United States Patent 19
`Higgins, III
`
`54
`
`75
`
`73
`
`21
`22
`51
`52
`
`58
`
`SHIELDED ELECTRONIC COMPONENT
`ASSEMBLY AND METHOD FOR MAKING
`THE SAME
`
`Inventor: Leo M. Higgins, III, Austin,Tex.
`Assignee: Motorola Inc., Schaumburg, Il.
`Appl. No.: 229,495
`Filed:
`Apr. 19, 1994
`Int. Cl. ... HOSK 1/OO
`U.S. Cl. ................................ 174/35 MS; 174/35 R:
`361/816; 361/818; 257/655; 257/660
`Field of Search .................................. 174/35 R, 260,
`174/35 MS; 361/816, 818; 257/655, 660,
`728, 659
`
`56)
`
`References Cited
`U.S. PATENT DOCUMENTS
`5,146,047 9/1992 Nagata et al. ..................... 174/35MS
`5,166,772 11/1992 Soldner et al. ......................... 257/659
`5,166,864 11/1992 Chitwood et al. .......
`... 361/386
`5,175,613 12/1992 Barker, III et al...
`... 257f73
`5,341,274 8/1994 Nakatani et al. ....
`... 361/88
`5,379,185
`1/1995 Griffen et al. ...
`... 36/709
`5,392,197 2/1995 Cuntz et al. .....
`... 36/818
`5,394,304 2/1995 Jones ................
`... 361/765
`5,513,078 4/1996 Komrska et al. ....................... 361/816
`FOREIGN PATENT DOCUMENTS
`2055413 5/1992 Canada ............................ H05K f16
`
`
`
`USOO5639989A
`Patent Number:
`11
`45 Date of Patent:
`
`5,639,989
`Jun. 17, 1997
`
`OTHER PUBLICATIONS
`William M. Hall; "Design Tech. for Control of Radiated and
`Conducted Noise in Portable Computing Equipment;”
`Northcon Conference, Oct. 1-3, 91; pp. 258-263 (Oct.
`1991).
`Howard W. Markstein; "Shielding Electronics From EMI/
`RFI;” Electronic Packaging & Production; pp. 40-44 (Jan.
`1991).
`Primary Examiner-Laura Thomas
`Attorney, Agent, or Firm-Patricia S. Goddard
`57
`ABSTRACT
`Electronic components are shielded from electromagnetic
`interference (EMT) by one or more conformal layers filled
`with selected filler particulars for attenuate specific EMI
`frequencies or a general range of frequencies. Shielding is
`accomplished through the use of a single general purpose
`shielding layer, or through a series of shielding layers for
`protecting more specific EMI frequencies. In a multilayer
`embodiment, a semiconductor device (50) is mounted on a
`printed circuit board substrate (16) as a portion of an
`electronic component assembly (10). A conformal insulating
`coating (24) is applied over the device to provide electrical
`insulation of signal paths (e.g. leads 54 and conductive
`traces 18) from subsequently deposited conductive shielding
`layers. One or more shielding layers (60, 62, and 64) are
`deposited, and are in electrical contact with a ground ring
`(56). In a preferred embodiment, the ground connections for
`the shield layers are separate from those used for power
`distribution within the devices.
`27 Claims, 7 Drawing Sheets
`
`exertexx-xx ceae
`
`are
`
`w
`
`s
`
`s
`
`sa s
`
`ay
`
`s
`
`
`
`
`
`
`
`
`
`
`
`
`
`t
`
`
`
`
`
`
`
`
`
`
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 001
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 1 of 7
`
`5,639,989
`
`24
`
`26
`
`12
`
`frn 15
`13
`Q Q.C.O.
`O.A.
`
`AaxYaYa YaaS Ya YaYa Yaaa.
