US20110075306A1

US20110075306A1 - Local integration of non-linear sheet i integrated circuit packages for esd/eos protection

Info

Publication number: US20110075306A1
Application number: US12/963,704
Authority: US
Inventors: Yves Leduc; Nathalie Messina; Charvaka Duvvury; Kurt P. Wachtler
Original assignee: Texas Instruments Inc
Current assignee: Texas Instruments Inc
Priority date: 2007-05-10
Filing date: 2010-12-09
Publication date: 2011-03-31
Also published as: US7872841B2; EP1990834B1; EP1990834A3; US20080278873A1; EP1990834A2

Abstract

A packaged semiconductor device (200) with a substrate (220) having, sandwiched in an insulator (221), a flat sheet-like sieve member (240) made of a non-linear material switching from insulator to conductor mode at a preset voltage. Both member surfaces are free of indentations; the member is perforated by through-holes, which are grouped into a first set (241) and a second set (242). Metal traces (251) over one member surface are positioned across the first set through-holes (241); each trace is connected to a terminal on the substrate top and, through the hole, to a terminal on the substrate bottom. Analogous for metal traces (252) over the opposite member surface and second set through-holes (242). Traces (252) overlap with a portion of traces (252) to form the locations for the conductivity switches, creating local ultra-low resistance bypasses to ground for discharging overstress events.

Description

This is a continuation of application Ser. No. 12/049,726 filed Mar. 17, 2008, the contents of which are herein incorporated by reference in its entirety.

FIELD OF THE INVENTION

The present invention is related in general to the field of electronic systems and semiconductor devices and more specifically to structure and method of local electrostatic discharge protection built into the package of the devices using a non-linear film.

DESCRIPTION OF RELATED ART

In the ongoing trend of scaling the silicon technologies towards the nanometer range, the ever-present challenge of protection against electrostatic discharge (ESD) and other destructive transient effects, such as cable discharge events, transient latch-ups, and electrostatic overcharge (EOS), has become even more difficult. The shrinking geometries of the integrated circuit (IC) reduce the intrinsic capability of the transistors to handle high ESD currents, and reduce the gate oxide breakdown voltage, but increase the metal resistance under ESD conditions.
Additional ESD challenges are presented by the requirements of high performance/high speed circuits for low leakage and low capacitance of the ESD protection structures; further by the integration, in the same chip, of 5 V USB (Universal Serial Bus) applications with 1.2 V operation (2 nm gate oxide technology); and by the customer demands for external interface IC protection to meet the component-level ESD protection simultaneously with the system-level ESD protection.
The system-level events (IEC, also called Human Metal Model) occur at 4 to 8 kV stress and are equivalent to tens of amps for 1 to 2 μs; integrated over time, their energy-under-the-pulse is thus many times more severe than the common Human Body Model (HBM). The current waveforms of different ESD stress events are compared in FIG. 1, which plots current in amperes (A) as a function of time in nanoseconds (ns). FIG. 1 shows an IEC pulse at 8 kV; the initial spike looks like a Charged Device Model (CDM) event followed by an HBM-like wave at nearly 15 A, representing energy under the pulse much larger than the common component level ESD pulse like the HBM event at 1 kV or 2 kV. Consequently, system-level ESD (IEC), which is important for device pins interfacing with the outside world, requires external protection devices that have to handle currents in the 30 A range for stress levels as high as 8 kV. Other ESD such as Cable Discharge Events (CDE) also introduce very high currents even at 1 kV for longer time periods of 1 μs.
The design of on-chip protection devices in standard technology has been frustrated by the impractical size requirements and the difficulty to make them compatible with circuit functions. For example, external interface pins which require system-level protection may use SCR (silicon controlled rectifier) devices, but they need to be free of latch-up issues during ESD stress; in addition, the large protection devices require wide metal widths for the encountered current densities—consequently, they become impractical.
In present technology, the most common approach to meet the stress challenge is to use an external protection method. Sometimes a dozen of small ESD protection devices are placed on the printed circuit board (PCB) to protect the system from ESD. A well known example of this class of protection is normalized as IEC 61000-4-2, where protection devices are placed around all connectors (battery, battery charger, SIM card, keyboard, microphone, earphone, LCD, USB, etc.) on both faces of the printed circuit board. In summary, an expansive and area-consuming approach.
U.S. Pat. Nos. 6,981,319 and 7,218,492, and U.S. Patent Application Publication 2007/0127175 describe devices and systems for electrostatic discharge suppression based on an electrostatic discharge reactance layer built from a polymer-based suppression material embedded with nanometer-size conducting particles. The material switches from insulating to conducting mode at the high voltages of an overcharge event. The device disclosed in the patents uses electrodes embedded in cavities on either side of the electrostatic discharge reactance layer; the electrodes have extensions, which overlap so that they determine the overall protective performance of the device. The structure of the embodiments, however, has the disadvantages that it does not lend itself to the industry trend of miniaturization and to the market need for fine pitch. Further, the method of fabrication is cumbersome and thus expensive.

