected to traces on the board but can be directly connected to a cable or wire to define the wafer and processor directly from the connector to the host device. Signal transmission line. Such a structure is helpful for considering high speed data applications (above 10 Gbps), and if NRZ coding is employed, the above would typically be beneficially used in systems operating above 15 Gbps. Since the receptacle connector can be entirely contained within the connector structure and does not need to be directly connected to a circuit board, the bottom wall of the cage can be continuous within the range of its fully sealed bottom of the housing, and thereby Can improve the EMI performance of the connector. Also eliminated the use of pressure fitting Needle mount connector. The pair of connecting elements in the form of sheets are arranged to fit into an opening in the rear of the receptacle connector. A primary ground plane is disposed between the two connecting elements to prevent signal interference (such as crosstalk) between the signal terminals of the two connecting elements. Accordingly, the plurality of connectors can each be independently mounted to a panel or a wall of the main device, or even interconnected with other connectors to form an integral body of a plurality of connectors suitable for vertical or horizontal stacking. Components. In addition, if desired, the connector can be positioned within the host device as a connector that can be supported on a circuit board, supported on a standoff or other support, or stand alone. This configuration defines the connector as having a high speed connector that forms a signal transmission line for high speed data applications of 10 Gbps or more that bypasses circuit traces on the circuit board of the master device .

Devices using the above connectors have very high data rates (10 Gbps and often 20 Gbps+) and often, connectors are used in conjunction with active cable assemblies, which tend to generate significant heat during data transmission. The connector can also include a heat sink assembly that extends into an interior of the housing and that is configured to contact a docking module that is inserted into the housing. The housing includes a plurality of walls that together define the interior in which a receptacle connector is received. Since the housings are often mounted along a panel of the main device, a heat sink assembly is configured to include: a heat transfer portion in contact with the docking module inserted into the housing; and a heat sink portion, The heat transfer portion is connected, which is also shown spaced apart from the heat transfer portion in a horizontal direction. In this manner, the heat sink advantageously extends behind the shielded housing and will include a plurality of fins facing downward. It is assumed that the air flow configuration can be correspondingly provided, which utilizes the open space behind the housing and can provide a reduction in the overall height of the main equipment.

In an embodiment, the termination of the at least one set of connectors is configured as a second end of the pair of wires to emulate a manner and spacing of the ordered geometry of the cable, thereby Crosstalk and other harmful factors are kept to a minimum at the connector location. In this configuration, all of the connector terminals have the same length. The free ends of the pair of signal terminals are arranged at a desired interval and include associated grounding members such that the grounding members associated with the respective pairs of wires can be terminated to respective grounding bodies of the connector to define the entire cable The length and an associated ground that extends over the connector. A ground plane through which the signal terminals can be coupled in a common mode by a common mode coupling can provide shielding and reduce crosstalk. The termination of the cable conductors and the connectors can be carried out in such a way that a specific desired geometry of the signal conductors and ground conductors of the cable is maintained through the cable as much as possible The termination area is up to the connector.

A single wafer package can be configured to include an integrated circuit mounted to a substrate. The substrate has a termination end region of the first end termination of the plurality of twinaxial cables. The length of the cable can vary and the length will be long enough for some cables to be easily and reliably terminated to a first external interface in the form of a single or multiple I/O connectors (which is an inlet connector) And a portion of an external connector of either or both of the outlet connectors). These connectors may preferably be mounted to one side of the main device in a manner that allows an external connector, such as a plug connector or a pluggable module, to interface with it. The assembly may have a cable extending between an inlet connector of the device and a wafer package formed as an integral component, or the assembly may also be included in a wafer package and an outlet connector of the device Additional cables extending between. The first end of the bypass cable can be configured as a connector that can be inserted into the first end of the chip package to have a "plug and play" capability. In this way, the outside The connector can be inserted into the master device as a single component or as a group of components (each component containing one or more signal transmission channels). The chip package can be supported within the housing of the device either alone or through standoffs or other similar attachments for a low cost, low speed motherboard.

Removing the signal transmission line from the wafer to the external connector of the motherboard in this manner releases the space on the motherboard, which space can accommodate additional functional components to provide additional equipment for the device The value and functionality while maintaining a low cost compared to devices that use motherboards for signal transmission lines. Moreover, incorporating the signal transmission line into the cable of the bypass assembly reduces the amount of power required to transmit high speed signals from the chip package to the external connector, thereby increasing the "green" value of the bypass assembly And reduce the operating costs of equipment using such a bypass assembly.

The cable extending between the connector and the chip package is preferably a "two-axis" type enclosed within a dielectric body in which two wires are disposed in each cable as a pair of signal conductors extending in the longitudinal direction of the wire. The pair of wires are preferably terminated at the proximal end of the cable to the socket connector and directly terminated at the distal end of the cable to the wafer package. The receptacle connector is preferably included in a connection structure (such as a housing, adapter frame, etc.) and mates with the attachment structure to define a shielded housing that is configured to receive an external connector (such as a Pluggable module). The second end of the cable conductor is directly terminated to the terminal of the receptacle connector and the grounding member, and the cable is preferably retained in the sheet-like support member to define a terminal strip on opposite sides of the card slot of the receptacle connector . A cable exits the connecting structure through a rear wall of the connecting structure.

Referring to the drawings, FIG. 1 is a perspective view of a computing device 50 (such as a switch, router, server, etc.), and wherein the top cover of computing device 50 is removed. Computing device 50 is comprised of a wafer 52 (which The wafer 52 can be part of a total wafer package 54 that can be managed by one or more processors or integrated circuits in the form of one or more discrete wafers packaged together. Computing device 50 has a pair of side walls 55, a first wall 56, and a second wall 57. A plurality of first connectors 80 (which may be in the form of input/output or I/O connectors) are disposed on the first wall 56 of the computing device 50 (which may be the front wall) such that the relative form of the pluggable module or the like is The docking connector can be plugged in to connect to the circuitry of computing device 50. The second connector 30 can be disposed on the second wall 57 (which can be a rear wall) for connecting the computing device 50 to a larger device (such as a server or the like, including the back for such a device) board). Computing device 50 includes a power source 58 and a cooling assembly 59 and a motherboard 62 having various electronic components (such as capacitors, switches, smaller wafers, etc.) thereon.

3A is a cross-sectional view of a prior art conventional wafer package and motherboard assembly for a conventional device. Wafer 52 can be an ASIC or any other type of processor or integrated circuit (such as an FPGA) and can be one or more discrete integrated circuits that are positioned together. Accordingly, the term wafer will be used herein as a generic term for any suitable integrated circuit. As shown in FIG. 3A, wafer 52 has contacts in the form of solder bumps 45 on its underside that connect wafers 52 to associated contact pads of a supported substrate 47. Substrate 47 typically includes plated through vias, microvias or traces 48 that extend through the body of substrate 47 to the underside of the body of substrate 47. These elements 48 are connected to contacts 49 which are disposed on the underside 47a of the substrate 47, and these contacts 49 may typically take the form of a BGA, PGA or LGA or the like. The wafer 52, the solder bumps 45, the substrate 47, and the contacts 49 cooperate to define a wafer package 52-1. The chip package 52-1 is butted to a motherboard 52-2 through a socket (not shown) made of FR4 material and used for a device. Motherboard 52-2 has multiple long guides Electrical traces 52a-52c, conductive traces 52a-52c extend from the chip package contacts 49 through the motherboard to other connectors, components, etc. of the device. For example, a pair of conductive traces 52a, 52b are required to define a differential signal transmission line, and a third conductive trace 52c provides an associated ground that follows the path of the signal transmission line. Each of such signal transmission lines is arranged through the motherboard or on the motherboard and this arrangement has certain disadvantages.