`
`SSSSYSSYSSYYYYY
`
`18
`
`NYAXAYYX YXaAKAs YAa
`
`S R R
`
`-10
`
`19
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`18
`
`18
`
`18
`
`18
`
`18
`
`18'
`
`or 4fe
`
`22
`to
`
`aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaayaa24/22/22/2ZZYZZYaYa YaYaaaaaa
`
`25
`
`A77G.4
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 002
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 2 of 7
`
`5,639,989
`
`
`
`N
`
`A
`
`T
`AVO
`V G
`
`F
`
`G
`
`S
`
`OOP)
`
`CD
`
`t O
`
`O
`
`n
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 003
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 3 of 7
`
`5,639,989
`
`
`
`Yarra/YYYYY
`
`zz
`
`ZYYYYYYYYY/YYYYY
`
`2
`
`O
`CN
`
`aaaaaaaaaaaay
`
`razz
`
`fa
`
`Xaaaaaaa
`
`YYYYYYYY-2
`
`aay
`
`221222/27aaZ212
`
`Yaaaaaaaaaaaa.
`
`79
`
`YAYAZYZYYYYY
`
`
`
`
`
`
`
`
`
`
`
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 004
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 4 of 7
`
`5,639,989
`
`Oz)^
`
`aw tyczzzzzzzzzz
`
`
`
`ZYaaaaaaaaZY2Y
`
`y 2
`
`Yaaaaaaaaaay
`
`??????????????/*
`
`t
`
`YYYYYYYY
`
`zz
`
`YaaaaaZZZY
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 005
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 5 of 7
`
`5,639,989
`
`- -
`
`O
`
`20
`
`22OOOPF
`40 2200PF
`INSERTION
`1000PF
`LOSS
`60
`47opf
`(db)
`22OPF
`100PF
`47PF
`
`80
`
`100
`1
`
`5
`
`10
`
`50 100
`FREQUENCY (MHz)
`
`500 1000 2000
`
`2000
`
`
`
`1600
`
`1200
`IMPEDANCE
`(0) 800
`
`400
`
`O
`1
`
`1000
`
`10
`
`100
`FREQUENCY (MHz)
`Az7G 27
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 006
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 6 of 7
`
`5,639,989
`
`10.0
`
`1.0
`
`LOSS
`TANGENT
`(%)
`
`.01
`1.O
`
`250
`
`200
`
`10
`
`100
`FREQUENCY (MHz)
`AVGS
`
`1000
`
`INITIAL
`PERMEABILITY
`(Hi)
`
`150
`
`100
`
`50
`
`-100
`
`O
`
`400
`
`500
`
`100
`200
`300
`TEMPERATURE (C)
`AVG9
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 007
`
`

`

`U.S. Patent
`
`Jun. 17, 1997
`
`Sheet 7 of 7
`
`5,639,989
`
`LOSS FACTOR
`
`(LOSS ANGEN)
`
`t
`
`-2
`10
`
`INITIAL
`PERMEABILITY
`
`1000
`
`
`
`100
`
`NITIAL
`PERMEABILITY
`(ii)
`
`OSS FACTOR
`
`10
`
`1000
`
`1
`
`10
`
`100
`FREQUENCY (MHz)
`A77GZO
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 008
`
`

`

`5,639,989
`
`1.
`SHIELDED ELECTRONIC COMPONENT
`ASSEMBLY AND METHOD FOR MAKING
`THE SAME
`
`O
`
`15
`
`20
`
`25
`
`30
`
`35
`
`45
`
`FIELD OF THE INVENTION
`The present invention relates to electronic component
`assemblies in general, and more specifically to electronic
`component assemblies which are shielded to protect against
`electromagnetic interference (EMI).
`BACKGROUND OF THE INVENTION
`Electromagnetic interference (EMI) is the generation of
`undesired electrical signals, or noise, in electronic system
`circuitry due to the unintentional coupling of impinging
`electromagnetic field energy. Any propagating electrical
`signal is comprised of an electric field (E-) and magnetic
`field (H-) component. The sinusoidal nature of the signal
`results in the tendency of circuit components, such as wires,
`printed circuit board conductors, connector elements, con
`nector pins, cables, and the like, to radiate a portion of the
`spectral energy comprising the propagating signal. Circuit
`elements are effective in radiating spectral components
`which have wavelengths similar to the radiating element
`dimensions. Long circuit elements will be more effective in
`radiating low frequency noise, and short circuit elements
`will be more effective in radiating high frequency noise.
