Showcasing the latest
Thursday, May 22, 2014

Technical innovations in design, process, materials, packaging and test have enabled widespread commercialization of breakthrough MEMS products. MEMS sensing applications will track growth in mobile, industrial, consumer and biomedical markets, enabled by innovations in functional sophistication, cost-reduction, and productivity.  

MEMS productivity gains arise primarily from Design, Process, Packaging and Test. The convergence of Design innovations, Process technologies, Packaging advances, and development of low cost Test technologies each plays a major role for wide scale adoption of breakthrough MEMS enabled products. Design offers greater sophistication of sensor function, specificity and sensitivity. Process advances enable previously impossible constructions. Packaging innovations provide complete solutions, more features in incredibly compact formats. Test improvements inspire trust and lower cost sensors. Integration of these elements is critical to meeting price points that catalyze volume growth and massive new market penetration.

This conference will showcase these technologies that form the basis for advancements in MEMS products, to achieve future capabilities, enabled by innovations unthinkable a decade ago.

Spend the week in California!

MEPTEC and the MEMS Industry Group (MIG) partnered this year to schedule their respective MEMS symposium and MIG’s M2M 2014 event so as to encourage cross attenance at both events. M2M will be held on May 20-21, the two days preceding the MEPTEC symposium. M2M is being held at the MEMS Sensor & Actuator Center (BSAC) on the UC Berkeley campus.

Advances in MEMS - Foundations

• Gene Burk, IMT
• Sean Cahill, Maxim Integrated Products
• Ira Feldman, Feldman Engineering
• MaryAnn Maher, SoftMEMS
• Mark Wendman

The cost to exhibit is $645.00 for MEPTEC and MEMS Industry Group members, and $745.00 for non-members and includes:

• One admission to the conference
• 6’ table, draped
• 2 chairs
•11”x17” custom table top sign with your logo and company Ndescription
• Logo, link to your URL and company description on special
NExhibitor page
• Company description in the symposium proceedings
• Printed and electronics versions of the symposium proceedings
• Marketing exposure through e-mail campaign

Click here to reserve a table now. For more information contact
Bette Cooper at or call 650-714-1570.

Sponsoring this event will provide a valuable opportunity to promote your company brand and product/service message to attendees, while supporting your business development and positioning goals. For benefits and pricing click on the Sponsorship Benefits link below. Click on the Sponsorship Application Form link below to sign up. For more information contact Bette Cooper at or call 650-714-1570.

Pricing for general admission is $495.00 for MEPTEC and MEMS Industry Group members and $595.00 for non-members and includes attendance, continental breakfast, refreshment breaks, lunch, and printed proceedings. A credit card is needed to hold the reservation.

Pre-registration is strongly recommended. There will be no guarantee of space or materials for on-site registrants.

Final confirmation including maps and directions will be sent before May 16.

Refunds for advance payment, less a $50 processing fee, will be given in full provided cancellation by phone or e-mail is received 10 business days before the event (Thursday, May 8). If you do not cancel by May 8 or are a no-show, the credit card provided to hold the reservation will be charged for the full amount.

A block of rooms has been reserved at the San Jose Holiday Inn (formerly the Wyndham Hotel) for a rate of $149.00. The hotel is conveniently located at 1350 North First Street, San Jose, CA in close proximity to the San Jose Airport.  To reserve online click here - the Group Code box already has the MEPTEC code, MTC. Enter dates between May 20 and May 24; the $149 rate will automatically come up. You can also call 408-453-6200 to reserve your room; be sure to mention MEPTEC in order to secure the special rate.The room block and special rate expires on May 14. After that date you may still be able to get a reservation but the special pricing will not be applied.

MEMS Design Innovations

MEMS product development often involves the co-design of MEMS devices, associated electronics, custom packaging and new fabrication processes requiring collaboration between domain experts--the “one product, one process, one package” dilemma. This session will focus on design for manufacturable MEMS based products and new trends in their design including design re-use and semi-custom design as well as new design techniques in fully custom designs. The increasing importance of system level design and software for MEMS-based products will also be highlighted.