The FR4 board material loss is increasing and above 10Ghz, which is starting to become a problem. In addition, these signal transmission line traces 52a-52c typically require steering, bending, and intersection to route the transmission line from the chip package contacts 49 to connectors or other components mounted on the motherboard 52-2. These changes in direction of traces 52a-52c can create signal reflection and noise problems as well as additional losses. Losses can sometimes be corrected by using amplifiers, repeaters, and equalizers, but these components also add to the final cost of manufacturing motherboard 52-2. This complicates the layout of the motherboard 52-2 as additional board space will be required to accommodate these amplifiers and repeaters and this additional board space is not available under the intended size of the device. Custom materials for circuit boards that reduce this loss are available, but the price of these materials dramatically increases the cost of the boards and the cost of the electronic equipment from which they are used. Still, long circuit traces require increased power to drive high speed signals flowing through the circuit traces, and as such, they prevent designers from attempting to develop "green" (energy saving) devices.

FIG. 3B is a cross-sectional view of wafer package 54 that can be used in computing device 50 of FIG. 1. The wafer 52 includes a high speed circuit, a low speed circuit, a clock circuit, a logic circuit, a power supply circuit, and other circuits connected to the substrate 53 of the chip package. Trace 54-1 of wafer package 54 leads to an associated contact pad 54-2, contact pad 54-2 is disposed within termination region 54-3, and termination region 54-3 is preferably disposed at or near edge 54-4 of substrate 53. The wafer package 54 can also include an encapsulant (such as epoxy) 54-5 (see FIG. 2A) that secures the wafer 52 within the package 54 to interface with associated cable connectors and Its components act as an integral component. As shown, the chip package 54 is partially coupled to the motherboard 62 via contacts 49, but such a connection means need not include a high speed signal transmission line located directly on the motherboard 62. However, as explained later with respect to Figures 22-29, if the high speed transmission line is traveling over a short distance, the supported circuit board can include a high speed transmission line; for example, the traces 52a-52c are just outside of the substrate 47. The connector is quickly terminated to a connector disposed on the circuit board 62.

Referring to Figures 1, 2D and 3B, if the cable 60 is to be directly terminated to the substrate 53, the cable 60 can be terminated to the package contact pad 54-2 through a suitable wire-to-board connector or the like, and these cables 60 is preferably a biaxial structure having two signal conductors 61 surrounded by a dielectric covering 61-1. Cable 60 can also include an associated drain wire 61-2, an outer shield 61-3, and a final outer insulation skin 61-4. The use of one or more drain wires is optional as this is an outer shield (which may be in the form of a conductive wrap, braided shield, etc.). In some cases, the two conductors can be wrapped in a single dielectric cladding. The spacing and orientation of the two wires constituting each such pair of wires can be easily controlled in such a manner that the cables are separated from and separated from the board and can be in a wafer, wafer set, A transmission line extending between the electronic component and a connector location on the circuit board or between two locations on the circuit board. In some embodiments, the ordered geometry of the cable as a signal transmission member makes it easy to maintain a signal transmission line with acceptable loss and noise compared to the difficulties encountered with the signal transmission line of the board. .

As noted above, cable 60 and its signal conductor 61 define a high speed signal transmission line that leads from wafer package 54 to first (inlet) connector 80 or second (outlet) connector 30. The ordered geometry of the cable 60 remains at a predetermined interval as a pair of signal conductors 61 to control the impedance across the cable 60. The use of cable 60 as a signal transmission line eliminates the need to lay high speed signal transmission lines in the form of traces on motherboard 62, thereby avoiding the high cost of foreign board materials and the losses associated with less expensive board materials such as FR4. .

As shown in Figures 2 through 2C, the cable 60 has opposite first and second ends 163, 164 that are coupled to the chip package 54 and the I/O connector 80 (which may be the first connector) or the back, respectively. A board connector 30 (which may be a second connector) to define a high speed signal transmission line bypassing the motherboard 62. The cable 60 maintains the ordered geometry of the signal conductor 61 over the entire length of the signal conductor 61 that traverses to and from the wafer 52 via the external interface. The ordered geometry of the plurality of cables 60 allows the plurality of cables 60 to be diverted, bent or crossed in their path without introducing signal reflections in the transmission line that can cause problems in the signal transmission lines of the board or The impedance is not continuous. The cable 60 is arranged in a first set of cables and a second set of cables, wherein the first set of cables extends between the first connector 80 and the wafer package 54, and the second set of cables are in the chip package 54 and the calculation The second connector 30 on the rear wall 57 of the device 50 extends between. The manner in which signal conductor 61 of cable 60 can be terminated to wafer package 54 can vary. As shown in FIG. 2C, the cable 60 can be terminated by a wire-to-board connector 66 that can be mated with contacts on the chip package 54 (which are removed for purposes of illustration). The wire-to-board connector 66 can also interface with the signal board connector 426 mounted on a surface of the signal circuit board 410 (shown in Figures 22-28). The heat sink 71 can be mounted on the surface of the wafer 52 as shown for heat dissipation, or can be integrated through the sealing body 54-5. Into the component. It should be noted that further details of possible configurations will be disclosed later for the embodiments shown in Figures 22-28.

The wafer 52, the substrate 53, the heat sink 71, and the wire-to-board connector 66 can be held together by a sealing body 54-5 or can be held together by other supporting structures as a single component, as shown in FIGS. 2 to 2C. Shown. This configuration allows a device designer to take full advantage of the space available on the motherboard 62 for additional components and circuitry, which increases the value of the computing device 50 without complicating the design of the board. These integrated components can be inserted into computing device 50 by simply inserting first connector 80 and second connector 30 into corresponding openings in front wall 56 and back wall 57 of computing device 50. An auxiliary connector for connecting the chip package 54 to other circuits of the computing device 50 can be provided as shown in Figure 3B. The components may also be provided in other forms such as, for example, 1) a waferless package 54, but with a wafer packaged substrate 53; 2) a wafer package 54 and having a first connector 80 or a second connector 30, in FIG. 2A Marked at 200, 201, respectively; and 3) a first connector 80 and a second connector 30 having openings arranged to extend onto the first wall 56 and the second wall 57 of the computing device 50, in FIG. Indicated at 202. In this manner, components 200, 201, 202 can all be plugged into a basic device to provide functionality to the device without the need to design such functionality to motherboard 62 of computing device 50. Additional details of a possible structure will be described later.

