Wireless communications is a rapidly growing segment of the communications industry, with the potential to provide high-speed high-quality information exchange between portable devices located anywhere in the world. Potential applications enabled by this technology include multimedia internet-enabled cell phones, smart homes, appliances, automated highway systems, distance learning, and autonomous sensor networks, just to name a few. Supporting these applications using wireless techniques poses significant technical challenge. As handsets move to meet broadband, the requirements of components are more astringent. Battery life has to be maximized, reception clarity in a multitude of environments has to be improved and at the same time the customers require a significant reduction in size. Although the industry has made significant strides in miniaturizing active high frequency components, advancements in packaging of high frequency devices and Surface Mount (SMT) passive components have lagged behind.
Mobile products are getting smaller and integration of components is required to meet this need. As more wireless communications products hit the market, the role of filters is becoming increasingly important. Enhancing the performance of passive high frequency filters along with active components while shrinking size and costs is a must. The industry managed to create design techniques that reduce the number of surface mount components by imbedding Rs, Cs, and Ls and today some of the passives are being integrated into semiconductor die and packaging. However, on-chip passives fabricated using IC semiconductor deposition processes do not deliver sufficient performance and pending on value, integrated on chip passives increase die size and thus cost. Likewise, ceramic technologies such as low temperature co-fired ceramic (LTCC), which uses multiple layers of thin ceramic material are nevertheless difficult to work with and have low high frequency and thermal performance. Although measurable progress has been made in embedding Passive components into polymer and epoxy base substrates large values of Inductor and Capacitance value continues to require large space in multi-layer circuits and interconnectivity between substrate layers increasing substrate complexity and cost.
Contemporary high frequency, RF, analog or mixed signal Packaging requires stringent electrical, mechanical, thermal and environmental performance. For instance, mechanical stress on components, heat dissipation of IC's, moisture resistivity, high frequency performance, they all have to be met simultaneously by the packaged module. In order to get quality high frequency performance of optimized discrete modules, passives must be scaled in size and integrated along active devices to fit the application. Passive components have lower parasitic inductance and capacitance, which is required for higher frequency operations. Today's high frequency modules operate in multiple frequency bands and encounter more interference in a crowded frequency spectrum than ever before. This, along with greater range and battery life demands, mandates superior efficiency of high frequency filters and matching networks. This heavy reliance on analog circuitry requires a large number of passive components to be used. Therefore, the industry solicits a robust single-package solution, where the integration of passives and actives alleviates and accommodates the massive volume of IC's and passive components needed for smaller and lighter products.
The foregoing summary, as well as the following detailed description of the technology, will be better understood when read in conjunction with the appended drawings. For the purpose of illustrating the technology, there are shown in the embodiments which are presently preferred. It should be understood, however, that the technology is not limited to the precise arrangements and instrumentalities shown. In the drawings:
Referring to
Most of the connections between the substrates found in related art technologies are accomplished using either solderballs 203 later reflowed to connect the two surfaces. Another method found in related art is the usage of solder columns 202, where cylinders of solder are disposed between the substrates and later reflowed. There are several disadvantages by using the aforementioned methods as interconnections between substrates. First, creep which is the term given to the material deformation that occurs as a result of long term exposure to levels of stress that are below the yield or ultimate strength. The rate of this damage is a function of the material properties of solder, the exposure time, exposure temperature and the applied load (stress). Creep is usually experienced in solder joints in all types of microelectronic packages when the devices is heated and cooled as a function of use or environmental temperature fluctuations. Such failures can be caused either by direct thermal loads or by electrical resistive loads, which in turn generate excessive localized thermal stresses. Depending on the magnitude of the applied stress and its duration, the deformation may become so large that a solderball 203 may experience brittle and/or ductile fracture, interfacial separation, fatigue crack initiation, propagation, creep, and creep rupture.
