As smartphones and tablets continue to drive the sales of MEMS devices for consumer electronics, most of the focus thus far has been on inertial sensors – MEMS accelerometers and gyroscopes. Additionally, MEMS based microphones and pressure sensors are seeing continually increasing adoption. RF MEMS technologies, on the other hand, have not been incorporated in any smartphone or tablet device in high volume until very recently. WiSpry, a leading RF MEMS component maker, is the first company ever to ship RF MEMS in volume for a smartphone OEM – the company’s first customer is Samsung and shipments began in September 2011. We recently spoke with Jeff Hilbert, WiSpry’s President and Founder, about the ongoing RF MEMS developments and opportunities for smartphone applications. In this detailed interview, Mr. Hilbert discusses the key business drivers for RF MEMS use in consumer electronics, current “pain points” for smartphone OEMs, most viable insertion points, competing technology alternatives, bill-of-material impacts, as well as the competitive landscape.
MEMS Journal: Are you shipping in volume now? How many units and for what type of a client?
Jeff Hilbert: WiSpry began shipping volume production in early September of last year. As disclosed by third-party reports, our initial customer was Samsung and the first phones incorporating our product continue to be on sale at multiple carriers. We are continuing to ship on a regular basis and are well on our way towards the 1M units shipped mark.
MEMS Journal: For this initial volume order, what is the insertion point in the overall system and what is primary function of the WiSpry device in this specific case?
Jeff Hilbert: The insertion point is in the RF front-end of the handset immediately adjacent to the antenna connection (highlighted box in diagram below). The product is a tunable impedance matching circuit that dynamically re-matches the power amplifier output to the antenna input under software control during phone operation. The product is a single-chip integrated RF-CMOS MEMS solution that provides multi-band matching capabilities across the entirety of the Smith Chart. Improvements of up to 3dB per frequency band are realizable. The solution is implemented as an application-specific standard product. As such, customization to different handsets and antennas is handled entirely in software.
MEMS Journal: How does an improvement of 3dB per frequency band translate to dollars? What is the business case and what are the benefits for the key players (e.g. carrier, handset maker, consumer, etc)?
Jeff Hilbert: For carriers, a 3dB improvement in the over the air (transmit and receive) performance of all of the phones on their network translates into an opportunity for huge savings in capital expenditures. It has been estimated that 1 dB less of RF performance can drive a requirement for 14% more cell sites to provide the same level of coverage which translates into hundreds of millions of dollars of capital expenditures. The same level of RF performance improvement provides handset OEM’s with potential benefits and savings in design complexity, time to market, and bill of materials costs. Consumers benefit through higher data rates, fewer dropped calls, longer battery life and potentially thinner and lighter handsets. Higher quality of service and increased consumer satisfaction lead to customer retention and overall market and market share growth for the handset OEMs and carriers.
The insertion point for WiSpry's device (shown in yellow) in the RF front-end of the handset is immediately adjacent to the antenna.
MEMS Journal: Again, for this insertion point, what is the "pain" that you are solving for the customer? In other words, why do they need your device?
Jeff Hilbert: We are addressing several rapidly growing pain points for our clients. The first is enablement of the large touch screen, thin form factor smart phones that are in high demand from consumers. These form factors result in very little internal volume being available for the antenna. Physically smaller antennas are being asked to perform (much) better in the face of: (1) coverage of a rapidly growing number of frequency bands, (2) more stringent signaling requirements supporting higher data rates, and (3) higher transmit and receive operating requirements being imposed by network operators.
By adding dynamic impedance tuning, we allow the antenna design to focus on efficiency while we take care of the matching with the end result of a better performing antenna that fits in the internal volume available. We are also able to provide reductions in time to market and the capability to ultimately standardize on a single bill-of-materials for the various models of the same phone deployed around the world.
MEMS Journal: Is this a replacement for another component, or a new component being introduced to the system?
Jeff Hilbert: The WiSpry product is a new function being introduced into the handset. It is an enablement and not a replacement for an existing component.
MEMS Journal: What the key economic drivers for this insertion point?
Jeff Hilbert: The key economic drivers are consumer satisfaction – availability of desired smart phone form factors; improved quality of service (fewer dropped calls) and improved power efficiency (which can take the form of better battery life). For the handset OEM, drivers include the ability to sell phones to network operators at the desired price, time to market advantage, and reductions in complexity, inventory, and cost.
MEMS Journal: For this specific insertion point, are there are other (MEMS or non-MEMS) alternative technologies and, if so, what are they? What are the pros and cons of each?
