Apparatus for electrically coupling bond pads of a microelectronic device

[0002] The present invention relates to methods and apparatuses for electrically coupling bond pads of a microelectronic device.

[0003] Computers and other electronic devices typically include a number of interconnected semiconductor devices. The semiconductor devices include a semiconductor chip or die containing internal circuitry. The dies are generally mounted in a package and connected to other semiconductor devices through external pins or contacts. However, the dies may also be connected directly to other circuitry, including another semiconductor die.

[0004] FIG. 1A is a side elevation view of a portion of a semiconductor die 20 having two bond pads 21 (shown as 21aand 21b) on a surface of the die. The bond pads 21 may be coupled to each other with circuitry 53 that is internal to the semiconductor die 20, as shown schematically in FIG. 1A. One bond pad 21ais electrically coupled with a wire 50 to a lead finger 43 of a conductive lead frame 40. In one conventional arrangement, one end of the wire 50 is bonded to the bond pad 21awith a “ball bond”60 and the other end of the wire 50 is bonded to the lead finger 43 with a “wedge bond”70. The semiconductor die 20 and the lead frame 40 may then be encapsulated in a plastic material (not shown) and ends 42 of adjacent lead fingers 43 (one of which is shown in FIG. 1A) may be bent downward to form connection pins. The pins may be inserted into corresponding sockets of another device (not shown) to couple the semiconductor die with the other device.

[0005] FIG. 1B is an enlarged side elevation view of a portion of the semiconductor die 20 shown in FIG. 1A, as the wire 50 is being attached to the bond pad 21a. The wire 50 can be attached with a wire bonding tool 30 (shown in FIG. 1B as a ball/wedge bonder 30a) by feeding the wire 50 downwardly through an aperture 31 of the ball/wedge bonder 30aand forming a wire ball 51 at the end of the wire 50. The ball/wedge bonder 30athen presses the wire ball 51 against the bond pad 21awhile the remainder of the wire 50 extends approximately normal to the bond pad 21a. The bonder 30athen applies heat and/or pressure to the wire 50 at the wire ball 51 to bond the wire to the bond pad 21a, forming the ball bond 60 shown in FIG. 1A. For example, the bonder 30acan use a thermosonic or thermocompression process to apply both heat and pressure to the wire 50. The bonder 30athen moves along the wire 50 to the lead finger 43 and presses the wire 50 against the lead finger 43. The bonder again applies heat and/or pressure to the wire 50 to both bond the wire 50 to the lead frame 40 (forming the wedge bond 70 shown in FIG. 1A), and separate the bonded portion of the wire 50 from a remaining portion of the wire.

[0006] FIG. 2A is a side elevation view of the semiconductor die 20 having the wire 50 connected between the bond pad 21 a and the lead finger 43 in accordance with another conventional arrangement in which a first wedge bond 70ais formed at the bond pad 21aand a second wedge bond 70bis formed at the lead finger 43. FIG. 2B is an enlarged side elevation view of a portion of the semiconductor die 20 shown in FIG. 2A as the wire 50 is being attached to the bond pad 21a.

[0007] Referring to FIG. 2B, the wire 50 can be attached to the bond pad 21awith a wedge/wedge bonder 30bby feeding the wire 50 through an aperture 31aof the wedge/wedge bonder 30band pressing the wire 50 against the bond pad 21a. The wedge/wedge bonder 30bthen applies heat and/or pressure to the wire 50 to bond the wire to the bond pad 21a, forming the first wedge bond 70ashown in FIG. 2A. The bonder 30bthen moves along the wire 50 to the lead finger 43 and presses the wire against the lead finger 43. The bonder 30bagain applies heat and/or pressure to the wire to bond the wire 50 to the lead finger 43, forming the second wedge bond 70bshown in FIG. 2A. In one conventional arrangement, the bonder 30bcan apply sufficient heat and/or pressure to both bond the wire 50 to the lead frame 40 and separate the bonded wire from the remaining supply of wire. In another conventional arrangement, the bonded wire can be separated from the remaining wire by clipping the wire next to the second wedge bond 70b.

