Jeff Hilbert Speaks About RF MEMS

WiSpry is a fabless, radio frequency (RF) semiconductor company headquartered in Irvine, California.  The company was formed in October 2002 through a spin-off of the RF/Wireless business unit of Coventor, Inc. and recently secured $7 Million Series A funding to drive growth of RF tunable filter and RF switch products for the cellular phone handset market. 

We spoke with Jeff Hilbert, President and CEO of Wispry and a co-founder of the company.  Jeff has over 25 years of executive management and technical experience in a number of leading semiconductor and MEMS companies including LSI Logic, Motorola, and Coventor.   

MEMS Investor Journal: RF MEMS are seen as potential replacements for some of the communications circuit components.  What are the main advantages of RF MEMS over the currently used components?

Jeff Hilbert: Compared to other technologies, components based on RF-MEMS provide superior RF performance and tunability, and do so over a much broader range of operating frequencies.  For example, an RF-MEMS switch provides significantly improved insertion loss, isolation and linearity, concurrently.  There is no need to trade off one parameter against another.  Wispry’s RF-MEMS products utilize electrostatic actuation providing additional (to the enhanced RF performance) power efficiency.  Die size is very competitive with that from other technologies and the products can cost less.  Further, through careful choice of materials and processing steps, it is possible to utilize RF-MEMS to monolithically integrate many of the current discrete front-end RF components that to date have stubbornly resisted integration.

MIJ: What are the main challenges in developing RF MEMS components?

JH: The major challenges are designing for manufacturability (DFM), packaging and reliability.   In general, RF-MEMS makes extensive use of existing semiconductor processing steps and materials but it does so in some unconventional ways.  Typically, RF-MEMS devices are much larger than their electrical counterparts, make extensive use of the Z-axis (they’re tall and the materials are thick), and depend upon material properties that are often “don’t cares” for transistors.  DFM involves developing test structures, process monitors, and metrology techniques along with applying statistical and parametric design methods to ensure that the design and process are well-centered to provide sustainable, high yields.  Packaging is a challenge in that the solution must provide a cavity for the moving mechanical structure(s) to operate and do so at low cost while maintaining the high level of RF performance achieved at the die.  Because of the many constraints on packaging, many RF-MEMS solutions are directed at wafer-scale approaches.   Reliability concerns can include mechanical reliability (how many times can the switch be actuated), failure due to shock, and material reliability (changes in properties over the lifetime of the end product).  In addition to the time and discipline required to collect reliability data, standard test procedures and methods for RF-MEMS are still evolving.

MIJ: What are the main challenges with integration of RF MEMS components into the overall system?

JH: One challenge is keeping the focus on the performance and benefits provided by the RF-MEMS technology, and not on the technology itself.  RF-MEMS is still a new and relatively untried technology in consumer electronics products.  As such, it is seen as somewhat risky.  This risk can be mitigated by ensuring the products are built by the same suppliers customers already use and that they have a familiar look and feel to them (standard packages, handling, power requirements, etc.).  A second challenge is ensuring the component is affordable for the end application on both a price and cost basis.  While there are many replacement opportunities at the component level, architectural changes are required to maximize the penetration and benefits provided by RF-MEMS in applications such as cellular phones.  Customers must drive such architectural changes so finding a lead customer willing to make a strategic commitment becomes key.  The components must either solve a problem significantly better than the alternative, or solve a previously unsolvable problem that is causing customer and market pain.  Finally, one must ensure that the components can actually be integrated when and to the degree the economics dictate.  This requires a careful choice of materials, process technology and division of functionality.

MIJ: Currently, are there any RF MEMS devices that have been deployed in mass quantities by any company?  If so, which device and which company?

JH: To the best of our knowledge in the wireless area the answer is “no”.  Most predictions indicate initial mass deployment in very late ’05 / early ’06.  When I refer to RF-MEMS, I am thinking of devices that have (mechanical) moving parts.  Some people will argue that the Agilent FBAR duplexer is a MEMS device.  We tend to think of this as a 3D microscale device.  If you want to include microstructures such as the FBAR in the definition of RF-MEMS, then the answer is “yes” as this filter has been deployed in quantities of many millions per year.

MIJ: Who are your main competitors and what is the key competitive advantage for WiSpry?

JH: Our main competitors are established front-end RF switch, module and filter suppliers.  Our key competitive advantage is products with a sustainable performance advantage at competitive economics coupled with the unique capability of making tunable, broadband RF solutions in silicon in traditional IC foundries.

MIJ: What about RF MEMS start-ups such as Discera and Xcom Wireless?

JH: Discera is focusing on a different application – frequency synthesis – then we are.  They could become a potential competitor at some point in the future but at present, they are more complementary than competitive.  Xcom Wireless is focusing on higher frequency and more defense related applications than Wispry.  If they move to lower frequency commercial applications in the future, they could become competitive as well.  Of the other RF-MEMS companies out there, MagFusion and Teravicta are the closest to Wispry.  We do not list them as direct competitors as we do not see them as having competitive solutions for our target market.  Their switches are too big, too slow, or too power hungry to make it in handsets.  Plus we believe we are the only RF-MEMS company with a technology platform that allows us to build multiple RF-MEMS products on a single process flow that allow for direct integration with active RF devices.  In our market, we do not believe customers will engage if all you have to offer is a single, discrete component replacement product.

Comments

[Hithesh K Gatty]

Nice thought put forward in the discussion, adressing crucial market issues in RF-MEMS. I am working on RF-MEMS. I need more info on 'different type of materials which can be used in RF-MEMS'.

mail to me at : hgatty@mems.ece.iisc.ernet.in

any related issues on RF-MEMS will be appreciated.

Thanx
Hithesh