摘要:
A wireless IC interconnect system and a source synchronous CDMA (SS-CDMA) bus interface facilitate interconnections between first and second IC locations. A signal conveyed using the wireless system is modulated and capacitively coupled to a transmission medium, and then capacitively coupled from the medium to a receiver which demodulates the modulated signal and provides the demodulated signal to the second IC location. Multiple signals can be conveyed simultaneously by modulating and demodulating them using multiple access algorithms such as CDMA and/or FDMA. The SS-CDMA bus interface utilizes source synchronous signaling and CDMA techniques to provide high bus concurrency and low channel latency. The interface is re-configurable, and provides multi-chip access in high-bandwidth multi-drop parallel interconnection applications. The interface employs spread spectrum multiple access schemes, which enable two or more data bits to be sent through the same channel simultaneously and successfully recovered at the receiver.
摘要:
The passive components of a transceiver, such as transmit/receive switches, antennas, inductors, capacitors and resonators, are integrated together on a common substrate to form an integrated passive transceiver section, which, in combination with other components, provides a highly reliable, low-cost, high-performance transceiver. Micro-electromechanical (MEM) device fabrication techniques are used to provide low-loss, high-performance switches and low-loss, high-Q reactive components, and enable the passive transceiver section's high level of integration. The passive components are preferably integrated on a low-cost glass substrate, with transceiver circuits containing active components fabricated on a separate substrate; the separate substrates are interconnected to implement the RF/analog and analog/digital interface portions of a transceiver. Additional MEM switching devices permit multiple, parallel signal paths to be switched in and out of the transceiver circuitry as needed to optimize performance.
摘要:
A high Q MEMS capacitor that can be continuously tuned with a large tuning ratio or reversibly trimmed using an electrostatic force. The tunable capacitor has a master/slave structure in which a control voltage is applied to the master (control) capacitor to set the capacitance of the slave (signal) capacitor to which an RF signal is applied via a suspended mechanical coupler. The master-slave structure reduces tuning error by reducing the signal capacitor's surface area and increasing its spring constant, and may eliminate the need for discrete blocking inductors by electrically isolating the control and signal capacitors. The trimmable capacitor provides an electrostatic actuator that selectively engages a stopper with teeth on a tunable capacitor structure to fix the trimmed capacitance.
摘要:
Power amplifiers having reactive networks (such as classes C, C-E, E and F) employ tunable reactive devices in their reactive networks, with the reactive devices respective reactance values capable of being adjusted by means of respective control signals. The tunable reactive devices are made from micro-electromechanical (MEM) devices capable of being integrated with the control circuitry needed to produce the control signals and other amplifier components on a common substrate. The reactive components have high Q values across their adjustment range, enabling the amplifier to produce an output with a low harmonic content over a wide range of input signal frequencies, and a frequency agile, high quality output. The invention can be realized on a number of foundry technologies.
摘要:
An integrated, variable gain microwave frequency power amplifier comprises a number of individual amplifier stages which contain microwave frequency active devices. Each stage is fed with a common input signal, and the individual stage outputs are connected to respective micro-electromechanical (MEM) switches which, when closed, connect the individual outputs together to form the power amplifier's output. The power amplifier's gain is determined by the number of outputs connected together. The preferred switch provides low insertion loss and excellent electrical isolation, enabling a number of amplifier stages to be efficiently interconnected to provide a wide dynamic range power amplifier. The switches are preferably integrated on a common substrate with the active devices, eliminating the need for wire bonds and reducing parasitic capacitances. A variable impedance network comprising a number of impedance matching networks selected using MEM switches can be connected to and integrated with a variable gain power amplifier to provide impedance matching that is appropriate for each of the power amplifier's possible output power levels.
摘要:
A method is described for transferring a thin film of arbitrarily large area from an original substrate to an alternate substrate. An etch stop layer is provided below an epitaxial layer, for example, grown on a semiconductor substrate. In a single transfer process, the epitaxial layer is bonded to a rigid host substrate having desirable thermal, electromagnetic, and/or mechanical properties. The original growth substrate is then removed from the transferred epitaxial layer using the etch stop layer. In a double transfer process, the epitaxial layer is first bonded to a rigid and porous temporary substrate using a thermally or chemically releasable resin, for example. The original growth substrate is removed using the etch stop layer so that the original substrate side of the epitaxial layer can be bonded to a rigid host substrate, as described above. The temporary substrate is then removed using the releasable resin to leave the transferred thin film attached to the host substrate.
摘要:
A method for planarizing surfaces in multi-layered semiconductor structures using elevated features in the form of semiconductor materials, such as for forming heterojunctions, or interconnection metal. A process of forming the features includes leaving residual photoresist on the features. After feature formation and definition of transistor or other structure locations, dielectric material is deposited across the structure. Remaining photoresist is subsequently removed along with dielectric deposited thereon leaving dielectric between the features. A layer of polyimide is spun on the structure and into depressions between the dielectric and features. Typically material deposition, etching, dielectric backfilling and spin-coating steps are repeated until a predetermined number of contact or conductivity regions or interconnection metal layers are formed in the desired multi-layered structure. In addition, intermediate etching steps may be employed for defining one or more transistor base or collector locations and metal or alloys deposited therein. Height variations in the resulting planar surface are controllable to within a fraction of a micron or less.
摘要:
A dual lift-off technique is used to provide self-alignment of the emitter area, the emitter contact, and the base contact of a heterojunction, bipolar transistor. A photoresist pattern which defines an emitter adjacent a base contact is formed on a suitable heterojunction bipolar semiconductor wafer. A base contact is formed by etching through the first semiconductor to the heterojunction and depositing metal on the second semiconductor. Dielectric is then deposited on the base contact. The photoresist is then lifted off with its dual covering of dielectric and metal. The emitter contact metal can then be deposited without requiring critical alignment because the base contact is covered with dielectric.
摘要:
An active inductor circuit includes a primary and a secondary coil and a drive circuit monolithically integrated on a common substrate to provide high-Q inductors. Each inductor circuit comprises a primary coil which carries a first current that varies with an RF input signal, and a secondary coil which carries a second current that varies with the RF input; an on-chip current source provides the second current. The inductor circuit is arranged such that there is a fixed phase difference of approximately 90° between the first and second currents, and such that the magnetic field induced by the second current compensates for energy that would otherwise be dissipated by the primary coil. When the second current is properly selected, the inductor circuit's input impedance is made purely imaginary, such that the circuit emulates an ideal inductor at a particular frequency.
摘要:
An integrated, tunable inductance network features a number of fixed inductors fabricated on a common substrate along with a switching network made up of a number of micro-electromechanical (MEM) switches. The switches selectably interconnect the inductors to form an inductance network having a particular inductance value, which can be set with a high degree of precision when the inductors are configured appropriately. The preferred MEM switches introduce a very small amount of resistance, and the inductance network can thus have a high Q. The MEM switches and inductors can be integrated using common processing steps, reducing parasitic capacitance problems associated with wire bonds and prior art switches, increasing reliability, and reducing the space, weight and power requirements of prior art designs. The precisely tunable high-Q inductance network has wide applicability, such as in a resonant circuit which provides a narrow bandwidth frequency response which peaks at a specific predetermined frequency, making possible a highly selective performance low noise amplifier (LNA), or in an oscillator circuit so that a precise frequency of oscillation can be generated and changed as needed.