`These circuit elements behave just like antennae which are
`designed for the transmission of the radiating wavelengths.
`Integrated circuits (ICs) which have output drivers that
`create pulses with high amounts of spectral energy are more
`likely than low power drivers to cause EMI due to the
`probable mismatch between the driver and line impedance,
`and the resistance to instantaneous signal propagation
`imposed by the parasitics of the conductor. For instance,
`CMOS (complementary metal oxide semiconductor) ICs
`which switch 5 volts in rapid rise times can have a large
`content of high frequency components and high spectral
`energy during operation. If the rise time of a propagating
`signal is less than the round trip propagation delay from
`Source to load, then the conducting medium will behave as
`a transmission line. In Such a connection, variations in the
`characteristic impedance Z, where: Z-(L/C)';
`L=conductor inductance; Ceconductor capacitance along
`the line will cause perturbations in the electromagnetic field
`associated with the propagating signal. These disturbances
`in the electromagnetic field result in reflections of portions
`of the signal energy at the points where the variation
`occurred. If the signal is not totally absorbed by the load at
`the end of the conductor length, due to unmatched imped
`ances or lack of proper line termination, the unabsorbed
`energy will be reflected back towards the source. These
`reflections give rise to radiated emissions. Proper termina
`tion and controlled impedance interconnections will reduce
`radiated noise significantly.
`The coupling of signal energy from an active signal net
`onto another signal net is referred to as crosstalk. Crosstalk
`is within-system EMI, as opposed to EMI from a distant
`source. Crosstalk is proportional to the length of the net
`parallelism and the characteristic impedance level, and
`inversely proportional to the spacing between signal nets.
`Proper interconnect layout design can reduce the incidence
`of crosstalk. Strong Sources of low impedance, H-field rich
`EMI are relatively high current and relatively low voltage
`components Such as power supply, solenoids, transformers,
`and motors. If the H-field possesses high intensity, the field
`can induce spurious current flow in other system compo
`
`2
`nents. Thus, noise radiated from within a system can inter
`fere with system performance by coupling with other system
`elements, not just adjacent conductor nets, as another form
`of within-system EMI.
`Electronic systems are becoming smaller, and the density
`of electrical components in these systems is increasing. As
`a result, the dimensions of the average circuit element is
`decreasing, favoring the radiation of higher and higher
`frequency signals. At the same time, the operating frequency
`of these electrical systems is increasing, further favoring the
`incidence of high frequency EMI. EMI can come from
`electrical systems distant from a sensitive receiving circuit,
`or the source of the noise can come from a circuit within the
`same system (crosstalk or near source radiated emission
`coupling). The additive effect of all these sources of noise is
`to degrade the performance, or to induce errors in sensitive
`systems. The prevalence of high frequency systems and
`portable electronics is creating a very complex spectral
`environment for the operation of sensitive electrical sys
`tems. The Federal Communications Commission (FCC) and
`the Federal Aviation Association (FAA) regulations on radi
`ated emissions are becoming increasingly difficult to meet
`without adding to system size, mass, or cost due to the need
`for the EMI shielding.
`EMI shielding has taken many forms. Sensitive or radi
`ating devices are often covered with a lid or enclosure which
`is connected to ground potential in the process of securing
`the cover in place. Shielding close to the source, where the
`field intensity is the highest, requires greater shield effi
`ciency to contain the field. Therefore, in many cases it has
`been more common to shield the sensitive, EMI receiving
`component. In some instances, entire circuit boards are
`covered with a grounded lid. Polymer thick film conductor
`materials, such as a screen-printable copper filled epoxy
`paste, are sometimes used to form a shield. In other known
`EMI shielding methods, individual ferrite components are
`often placed on device pins or in series with a circuit to
`attenuate unwanted noise which may be causing system
`errors, or acting as sources of radiated emissions. In another
`application of ferrite beads or elements, a ferrite component
`is used with a capacitor in order to form a low frequency
`inductance-capacitance (LC) band pass filter, effectively
`shorting unwanted signal frequency components to ground.
