Category Archives: Packaging and Testing

Cellucci named Zyvex COO


January 6, 2003

Jan. 6, 2003 — Thomas Cellucci has been promoted to chief operating officer at Zyvex Corp.

Cellucci, who joined the Richardson, Texas-based supplier of nano-based products and services in July, served as chief marketing officer and vice president of products. He has been focusing the firm on developing and commercializing tools and technologies for immediate and near-term small tech applications.

Cellucci most recently was president and chief executive of Etec Inc., a MEMS testing company in West Peabody, Mass. In 1999, he founded Cellucci Associates Inc., a consulting firm engaged in raising capital and providing business services to high-tech global firms.

Click here to enlarge image

MUNICH, Germany, Nov. 21, 2002 — Carsten Bahle of Wicht Technologie Consulting doesn’t mince words when it comes to the need to rethink MEMS packaging.

“The biggest stumbling block to commercial success is the lack of general, simple and effective packaging techniques,” he said in remarks made during a special forum on MEMS packaging held last week during the Electronica trade fair in Munich.

The reason, everyone agrees, is simple: cost. Packaging of a MEMS device generally takes up 50-90 percent of its cost, with 80 percent the norm. Although these microsystems are getting smaller all the time, that doesn’t mean packaging costs are shrinking as well. In fact, maintaining the high level of precision necessary at ever smaller scales tends to push prices up.

Click here to enlarge image

Despite that, there can be no scrimping on packaging when it comes to MEMS. Together with the functional unit (sensor, micromechanical component or integrated circuit), the surrounding package is the most important element of the device and plays several roles. It has to protect the sensitive functional unit from environmental factors that could affect its performance, like moisture, high temperature, vibration or corrosion. It also has to provide the component’s connection to the outside world through electrical, optical and other types of interfaces. Finally, it cannot hinder function in any way.

Yet if packaging is so important, it seems strange that it has it been treated like a neglected stepchild for so long.

“There are historical reasons for that,” said Erik Jung of the Fraunhofer Institute for Reliability and Microintegration (IZM), and one of Germany’s leading experts on packaging issues. “MEMS evolved from the microelectronic industry and took over an infrastructure that already existed. That is, design the device first, packaging is an afterthought.”

But that has been the wrong approach, he and others say.

“Simply put, design has to be tackled early on in the process,” said Katrin Persson of IMEGO, a microsensor systems company based in Gothenburg, Sweden. “Some 80 percent of breakdowns are due to packaging problems. We need to dedicate more attention to the outside of a component.”

Integral functions are part of the solution, according to Fraunhofer’s Jung. Packaging, he said, should be designed to do more than just protect. It should also add to the final component’s functionality.

“Packaging should be a value-added process,” he said.

For example, a microfluidic sensor package would add this value if it contained a tiny pipeline to bring the media to be measured to the device. Other concepts have the package forming part of the sensing structure itself, becoming part of the device’s own complex system instead of just a dead casing around it.

“We’re seeing people start to think in this direction,” Jung said. “Not as much as I’d like them to, but it’s a start.”

He said that over the past few years engineers on both the MEMS and the packaging sides of the equation have begun talking to each other earlier in the design process.

“But the minute commercialization begins,” he laughed, “the talking stops.”

Another area in which Jung wishes a lot more talking would be going on is in standardization, which he said would help solve some of today’s packaging and manufacturing problems. But that is not easy, given that MEMS environmental parameters are very diverse. Some devices must exclude light while others must allow it onto the die surface. Some packages exist in a vacuum, while others must pipe gases or liquids around a chip.

But it is highly desirable that the industry define a standard package for each application category. Come up with a reasonable standard regarding inputs and outputs, for example, and you could have one MEMS package that is appropriate for several different devices.

One group in Germany that Jung thinks could serve as a standardization model for the MEMS community is the AMA German Trade Association for Sensor Technology. The 400-member group association is developing sensor element modules with interfaces that are compatible with various types of components, such as transformers and microcomputers.

“It will save on resources. When you get, say, a new pressure sensor in the future, it will mean you don’t have to replace your entire sensor mechanism, just that module,” said Dirk Rein, a member of the AMA management.