Similarly, a solder column 202 may break in a similar manner and no longer perform its function. For example, excessive elastic deformations in slender structures 202 in electronic packages due to overstress loads may sometimes constitute functional failure, such as excessive flexing of interconnection wires, package lids, or flex circuits in electronic devices, causing shorting and/or excessive crosstalk. Other methods used in the related art refer to using the wirebonds 207 as means to interconnect in between IC's. Although, wirebonds are less subjective to creep, long wirebonds cause detrimental parasitic inductance in high frequency circuits due to length of the wires with themselves and to adjacent substrates creating adverse impedance into the circuit. Yet another method found in related art comprise of stacking multiple IC's (back to back-on top of each other) and eliminating the middle substrates. This method reduces the overall height of the 3D stack significantly but has resulted in catastrophic failure due to self-heating. The failure occurs where the heat generating semiconductor junctions lie. By eliminating the substrates, the heat generated in the IC's has no place to travel and self-heating will limit the overall operation of the device.
The present technology eliminates the aforementioned challenges by using two complementary but disjunctive approaches in 3D packaging. The mechanical challenge is resolved by the creation of an interconnector 300.
The electrical challenge in miniaturized or compact 3D packages of high performance, high frequency circuits is low inductance and low parasitic capacitance. This is because the size necessary to withstand significant current and dissipate heat increases inductance and capacitance over smaller low-power active components. Addressing this challenge is relatively easy on a single surface-mount substrate, but it becomes a significant challenge when the device is packaged in a 3D stack. Component and circuit miniaturization also mandates tighter packing of interconnects potentially introducing new parasitic coupling and distributed-element effects into circuits. The inventive approach may comprise the use of passive components 211, chip capacitors, inductors or resistors, as the interconnecting means between the substrates 208, 210, 209.
Examples of substrates for the 3D stack may comprise, but are not limited to: copper clad laminates, thermally stable copper-clad epoxy-glass laminates (FR4), flexible (FLEX) polyimide based substrates such as KAPTON®-DuPont and acrylic base adhesive PARALUX®-DuPont. Said substrate-materials are selected from the group consisting of liquid crystal polymer (LCP), polyolefin, fluropolymers such as polytetrafluorethylene (PTFE), polyvinylidene fluoride (PVDF), polyester products, including terphthalate and polyeylene terephthalate (PET), thermosetting resin bonds are based on the melamine formaldehyde systems and phenol formaldehyde systems. Typical characteristics of these materials comprise: Loss Tangent=0.01 (good in HIGH FREQUENCY) (excellent in analog circuits up to 2.0 ghz and digital circuits above 3.0 ghz) (stable with temperature variations), Er=4.0 to 4.5 (at 1.0 mhz) (specific Er is dependent on glass-to-resin-Er constant from 1.0 mhz to 3.0 ghz), CTEr=+220 ppm per degrees C, (high). Tg is 250 degrees C., Electro-Deposited Copper, Layer-to-layer thickness control=±0.002. Similar types of substrates for radio frequency and microwave applications with PCBs may include, but are not limited to, Rogers DURIOD® microwave laminates and teflon materials from manufacturers such as Arlon and DuPont. A ceramic approach has also been investigated and although not the preferred mode may include, but is not limited to, Ceramics such as Alumina (Al2O3), Aluminum Nitride (AlN), glass-ceramic composites such as LTCC, and HTCC.
Passive components (resistors, capacitors, and inductors) acting as connectors from substrate to substrate in a 3D device also may include: filters, diplexers, baluns, resonators and couplers piezoelectrics, and coils. For the disclosed inventive 3D highfrequency packages, it is ideal that most of the interconnections be considered as passive components. Passive components have the necessary form factors and performance capabilities for high frequency. Flat frequency response over wideband widths, repeatable performance over time and temperature, good directivity and low insertion loss may be important requirements. Passive components 211 continue to be important elements in RF design, with the same evolving cost and performance demands as active devices. Rather than amplifying signals as active components do, passive components 211 change the characteristics of a signal, letting only the desired frequencies pass through. Passive component manufactures are also moving toward high-temperature plastics and other packaging materials that can withstand the processing temperatures associated with lead-free board assembly. In addition to reducing the physical dimensions, capacitors have had much recent development work in the area of dielectrics—ceramics, glass, porcelain, plastics, and even silicon IC-type construction. Further component development will seek to reduce device size by exploiting existing chemistries, such as electrolytics, film, and ceramics, while lowering equivalent series resistance (ESR) and improving reliability.