The Changing Landscape of MEMS Manufacturing and Process Technologies

MEMS product development has often involved creation of new fabrication processes, fabrication sequences, and the use of new materials. Equipment manufacturers are providing new MEMS-specific equipment with enhanced capabilities and new materials are being utilized to enhance MEMS performance. In addition, traditional MEMS fabrication processes are being combined with new technologies such as 3D printing, meso-scale fabrication techniques, and nanotechnologies creating both great opportunities and great challenges for getting products to market in a timely, cost effective manner. This session will provide a look at the changing landscape of MEMS manufacturing and process development in the context of the “manufacturing renaissance” happening worldwide.

MEMS Packaging – Impact on Devices and Volumes

Electronics packages typically allow electrical connection while allowing the devices contained within to remain as isolated as possible from their environs. MEMS devices must interact with their surroundings, they are often subjected to direct contact with the environment, and usually must sense a particular variable without contamination from other potentially interfering inputs. Packages need to allow appropriate interaction with the measurand of interest, connect to power and signal I/Os, provide some isolation from undesired inputs, and create a test interface. The ideal package does all of the above and costs almost nothing. Low cost is achieved largely by exploiting existing high-volume technologies to the extent possible.

From another viewpoint, MEMS approaches have informed packaging efforts of traditional electronics. Stress isolation, deep trenching, wafer stacking, thermo-mechanical management techniques, and 2.5D and 3D package approaches have borrowed from the MEMS playbook.

This session will include speakers with insights into packaging of MEMS and the impact of MEMS on devices and their packaging at volume.

MEMS Device Testing Challenges

Testing of MEMS has been a traditional challenge as compared to CMOS devices. The need to measure a major physical input while eliminating minor inputs over temperature and with varying performance parameters has made this aspect of MEMS device production costly and difficult. The design of test systems follows closely the “one product, one test protocol, one test system” paradigm that is developed by the product company. Our speakers will address some of the issues and solutions that have been used in testing devices and which could be used in creating standardized testing of MEMS devices for high volume low cost applications in the future.

Several converging trends are transforming the entrepreneurial process for starting MEMS companies and transitioning MEMS devices into production and into the market. First, it is well-known that recent market set-backs have caused traditional VC funds to view any hardware start-ups with renewed scrutiny and skepticism. Hardware start-ups typically require large amounts of capital ($50-$100M) and many years (7-10), before getting close to a reasonable exit. This large investment in money and time is on top of the already inherently risky prospects for such a start-up being commercially successful. Secondly, MEMS is recognized, by investors, by foundries, and by large consumer electronics companies, as a very successful new product area because of the huge up-take of MEMS components in mobile devices during recent years. Third, key strategic issues in huge upcoming new consumer markets, such as wearables and IoT, are sensors and contextual awareness; areas which are uniquely solved by MEMS devices. And fourth, the sheer number of successful, high volume MEMS devices currently on the market, has created a huge pool of skilled MEMS developers and manufacturers which can be drawn upon for new devices and new start-up companies. All these factors dramatically influence how such companies get funding and how they operate. We will discuss all these issues as they relate specifically to new MEMS companies.

Kurt Petersen received his BS degree cum laude in EE from UC Berkeley in 1970. In 1975, he received a PhD in EE from the Massachusetts Institute of Technology. Dr. Petersen established a micromachining research group at IBM from 1975 to 1982, during which he wrote the review paper “Silicon as a Mechanical Material,” published in the IEEE Proceedings (May 1982). This paper is still the most frequently referenced work in the field of micromachining and micro-electro-mechanical systems (MEMS).

Since 1982, Dr. Petersen has co-founded six successful companies in MEMS technology, Transensory Devices Inc. in 1982, NovaSensor in 1985 (now owned by GE), Cepheid in 1996 (now a public company on NASDAQ: CPHD), SiTime in 2004 (still private), Profusa in 2008 (still private), and Verreon in 2009 (acquired by Qualcomm).

In 2011, Dr. Petersen joined the Band of Angels in Silicon Valley. The Band is an angel investment group which mentors and invests in early stage, high-tech, start-up companies. Today, he spends most of his time helping and mentoring such companies.