SUMMARY OF THE INVENTION

The structure of the prior art has distinct disadvantages. The structural complexity, especially the cavities of the electrostatic discharge layer, does not lend itself to design fine-pitched device terminals; also, the overall protection thickness does not facilitate device miniaturization. Further, the inherent resistance in the range of few hundred ma of the protection circuits makes the discharge of the high stress currents in the system-level ESD (IEC) events problematic. In addition, the fabrication method in prior art is cumbersome and thus expensive; it does not lend itself to mass production and low cost.
Applicants' investigations identified a method to protect the multitude of electrodes of an integrated circuit chip against electrical overcharge by assembling the chip onto a substrate, which is structured so that it includes a multitude of local, built-in fine-pitch protection shortcuts to bypass electrical overcharge events directly to ground, before they reach the electrodes. The shortcuts to ground are fine-pitched and exhibit, as measurements have shown, only a few mΩ resistance. They are thus well suited to discharge even the high IEC stress currents found in system-level ESD. The very low resistance compares favorably to the few hundred mΩ resistance inherent in the structures of the existing technology quoted above.
Further, the method for fabricating the substrate with the protection bypasses is low cost. The method lends itself the batch processing and mass production.
In one embodiment of the invention, the substrate has, sandwiched in an insulator, a flat sheet-like sieve member made of a non-linear material that switches from insulator to conductor mode at a preset voltage. The member is perforated with through-holes, otherwise both surfaces of the sheet are free of indentations.
Metal traces over one surface of the sieve member are positioned across a first set of the through-holes; each trace is connected to a terminal on the substrate top and, through the hole, to a terminal on the substrate bottom. Metal traces over the opposite member surface are positioned across a second set through-holes; each trace is connected to a terminal on the substrate bottom and, through the hole, to a terminal on the substrate top. The position of the latter traces overlaps with a portion of the first traces. These overlap areas are the locations for the conductivity switches.
It is a technical advantage that the switch from insulator to conductor mode is practically instant, since it is based on tunneling between nanometer-sized particles embedded in the member material.
The invention employs a flat sheet of the non-linear material, which extends practically throughout the package and can thus protect even the fine-pitched signal and power pins. The solution enabled by the invention saves significant PCB area and is much less expensive than traditional stand-alone protection devices.
The method of the invention is also less expensive than forming cavities in the non-linear material and embedding metal traces for overlaps in the cavities. Due to the minimized electrical paths, the structure of the invention can carry high discharge currents and offers much faster protection than the stacking of chips containing the ESD protections.
Another embodiment of the invention is a method for fabricating a semiconductor device with locally integrated protections against transients. The method includes providing a long tape, with over its whole length and width a thin (about 3 μm) flat sheet of non-linear material sandwiched between two metal layers. The non-linear material switches from an insulator to a conductor at a preset voltage.
The first metal layer is etched to create first traces over the non-linear sheet and gaps between the traces. The second metal layer is etched to create second traces over the non-linear sheet and gaps. The second traces partially overlap with portions of the first traces; the overlap areas are the locations for the conductivity switches.
An insulator foil with a metal layer facing outward is laminated on each of the first and the second trace, filling the gaps between the traces. This creates a flat tape-like substrate for the sites of a plurality of repetitive devices.
Sets of through-holes are drilled into the substrate from top and from bottom through the metal layers and foils to create access to the traces. These through-holes terminate at the traces on the sheet.
To access the first layer traces from the bottom and to access the second layer traces from the top, the through-holes need a set of through-holes in the sheet. Consequently, drilling perforates the sheet with through-holes and turns it into a flat sieve member. Metal (for example, copper) is then deposited to fill the through-holes and to thicken the metal layers.
The thickened metal layers on the substrate top and bottom are etched to create terminals for the metal-filled through-holes. The terminals are distributed, for each device site, between usage for signal and power, and for ground. The selection of the signal and power terminals on the substrate bottom is performed so that the trace, to which each terminal is connected, overlaps with a trace on the opposite surface of the sheet, and that the opposite trace is connected to ground. As a consequence, an electrostatic overcharge that hits the signal/power terminal will readily initiate a local switching of the sheet material to conductor mode and thus bypass the overcharge to ground without giving the transient a chance to damage the corresponding signal/power terminal and the electronic component on the substrate top. This advantage provides protection against transients even for very fine pitch center-to-center of the terminals.
The technical advances represented by certain embodiments of the invention will become apparent from the following description of the preferred embodiments of the invention, when considered in conjunction with the accompanying drawings and the novel features set forth in the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a plot of current in amperes (A) as a function of time in nanoseconds (ns) to compare waveforms of different electrostatic discharge (ESD) stress events. Human Body Model (HBM) and Charge Device Model (CDM) are used to test at the component level; system level events (IEC) are used to test at the system level.