Referring to FIG. 4, FIG. 7, and FIG. 8, a socket connector 70 is received in the first connector 80 and the inner connector 70 includes a body 108 formed of an insulating material. The body 108 includes a card slot 109 and a card slot. 109 opens to the front of the receptacle connector 70 and the inlet 67 of the first connector 80 (see Figure 5A). The card slot 109 is an example of a pair of interfaces, and although the card slot structure is preferred, other dockings The interface structure is also applicable. The card slot 109 is located above a polarizing channel 110 formed by the legs 110a, 110b of the card slot 109 supporting the bottom wall 68 (see FIG. 5) of the first connector 80, and prevents The correctly positioned opposite docking connector is inserted into the card slot 109. The main body 108 has a plurality of terminal receiving cavities 111 which are aligned on opposite sides of the card slot 109 and accommodate contact portions of the cantilevered terminals 115a, 115b of the two connecting elements 104a, 104b. The connecting members 104a, 104b support the terminals 115a, 115b in the form of respective single row terminals, as shown in Figures 4A and 8C. The two connecting members 104a, 104b each have a sheet-like structure and can be inserted into the body 108 from the rear to complete the assembly of the receptacle connector 70. As shown, the terminal arrays of the respective connecting elements 104a, 104b are thus positioned on opposite sides of the card slot 109.

8A to 8D show the basic structure of a connecting member 104 for the receptacle connector 70. A plurality of cables 60 and conventional wires 121 are arranged in an array extending in the lateral direction of the connector 70. The ends of the wires 121 and the cables 60 are stripped to expose the signal conductors 61 of the cable 60 and define the wires 121 and the free ends 121a, 60a of the cables 60. The free ends 121a, 60a are respectively used to terminate the terminals 115a. The corresponding tail 116 of 115b (see Figure 4A). In the illustrated embodiment, the two outer ends of the array are respectively positioned with pairs of cables 60, and the drain wires 61-2 of the cable 60 are simply bent upward and then bent again to lie flat The drain wire 61-2 is associated with the ground plate 125. The terminals 115a, 115b are held together by a holding strip 124 in a lateral array of their own spacing. This greatly preserves the geometry of the cable when it is terminated at the connector.

The illustrated receptacle connector 70 has a structure that enhances the signal integrity of the data signals through the receptacle connector 70 and provides a corresponding docking connection from the bypass cable conductors. An impedance transition between the circuits of a circuit card of the device. This transition is from 85 ohms to 100 ohms in a predetermined error level and is performed in multiple stages or three segments such that the transition transitions from an entry level impedance to a first transition impedance, followed by a second transition A gradual manner of shifting the impedance and ultimately transitioning to the final or third transition impedance is performed. In this manner, the impedance transition occurs over the entire length of the receptacle connector in a slightly gradual manner rather than at the tail or contact of the receptacle connector 70. In embodiments where no impedance transition is required, this transition can be omitted.

If a transition is set, the transition can be set by providing three different dielectric media through which the terminals 115a, 115b of the receptacle connector 70 extend. The first section medium is preferably a hot melt adhesive in which the impedance is increased by about 6 ohms from an incoming impedance of about 85 ohms, while the second section medium preferably comprises LCP (liquid crystal polymer). The impedance rises in the LCP by about another 6 ohms, and finally, the third section of the medium includes air, and the impedance in air rises to about 105 ohms, whereby the impedance transition is at an error level of about 5% ( Tolerance level). The change in the surrounding medium is also accompanied by a change in the width of the terminal (the width is widened or narrowed in different sections). The distance between the terminals 115a, 115b and the associated ground plane can also contribute to this selected impedance tuning. This transition occurs over the length of the receptacle connector 70 from the tail to the contact end to present a gradual increase over an overall length rather than just an increase at the terminal tail or contact.

Referring to Figures 8A-8D, the termination area of the cable 60/wire 121 and the terminals 115a, 115b is in a nest 130 that extends laterally and is comprised of an insulating material having a desired dielectric constant. form. The illustrated nest 130 has a U-shape and is positioned at a retaining strip of an adjacent terminal. 124. In this region, the drain wire 61-2 of the cable 60 can be bonded to the ground plate 125, which is above the cable 60 and vertically spaced from the tail 116 of the terminal and above the tail 116 of the terminal. The ground plate 125 has a plate body 125a having at least a portion of a flat surface that is contacted and soldered or otherwise joined by the drain wire 61-2.

The contact leg 126 is disposed as part of the ground plate 125 to form a contact portion 128 of the ground plate 125, which is preferably coupled to the tail portion 116 of the ground terminal of the receptacle connector 70. The contact leg 126 is vertically offset such that the ground plate 125 is spaced apart from at least a portion of the signal conductor 61 and extends over the at least a portion to the tail of the signal terminal in a row of terminals associated with a respective connecting element. As shown in FIGS. 8B-8D, each grounding plate 125 preferably includes three legs 126 that contact the socket connection in such a manner that the side of the tail of any two signal terminals is the two of the contact legs 126. The ground terminal of the device 70. From an impedance perspective, this arrangement allows the spacing between the signal terminals to approximately coincide with the signal conductor 61 of the twinaxial cable 60. In this manner, for the receptacle connector 70, a G-S-S-G pattern of the terminals 115a, 115b is held in the two rows of terminals on opposite sides of the card slot 109 (see Fig. 9).

Referring to Figures 8 and 8C, a rectangular frame 132 is disposed along the rear of each of the connecting members 104a, 104b and includes four walls 133 joined together about a bottom wall 134 to at least partially define a hollow interior recess 138. The front and rear walls 133 of the frame 132 are shown with a plurality of openings 135 that are configured to receive the cable 60 in a longitudinal extent of the cable 60 and the low power logic control wire 121 passing through the frame 132. And a low power logic control wire 121. The frame 132 is bonded to the nest 130 along the back of the nest 130 via an overmolded portion of the filled termination region. Frame 132 is made of a conductive material A material, such as a metal, is formed or may have an outer conductive coating such that when in the first connector 80, the connecting elements 104a, 104b are in electrical ground contact with the first connector 80. Thus, the frame 132 can provide a path to a ground or other reference plane. The frame 132 of the connecting elements 104a, 104b is positioned adjacent the nest 130 and behind the nest 130 and can be secured to the nest 130 as will be described later.

Referring to Figures 8 and 10, the wall 133 of the frame 132 can be provided with a vertical recess 136 as shown. These grooves 136 will engage the side walls 106a, 106b of the rear opening 106 of the first connector 80, and because the frame 132 is electrically conductive, the frame 132 can also mitigate EMI leakage from the rear opening 106 of the first connector 80. The open recess 138 of the frame 132 of the connecting member 104a, 104b through which the cable 60 and the wire 121 extend extends is filled with a dielectric material, such as a liquid crystal polymer ("LCP", which is a dielectric material that opposes the cable 60/wire 121. The frame 132 of the connecting elements 104a, 104b and the nest 130, which also houses a portion of the LCP, are fixed in the recess 138. In this manner, the sheet-like shape of the connecting elements 104a, 104b is defined and this unitary structure provides a biaxial A measure of strain relief of the cable 60.

The bottom walls 134 of the two connecting members 104a, 104b abut each other and are engageable with each other by way of a cylinder 140 engaging the holes 141 as shown in FIG. In this manner, the two connecting members 104a, 104b can be inserted into a rear opening of the body 108 such that the contacts of the terminals 115a, 115b are aligned with each other and received in the terminal receiving cavity 111 of the body 108 of the receptacle connector 70 to form An integrated connector assembly. As shown in FIG. 6, the connector assembly is pressed into the hollow interior space of the first connector 80 from below. An internal ground plane 142 is provided in the form of a flat conductive plate and is located between the two connecting elements 104a, 104b. The ground plane 142 extends from the rear end of the frame 132 of the connecting element to the nest The leading edge of 130. This ground plane 142 acts as a primary ground plane for preventing crosstalk between the pair of signal conductors within one connection element and the pair of signal conductors within the other connection element. The ground plate 125 serves as a secondary ground plate or busses for the signal conductor 61 of the cable 60 and its terminating portion with the terminal 115a.