Dr. Petersen has published over 100 papers, and has been granted over 35 patents in the field of MEMS. In 2001 he was awarded the IEEE Simon Ramo Medal for his contributions to MEMS. Dr. Petersen is a member of the National Academy of Engineering and is a Life Fellow of the IEEE in recognition of his contributions to “the commercialization of MEMS technology”.

Current Entrepreneurial Environment for MEMS

World’s Smallest Chip Scale Packaged MEMS TCXO for Mobile Applications
Niveditha Arumugam , Manager, Systems Design and Packaging
Engineering SiTime Corporation

Mobile applications require 32.768 kHz clock references with small form-factor, tight frequency stability, and ultra-low power at a competitive price. Legacy 32 kHz quartzbased technology has reached the limits of miniaturization, performance and cost.

SiTime’s MEMS technology has enabled a 32 kHz temperature compensated MEMS based oscillator (TCXO) in a chip-scale package (1.5mm x 0.8mm CSP) with 5 ppm frequency stability over -40C to 85C, making it the world’s smallest and best-in-class 32 kHz TCXO. The SiTime solution enables increased functionality, smaller size and longer battery life in wearables, smartphones and other mobile devices.

This talk will provide an overview of the innovations that enabled this game-changing product for mobile applications. Topics will include innovations in MEMS/CMOS technology, chip scale packaging, and test and calibration techniques.

High-Speed Testing of Pressure Sensors
Gary Casey, President, Casey Engineering

This subject is perhaps unique in that there is no one answer that one can expect all experts to converge upon. In fact, small differences in sensor type, pressure range, and customer requirements may result in a considerably different answer to the question of how to test the sensor in production. As an example, a differential pressure sensor might be tested in essentially the same way as an absolute or gage sensor – but if there is a common mode requirement, the optimum method will likely be dramatically different.

There are essentially 4 aspects of testing, how to handle and transport the sensor, apply pressure, supply electrical power, and finally how to measure the result. Often the high-volume methodology grows from prototype and low-volume methodologies, usually leading to difficulties that could have been prevented. For example, the pressure application must be rapid, accurate and stable; and the traceability to NIST should be a short and robust path. Examples of solutions to these challenges will be discussed.

MEMS-based Autofocus in Cameras
Vijay Chandrasekharan, Ph.D., Invensas

Miniaturization of consumer electronics and inclusion of MEMS in these devices leads to a synergy of constant improvement and comes with its own challenges. This paper discusses the development and challenges of the only MEMS autofocus actuator in the world. The actuation range is a 100 um for the target focus range and can move a payload upto 4 mg to reach the desired focus range. The MEMS based design helps focus in at least half the time, with 1/100th the power with 1/5th the tilt of the current focus solution in the market that use voice coils. Large pay load with a 100 um of motion poses a significant design challenge for a MEMS device for both modeling and mechanical reliability. The non-hermetically sealed injection molded plastic package is one of the key components in the reliability of the MEMS device and is pushes the limits of the injection molding process. The device has a shock survivability of > 10,000g and over 12 drops from heights > 1.0m in a smartphone.

Passive Micro Sensors:  When Power isn't an Option
Todd Christenson, Chairman, CTO and Founder, HT Micro

Microsensors requiring no electrical power to provide their monitoring function along with the MEMS approaches enabling them are discussed. Several aspects are contributing to the high volume exploitation of these zero-power devices. This family of electromechanical sensors in effect operates as their own energy harvesters and extracts energy from their physical surroundings to monitor the measurand. Their output is provided through ultra-reliable switch contacts which in turn require specific feature sizes and materials that are unique relative to semiconductor devices. Long lifetime requires that the micro-electromechanical structures reside within integrated hermetic packages. By their nature these sensors also must be able to be easily positioned and integrated into strategic or remote positions within a system. Tremendous volume demand and growth is apparent for such devices in situations such as awareness and intermittent monitoring. Examples including inertial and magnetic switches are described.