FIG. 2 shows a schematic cross section of a device portion including an electronic component assembled on a substrate fabricated according to the invention. The substrate has, sandwiched in an insulator, a flat sheet-like sieve member made of a non-linear material, which switches from insulator to conductor mode at a preset voltage.

FIG. 3 is a top view of the flat sieve member of an actual semiconductor device of about 12 by 12 mm side length. The diameter of a through-hole is about 80 μm.

FIG. 4 is a simplified circuit diagram illustrating the application of the member of FIG. 3 for effective ESD protection.

FIG. 5 is a plot of current in amperes (A) as a function of time in nanoseconds (ns) to illustrate the advantage of the present invention to discharge an IEC overcharge event effectively to ground using the ultra-low resistance of a local bypass built into the substrate.

FIG. 6 to FIG. 12 show schematic cross sections to illustrate certain process steps of a method for fabricating a semiconductor device with ESD protection according to the invention.

FIG. 6 depicts the steps of providing a sheet of non-linear material sandwiched between metal layers, and of providing insulator foils covered with a metal layer.

FIG. 7 depicts the step of patterning the metal layers on the sheet of non-linear material.

FIG. 8 shows the step of laminating the patterned sheet and the foils, forming a substrate.

FIG. 9 depicts the step of opening sets of through-holes into the substrate.

FIG. 10 shows the step of depositing metal to fill the through-holes and to add continuous metal layers.

FIG. 11 depicts the step of patterning the metal layers on both surfaces of the substrate to create terminals.

FIG. 12 shows the steps of attaching connections (bonding wire and solder bodies) to the substrate terminals, and of defining the local bypasses to ground.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