Experts predict that packaging issues could be the make-or-break issue when it comes to RF (radio frequency) MEMS switches, which are used in wireless applications. Current RF MEMS packaging increases a device’s size by a factor of 10. The final cost of the component breaks down to 5 percent device and 95 percent packaging. If RF MEMS are going to live up to their potential, the packaging price has to be brought down without sacrificing reliability, according to Robert Aigner, director of MEMS R&D at Infineon Technologies AG.

“Packaging is the turning point where RF MEMS will win or lose the battle against conventional components,” he said.

Aug. 29, 2002 — Palomar Technologies, a California-based maker of precision assembly equipment, said it has launched an automated MEMS packaging system.

The 3500-II picks and places parts into the package for bonding once they have been presented for assembly. The system eliminates the need for human handling of fragile MEMS devices and increases reliability and performance through automated part placement and attachment, the company said in a news release.

Palomar’s products offer automated assembly of optical, radio-frequency and microelectronic packages in telecommunications, automotive, aerospace and medical industries.

July 15, 2002 — Measurement Specialties Inc., a Fairfield, N.J.-based designer and manufacturer of MEMS sensors and sensor-based consumer products undergoing a restructuring, today announced it has delayed filing its annual report for fiscal year 2002. The American Stock Exchange has halted trading of the company’s stock. Measurement Specialties did not provide a reason for failing to file the report on time.

The company also said it had signed an agreement to sell its Milpitas, Calif. wafer fab facility, a manufacturing location for sensors and MEMS R&D, to Silicon Microstructures Inc., a subsidiary of Elmos Semiconductor AG. The transaction, totaling $5.22 million, is comprised of $1.85 million in prepaid credit for products and services and $3.35 million in cash to be paid at closing. The company expects to close the deal on Aug. 1, 2002.

Chairman Joseph Mallon said the two announcements today were unrelated. Last week, the company announced it had successfully negotiated a forbearance agreement with lenders through Nov. 1, 2002. Mallon said the company will provide details of the agreement in an upcoming conference call.

By Richard Acello
Small Times Correspondent

SAN DIEGO, April 22, 2002 — OMM Inc., a four-year-old San Diego firm, is tapping into its reserves of patience as it waits for the telecom industry to adopt its MEMS-powered technology for optical switches.

MEMS-based switches route voice and data traffic from one fiber optic cable to another by steering light through a lens, reflecting it off a movable mirror and redirecting the light back into any one of a number of output ports. OMM’s latest product is a 32 x 32 switch that Lawrence Gasman, Communications Industry Researchers president, called a culmination

Phil Chapman

Vital facts about
OMM Inc.

of MEMS two-dimensional technology.

In a 2-D system, the micromirrors move in a planer motion. In an evolving technology called 3-D, the mirrors can tilt and swivel. In-Stat/MDR MEMS analyst Marlene Bourne compared the difference in the two technologies to that between a seesaw and a gyroscope.

Bourne said OMM is the acknowledged leader in the 2-D arena. “They largely have it to themselves.”

Since MEMS-based technology is still a novel approach to switching in the telecom market, the barriers to adoption of 3-D are even higher than 2-D, said Bourne. Even with 2-D, adoption barriers are significant. “The technology has proven itself, but the market isn’t ready,” said Bourne. The purchase lag includes time for demonstration and time to bring the carriers on board.

Nevertheless, OMM’s investors have poured more than $150 million into its development. Roughly half has come from a consortium of its customers, and the rest from venture firms, says Phil Chapman, OMM’s chief executive.

“No one is doing what we do yet,” Chapman said. “There are probably 30 companies who have announced product for the last two years but no one has been able to commercialize it.”

The CEO said the current telecom slowdown is “kind of irrelevant to us.” That’s because slowdown or no, carriers are in need of equipment that lower their costs, and raise efficiency. “The current networks are not particularly efficient. (They were) designed in the days of voice communications; now with data transmission everything is packetized. The industry going through massive set of issues. Prices they can charge their customers have crashed. They are busy in field trials of trying out next generation of equipment and we are sitting in the heart of it. Our goal is to make sure we’re in the right place at the right time.”