Critical Success Factors for the Commercialization of MEMS:
the 2013 MEMS Industry Report Card
Roger H. Grace, President Roger Grace Associates

Barriers to the commercialization of every industry and technology exist… Microelectromechanical Systems (MEMS) are no exception. As we have passed more than a half-century of MEMS research, design, and development, many people ask… “Why has it taken so long for the MEMS technology to realize its potential or reach true commercialization?” The presentation addresses 14 critical success factors that I have developed and consider to be the major barriers to the commercialization of MEMS and the progress made to date to overcome these barriers through the introduction of a “MEMS Industry Report Card” for the year 2013. Some of the topics addressed are: R&D, Marketing, Infrastructure Development, Management Expertise, Creation of Wealth and Profitability. The report card has been updated and presented yearly since 1998 (where it was first given at the seminal Hilton Head Conference) and continues to be given yearly to many enthusiastic audiences worldwide.

Development of MEMS micro-relays
Amit Kelkar, M Sr. PMTS (Sr. Scientist), Maxim Integrated Products

RF MEMS micro-relays offer significant advantages over conventional electromechanical or reed relays – such as smaller size, better switching performance and lower cost. A novel process flow for fabricating MEMS micro-relays is presented here. First, a bottom wafer with contact metallization is bonded to a spring wafer with complementary contact metallization. Next, the spring wafer is micro-machined to the desired thickness and MEMS structures are formed using silicon reactive ion etch. Finally, a cap wafer is bonded to the bottom+spring wafer to form hermetically sealed cavities with MEMS micro-relays. Package sizes of 3 x 3 mm (WLCSP) and 5 x 5 mm (LGA) are successfully demonstrated. Two critical vectors for micro-relay performance are investigated – low stiction and stable on-resistance. Ruthenium is used as the contact metal for low stiction and Gold is used as the bonding metal for achieving a hermetic seal and enabling a stable on-resistance through 100 million switching cycles. With the device and process developed here, excellent cycling reliability consistent with benchmark conventional relays is shown.

MEMS Ultrasonic Rangefinding Solution for 3D Gesture Human Machine Interface (HMI)
Michelle Meng-Hsiung Kiang , Ph.D., CEO and Co-founder, Chirp Microsystems, Inc.

Bulk piezoelectric ultrasonic transducers are widely used in industrial applications and medical imaging. The situation is somewhat analogous to the state of inertial sensors some ten to twenty years ago before the widespread commercialization of MEMS gyros and accelerometers. Micromachined ultrasonic transducers are poised to enter existing applications and introduce entirely new applications that take advantage of their small size, low cost, low power consumption, and compatibility with IC manufacturing methods.

This talk will discuss some of the challenges faced in the design and manufacture of piezoelectric micromachined ultrasonic transducers (PMUTs) for human machine interface (HMI) in energy constraint applications such as consumer electronics.

Standardization, Miniaturization and Cost Efficiency of MEMS and Sensor Packaging through Advanced Packaging Technologies
Charles Lee, Technical Manager, ASE Singapore Pte Ltd

Abstract to follow

Through Wafer Dicing without Saws or Lasers: High Productivity and Design Flexibility
David Lishan, Principal Scientist, Director-Technical Marketing, Plasma-Therm LLC

Wafer singulation technology is evolving. Although saws are widely accepted, the technology is running into performance boundaries regarding die strength (chipping), cutting speed for thin wafers, throughput for small die, large street widths, and are constrained to orthogonal streets. Laser dicing has been introduced to address some of these issues but limitations associated with induced damage and material incompatibilities have affected adoption.

This presentation will introduce a through wafer dicing using plasma etching that eliminates the constraints of saws and lasers, a new approach that utilizes standard dicing tapes and tape frames and enables non-orthogonal streets for layout flexibility, reduced street widths, chip/crack-free edges, and low stress corners. In addition, plasma etching dramatically increases singulation throughput as wafers become thinner to address the demands of smaller consumer form factors and volumes of smaller die (e.g. < 2mm2).

This new approach results in significantly higher yields, productivity increases, and encourages production of new die shapes. Examples highlighting this new technology will be presented along with various options for process flows.