FIG. 2 illustrates a schematic cross section of an embodiment of the invention, which is an electronic device generally designated 200 with protection against transients. Device 100 includes an electronic component 201 connected by solder bodies 210 to a substrate 220. Component 201 may be a semiconductor chip, or may be another device in need for protection against electrostatic discharge, system level transients, cable discharge events, transient latch-ups, or any other electrostatic overcharge events. As FIG. 2 shows, chip 201 has metallic electrodes for external connections. According to their electrical function, the electrodes are grouped in sets: First electrodes 202 serve electrical signal and power, and second electrodes 203 serve electrical ground potential (or supply/ground; zero power being equivalent to ground). One of the first electrodes is schematically shown with its own protection device 204.
FIG. 2 also shows a substrate 220, onto which the component 201 is assembled, and which has built-in local ESD protection shortcuts for the first electrodes 202. The shortcuts are integral with the substrate and are operable to bypass electrical overcharge events directly to ground. Substrate 220 includes insulator material 221 (such as epoxy glass, ABF, etc.) preferably in the thickness range from about 20 to 60 μm. For some devices it may be thinner or thicker, but driven by handheld and wireless product applications, the overall industry trend is towards thinner thicknesses. Substrate 220 has a top surface 220 a and a bottom surface 220 b. Both top and bottom surfaces have metallic terminals grouped in sets: Top surface 220 a has first set terminals 231 and third set terminals 233; and bottom surface 220 b has second set terminals 232 and fourth set terminals 234.
Substrate 220 further includes a flat sheet-like sieve member 240 sandwiched in the insulator 221. Sieve member 240 extends throughout the length and width of the device. Sieve member 240 is made of a non-linear material, an insulating polymer compound embedded with nanometer-size conducting particles. The compound switches from insulator to conductor mode at a preset voltage, which is mainly determined by the member thickness. For device 200 in FIG. 2, the member 240 is flat and sheet-like in the thickness range between about 3 and 10 μm.
Comparing to known technology, a substrate according to the method of this invention enables the thickness of the sheet to be reduced by about one order of magnitude and still provide adequate ESD protection to the electronic component. The non-linear material is commercially available in industry from, for example, Electronic Polymers, Inc., Round Rock, Tex., U.S.A.
Sieve member 240 has a first surface 240 a and a second surface 240 b. Both surfaces are free of indentations such as cavities, grooves and trenches so the process of forming a substrate is much simplified comparing to known technology. As FIG. 2 illustrates, sieve member 240 is perforated by through-holes, which are grouped in sets: first set through-holes are designated 241; second set through-holes are designated 242.
Referring to FIG. 3, the flat sheet-like sieve member 301 of an actual semiconductor device is depicted in top view in order to illustrate an example of the high number of through-holes 302, and the extension of the member over the entire length and width of the device. The actual size of the square-shaped sieve member in this example is 12 by 12 mm, and the diameter of the through-holes is about 80 μm. In other devices, the through-holes may have preferred diameters between about 80 to 150 μm. The number and distribution of the through-holes depends on the number and position of device terminals to be protected against ESD events.
On the first surface 240 a of sieve member 240 are first metal traces 251, preferably made of copper in the thickness range from about 10 to 25 μm. Each trace 251 is positioned across a first set through-hole 241 through sieve member 240. Further, each trace 251 has a conductive connection to a third set terminal 233 on the first substrate surface 220 a, and, through the respective first set through-hole 241, a conductive connection to a second set terminal 232 on the second substrate surface 220 b. In other words, both the connection to the third set terminal 233 and the connection to the second set terminal 232 terminate at trace 251.
On the second surface 240 b of sieve member 240 are second metal traces 252, preferably made of copper in the thickness range form about 10 to 25 μm. Each trace 252 is positioned across a second set through-hole 242 through sieve member 240 so that trace 252 partially overlaps with a portion of at least one first trace 251. The overlap areas are the locations for the conductivity switches described above, which provide the local bypass to ground for transient events. Further, each trace 252 has a conductive connection to a fourth set terminal 234 on the second substrate surface 220 b, and, through the respective second set through-hole 242, a conductive connection to a first set terminal 231 on the first substrate surface 220 a. Again, both the connection to the fourth set terminal 234 and the connection to the first set terminal 231 terminate at trace 252.
FIG. 2 depicts a flip-chip attachment of a component 201 onto substrate 220. Solder bodies 210 connect first electrodes 202 and second electrodes 203 of component 201 with the first set terminals 231 and the third set terminals 233 of substrate 220. The connection is so that each first electrode 202 is protected by a low-resistance local bypass built into the substrate 220, which discharges any transient pulse to ground by switching the conductivity of the non-linear material between the overlapping traces.
Solder bodies 260 may be used to connect the second set terminals 232 to external electrical signal and power, and to connect the fourth set terminals 234 to external electrical ground. Instead of the solder bodies 210 or the solder bodies 260, wire bonding, pressure contacts, or other interconnection means may be used for achieve electrical connection. When bonding wires are used, it is preferred to also employ an encapsulation material such as a molding compound in order to protect the wires and the component.
