UltraCMOS and DuNE: Poised for Success
Microwave Journal talks with Rodd Novak, Vice President of Marketing at Peregrine Semiconductor. Novak is responsible for the design and execution of Peregrine's global product marketing strategy and directs the company's strategic business development activities. Since joining Peregrine in 2002, Novak has contributed to the company's focus on the RF wireless and broadband markets. Prior to joining Peregrine Semiconductor, he served as corporate marketing manager for CTS Corp. and as an electrical engineer at Westinghouse Corp., and later moved to a marketing role with the commercial business unit of its subsidiary, Xetron Corp. He holds a BSEE degree from the University of Cincinnati and and MBA degree from Xavier University
MWJ: Peregrine is a San Diego-based company with nearly 265 employees worldwide. How did the company get its start?
RN: In 1979, inventors associated with Hewlett-Packard and CalTech University were awarded a foundational patent on their method for producing a low-defect layout of silicon-on-sapphire (SOS) wafer. However there existed known technical flaws which HP could not resolve. Further research on honing the method, and its significance on the RF industry, were shelved.
Simultaneously, under the US Government’s technology transfer policy, government entities retained the right to pursue further research work on any patent for government use. As a government employee and head of R&D at a US Navy research laboratory, NELC, Dr. Ronald Reedy and his team resurrected the patent and began research on the recipe details of the SOS process. Within 10 years, Dr. Reedy was certain they had resolved the technical flaws, and an idea for a commercial venture was born.
During this time, the partners acquired the foundational patents from Hewlett-Packard and a new company took flight. Peregrine Semiconductor Corporation was opened for business with foundry support from market leaders such as Intel, Xilinx, TRW, IBM and Union Carbide. The initial SOS device – the world’s first 1 GHz CMOS RFIC – was delivered in 1993.
Throughout the late 1990s, Peregrine Semiconductor honed its processes and identified key attributes of the SOS process – today known as UltraCMOS™ Technology – which catapulted performance of its RF IC devices. No longer would design engineers need to rely on cumbersome pin-diode solutions or temperamental GaAs ICs. Competitive SOI solutions made progress, however could never quite reach the performance levels demonstrated by the UltraCMOS process. Early adopters of the technology included space and defense engineers for satellite applications. Today, high-volume commercial UltraCMOS wafers are processed in multiple CMOS foundry facilities around the world, and more than 300 million UltraCMOS RF ICs have shipped since the company’s inception. The Company’s product portfolio includes the highest performance monolithic RF ICs on the market today and its customer roster includes many of the world’s largest RF module and wireless application OEMs. As further technological advancements are made, lower component costs and greater integration of RF signal chain will occur.
MWJ: Peregrine’s web site has a lot of content on your patented UltraCMOS technology, which is a Silicon-on-Sapphire (SoS) process. SOS devices have drawn a lot of attention because of their high-speed and low power consumption yet initially had some trouble reaching its commercial potential because of the difficulty in forming the silicon (Si) layer with sufficiently-low crystallization defects due to the unconformity between the crystal lattice of the sapphire (Al2O3), which is the supporting substrate of the SOS wafer and the Si crystal lattice. Could you describe the process technology Peregrine has developed to address this challenge?
RN: One of Peregrine’s core technologies is its ability to create a high-quality thin Si layer atop the sapphire substrate. Having been pursued for 40 years without success by others, Peregrine’s solid-phase epitaxial re-growth process creates a defect-free lattice in which high performance transistors can be manufactured with traditional CMOS high yields.
MWJ: So the result is a very thin Si film with few crystalline defects. I understand that the SOS devices are then fabricated by forming the metal-oxide semiconductor field-effect transistors on the Si film on top of the sapphire substrate. What are the big performance advantages over bulk silicon devices at this point? How do die-size and integration contribute to the advantages?
RN: The most important advantage of UltraCMOS over bulk silicon technology is its insulating substrate. The insulating substrate virtually eliminates the parasitic drain capacitance that is present in bulk silicon. This does several important things. The simple thing is that it leads to dramatic improvement in transistor performance because this capacitor does not need to be charged and discharged on every cycle. The insulating substrate also provides better isolation between circuit elements. More subtly, because the drain capacitor in bulk silicon is a depletion capacitance between two semiconductor layers it behaves in a non-linear manner. Eliminating this non-linearity enables UltraCMOS to have significantly better harmonic performance than bulk silicon.
UltraCMOS benefits from its similarity to bulk silicon by its use of the same basic CMOS technology and the fundamental benefits of standard CMOS processing; that is: high manufacturing yields, low power consumption but high power operation and high levels of integration. UltraCMOS benefits from all the process refinements that have been developed for CMOS on bulk silicon, and from the existing manufacturing infrastructure that was assembled around silicon.