`Many enclosed systems powered by external alternating
`current wiring are also shielded from EMI by the incorpo
`ration of internal shields. As one example, a metal cabinet
`housing which encloses the system may be designed to
`function as a shield. However, metal housings are often too
`expensive or heavy for portable applications. To avoid some
`of the weight and expense, the inside of a molded plastic
`housing may be coated with a thin metal film. Sometimes
`metal-filled paints are applied to the housing. If cost permits,
`metal-filled plastic is sometimes used to form the housing.
`In most cases these different types of shields are connected
`to ground potential. Any break in the shield will form an
`aperture through which radiation will emit. Thus, great care
`is taken to use conductive gaskets to seal access areas. Also
`within housings, a conductive metal screening may be
`inserted in the airflow path of a fan-powered cooling system
`to help reduce radiated emissions from the cooling or
`exhaust port.
`A common feature of these and other prior art EMI
`shielding methods is that the prior art methods are attempt
`to shield EMI in a broad sense using an “all-encompassing”
`shield material. However, such broad approaches are not
`Sufficiently effective. As an example, a metal housing may
`shield the enclosed electronic devices from various E-field
`
`50
`
`55
`
`65
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 009
`
`

`

`5,639,989
`
`10
`
`15
`
`20
`
`25
`
`30
`
`3
`components of EMI, but not H-field components.
`Accordingly, an alternative shielding method which more
`specifically targets shielding of particular EMI frequency
`ranges would be welcome. Particularly, such a method
`should effectively prevent radiative emissions and protect
`system components from a selected spectrum of impinging
`electromagnetic radiation without significant additions to
`system size, mass, or cost. Fulfilling this need is a growing
`concern as portable electronics and communications become
`increasingly popular.
`SUMMARY OF THE INVENTION
`In one form of the invention, an electronic component
`assembly is shielded. The assembly has a wiring substrate
`having a plurality of conductive signal traces and at least one
`ground element formed on a surface thereof. The assembly
`also has a semiconductor device mounted to the wiring
`substrate and electrically coupled to the plurality of signal
`traces. A conformal insulating layer is formed over the
`semiconductor device and over the signal traces, such that a
`portion of the at least one ground element is uncovered by
`the insulating layer. A conformal shielding layer is deposited
`over the insulating layer and in contact with the uncovered
`portion of the at least one ground element. The shielding
`layer is a precursor material filled with a first and a second
`plurality of particles, wherein the first plurality of particles
`is selected to attenuate electromagnetic signal frequencies
`within a first range and the second plurality of particles is
`selected to attenuate electromagnetic signal frequencies
`within a second range of frequencies which is at least
`partially exclusive of the first range.
`BRIEF DESCRIPTION OF THE DRAWINGS
`FIG. 1 is a cross-sectional view of a portion of a shielded
`electronic assembly in accordance with the present
`invention, as viewed along the line 1-1 of FIG. 2.
`FIG. 2 is a top-down view of a larger portion of the
`electronic assembly of FIG. 1, withput shielding layers
`being illustrated.
`FIG. 3 is a cross-sectional view of a portion of another
`shielded electronic assembly in accordance with the present
`invention which includes multiple shielding layers, as would
`be viewed in a cross-section along the line 3-3 of FIG. 2.
`FIG. 4 is a cross-sectional view of conductive traces of a
`substrate shielded in accordance with the present invention.
`FIG. 5 is a cross-sectional view of yet another electronic
`assembly which includes a shielded housing in accordance
`with the present invention.