FIG. 4 and FIG. 5 illustrate the impact of an ultra-low resistance of the local bypass to shortcut the impinging transient pulse to ground before it can endanger a component. In FIG. 4, a substrate terminal designated 401 is connected through path resistance 402 to the non-linear overcharge suppression material 403 and ground 404. When an electrostatic pulse strikes terminal 401, a peak current I_mflows through the path resistance 402 and the conducting non-linear material 403 to ground 404. Terminal 401 is connected to component/chip electrode (I/O) 410. For illustration purposes, the I/O also has a conventional clamp 411, which is connected to ground 404. Clamp 411 has a resistance, which allows a current I_cto flow to ground 404.
FIG. 5 plots current in amperes (A) through both paths as a function of time in nanoseconds (ns) for two resistances. In the solution of the present invention, the ultra-low resistance of the local bypass puts the value of path resistance 402 in the range of few mΩ(R₁about 2 to 20 mΩ. The current follows the high-energy curve 501 in FIG. 5 for the bypass, with I_maround 30 A, and the low-energy curve 502 for the I/O clamp, with l_conly about 4 A. In contrast, the solution of the known technology puts the value of the path resistance 402 in the range over 100 mΩ(R₂between about 0.1 and 2Ω). In such case, the current follows curve 503 for the connection to ground, with I_mabout 20 to 25 A, but the current 504 through the I/O clamp is considerably more stressful with l_cabout 10 A, which is more likely to damage the electronic component on the substrate.
Another embodiment of the present invention is a method for fabricating a semiconductor device with protection against transient pulses. The method starts with providing a flat tape 601 as depicted in FIG. 6. The tape includes a sheet 602 of a non-linear material sandwiched between a first metal layer 603 and a second metal layer 604. The non-linear material is an insulating polymer compound embedded with nanometer-size conducting particles that allow the material to switch from insulator to conductor mode at a preset voltage. Sheet 602 is preferably between 3 and 10 μm thick. Metal layers 603 and 604 are preferably copper in the thickness range from about 25 to 100 μm. The sheet and the metal layers extend over the entire tape length and width.
Referring now to FIG. 7, in the next process steps, the first and second metal layers are patterned in a photolithographic process. The first metal 603 layer is etched to create first metal traces 703, which are separated by first gaps 713. The second metal layer 604 is etched to create second metal traces 704, separated by second gaps 714. The traces are so designed that the second traces 704 extend across the first gaps 713 and partially overlap with portions of the first traces 703. These overlap areas are designated 724 in FIG. 7; they are the locations for the conductivity switches of the non-linear sheet with the areas.
In the next process step, shown in FIG. 6, a first insulator foil 610 is provided, which is plated with a third metal layer 611. The insulator material may be epoxy glass, ABF, or related compounds, and is preferably between about 20 and 60 μm thick. Metal layer 611 is preferably copper between about 25 and 100 μm thick. In FIG. 7, first foil 611 is shown just before being placed on the first traces 703, with the third metal layer 611 facing outwardly.
A second insulator foil 620 is also provided, which is plated with a fourth metal layer 621. The insulator is preferably a material like epoxy glass or ABF and has a thickness in the range from about 25 to 100 μm. In FIG. 7, second foil 620 is shown just before being placed on the second traces 704, with the fourth metal layer 621 facing outwardly.
In the next process step, illustrated in FIG. 8, the insulators 610 and 620 are laminated unto the metal traces 703 and 704, respectively. In this process, the gaps 713 between traces 703 and gaps 714 between traces 704 are filled with the insulating material. As a result of the step, a substrate 801 is formed, which has a flat sheet 602 of non-linear material with partially overlapping metal traces 703 and 704 sandwiched between the insulator 610 and 620, wherein the non-linear material sheet extends throughout the length and width of the tape. The substrate offers, after completing the following process steps, a plurality of sites for the assembly of components into a series of repetitive devices.
In FIG. 9 results of further process steps are depicted, in which through-holes in the substrate are opened to provide connections to the embedded metal traces. A preferred method of opening the through-holes is laser drilling; alternatively, plasma etching or any suitable drilling technique (such as mechanical) may be employed. The preferred diameter of the through-holes is between about 3 and 10 μm—the small size can provide a high density, fine-pitch center-to-center array of through-holes.
To facilitate the description of the function of the through-holes, they are grouped in sets. The first set of through-holes is designated 901; they are aligned with the first gaps 713 (see also FIG. 7). Through-holes 901 extend through the metal layer 611, the insulator foil 610, the insulator-filled gaps 713, and the sheet 602 of non-linear material; terminate at the metal traces 704. Consequently, first through-holes 901 perforate sheet 602 with a first set of through-holes.
The second set of through-holes are designated 902; the through-holes are aligned with the second gaps 714 (see also FIG. 7). Through-holes 902 extend through the metal layer 611, the insulator foil 620, the insulator-filled gaps 714, and the sheet 602 of non-linear material; and terminate at the metal traces 703. Consequently, second through-holes 902 perforate sheet 602 with a second set of through-holes.
FIG. 3 depicts a functional sheet 602 with the combined first and second sets of through-holes.
The third set of through-holes is designated 903. They extend through metal layer 611 and insulator foil 610, and terminate at the metal traces 703.
The fourth set of through-holes is designated 904. They extend through metal layer 621 and insulator foil 620, and terminate at the metal traces 704.