High frequency packaging or applied electromagnetic engineering, is the design of guided-wave structures such as waveguides and transmission lines, transitions between different types of transmission lines, and antennae all require control of the underlying electromagnetic fields. Power amplifier combiners and transmission line components are challenging when packaged under a single compact device. Fields are usually contained in the art by including metal planes 208 and metal vias 204 trough the structure as seen in
In wireless products, it is not unusual for engineers to go through two or three design revisions, therefore the faster they test components, make changes, the faster they can get to market. High frequency packaging is very dimension and structure specific. Any minor change in the physical design proportions and parameters allow for distortion in signals and input values not counted for in the models. Failures due to inadequate thermal design may be manifested as components running too hot or too cold. This will cause operational parameters to drift beyond specifications. Operation at higher frequencies requires predicting with accuracy the performance of the complete circuit. As expected, tradeoffs are required, especially at the smallest and largest sizes of devices. Rapid rework of high frequency circuits has been a major impediment for 3D structures, both in laboratory device and production test devices.
The disclosed technology allows for the rework process of components, for engineering and prototype, to be quickly removed from the top substrate without damaging the interconnections between the substrates 210, 208, 209 and without causing destructive damage. This may be achieved by “wetting” the bottom side of the top substrate 210 in
Thermal performance failures can arise due to incorrect design of thermal paths in an electronic assembly. This includes incorrect conductivity and surface emissivity of individual components as well as incorrect convective and conductive paths for heat transfer. Thermal overstress failures are a result of heating a component beyond critical temperatures such as the glass-transition temperature, melting point, fictive point, or flash point.
For power-handling components like resistors, size reduction amounts to raising the current-handling ability of a given device. That typically requires improvements in material composition and packaging to dissipate more heat in a small area—the same challenge faced by power semiconductors and high-speed processors IC's 206.
Recently, much work has been undertaken to properly characterize the effects of packaging. This concept is being extended to the layout surrounding the components on 3D Multi-chip modules (MCMs), layered board techniques and passives-in-package (PiP) devices. Adequate design checks require proper analysis for thermal stress, and should include conductive, convective, and radiative heat paths. Integrated modules have to be focused on manufacturability, to reduce cost while implementing the performance enhancements required for new high-speed/high-frequency applications. Two approaches are used to deal with the additional parasitic reactance of power devices—keeping the size as small as possible using materials with high thermal performance, and accurately modeling the component so compensation for its effects can be designed into other portions of the circuit.
Reliability specifications require testing of the products, including the expected accelerated life tests at elevated temperatures along with vibration and electrical stresses. High power applications require materials and testing to handle high voltages, high DC currents, high frequency currents, high temperatures and thermal cycling. Like the rest of high frequency electronics, there is a divergence in emphasis—small size and low cost vs. highest performance. In handsets impedance matching techniques require directional couplers to sample input or output signals and inject correction signals. Consistent performance at reasonable cost is the key.
Reliability experiments where performed using the disclosed technology in order to validate the models described in
Another reliability test was performed to determine the ability of component and solder paste interconnections to withstand mechanical stresses induced by alternating high and low temperature extremes. Permanent changes in electrical and/or physical characteristics can result from this mechanical stress. Cross section and SEM were added for check solder joint cracking. Test method—Machine: BAMCO #5,—Temperature: −40° C. to +100° C.—Transfer time: 5 minutes—Dwell Time: 15 minutes—Number of Cycle: 1000 cycles. The result was 100 (%) Yield.
It will be appreciated by those skilled in the art that changes could be made to the embodiments described above without departing from the broad inventive concept thereof. It is understood, therefore, that this technology is not limited to the particular embodiments disclosed, but it is intended to cover modifications within the spirit and scope of the present technology.
This application is a continuation-in-part of U.S. patent application Ser. No. TBD, Attorney Docket Number WJT08-0105/JSF01-0102, filed Feb. 14, 2006, entitled, WIRELESS RF CIRCUITRY OPTIMIZED FOR 3D PACKAGING TECHNOLOGIES, which claimed the benefit of priority under 35 U.S.C Section 119 from U.S. Provisional Application Ser. No. 60/653,162, filed Feb. 15, 2005, entitled, “Wireless RF Circuitry Optimized for 3D Packaging Technologies”.
Number | Date | Country | |
---|---|---|---|
60653162 | Feb 2005 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 11353950 | Feb 2006 | US |
Child | 11361513 | Feb 2006 | US |