`FIG. 6 is a graph showing the insertion loss (attenuation)
`characteristics of a typical capacitive EMI filter material as
`a function of frequency and the capacitance of the filter
`component.
`FIG. 7 is a graph demonstrating the idealized impedance
`of a typical surface mount ferrite component, and also
`indicating the resistive and reactive impedance components.
`FIG. 8 is a graph illustrating the loss tangent as a function
`of frequency for a typical nickel-zinc, soft ferrite EMI
`55
`shielding component.
`FIG.9 graphically depicts the effect of temperature on the
`initial permeability of amaterial used in a typical softferrite
`component.
`FIG. 10 is graph showing the effect of frequency on the
`initial permeability and loss factor of a typical soft ferrite
`component.
`DETALED DESCRIPTION OF APREFERRED
`EMBODIMENT
`The present invention utilizes a particulate-filled polymer
`system as an EMI shield, wherein the particulates used as
`
`35
`
`45
`
`50
`
`65
`
`4
`fillers are chosen to attenuate a specific EMI frequency
`range or ranges, as determined by the needs of a given
`environment. The polymer system uses highly conductive
`metal, ferromagnetic conductive metal, insulative and lossy
`dielectrics, and/or ferromagnetic materials with low to high
`volume resistivities as fillers. Various filler types can be
`mixed with a polymer to form one general purpose EMI
`shield coating. Alternatively, multilayer coatings can be
`formed, wherein each layer may be formulated to contain
`specific mixtures of fillers to target specific E or H field
`frequencies for attenuation. The layers comprising the mul
`tilayer structure are applied in a sequence designed to
`protect the underlying sensitive circuit fromimpinging EMI,
`or to prevent EMI from being emitted from a noisy circuit.
`The two types of structures, general purpose or multilayered,
`may be formulated to attenuate a broad spectrum of EMI
`without knowledge as to what EMI frequencies may
`impinge the device or circuit or which may radiate from the
`noisy circuit component. Alternatively, the coatings can be
`specifically formulated to provide high attenuation of spe
`cific field types and frequencies which are being radiated or
`being received by a sensitive circuit. The prior art of bulky
`metal sheet shields and the inadequate shielding of thin film
`coated plastic system enclosures is replaced in the present
`invention by a highly efficient, low mass shield coating(s).
`Portable electronics will benefit from such EMI shielding,
`allowing lower mass and Smaller size. Direct application of
`such coatings over printed circuit boards (PCBs) populated
`with active devices will reduce crosstalk between long
`parallel conductors.
`In the general purpose structure, an emitting or sensitive
`receiving circuit is first coated with a thin layer of an
`unfilled, highly insulative, polymer. Then, a second coating
`layer is applied. This second layer is compounded to be a
`general purpose, broad spectrum shield material. The second
`layer should be highly electrically conductive in order to
`conduct away the current induced in the layer by the
`absorption of the fields. High conductivity (low impedance)
`also permits a high degree of reflection of the high imped
`ance E-field components of the impinging field due to the
`great difference in field and shield layer impedances. It
`should also contain dielectric materials which are highly
`lossy at E-field frequencies present in the EMI field. The
`second layer may also contain ferromagnetic materials
`which are lossy at H-field and E-field frequencies present in
`the EMI. The high permeability of the ferromagnetic fillers
`will couple the H-field components via the establishment of
`a level of magnetism, and conduct the field lines away from
`the device. Conductive ferromagnetic materials are also
`useful in a general purpose shield layer due to their effi
`ciencies in coupling H-field energy while aiding in the
`creation of a low resistance path for the conduction of
`induced currents out of the system to ground.