CMOS has long been regarded as the technology of choice for integration, particularly in high-volume applications where cost is a major driver. During the last few years, advances in CMOS technology have led to its use in analog devices, and then into the intermediate frequency (IF) and radio frequency (RF) domains that were once dominated by BiCMOS and GaAs. Peregrine’s UltraCMOS combines the benefits associated with both bulk CMOS and the exotic technologies, and is bringing “Moore’s Law” to the RF signal chain.
MWJ: Which application areas do they compete in most strongly and why?
A: These significant performance advantages exist over competing processes such as GaAs, SiGe, BiCMOS and bulk silicon CMOS in applications where RF performance, low power and integration are paramount. Additionally, because UltraCMOS devices are fabricated in standard high-volume CMOS foundries, products benefit from the fundamental reliability, cost effectiveness, high yields, scalability and integration of CMOS, while achieving the peak performance levels historically expected from SiGe and GaAs.
MWJ: Could you talk a little about the needs of the multi-band handset market, what it means for switch technology and why you believe UltraCMOS offers a superior solution to GaAs?
RN: As the industry drives forward with multi-band applications, these performance benefits of UltraCMOS SOS have come not a moment too soon.
The bottom line is that the RF signal path in a mobile handset has become very crowded. The complexity of cellular phones has grown at a rapid pace, moving from dual-band, to tri-band, and now quad-band. In addition, these phones also need to handle a variety of signals for peripheral radios, such as Bluetooth, WiFi, and GPS. This trend is expected to continue as WiMAX and LTE (4G) capabilities are added. In a mobile handset, the antenna switch is the gatekeeper that controls antenna access for all of the radio signals. Currently, designers are specifying new single-pole, nine-throw (SP9T) switches, and we can reasonably expect that SP10T is not far away.
MWJ: Peregrine has brought to market a wide variety of RF switches, in addition to the high-throw-count switches, with significant performance advancements. What is driving these new products?
RN: Since the mid-1990s, integration, standardization and modularization have provided the GSM handset industry a roadmap of reduced size, increased performance and reduced cost. Competing solutions revealed strengths in the different process technologies: GaAs and pin diodes emerged dominant in the front-end module. However, the advent of the WCDMA platform introduced complexities that stumped module and IC manufacturers alike. Competing for the same slots are various options based on multiple technologies – forced together and promoted as “integrated,“ these solutions introduce “stacked margin” costs and invite second sourcing dilemmas. Further, technical issues surrounding Intermodulation Distortion (IMD) and the antenna switching function leave handset manufacturers continuing their search for an elegant result that meets 3GPP specifications. That search has ended. Peregrine’s UltraCMOS process provides for monolithic integration, and its portfolio of multi-throw antenna switches not only delivers the RF performance of the incumbent technologies, but also enables long-term roadmaps for the design of multi-band, multi-platform mobile communications. As technical advancements such as Peregrine’s HaRP enhancements are made, lower component costs and greater integration of the front-end will occur.
MWJ: These high-throw, high-linearity, low-loss switches operate between the PA or FEM and the antenna. Do you work closely with PA, FEM and/or antenna vendors to enhance switch performance? With the handset developer?
RN: Yes. Peregrine has been engaged with nearly all of the industry’s most influential vendors and end-use manufacturers to ensure that our roadmaps are aligned with the needs of the marketplace.
MWJ: What is HaRP™ technology? How does it impact device linearity and why is this so important for multi-band handsets?
RN: The HaRP technology inventions are patented process and design advancements which dramatically improve harmonic results, linearity and overall RF performance.
These HaRP enhancements are applied to Peregrine’s high power RF switch product line which enabled a long awaited breakthrough in Intermodulation Distortion (IMD) required 3GPP standards body for GSM/WCDMA applications.
The next generation of cellular phones will integrate multi-band and multi-mode systems in the same handset. The WCDMA system standard places high linearity requirements for the RF front-end because some or all WCDMA bands must be routed through a specially designed multi-mode antenna switch. The antenna switch, which is usually connected directly to the antenna port without any filtering, must be linear enough to cope with any unwanted outside signals introduced to the antenna port without degradation of the mobile phone receiver performance. As the current trend is towards for example 4xGSM (850,900,1800,1900) and 3xWCDMA (850, 1900, 2100) front-ends, it becomes necessary to route some or all WCDMA bands through a specially designed multi-mode antenna switch. Peregrine’s products provide for an ever-increasing number of RF paths to connect to the antenna through a single CMOS device.