Referring to Figures 5, 8, and 10, the recess 136 in the wall 133 of the connecting member 104a, 104b engages the side wall 106a, 106b of the rear opening 106 of the first connector 80 and is disposed in the body 108 of the first connector 80. The opposite two outer buckles 144 are received in the openings 146 of the side walls 64a, 64b of the first connector 80. The buckle 144 can be oversized in size to deform when the receptacle connector 70 is inserted into the housing 63. The recess 136 can be shaped with a tip 148 that is directed inwardly or inwardly to ensure reliable contact with the first connector 80.

The two EMI absorbing pads 102a, 102b can be disposed on opposite surfaces of the connecting elements 104a, 104b of the receptacle connector 70 before the receptacle connector 70 is pressed into the interior 78 of the first connector 80 from below. As previously described, the connecting members 104a, 104b vertically define the recesses 136 such that the connecting members 104a, 104b can engage the side walls 106a, 106b of the rear opening 106 of the first connector 80 and cooperate with the EMI absorbing pads 102a, 102b. EMI leakage protection on the four sides of the connecting elements 104a, 104b. In effect, the rear wall 65 of the first connector 80 is combined with the electrically conductive connecting elements 104a, 104b to form a fifth wall that prevents EMI leakage. The EMI absorbing pads 102a, 102b seal the space between the connecting elements 104a, 104b and the opposing surfaces of the housing 63. These EMI absorbing pads 102a, 102b occupy an open space above and below the connecting elements 104a, 104b (the two spaces are typically left blank in conventional connectors).

The EMI absorbing pads 102a, 102b are preferably aligned and the cables, wires 60, 121 at the connecting elements 104a, 104b terminate over the area of the tail 116 of the terminals 115a, 115b of the receptacle connector 70. The bottom EMI absorbing pad 102b is held between the bottom wall 68 and the bottom connecting member 104b, while the top EMI absorbing pad 102a remains in place between the top connecting member 104a and the rear cover 90 of the housing 63. This is achieved by the ribs 103 formed on the bottom of the back cover 90 and extending downwardly into contact with the pad 102a, as shown in FIG. The connecting elements 104a, 104b, the EMI absorbing pads 102a, 102b are thus interposed between the top wall 79 and the bottom wall 68 of the housing and the EMI absorbing pads 102a, 102b help ensure that EMI is reduced along the rear wall 65 of the housing 63. The rear opening 106 leaks.

In the case where the cable 60 is directly terminated to the terminals 115a, 115b of the connector 70, the first connector 80 is configured to be mounted off and mounted on a circuit board and mounted on a panel or to become a master device The setting of a free-standing connector inside. The first connector 80 need not be mounted to a motherboard 62 in one end, but can be mounted to a host through fasteners extending through the opening in the motherboard 62 and into the screw bosses 100. Board 62. Thus, the beneficial sealing of the bottom of the first connector 80 and the need to eliminate the right angle connector not only eliminates the need to mount the first connector 80 on the motherboard 62, but also facilitates vertical and horizontal A plurality of first connectors 80 are arranged to be stacked.

Accordingly, the wires 121 of the receptacle connector 70 can be directly coupled to components of the computing device 50, such as a processor or a wafer package or the like disposed on the traces on the bypass circuit board. Because of the direct connection, the receptacle connector 70 need not be mounted to a circuit board, but can be enclosed within a structure such as the disclosed first connector 80 and the mounted panel. The first connector 80 can be disposed on an adapter frame that can be constructed similarly to the housing if desired. Still, the first connector 80 can be used as an internal connecting sleeve, positioned in the calculation A plug connector is housed in the device 50. One end of the cable 60 of the first connector 80 terminates at the tail 116 of the terminals 115a, 115b of the connection elements 104a, 104b such that the cable 60 can be terminated at its other end to the chip package 54 or processor of the computing device 50. . An integrated bypass assembly such as this can be installed and removed or replaced as a single piece, bypassed on the motherboard 62 and eliminating the loss associated with the FR4 material, thereby simplifying the design and reducing the host The cost of the board 62.

Referring now to FIGS. 4-9, a first connector 80 is shown in FIGS. 5 and 5A that can be used as an external interface for receiving the first connector 80 of the computing device 50. The first connector 80 is disposed on the front wall 56 of the computing device 50 (see FIG. 1) and houses opposing docking connectors (such as pluggable electronic modules, etc.) in the form of plug connectors. The first connector 80 includes a conductive housing 63 that includes two side walls 64a, 64b, a rear wall 65, and a top wall 79 and a bottom wall 68. All of these walls together define an interior 78 that houses a corresponding external docking connector that interfaces with a receptacle connector 70. All of the walls of the first connector 80 can be formed together as a single piece of adapter frame, or all of the walls of the first connector 80 can be separate components that are joined together to form an integrated assembly. It should be noted that the first connector 80 is not limited in its operation to only accommodate the pluggable module, but can accommodate any suitable connector with a suitable configuration.

The side walls 64a, 64b, the rear wall 65, the top wall 79, and the bottom wall 68 of the housing are both electrically conductive and provide shielding of the connection portions made within the first connector 80. In this regard, the first connector 80 is provided with a conductive bottom wall 68 that completely seals the bottom of the housing 63, whereas the known housing and frame are open at the bottom where they are mounted on the circuit board. The first connector 80 is shown to include a receptacle connector 70 that causes the wires to be directly connected to their terminals 115a, 115b and thus does not need to be terminated to the computing device Traces on the motherboard 62 of 50. Prior art connectors surrounded by a housing or frame are of the right angle type, which means that the connectors extend from their mating faces at right angles to the traces of the board and connector terminations. Right-angle connectors terminated on a board can cause signal integrity problems at high speeds due to varying lengths of terminals and bending of the terminals, such as capacitance increases in curved corners and characteristic impedance of the system in connectors and connections Jumps or dips at the interface of the device and the board. Additionally, the cable 60 exiting the rear wall 65 of the first connector 80 can potentially eliminate the need for press-fit pins as a means of mounting the connector to the board because common mounting holes can be used for Threaded fasteners, thereby simplifying the overall design of the motherboard 62 of a computing device 50. As shown, the receptacle connector 70 terminates in the wires of the cable 60 and exits the rear wall 65 of the first connector 80, thereby avoiding the aforementioned problems.

As shown in FIGS. 5-6B, the bottom wall 68 of the inlet connector 80 seals the bottom of the first connector 80. The bottom wall 68 is shown formed from a sheet of sheet metal having a bottom plate 72 and two side attachment wings 73 that extend along the outer surfaces of the side walls 64a, 64b of the housing, respectively. The opening 74 in the connecting wing 73 engages the snap or tab 76 on the side walls 64a, 64b and holds the bottom plate 72 in place. Additional means of attachment may include an inner wing 75 that is also bent upwardly from the bottom plate 72 but positioned along the edge of the bottom plate 72 to extend along the inner surface of the side walls 64a, 64b to the interior 78. Two such inner wings 75 are shown in Figures 11B and 13A and each include a contact piece 75a that extends inwardly for contacting an opposite side of an opposite connector inserted into the interior 78. . Two flanges or flanges 77a, 77b may also be provided at opposite ends of the bottom plate 72, respectively, extending at an angle relative to the bottom plate 72 to engage the front and rear walls 65 of the housing 63 to form conductive contacts at these locations. It also provides EMI shielding. Covering the entire bottom with a bottom wall 68 significantly reduces EMI in this area. Multiple standoffs 69 can be formed if needed At the bottom wall 68. The multi-point contact between the bottom wall 68 and the housing 63 provides the receptacle connector 70 with a reliable EMI shield gasket along the entire bottom of the first connector 80.