`In the multilayer embodiment, once again a highly insu
`lative polymer coating is first applied to an EMIemitting or
`EMI receiving circuit. Next, a series of individual coatings
`are applied, each chosen to shield a specific subset of
`interference frequencies. For example, a second layer may
`be formulated to be highly electrically conductive through
`the use of highly conductive filler particles. Highly conduc
`tive ferromagnetic powders can also be added to increase the
`permeability of this layer, thereby increasing shield absorp
`tion of impinging fields. A third layer could be designed to
`absorb E-field radiation by incorporating powders which are
`highly lossy at the E-field frequencies in the EMI field. A
`fourth layer could be formulated to attenuate H-fields fre
`quencies through use of ferromagnetic powders which are
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 010
`
`

`

`5,639,989
`
`10
`
`5
`
`35
`
`20
`
`S
`highly lossy at these frequencies. A fifth highly conductive
`layer, similar to the second layer, may be desired as the
`external layer in order to reflect the bulk of the impinging
`E-field energy, and to permit conduction of any induced
`currents. The number and sequence of these layers can be
`modified to best optimize the structure to provide either
`protection from external fields, or to prevent emissions from
`being radiated from the coated circuit, depending on a
`particular system's requirement for shielding.
`In both embodiments, general purpose and multilayered,
`ground planes, ground rings, or smaller ground connections
`may be dispersed about the surface or the periphery of the
`device to be shielded. In the case of shielding a printed
`circuit or wiring board (PCB or PWB), etched copper
`conductor features that are connected to ground may be
`present on the surface of the PCB. A first, highly insulative,
`polymer layer will not cover these ground potential ele
`ments. All of the layers applied subsequent to the pure
`polymer layer will have sufficient electrically conductive
`filler, along with the E-field and/or H-field absorbing fillers,
`so that all layers will be able to conduct induced currents to
`ground, either through a direct connection or indirectly
`through underlying layers. The ground connections used for
`sinking the EMI noise currents preferably are connected to
`System ground far away from power supply ground connec
`25
`tions to active devices. If this is not done properly, the EMI
`noise that is routed to ground, preventing system noise, may
`reintroduce noise into the system via the power supply
`system ground, significantly degrading net shield effective
`CSS
`30
`Examples of a general purpose embodiment and a mul
`tilayer embodiment of the present invention are illustrated in
`FIGS. 1-4. FIG. 1 is a cross-sectional view of a shielded
`electronic assembly 10 in accordance with the present
`invention. Electronic assembly 10 includes a semiconductor
`device 12. Device 12 has a semiconductor die 13 which is
`connected to an underlying wiring substrate 16 by a plurality
`of solder balls 14. Solder balls 14 may be formed on an
`active surface of die 13 using one of any known bumping
`procedures. The solder balls act as electrical input/output
`(I/O) connections to conductive traces or pads 18 of the
`substrate. Substrate 16 can be formed from a printed circuit
`board (PCB), a flexible circuit, a ceramic substrate, a thin
`film multichip module substrate, or similar substrate mate
`rial. Furthermore, substrate 16 may be a rigid or flexible
`support for electronic components. Substrate 16 also
`includes an EMI ground ring 19. FIG. 2, which is a top view
`of assembly 10, more clearly illustrates the ring shape of the
`EMI ground ring. FIG. 1's cross-sectional view is taken
`along the line 1-1 of FIG. 2. Substrate 16 also includes
`plated through-holes 20, also known as vias, for routing
`Selected traces, ground rings, pads, or the like, to internal
`conductive layers or planes. As illustrated, substrate 16
`includes several internal conductive layers, specifically an
`X-direction signal layer 22, a Y-direction signal layer 23, a
`power plane 25, and an EMI ground plane 27. Through
`holes 20 extend through substrate 16 to contact appropriate
`internal conductive layers. For instance, as illustrated in
`FIG. 1, EMI ground ring 19 is connected to the EMI ground
`layer 27 by through-hole 20. Conductive traces or pad 18 are
`connected to appropriate internal signal layers 22 and 23 or
`power plane 25.