[0008] As discussed above, two or more bond pads 21 may be connected within the die 20 by internal circuitry 53. The internal circuitry 53 may include very small conductive lines. One drawback with this arrangement is that the conductive lines may have a high resistance, increasing the current necessary to transmit signals between the bond pads, and increasing the heat generated by each semiconductor die. In addition, internal circuitry 53 is inaccessible once the die has been manufactured. Accordingly, another drawback with conventional arrangements is that they may lack the flexibility for interconnecting bond pads that are not connected by the internal circuitry at the time of manufacture.

[0009] Yet a further drawback with the conventional methods and devices discussed above is that it may be difficult to route wires between the lead frame 40 and bond pads that are not proximate to the lead frame 40. For example, if one or more of the wires 50 is particularly long, so as to reach a particular bond pad, the wire may be more likely to break or contact other adjacent wires, creating a short circuit that can affect the operation of the semiconductor device.

[0010] The present invention is directed toward methods and apparatuses for electrically coupling bond pads of a microelectronic device. In one aspect of the invention, the apparatus can include first and second spaced apart bond pads on a surface of a microelectronic device. The microelectronic device can further include a conductive member connected to and extending between the first and second bond pads. The conductive member can be positioned on or above the surface of the microelectronic device. In one aspect of the invention, the conductive member can include a wire, and, in another aspect of the invention, the conductive member can include a flowable conductive material, such as a conductive epoxy. In still another aspect of the invention, the microelectronic device can include an insulating material between the conductive member and the surface of the microelectronic device.

[0011] In yet another aspect of the invention, the apparatus can include a microelectronic device having at least one bond pad for receiving wire connections. The microelectronic device can further include two wires connected to the same bond pad, for example, a first wire connected at one end to the bond pad with a first bond and a second wire connected at one end to the first bond with a second bond. Either or both of the first and second bonds can be a wedge bond or a ball bond, and the opposite ends of the first and second wires can be connected to other bond pads of the microelectronic device, or to external structures.

[0012] FIG. 1A is a side elevation view of a portion of a semiconductor die and lead frame in accordance with the prior art.

[0013] FIG. 1B is an enlarged partial cross-sectional side elevation view of a portion of the semiconductor die and the lead frame shown in FIG. 1A adjacent a wire bonding tool in accordance with the prior art.

[0014] FIG. 2A is a side elevation view of a semiconductor die and lead frame in accordance with another embodiment of the prior art.

[0015] FIG. 2B is an enlarged partial cross-sectional side elevation view of a portion of the semiconductor die and the lead frame shown in FIG. 2A adjacent a wire bonding tool in accordance with another embodiment of the prior art.

[0016] FIG. 3A is a side elevation view of a portion of a microelectronic device assembly having a bond pad with a ball bond and a wedge bond on top of the ball bond in accordance with an embodiment of the invention.

[0017] FIG. 3B is a side elevation view of a portion of a microelectronic device assembly having a bond pad with two wedge bonds attached to the bond pad in accordance with another embodiment of the invention.

[0018] FIG. 3C is a side elevation view of a portion of a microelectronic device assembly having a bond pad with a wedge bond and a ball bond on top of the wedge bond in accordance with still another embodiment of the invention.

[0019] FIG. 3D is a side elevation view of a portion of a microelectronic device assembly having a bond pad with two ball bonds attached to the bond pad in accordance with yet another embodiment of the invention.

[0020] FIG. 3E is a side elevation view of a microelectronic device assembly in accordance with another embodiment of the invention.

[0021] FIG. 4 is a cross-sectional side elevation view of a portion of a microelectronic device assembly having two bond pads connected to each other with a flowable conductive material in accordance with still another embodiment of the invention.

[0022] FIG. 5 is a top plan view of a portion of a microelectronic device assembly having two crossing conductive members extending between bond pads of the assembly.

[0023] FIG. 6 is a top plan view of a microelectronic device assembly having conductive members extending between bond pads of the microelectronic device assembly in accordance with still another embodiment of the invention.

[0024] FIG. 7 is a top plan view of a microelectronic device assembly having a conductive member extending beneath a support portion of a lead frame in accordance with another embodiment of the invention.

[0025] FIG. 8 is a top plan view of a wafer having a plurality of interconnected microelectronic dies in accordance with yet another embodiment of the invention.