Chapman said the telecom industry is conservative by nature, but “at this point MEMS reliability issues have been addressed.” The slow road to adoption is a fact of life in a healthy or unhealthy market. “But this is where they need to go.”


null

COMPANY FILE: OMM Inc.
(last updated April 22, 2002)

Company
OMM Inc.

Headquarters
9410 Carroll Park Dr.
San Diego, CA 92121

Branches, divisions and subsidiaries
Sales offices in Boston and Munich, Germany.

History
OMM was founded in June 1997 with the goal of solving a challenging optical network problem: creating a photonic signal switch that “skipped a step” of converting optical signals to electrical and back again. Beginning with its first product shipments in late 1999, OMM has since delivered its switching subsystems to more than 60 global customers. In mid-2001, OMM announced that it was discontinuing its 3-D MEMS efforts and focusing strictly on its 2-D products, which had already passed stringent Telcordia environmental and reliability standards.

Industry
Photonics. OMM designs and manufactures what it calls the first scalable MEMS-based all-optical switching subsystems. These 2-D switches connect inputs to outputs in different types of networks. They are encased in hermetic packaging based on OMM’s proprietary technology.

Selected small tech-related products
32×32, 16×16, 8×8, 4×4, OADS-8, 2×16 optical switches. OMM’s MEMS-based optical switching subsystems include arrays of digital micromirrors. The micromirrors function as part of a low-loss system that eliminates the need for optical-to-electronic signal conversion.

Management

  • President and CEO: Phil Chapman
  • VP Manufacturing: Jim Hartman
  • Director of Engineering: John Gritters
  • Employees
    155

    Investment history
    OMM has closed eight rounds of institutional venture financing totaling over $150 million, most recently in November 2001 with a $22.3 million round. More than 11 institutional investors have participated, including Bessemer Venture Partners, Sevin Rosen Funds and Atlas Venture. Originally poised to go public in the late winter of 2001, OMM withdrew its IPO in early March of that year due to unfavorable market conditions.

    Selected strategic partners

  • Nortel Networks
  • Alcatel
  • Siemens AG
  • Solectron Corp.
  • Sycamore Networks Inc.
  • Lumentis AB
  • Revenue
    Annual sales: $6.933 million

    Barriers to market
    Packaging to meet stringent Telcordia reliability standards, a “very tough set” of requirements that include sealing, cycles of operations and temperature extremes.

    Competitors

  • Agilent
  • Agere
  • Cisco
  • Corning
  • JDS Uniphase
  • Lucent
  • Nortel Networks
  • PMC-Sierra
  • Applied Micro Circuits
  • Goals
    Short-range: Successful field trials and early adaptors. Long range: To see all-optical switching deployed widely within the global communications structure within the next couple of years.

    Why are they in small tech
    “I’ve been in the semiconductor field for the last 10 years, this is a new and exciting area,” said Phil Chapman, president and CEO. “We get to decide how to price it, how to package it, and everyone else has to follow what we set up, and you don’t get to many opportunities in life to do that.”

    What keeps them up at night
    “The timing of the adoption of this technology,” Chapman said. “Being early is better than being late, but it is frustrating while you wait for the demand to catch up.”

    Contact information
    URL: http://www.omminc.com/home.html
    E-mail: [email protected]
    Phone: (858) 362-2800
    Fax: (858) 362-2999

    Recent news
    Report: Role of 2-D MEMS to grow in optical switches
    Siemens approves OMM’s switches for optical system
    OMM launches two new MEMS switches


    null

    Reprints of this article are available here.

    By Tom Henderson
    Small Times Senior Writer

    ANN ARBOR, Mich. — The first Engineering Research Center in the U.S. devoted exclusively to MEMS has started 42 doctoral research projects in its first nine months.

    Director Kensall Wise hopes that some of them will lead to large-scale treatment for deafness in the next 20 years.