Referring now to Figure 5, the top wall 79 preferably includes an access opening 81 that communicates with the hollow interior 61 and is aligned with the receptacle connector 70 (primarily the area of the front of the receptacle connector 70). A heat transfer element 82, shown as a heat sink in the form of a heat sink fin, can be configured to have a base 84 that extends at least partially into the access opening 81. The base 84 has a flat bottom contact surface 85 that contacts an opposing surface of a module that is inserted into the interior 78 of the housing. Two retaining members 86 are shown coupled to the top wall 79, and each retaining member 86 has a pair of preloaded contact arms 88 that apply a downward retaining force to a top plate 87 of the heat transfer element 82. An EMI gasket 89 is disposed to extend around the perimeter of the inlet opening 81 and is sandwiched between the top wall 79 and the heat transfer element 82.

The first connector 80 also includes a rear cover 90 that extends over the rear of the interior 78 to cover a portion of the receptacle connector 70. A recess 91 may be formed in the rear cover 90 to receive a chevron-shaped EMI gasket 92 interposed between the opposite surfaces of the rear cover 90 and the top wall 79. It can be seen that the rear cover 90 includes an opening in the form of a recess 94. The top wall 79 (see Fig. 13A) may include an engagement hook portion 95 as shown, the engagement hook portion 95 being received in the recess 94 so that the top wall 79 can slide forward so that the front edge portion of the top wall 79 abuts The front flange of the first connector 80 is such that the top wall 79 is joined to the housing 63, and the housing 63 can include a protruding tab 96 formed therewith that engages a corresponding recess of the top wall 79. 97 (Fig. 5A and Fig. 13A). A fastener 99 (which may be a screw or other form of fastener) may be used to secure the top wall 79 to the housing 63 by engaging a threaded bore formed in the threaded boss 100 supported by the housing 63. In this manner, the housing 63 is sealed in a manner that significantly reduces EMI leakage.

Referring to FIG. 15 to FIG. 16A, since the socket connector 70 is directly connected to the cable 60, the first connector 80 is not required to be mounted to the motherboard 62 through direct termination, but can be supported or penetrated through other structures. The opening 122 of the motherboard 62 is inserted into the screw 120 with the screw boss 100 to be mounted. Sealing the bottom of the first connector 80 and eliminating the right angle connector will not require mounting the first connector 80 on the motherboard 62 and facilitating the placement of the plurality of first connectors 80 in the vertical and horizontal directions. Figures 15 and 16 show only two different stacked forms. 15 and 15A show a pair of first connectors 80 in which their inlets 67 are all oriented horizontally and stacked vertically. The two first connectors 80 are shown as being supported by a bottom screw 120 in an upward manner through an opening 122 in the motherboard 62 to engage the threaded boss 100 for support on a motherboard 62. A set of intervening screws 123 are provided to engage the threaded bosses 100 of the lower housing, and these intervening screws 123 have threaded male ends and threaded female ends 123a. The female end 123a engages a screw 127 that extends into the top of the housing with the screw boss 100 of the upper housing. Thus, the plurality of first connectors 80 can be stacked in such a manner that complex high speed connection traces formed on the motherboard 62 and terminating in the receptacle connector 70 are not required.

16 through 16A illustrate another manner in which a plurality of first connectors 80 of the present invention can be arranged. This arrangement includes three housings that are erected along the front of the computing device 50 but raised off a horizontally aligned row on the motherboard 62. Figure 16A shows a mounting nest 150 having a base 152 defining a recess and two extending sidewalls 153 that receive a first connector 80. The mounting nest 150 has two attachment flanges 154 that can be mounted to a panel 156 by fasteners as shown extending through openings 155 in the base 152. A fastener can be used to attach the first connector 80 to the nest 150 and the fastener extends through the opening 155 of the base and into the boss with the screw hole Within 100. The top wall 79 of the first connector 80 can be mounted to the housing 63 using the intervening screws 123 described above such that the adjacent first connectors 80 can be assembled into an integrated arrangement and extended through the screws 120. The base 152 of the nest 150 enters the female end 123a of the opposing intervening screw 123 or the threaded boss 100 that enters the first connector 80. The plurality of first connectors 80 may also be closely spaced together as shown in FIGS. 14 to 15B because the heat transfer elements 82 have their fins extended toward the rear of the housing 63 as described below.

Accordingly, a free-standing connector/housing is configured to be mounted to an outer wall (such as a panel or bezel) of a main device or to a circuit board without requiring any end traces to be positioned below the housing. Such a self-standing connector does not have to be mounted on a circuit board, but can be mounted on a panel. The connector can take the form of an adapter frame, a shielded cover or a similar housing type. Still further, the connector can be used as an internal connection sleeve to provide an internal connector that is positioned within the main device and that houses a plug connector. The cable of the connector terminates at the proximal end of the cable at the end of the terminal of the connecting element, and the cable can be terminated at its distal end to the chip package or processor of the host device. An integrated bypass assembly such as this can be installed and removed or replaced as a single piece, bypassed on the board and eliminates the loss associated with the FR4 material, thereby simplifying the design and lowering the board the cost of.

The docking connector for connection to the I/O connector generates heat during operation and this heat must be removed to maintain efficient transmission and reception of signals during operation. High temperature can not only negatively affect the performance of the module, but also negatively affect the performance of the device using the module, so it is important to remove the heat generated by this operation. This heat removal is typically achieved by using a heat sink, the heat sink A solid base is included that contacts a selected surface of the module, typically the top surface. These heat sinks also have a plurality of fins that project upwardly from the base into the interior space of the device. The plurality of fins are spaced apart from one another such that air can flow over the fins and around the fins in a manner that dissipates heat from the fins into the surrounding interior atmosphere. The plurality of fins are mounted above the heat sink and the module and extend upwardly by a prescribed height to achieve a desired degree of heat exchange. However, the use of such a heat sink does not allow the designer to reduce the height of the device in which the module is employed, thereby reducing the likelihood of reducing the overall height of such a device.

In this regard, as shown in FIGS. 17 to 19G, a heat sink assembly 240 is disposed to include a heat transfer portion 241 having a solid base portion 242 that hangs down into the interior space 226 of the housing 222. . The base 242 of the heat transfer portion 241 is complementary in shape to the opening 232 of the housing 222 such that the base 242 can extend through the opening 232 and into the interior space 226 to interface with the front opening 230 of the interior pod 229 of the housing 222. The top or upper surface of a module is in thermal contact. The base 242 can further include a skirt 244 that extends around at least a substantial portion of the perimeter of the base 242 and preferably extends around the entire perimeter of the base 242. This skirt 244 is received in a corresponding recess 246 formed in the top surface 233 of the housing 222 and preferably formed around the opening 232. An annular and electrically conductive EMI gasket 247 is disposed to fit within the recess 246 and surround the opening 232. The EMI gasket 247 has a plurality of resilient fingers 248 that provide a conductive seal between the skirt 244 of the heat transfer portion and the recess 246 at the top of the housing to prevent EMI from passing through the opening 232 leaked. The EMI gasket 247 fits within the recess 246 and surrounds the opening 232 while the plurality of resilient fingers 248 extend radially outwardly as shown and contact the bottom surface of the skirt 244. The opening 232 at the top of the housing 222 is considered a contact The opening is because it allows the heat transfer portion 241 to extend into the interior space 226 of the housing 222 and is in thermal transfer contact with any of the modules inserted into the interior space 226 through a thermal contact surface 250 (see Figure 19C).