[0026] The present invention is directed toward methods and apparatuses for conductively linking bond pads of microelectronic devices. The apparatus can include a microelectronic device having wires bonded between bond pads of a single device or between the bond pads of two devices. The wires can be bonded to the bond pads using various combinations of ball bonds, wedge bonds and other types of bonds. Many specific details of certain embodiments of the invention are set forth in the following description and in FIGS. 3A-8 to provide a further understanding of such embodiments. One skilled in the art, however, will understand that the present invention may have additional embodiments and that they may be practiced without several of the details described in the following description.

[0027] FIG. 3A is a side elevation view of a portion of a microelectronic device assembly 110 in accordance with an embodiment of the invention. The assembly 110 can include a microelectronic device 120 having bond pads 121 (shown as a first bond pad 121aand a second bond pad 121b) on an upper surface 123 of the microelectronic device 120. The bond pads 121 can be connected to each other or to external structures with conductive members 150 (shown as a first conductive member 150aand a second conductive member 150b) that are positioned external to the upper surface 123 of the microelectronic device 120. In one embodiment, the conductive members 150 can include wires made from aluminum, gold or other conductive materials, and in other embodiments, the conductive members 150 can include other structures and materials.

[0028] In one embodiment, the first conductive member 150acan be connected between the first bond pad 121aand an adjacent conductive structure, such as a lead frame 140. In one aspect of this embodiment, the first conductive member 150acan be connected to the first bond pad 121awith a ball bond 160aand can be connected to the lead frame 140 with a wedge bond 170a. The ball bond 160acan be made by forming an enlarged region at one end of the first conductive member 150aand then engaging the enlarged region with the bond pad 160aand applying heat and/or pressure to the first conductive member 150aand/or the bond pad 121a. The wedge bond 170acan be formed by engaging a side surface near the other end of the first conductive member 150awith the lead frame 140 and applying heat and/or pressure to the first conductive member 150aand/or the lead frame 140. Suitable devices for forming both the ball bond 160aand the wedge bond 170aare available from Small Precision Tools of Petaluma, Calif.

[0029] The bond pads 121aand 121bcan be connected to each other with the second conductive member 150b. Accordingly, the second conductive member 150bcan include a second ball bond 160bat the second bond pad 121band a second wedge bond 170bat the first bond pad 121a. In one embodiment, a lower surface 171 of the second wedge bond 170bcan be adjacent an upper surface 162 of the first ball bond 160a. In another embodiment, the second wedge bond 170bcan be adjacent the first ball bond 160ain a side-by-side arrangement, for example, where the first bond pad 121ais large enough to permit such an arrangement.

[0030] An advantage of the assembly 110 shown in FIG. 3A is that the second conductive member 150bcan provide a conductive link between the bond pads 121aand 121bwithout changing the internal circuitry of the microelectronic device 120. This can be advantageous where no internal connections couple the bond pads 121aand 121bat the time the microelectronic device 120 is manufactured, and where the conductive link provided by the second conductive member 150bcan be used to modify or repair the microelectronic device 120 after it has been manufactured. Alternatively, where a conductive path does exist between the two bond pads 121aand 121b(for example, a metal layer or other internal circuitry 153 shown schematically in FIG. 3A), the second conductive member 150bcan supplement the internal circuitry 153. In one embodiment, the second conductive member 150bcan be relatively large when compared to the internal circuitry 153 and can accordingly reduce electrical resistance between the two bond pads 121aand 121b, increasing current flow and signal speed between the bond pads and decreasing the amount of heat generated by the microelectronic device 120.

[0031] Another advantage is that in one embodiment, at least some of the internal connections of the microelectronic device 120 can be eliminated and replaced with the external second conductive member 150b. This arrangement can make volume available within the microelectronic device 120 for other circuit elements, potentially increasing the capability of the microelectronic device 120.

[0032] Still another advantage is that a variety of microelectronic devices 120 can be made compatible with a single lead frame configuration, or conversely, a variety of lead frames 140 can be made compatible with a single microelectronic device configuration. For example, where the portion of the lead frame 140 shown in FIG. 3A is configured to be proximate to the first bond pad 121a, and a later version of the microelectronic device 120 requires that the signal produced at that portion of the lead frame 140 be transmitted to the second bond pad 121b, the second conductive member 150bcan be used to route the signal to the second bond pad 121b.

[0033] Yet another advantage is that the second conductive member 150bcan be selectively sized to affect the capacitance of circuits that include the first and second bond pads 121aand 121b. For example, the distance between the second conductive member 150band the microelectronic device 120 can be varied and/or a dielectric material 152 can be positioned between the second conductive member 150band the microelectronic device 120 to adjust the capacitance of selected circuits in the microelectronic device 120.