    In a confidential report issued last week to the 19 members

    null

    A microelectrode array for stimulating
    the central nervous system.
    of its Industrial Advisory Board, the ERC detailed its progress in a wide range of devices and technologies. They include cochlear implants to treat deafness, environmental sensors, micro devices for the delivery of tiny amounts of drugs to fine-tune the treatment of Parkinson’s and implantable sensors to monitor arterial pressure in humans.

    The ERC also plans to create a master’s degree program in microsystems, and to develop educational programs for junior high and high school students.

    The ERC is one of 17 centers approved and funded nationwide by the National Science Foundation and the only one dedicated to research in MEMS, according to Lynn Preston, who oversees the ERCs as deputy director of the National Science Foundation’s Division of Engineering Education and Centers.

    Partner universities include the University of Michigan, where the ERC is headquartered, Michigan State University and Michigan Technological University. The focus is on wireless integrated microsystems (WIMS), with two major thrusts being the creation of implantable MEMS devices for the treatment of neurological disorders and external sensing devices to be used for environmental applications.

    The WIMS ERC was funded for five years for $29.5 million, about $16 million of that coming from the NSF. The NSF will review the ERC after three years. If suitable progress has been made, funding will be renewed for another five years, said Preston.

    The ERC is also funded by the member universities, other federal agencies and the state of Michigan.

    The WIMS program at U-M began Sept. 1, 2000. U-M’s other ERC, in reconfigurable machining systems, is six years old. Only four other universities head up two ERCs — the University of Illinois, Carnegie Mellon, Georgia Institute of Technology and Massachusetts Institute of Technology.

    “When we give a school a second ERC, the proposal obviously has to bubble right to the top. It can’t be at the edge,” said Preston. “And we have a pretty good history with the university.

    “This was one of the strongest proposals we’ve seen in a lot of competitions. And we heard that throughout the review process,” said Preston.

    “What made it so strong? They’ve taken on systems-level challenges. They’re putting together devices, components and control systems. For example, they’re doing that with their cochlear implants.

    “It’s sufficiently complicated, but it’s doable in a few years. I think they knew they could get some integrative aspects going quickly . . . And they had strong commercial partnerships in place.”

    The only other ERC funded last year was one in sub-surface sensing and imaging systems, a joint effort by Boston University, the University of Puerto Rico-Mayaguez and Rensselaer Polytechnic Institute.

    A PROGRESS REPORT

    “We’ve been working 60-80 hour weeks the last month,” said Wise, a MEMS pioneer and U-M prof since 1974. “The faculty works harder now than it does during the school year. To me, teaching is R&R. That’s fun. This is work.” The U-M school year ended in April.

    Wise and ERC staff met for two days last week for a semi-annual conference with the center’s 19-member Industrial Advisory Board. Forming alliances with business has been a central tenet of the ERCs since the NSF launched them in the fall of 1984.

    Larger, out-of-state businesses pay the WIMS ERC $50,000 a year to become partners. Four in-state businesses with fewer than 50 employees pay $10,000. In return for that money and program input and advice, the firms get access to top students and faculty, a heads-up on promising projects, early access to intellectual property and a break on royalties for future collaborations.

    U-M’s industrial partners — a combination of national heavyweights such as 3M, Honeywell, Intel, Motorola, National Semiconductor and General Motors as well as smaller local companies such as Advanced Sensor Technologies, ISSYS and Ardesta LLC — were brought up to date on progress in the first nine months. Small Times Media is owned by Ardesta.

    Accomplishments include:

    * The ERC has 28 faculty, 27 undergraduates, two masters students, 66 doctoral students and four post-docs, with plans to expand to one undergraduate paired with each doctoral student.

    * Forty-two doctoral projects have begun, including sensors that use metal nanoclusters, gas sensing devices with environmental applications, diamond film coatings, sensors for remote rain monitoring, micropumps for precise drug delivery of tiny (100-picoliter) amounts of drugs for treatment of epilepsy and Parkinson’s, and hermetic packaging technologies for WIMS devices.

    * WIMS researchers have made a prototype wireless pressure sensor, the forerunner of devices that may one day be planted in stents to allow the measurement of intra-arterial pressure simply by waving a radio-frequency wand over the implant site.