The heat transfer portion 241 has a solid base 242 that preferably includes a flat thermal contact surface 250 (at the bottom of the base 242) that is configured to enter the contact opening of the housing 222 and is effective and reliable The way of thermal contact contacts the top surface of a module within the interior pod 229. The base 242 can include a beveled guide on its contact surface 250 to facilitate insertion of a module. The heat sink assembly 240 further includes a distinct heat dissipating portion 252 that transfers heat generated by the module and transmitted through the thermal contact surface 250 to an opposite top surface of the module to the heat transfer portion 241. Out. As shown in FIG. 18, this heat radiating portion 252 is significantly different from the heat transfer portion 241 and spaced apart from the heat transfer portion 241 in a longitudinal or horizontal direction.

The heat sink portion 252 includes a base portion 254 that extends from the heat transfer portion 241 along a similar longitudinal axis in a cantilevered fashion. A plurality of vertical fins 256 are disposed on the base 254 and extend vertically downward from the base 254 of the heat sink. As shown, the plurality of fins 256 are spaced apart from each other in the longitudinal direction (horizontal direction) to define a plurality of refrigerating passages 258 between the plurality of fins 256, the plurality of refrigerating passages 258 being longitudinally spaced apart The heat transfer portion 241 further extends longitudinally relative to the module. In order to maintain the heat transfer portion 241 in contact with a corresponding module and resist any torque that may occur due to the weight and/or length of the heat sink portion 252, a plurality of retainers 260 are shown. The retaining members 260 are mounted to the top surface 233 of the housing 222 by means of fasteners, such as rivets. The fasteners can also be formed as part of the housing 222, essentially a vertical cylinder 263, which is received by the vertical cylinder 263. In a corresponding opening 264 provided on the base 265 of the holder 260. The free ends of these vertical cylinders 263 can be "dead-headed" or "mushrooms" Mushroomed to form a connection between the retaining member 260 and the skirt 244. It is seen that the retaining member 260 has a pair of cantilevered resilient arms 267 associated therewith, the resilient arms 267 from the base 265 Longitudinal extension, as shown. The resilient arm 267 has a flexible arm that is formed to be biased downward in advance. The resilient arm 267 terminates at the free end 268 and the free end 268 extends at a downward angle to contact The skirt 244 of the heat transfer element. Four such contact points are provided to the heat sink assembly 240, as shown, and the four contact points will define a quadrilateral if joined by an imaginary line. However, these of the resilient arms 267 The point of contact can vary at the location shown, depending on the range of spaces available on the skirt 244 of the heat sink assembly 240.

The resiliency of the resilient arms 267 allows the designer to obtain a desired length by constructing the length of the resilient arms 267, the depth of the resilient arms 267 to sag into the recesses 246, and the height of the resilient arms 267 that are coupled to the stubs 269 of the retaining members 260. Contact pressure. The retaining member 260 is securely coupled to the skirt 244 such that the sides of the housing 222 need not be formed and employ attachments that will take up space and affect the spacing between the housings 222. The vertical cylinder 263 also has a low height such that the frame of the holder 260 does not excessively increase in any direction, including the vertical direction. The resilient arms 267 are relatively short in length and thereby contact the heat transfer portion 241 near its four corners to apply a reliable contact pressure to the heat transfer portion 241 to maintain the heat transfer portion 241 with any of the modules. Good heat transfer contact.

Uniquely, the plurality of fins 256 are not in intimate contact with the heat transfer portion 241 of the heat sink assembly 240. Instead, it is positioned at the heat dissipation portion 252 and extends downward from the heat dissipation portion 252. The plurality of fins 256 are longitudinally spaced away from the heat transfer portion 241 and its base 242. The plurality of fins 256 are also arranged on a series of planes, which are shown as a vertical plane F, a vertical plane F and a horizontal plane H1 (heat transfer section) The skirt 244 of 241 extends along the horizontal plane H1 and the horizontal plane H2 (the thermal contact surface 250 extends along the horizontal plane H2) intersects. As shown in Fig. 19C, not only the vertical plane F intersects the two horizontal planes H1, H2, but the plurality of fins 256 themselves extend for the height to intersect the two horizontal planes H1, H2. In addition, adjacent fins 256 are separated by an intervening refrigeration passage or air passage 258 through which air can circulate. The plurality of fins 256 and the plurality of refrigerating passages 258 extend perpendicular to a longitudinal axis of the heat sink assembly 240. In this manner, the plurality of fins 256 can occupy a space R behind the housing 222 and above the wires 272 that terminates with a receptacle connector 271 that is supported within the housing 222. Positioning the plurality of fins 256 in this manner will reduce the overall height of the apparatus employing such a housing structure to a height that is generally about the fins that will protrude upwardly from the housing. It is desirable to have the plurality of fins 256 not touch the wire 272 under this orientation. In this regard, the height of the fins 256 is preferably less than the height of the housing 222 as shown.

The heat transfer portion 241 and the heat radiating portion 252 are shown integrally formed as one piece to promote heat transfer from the heat transfer portion 241 to the heat radiating portion 252. However, it is conceivable that the heat transfer portions, the heat radiating portions 241, 252 may be separately formed and then joined together if necessary. In order to further increase the heat transfer from the heat transfer portion 241, a heat transfer member 274 is disposed to extend longitudinally along the heat transfer portion 241 and the heat radiating portion 252 and to contact the heat transfer portion 241 and the heat radiating portion 252. Such a transfer element 274 is shown in the drawings as a heat pipe 275 having a rectangular or elliptical cross-sectional shape including a major axis and a minor axis defining such a shape (Fig. 19D). The rectangular shaped heat pipe 275 increases the amount of contact area between the heat pipe 275 and the heat transfer portion and the heat radiating portions 241, 252 of the heat sink assembly 240. Other non-circular structures (such as a rectangular inner cavity) or even cylindrical may be employed. Heat pipe 275 The trenches 278 are received in a common trench 278 that also extends longitudinally along the heat spreader assembly 240 and extends along the contours of the heat transfer, heat sinks 241, 252. Accordingly, the heat pipe 275 has an offset shape having two distinct portions extending at different heights or height levels of the heat sink assembly 240, namely an evaporation zone 279 and a condensation zone 280, respectively.