[0034] FIG. 3B is a side elevation view of a portion of a microelectronic device assembly 210 having two wedge bonds 270aaand 270battached to the first bond pad 121ain accordance with another embodiment of the invention. As shown in FIG. 3B, one end of a first conductive member 250acan be attached to the first bond pad 121awith the first wedge bond 270aa, and the other end of the first conductive member 250acan be attached to the lead frame 140 with a second wedge bond 270a. A second conductive member 250bcan be connected between the first bond pad 121aand the second bond pad 121bwith a third wedge bond 270band a fourth wedge bond 270bb. In one embodiment, a lower surface 271 of the third wedge bond 270bengages an upper surface 272 of the first wedge bond 270aa. In another embodiment, the first wedge bond 270aacan be placed on top of the third wedge bond 270b, or the wedge bonds 270aaand 270bcan be placed side by side, as discussed above with reference to FIG. 3A.

[0035] An advantage of the arrangement shown in FIG. 3B is that it may be easier to form wedge bonds than it is to form ball bonds because the wedge bonds do not require forming an enlarged portion at the end of the wire. Conversely, an advantage of the arrangement shown in FIG. 3A is that once the ball bond is formed at one end of the conductive member 150, the other end of the conductive member can be positioned in any direction relative to the ball bond, whereas in the arrangement shown in FIG. 3B, the tool that forms wedge bonds at both ends of the conductive members 250 can generally move in only one direction.

[0036] Furthermore, the ball bonds may be less likely than the wedge bonds to damage the structure to which they are attached. Accordingly, it may be advantageous to apply a ball bond to the relatively small bond pads 121 of the microelectronic device 120, and apply wedge bonds to the larger and more robust lead frames 140 (as shown in FIG. 3A). However, this difference between the arrangements shown in FIG. 3A and FIG. 3B may be less important where the tool forming the wedge bond clips the conductive member adjacent the wedge bond rather than deforming the conductive member to separate it from the wedge bond. In either case, the bond that is formed first on the bond pad 121, whether it is a wedge bond or a ball bond, can provide a support structure that protects the bond pad 121 from damage which might otherwise be incurred by the bonding tool.

[0037] In other embodiments, other combinations of ball bonds and wedge bonds may be used to connect a bond pad 121 on one portion of a microelectronic device 120 with a bond pad 121 on another portion of the microelectronic device 120. For example, FIG. 3C is a side elevation view of a portion of a microelectronic device assembly 310 with a first conductive member 350acoupled to the lead frame 140 with a first wedge bond 370aand coupled to the first bond pad 121awith a second wedge bond 370aa. A second conductive member 350bcan be coupled to the first bond pad 121awith a ball bond 360bon top of the second wedge bond 370aa, and can be coupled to the second bond pad 121bwith a third wedge bond 370b.

[0038] FIG. 3D is a side elevation view of a portion of yet another microelectronic device assembly 410 having a first conductive member 450aconnected at one end to the first bond pad 121awith a first ball bond 460aand connected at the opposite end to the lead frame 140 with a first wedge bond 470a. A second conductive member 450bcan be coupled to the first bond pad 121awith a second ball bond 460bthat is positioned on top of the first ball bond 460a. The second conductive member 450bcan then be connected to the second bond pad 121bwith a second wedge bond 470b.

[0039] As shown in FIG. 3D, the assembly 410 can also include an insulator 452 positioned between the second conductive member 450band the upper surface 123 of the microelectronic device 120. In one embodiment, the insulator 452 can include a polyimide and in other embodiments the insulator 452 can include other non-conductive materials. The insulator 452 can reduce the likelihood of contact between the second conductive member 450band conductive structures (not shown) on the upper surface 123 of the microelectronic device 120.

[0040] In still a further embodiment shown in FIG. 3E, one or both of the bond pads 121 can be coupled directly to an external device without the first conductive member 450a. For example, in a “flip chip” embodiment, the microelectronic device 120 can be inverted and placed face down on a substrate 145 such that the upper surface 123 of the microelectronic device 120 faces the substrate 145 with the second ball bond 460band the second wedge bond 470bconnected directly to contacts 146 on the substrate 145. In other embodiments, this type of arrangement can include any of the bonding configurations shown in FIGS. 3A-3C.