In the next process step, depicted in FIG. 10, the through-holes are filled with metal, preferably copper. As FIG. 10 shows, metal 1000 is deposited to fill the through-holes of the sets 901, 902, 903, and 904, and, in the same process, to add a continuous (fifth) metal layer 1011 on top of the (third) metal layer 611, and a continuous (sixth) metal layer 1021 on top of the (fourth) metal layer 621. In some embodiments, where the deposited metal (1011, 1021) has the same composition as the metal (611, 621) laminated on the insulator foils, layer 611 together with layer 1011, and layer 621 together with layer 1021 become metal layers of uniform composition, as depicted in FIG. 2. In other embodiments, though, the metal layers maintain their individual characteristics, as the shading of FIG. 10 suggests.
The preferred method of depositing metal is plating. Many embodiments add process steps after the deposition, which prepare the surface of metal 1011 to facilitate the later process of attaching solder bodies. For example, one may add thin layers of nickel or nickel and palladium on copper, or attach wire bonds. Further, it is convenient for many embodiments, to add process steps to prepare the surface of metal 1021 to facilitate the later process of attaching solder bodies.
In the process step depicted in FIG. 11, the metal layers 1011 and 611, and the layers 1021 and 621, are patterned to create terminals for the respective metal-filled through-holes. The terminals to be formed from layers 1011 and 611 are destined to be connected to the electrodes of that electrical component, which needs to be protected against overcharge events. The terminals to be formed from layers 1021 and 621 are destined to be connected to external signal, power, and ground; they are thus referred to as “external” terminals. Preferred patterning techniques include etching by plasma and by chemical processes. The patterned terminals are grouped in sets according to the sets, to which the respective metal-filled through-holes belong, which the terminals serve.
The terminals patterned from the metal layers 1011 and 611, which are connected to a through-hole of the first set (extending through the member 240), represent a first set. In FIG. 11, they are designated 231 in accordance with the designations of FIG. 2.
The terminals patterned from the metal layers 1011 and 611, which are connected to a through-hole of the third set (not extending through the member 240), represent a third set. In FIG. 11, they are designated 233 in accordance with the designations of FIG. 2.
The terminals patterned from the metal layers 1021 and 621, which are connected to a through-hole of the second set (extending through the member 240), represent a second set. In FIG. 11, they are designated 232 in accordance with the designations of FIG. 2.
The terminals patterned from the metal layers 1021 and 621, which are connected to a through-hole of the fourth set (not extending through the member 240), represent a fourth set. In FIG. 11, they are designated 234 in accordance with the designations of FIG. 2. The pattering of the terminals completes the fabrication of substrate 220.
As depicted in FIG. 12, certain second traces 1252 and their respective fourth set terminals 1234 and first set terminals 1231 are selected for connection to electrical ground. The remaining second traces and the first traces, together with their respective terminals (of the fourth set 1234, second set 1232, first set 1231, and third set 1233) are selected for connection to electrical signal and power.
As a result of the selections, any overcharge (symbolized by lightning signs 1270) hitting an external substrate terminal for signal/power switches the sheet of non-linear material locally from insulator to conductor mode (symbolized by lightning signs 1271) and is discharged to ground after traveling along a short and localized path of only few mΩ resistance.
In the next process step, a plurality of electronic components is provided, such as semiconductor chip 201 in FIG. 2. A component has electrodes 202 for electrical signal and power, and electrodes 203 for electrical ground (see FIG. 2). The component electrodes 202 are connected to the site terminals selected for signal and power using solder bodies 210 as shown in FIG. 2; or bonding wires 1280 as shown in FIG. 12. The component electrodes 203 are connected to the respective site terminals selected for ground, again using solder bodies 210 as shown in FIG. 2 or bonding wires 1280 as shown in FIG. 12.
The connecting steps are repeated for each component, until the assembly of each component on a respective substrate site is completed.
For embodiments with bonding wires it is preferred to protect the wire-assembled component in encapsulation compound, preferably using a molding technology.
It is preferred for many embodiments to add solder bodies 260 onto the external substrate terminals; see FIG. 12. Alternatively, terminals 1232 and 1234 may be used as pressure contacts for connecting to external parts, since the terminals are flat and positioned in a plane.
Finally, the tape with the assembled components is singulated into discrete devices. A preferred technique is sawing.
While this invention has been described in reference to illustrative embodiments, this description is not intended to be construed in a limiting sense. Various modifications and combinations of the illustrative embodiments, as well as other embodiments of the invention, will be apparent to persons skilled in the art upon reference to the description. As an example, the component may be a single semiconductor chip or a stack of chips; the component may belong to a particular product family, such as memory, or it may include a composite functionality.
As another example, the insulator material of the substrate, thin as it generally is, may be more or less flexible, even stiff. As yet another example, the pitch center-to-center of the metal-filled through-holes can be reduced so that the invention is effective for many semiconductor device technology nodes and is not restricted to a particular one. It is therefore intended that the appended claims encompass any such modifications or embodiments.