The heat pipe 275 is a hollow member having a lumen 282 defined by a side wall 283 that is sealed at its ends and contains a two-phase (e.g., vaporizable) fluid within the lumen 282 of the heat pipe 275. Examples of two-phase fluids that may be present in embodiments of lumen 282 include purified water, chlorofluorocarbons, and the like. Heat pipe 275 and its side walls 283 can be constructed of aluminum, copper or other thermally conductive material. The inner cavity 282 preferably includes an evaporation zone 279, an adjacent heat transfer portion 241, and a condensation zone 280 adjacent the heat sink 252. Heat is transferred from the heat transfer portion 241 through the bottom wall and side walls 283 of the heat pipe 275 into the inner cavity 282, which enables the two-phase fluid present in the evaporation zone 279 to evaporate. The steam can then be condensed into a liquid in the condensation zone 280. In the illustrated embodiment, the steam exotherms as it condenses, and the heat transmits through the sidewall 283 of the heat pipe 275 out of the cavity 282 into the base 254 of the heat sink 252 and its associated fins 256. The inner chamber 282 can include a wick 284 to facilitate condensed liquid travel back along the wick to the evaporation zone 279 (Fig. 19D). The wick 284 may employ a groove-like passage on the inner surface of the inner cavity 282 or an extent of a mesh or the like.

As shown, the heat transfer portion 241 and the heat sink portion 252 of the heat sink assembly 240 extend longitudinally but extend at different height levels, with the heat sink portion 252 being elevated relative to the heat transfer portion 241. This difference in height level facilitates the movement of liquid vapor upward from the heat transfer portion 241 to the heat sink portion 252 to some extent, but its main purpose is to accommodate heat dissipation in the horizontal range of the housing 222. The portion 252 does not have to modify the housing 222 to receive the heat sink 252. If it is desired to extend the heat sink 252 at the same level of height as the heat transfer portion 241, portions of the back wall 224 and top surface 233 that are near them will need to be modified. A groove or recess 246 may be formed in the two walls 224, 233 to receive a region of the heat sink assembly 240 between the heat transfer portion 241 and the heat sink portion 252. Moreover, although most of the heat pipes 275 are illustrated, it will be appreciated that a plurality of heat pipes (such as a pair of heat pipes 290, as shown in phantom in FIG. 19E) may be arranged to be disposed within the grooves 278 of the heat sink assembly 240. In this case, the pair of heat pipes 290 can be sealed in a medium that facilitates heat transfer to compensate for direct contact between a pair of heat pipes 290 and a single rectangular shaped heat pipe 275 as shown. The amount of loss on. A thermal paste or other compound can be placed on the heat pipe 290 to improve heat transfer.

This heat sink assembly 240 thermally engages the housing 222 and uniquely transfers heat to the area behind the housing 222. With this configuration and its depending fins, the apparatus employing such a heat sink assembly 240 can have a reduced height to allow additional equipment to be present in the closets and stacks. The plurality of fins are positioned such that the space between all of the fins is used for refrigeration because none of them have a light pipe or any other element extending through them. The heat sink of the heat sink extends horizontally but is spaced above the motherboard of the device so that the designer can use this open space for additional functional components without increasing the lateral dimensions and depth of the master. An example of a manner of a connector having a heat sink integrated with a connector can be arranged and mounted for a master device as shown in FIGS. 15 to 16A.

As described above, the housing 63 of the first connector 80 can be allowed in a manner that the first connector 80 is not supported by a circuit board but can be supported by other structures such as a conductive frame or an insulating frame. The first connector 80 is positioned in a manner to be positioned. In such a configuration, all of the terminals are preferably connected to conductors in a cable so that signals (high speed and low speed) can be properly guided. In many such embodiments, the housing of the connector can be formed from a die cast material. Such a structure is not necessary as conventional stamped metal structures would also be suitable.

Although a housing can be supported by a separate structure, such as a frame, it should be noted that the connector can mount a circuit board in a more conventional manner. For example, the first connector 80' of Figure 20 is formed from a stamped material and is configured to be press fit to a circuit board and to provide a plurality of ports 80a' in a group structure. The housing 63' includes a plurality of partition walls 63a' that help define the plurality of ports 80a', and the housing 63' includes a plurality of tail portions 63b' configured to be pressed into a circuit board. For certain system configurations that wish to extend a circuit board to an edge of a box, it may be beneficial to have a first connector 80' configured such that it can be press fitted as shown. The method is installed on the circuit board. In such a configuration, it is contemplated that the high speed signal terminals are connected to the cable but the low speed signals are in a desired manner (e.g., a connector such as that given in PCT Application No. PCT/US2014/054100, the entire disclosure of which is incorporated herein by reference. It can be mounted on a cable and connected to a circuit board.

FIG. 21 illustrates an embodiment of a computing device 300. The computing device 300 includes a base 301 having a first wall 305a and a second wall 305b. As shown, the first wall 305a is a front wall and the second wall 305b is a rear wall (the first wall 305a and the second wall 305b are opposite), but the first wall 305a and the second wall 305b need not be in a relative The side and thus many of the structures are feasible.

The base 301 supports a main circuit board 312 that extends to the adjacent front wall 305a. A chip package 340 is disposed on the circuit board 312 and the plurality of first connectors 321 are transmitted through the high speed A plurality of cables 322 of signals are coupled to the wafer package 340. A second connector 331 is positioned at the second wall 305b and is coupled to the wafer package 340 through a plurality of cables 332. Power is supplied to the main circuit board 312 (and various electronic components supported by the main circuit board 312) via a power connection 330. It should be noted that the distance between the first wall 305a and the second wall 305b can exceed 20 inches, but problems can arise if the main circuit board 312 is used to transmit high speed signals, especially when the data rate is close to and exceeds 25 Gbps. The use of non-return-to-zero (NRZ) coding would require a signal transmission frequency of approximately 12.5 GHz and such frequencies would be less compatible with conventional FR4-based boards. In addition, as data rates approach 40 Gbps (and signal transmissions approach 20 GHz), even the use of foreign materials may not be sufficient to allow for the use of conventional board technology.

22-27 illustrate another embodiment of a computing device. Specifically, the computing device 400 includes a base 401 having a first wall 405a and a second wall 405b. A main circuit board 406 is provided, and the main circuit board 406 includes electronic component circuits 407. The electronic component circuit 407 can include a power supply, a cooling fan, and various digital signal transmission wafers that do not need to operate at high frequencies to support the required bandwidth or only travel short distances. A support connection 408 (shown as a cable assembly for assembly but also a two board connector) can provide control signals and/or power signals to a signal Plate 410, signal board 410 supports one or more wafer packages 440.

A plurality of first connectors 421 are positioned along the first wall 405a to provide a plurality of ports 421a to the first wall 405a. Similarly, a plurality of second connectors 431 are positioned along the second wall 405b to define a suitable docking interface along the second wall 405b. However, as can be appreciated, the signal board 410 does not extend to the first connector 421 or second connector 431. Rather, the plurality of first connectors 421 are connected to the wafer package 440 via a plurality of cables 422 while the plurality of second connectors 431 are connected to the wafer package 440 via a plurality of cables 432. The first connector 421 can be formed in a manner similar to that described above and thus the intrinsic details of the connector 421 will not be repeated for the sake of brevity. Rather, it will only be noted that the above-described features of such a connector can be used herein in the desired manner and in the desired combination to provide a connector that meets system requirements.

Referring to Figures 16 and 25, as can be appreciated, the plurality of first connectors 421 can be supported by a front circuit board 411. Alternatively, the front circuit board 411 may be omitted and a frame such as the mounting nest 150 (described above) may be used to support the plurality of first connectors 421. If the front circuit board 411 is employed, the connector support 411a It can be used to fix the first connector 421 to the front circuit board 411.