[0041] FIG. 4 is a cross-sectional side elevation view of a portion of a microelectronic device assembly 510 having two bond pads 121aand 121bdirectly connected with a flowable conductive material 554 in accordance with still another embodiment of the invention. In one embodiment, the conductive material 554 can include an epoxy having a suspension of conductive particles, such as silver, that create a conductive path between the bond pads 121. The epoxy can be applied in a liquid or paste form that then hardens to form a permanent connection between the bond pads. In other embodiments, other materials can be used to connect the bond pads 121, so long as they provide a conductive path between the bond pads.

[0042] The microelectronic device assembly 510 shown in FIG. 4 can also include an insulator 552 between the conductive material 554 and the upper surface 123 of the microelectronic device 120. As discussed above with reference to FIG. 3D, the insulator 552 can prevent electrical contact between the conductive material 554 and structures on the upper surface 123 of the microelectronic device 120.

[0043] An advantage of the assembly 510 shown in FIG. 4 is that it may be easier to apply the flowable conductive material 554 to the microelectronic device 120 than it is to attach the conductive members shown in FIGS. 3A-3D. For example, an applicator used to dispense the conductive material 554 can travel in any direction once initial contact has been made with one of the bond pads 121. Furthermore, the conductive material 554 may not require heat or pressure to form a bond with the bonds pads 121, reducing the likelihood of damaging either the bond pads 121 or the microelectronic device 120 in the region of the bond pads. Conversely, an advantage of the conductive members shown in FIGS. 3A-3D is that they may be more conductive than the conductive material 554, thereby providing faster signal speeds and less heat than the conductive material 554.

[0044] FIG. 5 is a top plan view of a portion of a microelectronic device assembly 610 having a microelectronic device 620 with two crossing conductive members 650 (shown as a first conductive member 650aand a second conductive member 650b) connected between four bond pads 621 (shown as 621a-621d) of the microelectronic device 620. In one embodiment, the first conductive member 650acan be coupled between first and second bond pads 621aand 621b, and the second conductive member 650bcan be coupled between third and fourth bond pads 621cand 621dsuch that two conductive members 650 cross over each other. In one aspect of this embodiment, one or both of the conductive members 650 can include an insulating layer 655 to prevent inadvertent electrical contact between the conductive members 650. Alternatively, insulators such as those shown in FIGS. 3D and 4 can be positioned between the first and second conductive members 650aand 650b, and/or between the conductive members and an upper surface 623 of the microelectronic device 620.

[0045] In one embodiment, wire bumps 656 can be disposed on the bond pads 621 beneath the conductive members 650 to reduce the likelihood of damaging the bond pads 621 when the conductive members 650 are coupled to the bond pads 621. Details of the wire bumps 656 and the application of the wire bumps to the bond pads 620 are disclosed in copending U.S. patent application Ser. No. 08/840,604, incorporated herein in its entirety by reference. Alternatively, where the bond pads 621 are connected to external devices, such as a lead frame, the conductive members 650 can be coupled to the bond pads 621 by attaching the conductive members on top of the bonds of other conductive members, in a manner generally similar to that shown in FIGS. 3A-3D. In still another embodiment, the conductive members 650 can be bonded directly to the corresponding bond pads 121 without any intermediate structures.

[0046] FIG. 6 is a top plan view of a microelectronic device assembly 710 that includes a microelectronic substrate 720 having bond pads 721 (shown as a first bond pad 721aand a second bond pad 721b). The assembly 710 further includes a lead frame 740 having lead fingers 743 (shown as a first lead finger 743aand a second lead finger 743b) positioned adjacent the microelectronic device 720. Each lead finger 743 may be coupled with a conductive member to a corresponding bond pad 721. For example, the first lead finger 743acan be coupled to the first bond pad 721awith a first conductive member 750a, and the second lead finger 743bcan be coupled to the second bond pad 721bwith a second conductive member 750b.

[0047] Alternatively, the second bond pad 721bcan be coupled to the first lead finger 743avia a third conductive member 750aa, shown in dashed lines in FIG. 6. For example, the first lead finger 743acan be connected to the first bond pad 721a, as discussed above, and the first bond pad 721acan be connected to an auxiliary bond pad 721cwith a conductive element 753, such as a flat metal layer 753. The conductive element 753 can be positioned either above, on, or below an upper surface 723 of the microelectronic device 720, and the auxiliary bond pad 721ccan be connected to one end of the third conductive member 750aa. Accordingly, the conductive element 753, the auxiliary bond pad 721c, and the third conductive member 750aacan provide a connection between the second bond pad 721band the first lead finger 743a.