Claims

1. A device comprising:

a semiconductor chip having on a first chip surface electrodes for ground, power, and for input and output signals;

a sheet of non-linear material of an uniform first thickness and a first surface and a second surface;

a first metal pattern having traces between gaps attached to the first surface of the sheet of non-linear material and electrically coupled to the semiconductor chip;

a second metal pattern have traces between gaps attached to the second surface of the sheet of non-linear material;

the sheet of non-linear material being switchable from being insulative to being conductive at a preset voltage across the first thickness;

a first trace of the first metal pattern overlapping a first trace of the second metal pattern across the sheet of non-linear material; and

the first trace of the first metal pattern electrically connected to a first electrode and the first trace of the second metal pattern electrically connected to a ground electrode.

2. The device of claim 1, in which the sheet of non-linear material is perforated with through-holes.

3. The device of claim 2, further comprising conductive members extending through the through-holes.

4. The device of claim 3, in which each conductive member is physically attached to a trace.

5. The device of claim 1, in which the first electrode is an electrode for power.

6. The device of claim 1, in which the first electrode is an electrode for input signal.

7. The device of claim 1, in which the first electrode is an electrode for output signal.

8. The device of claim 1, in which the semiconductor chip has a plurality of ground electrode.

9. The device of claim 8, in which the plurality of ground electrode are physically connected by traces.

10. The device of claim 1, in which the non-linear material between the overlapping portion of the first trace of the first metal pattern and the first trace of the second metal pattern switches from insulative to conductive when the first electrode receives a voltage spike higher than the present voltage.

11. The device of claim 10, in which the switched conductive material reverts to insulative after the spike dissipates from the first electrode.

12. The device of claim 1, further comprising insulated material laminated to the metal traces and co-extensive with the sheet of non-linear material.