Why cognitive silicon is a crucial companion to MEMS sensor and energy harvesting devices
Guy Paillet, CEO, General Vision, Inc.

MEMS have revolutionized the sensor field by allowing miniaturization, lower power and lower cost.

ICSD (Intelligent Connected Sensory Devices) in most cases are missing the intelligence. What developers call sensor intelligence is very often "nave" signal processing leading to very simple, almost always, threshold bases decision. Nonlinear classification, which is the essence of the biological brain and is related to statistical data analysis (probabilistic or deterministic) is too "power hungry" to be miniaturized and powered by energy harvesting with standard methods (CPU, DSP, GPU). On the contrary, the biological brain is extremely frugal and low frequency. In order for a CPU to pull the same recognition (aka classification) and learning capability than the CM1K Neuromorphic Memory, such a hypothetical device would need to deliver more than 100,000 MIP at 500 mW with a frequency of 27 MHz. As an example, the very best Intel i7 processor delivers 180,000 MIPS but at 3.9 GHz and 47 W (not including the fan); therefore classification functions are poorly implemented by Von Neumann architecture.

Sensor and sensor fusion are very dependent of being able to filter unnecessary signals in order to convert signal to information and consolidate this information into actionable insights. This is the promise of merging neuromorphic silicon into MEMS technology for both the sensor and the energy harvester. A concept will be presented outlining fully autonomous wireless condition monitoring and anomaly detection (vibration). As IBM CEO Virginia Rometty stated, cognitive computing is the third wave of computing.

Reducing Development Time for MEMS Devices
Scott Smyser, Executive Vice President, Worldwide Marketing and Business Development

Increasing market demands and tough competition are driving MEMS companies to develop new products at an increasing rate. In this fast-paced environment, minimizing time-to-market is becoming critical. The need to quickly develop new sensor prototypes and quickly ramp into production calls for existing, proven, off-the shelf sensor signal conditioners.

Typically, MEMS product companies will develop their signal conditioner in conjunction with a MEMS sensor either by internal development or by an external ASIC design provider. The ASIC is specifically equipped with an analog front-end that converts the analog output of the MEMS to the digital domain. Consequently, the sensor and ASIC must be highly compatible to account for the analog output signal level with respect to offset, gain, sensitivity, linearity and other factors. Developing one side of this equation is challenging, but doing both in parallel is wrought with uncertainty as both the MEMS and ASIC can only be simulated and not fully tested until they both exist in silicon. The result is often a race to continuously revise the MEMS, ASIC or both until they “fit” sufficiently to be launched as a final product. This “fit” comes in addition to all other design considerations that must be given to an integrated sensor device with a physical sensor microstructure and a complex mixed-signal ASIC.

Repeated ASIC and MEMS tape-outs are costly, resource intensive and time-consuming. The risk is always that the final integrated product misses its design specification and market. Si-Ware Systems (SWS) has developed unique development platforms that combine high performance ASICs with configurability and programmability to interface with a variety of MEMS devices. With these platforms, time, risk, and cost are dramatically reduced for a MEMS product development.

Design of MEMS Piezoelectric Vibrational Energy Harvesters for Industrial and Commercial Applications
Kathleen M. Vaeth, Vice President of Engineering,
MicroGen Systems

The increasing availability of small, lightweight, and affordable MEMS sensors and RF transmitters has opened up the possibilities of smart, integrated wireless sensor networks for monitoring and control of industrial equipment, building infrastructure, and transportation vehicles.  Energy harvesting technology is key to enabling increased intelligence and proliferation of these networks, which are currently limited by available battery power.  In this paper, we report the design, fabrication, modeling, and characterization of MEMS-based piezoelectric vibrational energy harvesters with power output suitable for supporting wireless sensors and sensor networks.  Finite element modeling of the harvester resonant frequency and peak AC power output as a function of deflection shows good agreement with experimental data over a range of frequencies, and mechanical analysis of the harvester movement provides guidance in design for robustness. Demonstrations of harvesting vibrational energy to power real world objects will be presented, including a wireless sensor network for monitoring building temperature. The data obtained and lessons learned from the installation will be discussed.