    * A set of prototype electrode arrays has been designed for a wireless cochlear implant, the first of what Wise hopes will be a series of neural prostheses.

    “Most deafness will be treatable, probably in the next two decades,” said Wise. “Though some conditions that wipe out both the cochlea and the auditory nerve will be tougher.”

    Other devices may let the blind see and reduce the effects of Parkinson’s, paralysis and epilepsy.

    EDUCATIONAL THRUSTS

    Education linking with schools has been a major goal of the ERC program.

    * In late June, high school science teachers from four Michigan communities — Okemos (near Michigan State), Houghton (Michigan Tech), Ann Arbor and Dearborn will meet with ERC officials for a workshop whose aim is to develop ways of using WIMS to make science more fun for high school students.

    “This stuff is pretty cool,” said Wise. “You show kids thermal flow meters — they’re pretty easy to understand. And infra-red imagers are cool. This stuff is pretty visual; there’s a lot of interesting principles you can demonstrate using WIMS.”

    * At MSU, beginning in July, there are four one-week intensive residential programs for promising minority science students from the Detroit area.

    The free program brings in 25 kids each week from the eighth and ninth grades and covers Web design, engineering design, mathematics and selected topics in WIMS.

    The program, funded in part by the ERC this year, began last summer and involves other Michigan universities, including Wayne State, the University of Detroit-Mercy and Lawrence Tech University.

    * A curriculum for a Master of Engineering degree in integrated microsystems is being developed, built around a core of WIMS/MEMS courses.

    AN IMPRESSED INDUSTRIAL CHAIRMAN

    Nader Najafi, CEO of ISSYS, Inc., an Ann Arbor-based maker of pressure and flow sensors, is chairman of the ERC’s Industrial Advisory Board.

    “All the industrial members were very much impressed,” he said of the progress report, adding particular praise for the ERC’s educational programs. Najafi was with about 35 industry leaders at a MEMS Industry Group conference in Pittsburgh on April 24-25. The announcement that the ERC is creating a master’s program in microsystems addresses a major concern raised by that group.

    “At the MEMS Industry Group meeting, we all expressed a need for a dedicated master’s degree in MEMS,” he said. “So, this is very exciting.”

    He said the ERC hopes to make it easier for students to get internships in industry, and to arrange travel by professors and students to visit member companies. Their goal: help students relate what they learn in the classroom with real-life applications in the workplace.

    “We want to make a bridge in this gap of technology transfer,” he said.

    A key to that bridge, he says, is the appointment of Joe Giachino as director of external affairs and industry liaison for the ERC. Giachino is a former senior technical specialist at Visteon, the giant auto parts supplier, focusing on MEMS and microsystems. He brings an industrial/commercial perspective to the ERC.

    “With the ERCs, what we’re doing is something that industry asked us to do,” said Preston. “The ERCs mimic in academia something closely resembling what happens in industry. Our students prove to be better students than those who work with a single professor, write a dissertation and graduate.”

    Najafi said that NSF’s commitment to the WIMS ERC proves small tech has arrived. “The NSF only invests in technologies that can give the U.S. a technological advantage. Getting an ERC for MEMS is s very good sign that MEMS is now considered a viable economic strategy.”


    null

    ERC FACTS AND FIGURES

    * The NSF launched its ERC program in 1984. It has established 37 ERCS in all; 16 successfully finished their 10-year terms, 17 are under way and four were terminated early upon review.

    * Industry partners now account for 28 percent of funding, with the NSF contributing 31 percent, other federal agencies 24 percent, academia 11 percent and state government 5 percent.

    * Fifty for-profit companies have been spun off from ERC research. * ERCs have produced more 562 inventions and secured 330 patents. * In 1999, more than 200 different technologies were transferred to industrial partners.

    * Participating ERC students have received 2,282 Ph.Ds, 2,171 master’s degrees and 2,138 bachelor’s degrees.

    * The centers have resulted in 18 new degree programs, 648 new or modified courses, 84 new textbooks, and 9,777 publications in refereed journals.