Referring to FIGS. 25-27, the signal board 410 supports a wafer package 440, the wafer package 440 includes a wafer 445, and a plurality of signal board connectors 426 are disposed around the wafer package 440. In one embodiment, such as shown in Figure 27, the signal board connectors can be positioned on at least two sides of the wafer 445 and can also be positioned on four sides to increase system capacity. At least two signal board connectors 426 can be disposed on each side of the wafer 445 if desired. Of course, if the signal board 410 (or corresponding structure) has a wafer 445 on both sides (e.g., the top side and the bottom side), a further increase in density can be provided.

Referring to Figures 24, 25 and 27, in operation, signal board 410 will be prepared (this may include a reflow operation) and then docked to first connector 421. In order to consider such a mounting process, the cable 422 has a first end 422a terminated to the first connector 421 and a second end 422b terminated to the first board connector 423, and the plurality of first board connectors The 423 can interface with a plurality of signal board connectors 426. Similarly, the cable 432 can have a first end 432a that terminates in the second connector 431 and A second end 432b is terminated to the second board connector 433. Such a structure allows the signal board 410 to be mounted in the base 401 (see FIG. 22) before being connected to the first connector 421 and the second connector 433. In operation, the signal board connector 426 will desirably include a base 426a that supports a plurality of terminals 429 that are soldered to the signal board 410. Of course, in an alternate embodiment, a signal board connector 426 can be press fit onto the signal board 410. Since it is well known to mount a connector on a circuit board using soldering or press-fitting, a further description of the details of such a connector will not be provided herein.

Referring to Figure 26, regardless of how the signal board connector 426 is mounted on the signal board 410, the connector 426 provides a dockable interface for the signal board 410. Signal board 410 supports a wafer 445 (wafer 445 will typically be on some nature carrier or substrate). The wafer 445 can be coupled to the signal board 410 as described above and can be supported by a heat dissipation frame 442 that supports a cooling block 441. The refrigeration block 441, which may be in the form of a conventional heat sink, includes a refrigeration base 441a that will preferably be pressed against the wafer 445 and may include a cooling fin 441b to increase surface area and increase refrigeration of the wafer 445. As can be appreciated, a thermally conductive compound 443 thermally bonds the wafer 445 to the refrigeration base 441a.

Referring to Figures 22 and 26, wafer 445 as described above can be any desired combination of an ASIC, DSP and/or controller and processor whereby wafer 445 is coupled via a signal board 410 via a very short path. The first connector 421 and the second connector 431. Because signal loss is related to the traveled distance along the signal board 410, shortening the distance as shown allows for much lower losses than conventional systems while allowing the use of conventional board materials and construction methods.

28A and 28B illustrate an alternative to a signal board connector and a first board connector. Example. Specifically, a first board connector 523 includes a base 523a that supports one or more rows of cables 522. The base 523a of the first board connector 523 is butted to a signal board connector 526. The signal board connector 526 includes a base 527a in a housing 527b. The housing 527b has a leg portion 527c. The 527c helps to secure the housing 527b to the corresponding signal board. A pressure member 528a secures the first board connector 523 in the signal board connector 526 and includes an actuating element 528b that, when actuated, causes the first board connector 523 to be removed from the signal board connector 526. Move out.

Of course, other configurations of the first board connector and the signal board connector are also possible, and thus the illustrated embodiment is not intended to be limiting unless otherwise stated. Additionally, the first board connector and the second board connector may be the same or different to ensure that the correct connector cannot be installed in the wrong location. To further protect against potential installation problems, the first board connector and the second board connector can each be keyed differently so that only the required structure can be installed.

FIG. 29 shows an alternate embodiment of a wafer package 640. A support layer 614 is configured to support a flexible circuit 690, which in turn supports a substrate 646, and a substrate 646 supports a wafer 645. Wafer 645 and substrate 646 can be as described above. Support layer 614 can be a circuit board or some other material. Although the support layer 614 is shown as being comparable in size to the wafer 645 and the substrate 646, in practice, it is contemplated that the support layer 614 will be larger than the wafer 645 and the substrate 646 to provide additional support. As can be appreciated, instead of using a circuit board to connect the wafer to a connector, the illustrated embodiment uses a flexible circuit to connect the signal board connector 626 to the wafer 645. The entire wafer package 640 can be supported by a frame (not shown) or by a support layer 614, depending on the desired structure, and the entire wafer package 640 can include a heat sink to help dissipate heat from the wafer 645. In such a system, the support layer 614 can be The heat sink provides a mounting location similar to that described above.

The flexible circuit 690 can be made in an intricate pattern because they can be multi-layer thick and the flexible circuit 690 can be terminated to a plurality of signal board connectors 626 (the signal board connector 626 can be configured to be coupled to the corresponding first board connector) So that the rest of the system is basically the same). Thus, a soldered connection between the flexible circuit 690 and the substrate 646, or wafer 645 is feasible. Moreover, as can be appreciated, a flexible circuit 690 can provide high performance relative to a circuit board. It should be noted that although the substrate 646 is shown as a separate component, in one embodiment, the flexible circuit 690 can be substituted for the substrate 646, and, therefore, the use of the substrate 646 is optional in the wafer package 640. Thus, as can be appreciated, a signal board connector 626 can, but need not, be mounted on a circuit board.

It should also be noted that a flexible circuit that is often formed by KAPTON but can also be a rigid-flex circuit can be made of polyimide, acrylate adhesive, epoxy adhesive, fluoropolymer. A solution of a binder is formed. Thus, the choice of materials is not intended to be limiting, as the design of the system allows even relatively poor materials like FR4 to be effectively used. It is expected that the system will provide less loss than the 2dB loss due to the board, and in some cases, the loss between the wafer and the signal board connector can be less than 1 dB.

The description provided herein describes various features by means of its preferred and exemplary embodiments. Numerous other embodiments, modifications, and variations can be made by those skilled in the art in the scope of the appended claims.

Claims (6)

A computing device comprising: a housing having a first wall; a chip package including a wafer electrically coupled to a first signal board connector; a first connector supported adjacent the first wall The first connector includes a housing having a pair of interfaces and a base, the base supporting a first pair of terminals, each of the terminals of the first pair of terminals each having a contact in the docking interface The first connector is supported by a tray; a first cable has a pair of conductors, and the first cable includes a first end and a second end, the first of the first cable Endly connected to the first pair of terminals; a first board connector, the first board connector is terminated to the second end of the first cable and configured to interface with the first signal board connector And an electronic component circuit that supplies power to the wafer. The computing device of claim 1, wherein the wafer is supported by a signal board and the first signal board connector is mounted on the signal board. The computing device of claim 2, wherein the housing comprises a second wall, the computing device comprising a second connector supported adjacent the second wall, the second connector terminating a second end of the second cable, the second cable has a second end connected to a second board connector, wherein the signal board supports a second signal board connector, A second signal board connector is electrically coupled to the wafer, and the second board connector is configured to interface with the second signal board connector. The computing device of claim 1 wherein the wafer package comprises a substrate supporting the wafer. The computing device of claim 4, wherein the substrate is supported by a signal board, The signal board does not extend to the first connector. The computing device of claim 1, wherein the wafer is electrically connected to the plurality of first signal board connectors, and the plurality of first signal board connectors are each docked via a respective first board connector A different first connector.