`As illustrated in FIG. 1, Device 12 is attached or mounted
`to substrate 16 by a direct chip attach (DCA) method, such
`as those known in the art. DCA of die 13 to substrate 16
`includes the application of an underfill material 15 between
`the die and the substrate. Underfill material 15 is initially
`
`6
`dispensed about the perimeter of die 13. Surface tension and
`surface energy effects cause underfill material 15 to be
`drawn under the die 13, where all air is displaced and the
`material 15 wets all contacted surfaces. Subsequently, the
`assembly is processed to effect the cure of underfill material
`15, whereupon the material adhesively bonds die 13 to
`substrate 16. Underfill material 15 provides substantial
`mechanical reinforcement of die 13 to substrate 16
`assembly, reducing the stress transferred to solder balls 14
`during subsequent processing and the temperature cyclingto
`which the assembly is exposed during testing and product
`usage. Underfill material 15 can also provide for environ
`mental protection of the interconnect balls 14 and the
`circuitry on die 13. Particular materials suitable for use as an
`underfill material in accordance with embodiments of the
`present invention include those already commercially avail
`able for such purpose.
`In accordance with the present invention, after die 13 is
`properly attached to substrate 16 and the appropriate elec
`trical connections to the substrate are made, a conformal
`insulative layer or coating 24 is dispensed and cured over the
`die and other regions of substrate 16 which will be shielded
`to prevent emission or coupling of EMI. Coating 24 is
`applied directly on the die and portions of the substrate to
`prevent electrical short circuiting to a subsequently depos
`ited conductive layer(s). Thus, coating 24 should be a highly
`insulative material. A secondary purpose of layer 24 is to
`increase the reliability of device 12 by increasing resistance
`to environmental stresses. Coating 24 in a preferred embodi
`ment is a pure polymer having a low modulus of elasticity,
`Such as a silicone gel or elastomer, polyurethane, epoxy,
`polysiloxane, acrylic, and the like. Coating 24 could also be
`a filled polymer system where the particulate filler was a
`highly insulative material such as specific ferrite and/or
`ceramic dielectric material. Coating 24 may be applied by
`Syringe dispensing, spraying, dip-coating, curtain coating,
`Screen or stencil printing, or by any other appropriate means.
`In order to avoid coating circuit regions which are to be left
`uncoated, in other words those to which contact by the
`Subsequently deposited conductive layer is required, it may
`be necessary to remove the coating 24 from such regions
`through photolithographic, laser processing, mechanical
`ablation, or other appropriate means. For example, as illus
`trated in FIG. 1, ground ring 19 should be uncoated by
`insulative layer 24 so that contact between the ring and a
`Subsequently deposited shield layer 26 (described below)
`can be made.
`In a preferred embodiment for a portable electronics
`application, coating 24 will generally range in thickness
`from 1 to 250 pum; however, this range is not intended to
`limit the invention since other applications may require
`thinner or thicker coatings. If surface conductors, such as
`signal traces, are in direct contact with coating 24, the effect
`of added capacitance should be considered in the electrical
`design of the system since the characteristic impedance and
`capacitive line loading would be affected. These consider
`ations may demand that coating 24 be a pure polymer system
`(without fillers) which would addminimal capacitance to the
`exposed conductor lines. Furthermore, any fillers with are
`incorporated in coating 24 must be chosen so as not to
`attenuate the principal frequencies of the signals running
`through the system conductors, such as signals through
`conductive lines or pads 18.
`Next, a shielding layer 26 is deposited over insulative
`coating 24 of the assembly and in contact with EMI ground
`ring 19, as illustrated in FIG. 1. Shielding layer 26 is a
`particulate filled conformal coating. Like coating 24, shield
`
`45
`
`50
`
`55
`
`65
`
`Momentum Dynamics Corporation
`Exhibit 1023
`Page 011
`
`

`

`5,639,989
`
`10
`
`15
`
`30
`
`35
`
`45
`
`25
`
`7
`ing layer 26 may be applied by Syringe dispensing, spraying,
`dip-coating, curtain coating, screen or stencil printing, or
`any other appropriate means. Likewise, any portions of
`substrate 16 or device 12 which are to be left uncoated by
`shielding layer 26 may be processed so as not to coat
`undesired surfaces, or material 26 may be removed from the
`desired areas by photolithographic or laser processing,
`mechanical ablation, or equivalent methods. Having shield
`ing layer 26 deposited across the entire assembly should not
`pose problems if the appropriate underlying devices are
`protected by insulative coating 24. Generally, shielding layer
`26 will to deposited to a thickness of about 5-500 um,
`although other applications may dictate thinner or thicker
`coatings.