Comparatively, GaAs-based multi-mode RF switches tend to experience increased insertion loss and die size with linearity improvements. This is because the conventional circuits need multi-stacked FETs or multi-gate FETs and large gate width to achieve low distortion, which results in large parasitic off-capacitances with degraded insertion loss. Today, no GaAs-based switch can achieve the IP3, harmonics and ease of use of Peregrine’s UltraCMOS HaRP-enhanced switches.
MWJ: In addition to RF and Cellular switches, Peregrine’s products include RF digital step attenuators, PLL frequency synthesizers, and pre-scalers for the wireless RF, broadband, space, defense and avionics markets. Do you see UltraCMOS being applied to new types of devices?
RN: Because of the highly flexible nature of UltraCMOS and the exceptional RF performance it delivers, we believe the product portfolio is virtually unlimited.
MWJ: I notice you have recently added a number of new 7-bit DSAs to the portfolio. Which applications do RFIC digital step attenuators (DSA) primarily target and what specifications are driving these new products?
RN: Peregrine's highly linear, monolithic RF DSA family is ideal for applications such as wireless base station, fixed wireless, microwave, VSAT, broadband telecom, military, space, test instrumentation, optical equipment and any RF system requiring a wide selection of resolutions, minimum attenuation step sizes and termination impedances. UltraCMOS™ DSAs operate on a single 3-volt supply and incorporate a unique initial attenuation state at power-up, industry-leading linearity over frequency and temperature, a versatile parallel or serial CMOS control interface, and a standard programming sequence and footprint across the family.
MWJ: Integrating the power amp onto Silicon seems to be the industry’s Holy Grail. Is this an area that Peregrine might be able to address in the future?
RN: UltraCMOS technology has demonstrated the ability to support both high voltage and high power applications with its switch products. Utilizing UltraCMOS’s unique ability to stack transistors, power amplifier appears to be within the reach of UltraCMOS.
MWJ: Let’s also spend a few minutes on the most recent introduction: DuNE™ Technology and Peregrine’s new Digitally Tunable Capacitors. Explain the market demand for these products. Which mobile applications require antenna tuning and why?
RN: Because of the increase in new features, functionality and industrial design requirements, the space available for the mobile system antenna is shrinking at a rapid rate. As antennas are wrapped and repathed, they lose efficiency. Some of this lost performance can be recovered with antenna tuning, in which the system uses dynamic impedance tuning techniques to optimize the antenna performance for both the frequency of operation and the environmental conditions.
One of the most significant challenges facing the mobile handset designers is the poor antenna performance for multi-band multi-mode handsets. Dynamically tuning the antenna to compensate for the increasing bandwidth requirements and environmental effects will significantly improve the antenna performance. Until now, no tunable element met the needs of the mobile products industry in power handling, reliability, high volume production and integration.
Further, as the market demands new wideband services in the handset, such DVB-H and ISDB-T for mobile TV, the use of antenna tuning becomes a necessity. Achieving broadband reception is directly related to the physical size of the antenna and the handset; any antenna mismatch degrades range and reception quality. To date, designers have typically had only two antenna choices: a passive internal antenna with poor performance, or the external whip antenna. The embedded, internal antenna is demanded by the market, yet poor performance is not accepted. As a result, interest in “tunable” internal antennas – those which cover a narrow section of the bandwidth and can be “retuned” as the receive channel changes – is increasing.
MWJ: How have mobile application designers dealt with the need for antenna tuning in the past? What makes the invention of DuNE DTCs so significant?
RN: For cellular antenna tuning, several first generation open-loop antenna tuning systems are being used in handsets, and development teams are beginning to design adaptive closed-loop antenna tuning systems. Micro-electromechanical systems (MEMS) and ferroelectric materials technologies (such as BST) have been used to implement tunable antennas and filters, however they are not yet proven for high-volume production and both typically require a high bias voltage (up to 30V or higher) to tune, requiring a separate CMOS charge pump and controller chip.
Mobile TV applications such as DVB-H and ISDB-T also require very broadband reception and delivering good coverage to an already complex system, all in a small space. When the DVB-H antenna is embedded in the GSM handset, the tunable element for DVB-H antenna tuning needs to be able to handle high levels of coupled GSM power without shift in capacitance or causing nonlinearities. Typically better than +26dBm of linear power handling is required to meet GSM/DVB-H interoperability requirements. The existing design options for mobile TV antenna tuning (such as varactor diodes or bulk-CMOS switched capacitor banks) do not meet these power handling requirements. Using a high-linearity switch (e.g. SP4T) to implement switchable tuning network can meet interoperability requirements, but results in very rough tuning.
The first devices in the DuNE DTC portfolio are designed to meet the stringent broadband requirements for DVB-H and ISDB-T mobile TV; multi-mode, multi-band GSM/WCDMA cellular handsets; and the power handling requirements for interoperability between the two applications.