[0048] An advantage of the assembly 710 shown in FIG. 6 is that the auxiliary bond pad 721callows the second bond pad 721bto be coupled to the first lead finger 743awithout routing the third conductive member 750aaover other lead fingers or over other bond pads. Accordingly, the assembly 710 can reduce the likelihood that the third conductive member 750aawill contact or otherwise interfere with other bond pads or lead fingers of the assembly 710.

[0049] Another advantage of the assembly 710 shown in FIG. 6 is that the auxiliary bond pad 721ccan be positioned on the microelectronic device 720 such that the length of the third conductive member 750aais less than it would be if it were to extend all the way to the first lead finger 743a. Accordingly, the third conductive member 750aamay be less likely to make electrical contact with other structures of the microelectronic device 720, and may be less likely to become entangled or otherwise engaged with such structures, which might damage the third conductive member 750aa. Furthermore, by placing the auxiliary bond pad 721cnear the second bond pad 721b, the third conductive member 750aamay be more nearly “vertical”(as seen FIG. 6) near the second bond pad 721b, which may reduce stresses at the bond between the third conductive member 750aaand the second bond pad 721b.

[0050] FIG. 7 is a top plan view of another embodiment of the assembly 710 in which the lead frame 740 includes a bar or a bus 741 that extends over the upper surface 723 of the microelectronic device 720. Accordingly, the microelectronic device 720 can include a metal layer 753athat extends between the first bond pad 721aand the auxiliary bond pad 721cbeneath the upper surface 723 of the microelectronic device 720 so as not to contact the bus 741. For example, in one embodiment, the metal layer 753acan be disposed beneath the upper surface 723 of the microelectronic device 720 during fabrication of the microelectronic device.

[0051] The microelectronic device 720 can further include third and fourth bond pads 721dand 721ethat can be selectively coupled to the auxiliary bond pad 721cusing any of the techniques described above with reference to FIGS. 3A-5. For example, in one embodiment, the third bond pad 721dcan be connected to the auxiliary bond pad 721cto produce a desired operating characteristic, such as an operating voltage, in the microelectronic device 720. Alternatively, the fourth bond pad 721ecan be connected to the auxiliary bond pad 721cto produce a different operating characteristic in the microelectronic device 720. In still another arrangement, both the third and fourth bond pads 721dand 721ecan be connected simultaneously to the auxiliary bond pad 721c. Accordingly, an advantage of the assembly 710 shown in FIG. 7 is that the combination of the metal layer 753aand the auxiliary bond pad 721ccan be used to route signals from the first lead finger 743ato one or more bond pads that are not adjacent the lead finger 743a. In alternate embodiments, the metal layer 753acan be replaced with insulated conductive members, such as those shown in FIG. 5, or other structures that similarly extend between the first bond pad 721aand the auxiliary bond pad 721cwithout making electrical contact with the bar 741.

[0052] FIG. 8 is a top plan view of a microelectronic wafer 822 having a plurality of microelectronic dies 820, including a first microelectronic die 820aand a second microelectronic die 820b. The first die 820acan include a first bond pad 821aand the second die 820bcan include a second bond pad 821b. A conductive member 850 can be connected between the bond pads 821aand 821bto interconnect microelectronic dies positioned on the wafer 822.

[0053] An advantage of the embodiment shown in FIG. 8 is that two or more microelectronic dies 820 can be connected to each other to form a larger device before the dies 820 are separated from the wafer 822. For example, several memory dies can be connected to form a device having a larger memory capacity. Accordingly, the dies 820 form building blocks and the same die design can be used for both relatively small and relatively large memory devices. By connecting the dies of the relatively large device before the dies are separated, the individual dies need not be separated at all, reducing the likelihood that the dies will be damaged during separation. Furthermore, this arrangement virtually eliminates the possibility that the dies will be misaligned relative to each other when they are connected together.

[0054] From the foregoing it will be appreciated that, although specific embodiments of the invention have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the invention. Accordingly, the invention is not limited except as by the appended claims.