`In the embodiment illustrated in FIG. 1, shielding layer 26
`is a general purpose EMI shield material which incorporates
`a variety of filler materials for attenuating a broad range of
`frequencies which may impinge device 12. Preferably, the
`shielding layer is a polymer precursor material with filler
`materials selected to attenuate a specific frequency or range
`of frequencies estimated to be an EMI problem in a given
`20
`system and environment. One or more filler materials may
`be used. In using more than one type of filler material, the
`attenuation characteristics of the materials preferably attenu
`ate either exclusive or at least overlapping EMIfrequencies
`to enhance shield effectiveness. For instance, one filler
`material may be chosen to attenuate E-field frequencies and
`another filler material for H-field frequencies. The filler
`particle size distributions and morphologies can be specified
`to provide the desired particle packing and fill volume in the
`shield layer. Further, if the wavelengths of the impinging
`radiation are known, the particle sizes can be adjusted to
`cause scattering and diffraction of the impinging radiation.
`As a general purpose EMI shield, layer 26 should contain, at
`least, particulate fillers which impart high electrical
`conductivity, such as Cu, Ag, and alloys of these metals,
`among others, to allow induced currents to be conducted
`away from the circuit. As a general purpose shield, layer 26
`may include additional fillers of some or all of the following
`general types of particulate fillers: conductive and ferromag
`netic metals such as Ni, Co, Fe, Fe/Co alloys, Fe?Ni/Co
`alloys, and other alloys of these metals; lossy ferrite mate
`rials such as Niferrite or Mnferrite, or magnetic oxides; or
`dielectric materials such as barium titanate, strontium
`titanate, niobate materials, Zirconate materials, or magnetic
`oxides, among others, which are lossy at the frequencies of
`the EMI.
`To allow induced currents to be conducted away from the
`circuit. shielding layer 26 must be electrically connected to
`a reference potential, such as ground potential. As illustrated
`in FIG. 1, shielding layer 26 is in direct contact, and
`therefore electrically coupled to, EMI ground ring 19 of
`substrate 16. The ground ring is coupled to internal EMI
`ground plane 27 by one or more plated through-holes 20.
`Ground ring 19 need not be in the form of a ring surrounding
`the device, as will become apparent below in the discussion
`of FIG. 2. It is preferable that ground ring 19 is connected
`to an EMI ground plane 27 which is not used for general
`power supply distribution (e.g. ground) for device 12. This
`is to prevent power supply noise from coupling into the
`shield, and to prevent EMI-induced signals in the shield
`from coupling into the power supply system. The introduc
`tion of power supply noise into shielding layer 26 layer may
`cause EMI to radiate from the shield into die 13. Indepen
`dent grounding of shielding layer 26 will also help to prevent
`currents in shielding layer 26, induced by coupled EMI.
`from entering the device's power supply system where the
`EMI-induced current could add to system noise.
`
`50
`
`55
`
`65
`
`8
`FIG. 2 is a top-down view of a larger portion of electronic
`assembly 10, illustrating various electronic components and
`semiconductor devices which can be shielded in accordance
`with the present invention. As illustrated in FIG. 2, no
`shielding layers are present so that the devices and substrate
`features are apparent. FIG. 2 shows three types of integrated
`circuit devices which are commonly mounted on the surface
`of a substrate, such as substrate 16, one of which is device
`12 as described above in reference to FIG. 1. It is important
`to note, however, that FIG. 2 does not illustrate all possible
`types of electronic components w