Jeff Hilbert: There are other alternative technologies being used to address the same insertion point. For lower performance applications where a limited amount of tuning or frequency band coverage is the objective, semiconductor switches are being used to switch antenna feed points, switch between antennas, or switch banks of fixed capacitors. These approaches are relatively simpler to integrate and can be provided at a lower cost but also result in higher losses and provide 50% or less bandwidth or coverage, resulting in an overall lower performance solution.
Another technology being deployed is based on ferroelectric materials that are being used to implement a more traditional analog varactor function based on a voltage tunable dielectric film. Compared to MEMS, existing implementations offer lower tuning ranges, less linearity, lower Q values, and require a larger footprint, in part due to the need to implement a control loop for the varactor. Because of these limitations, current film-based approaches require significant hardware customization for each phone / antenna but once completed, adoption may be more straightforward for the client than it is for a more flexible, higher performance solution like ours. At present, there is no other production-ready MEMS-based solution available in the market that we are aware of.
MEMS Journal: What is the main drawback of the MEMS based approach?
Jeff Hilbert: MEMS will generally face more challenges than the other technologies in areas such as speed and reliability. However, with careful design and a thorough understanding of the end customer requirements, these need not be limiting factors to adoption and success. The performance, flexibility, and capabilities enabled by MEMS technology more than compensate for these design and development challenges.
MEMS Journal: How big is the market for this insertion point and what is your estimate based on?
Jeff Hilbert: The initial target market for this insertion point is the smart phone segment of the mobile handset market. In 2012 the target market is estimated to be on the order of 700M to 900M units growing to approximately 1.8B units annually by 2015, and to approximately 2.8B units annually by 2020. Depending on the level of performance, this equates to a total market opportunity in the range of $350M to $1.35B this year for this insertion point. Additional opportunities exist in other types of wireless devices including tablets, readers, laptops, and data dongles. These estimates are based on market data for the mobile handset market from numerous analysts and existing price points in the market.
MEMS Journal: Which other insertion points are you pursuing next? What are the main business drivers for each?
Jeff Hilbert: The next insertion point will be RF filtering. We introduced our first, and as far as we are aware, the industry’s first and only tunable RF filter product in February. Going forward we will bring a growing family of tunable RF filtering solutions to the market. The main economic drivers are similar to those for the tunable impedance matching product – reduction in BOM cost and complexity, space, size and power constraints, reductions in product development cycle times, and improved quality of service as well as consumer satisfaction.
MEMS Journal: Which companies are the main players in the RF filtering space? For each, what is their core technology?
Jeff Hilbert: There are a large number of main players in the RF filtering space (as it applies to cellular). Companies include Murata, Skyworks, RFMD, TDK-EPCOS, Triquint, Avago, Renesas, Hitachi Metals, and Sony to name a few. There are a diversity of filtering requirements in the RF front end and different players are covering different subsets of these requirements often driven by their own or their customers’ integration strategy and architectural decisions.
A diversity of technologies are employed to implement these filters including various semiconductor technologies, ceramics, and acoustics. In particular, acoustics based techniques are used to implement filters such as SAWs, BAWs, and FBARs which provide high performance, band-wide filtering functions. However, to achieve the required performance, these devices are frequency/ function/ application specific and are not adjustable. As the complexity of the RF front-end of the handset expands with more and more frequencies and modes of operation, the number of such discrete filters required increases rapidly. RF-MEMS provides the capability to make such filters software programmable, or tunable, allowing for a high performance solution that is realizable within the cost, power and form factor constraints that are increasingly dominating future smartphone designs.
MEMS Journal: Aside from WiSpry, which RF MEMS companies do you find the most promising?
Jeff Hilbert: We really don’t see any other RF- MEMS companies in the consumer electronics market today. Some people consider devices such as BAWs and FBARs as MEMS but we do not include them our definition, which is centered on micro-scale devices exhibiting mechanical motion.
We are well aware of companies such as TDK-EPCOS and Cavendish Kinetics who are reported to be developing MEMS based products or who have MEMS technology in-house. Everybody’s technology has some potentially attractive features on paper but the real challenge is converting such features into solutions that customers need and will buy.
To really understand the end applications and market requirements, you have to be in the market to gain access to the customers and build the expertise required to field winning solutions. From this perspective we believe we have a significant and sustainable lead on any potential competitors we are aware of. Undoubtedly, there are new RF MEMS companies that will emerge and the competitive landscape will evolve.
*********************************************
This article is a part of MEMS Journal's ongoing market research project in the area of RF MEMS. If you would like to receive our comprehensive market research report on this topic, please contact Dr. Mike Pinelis at [email protected] for more information about rates and report contents.
Copyright 2012 MEMS Journal, Inc.
Comments