Size Matters
By Marc Weingarten, Tue Apr 09 00:00:00 GMT 2002

Less is more, and MEMS technology promises to miniaturize your mobile.

Mobile phone users, it seems, are obsessed with miniaturization. A kind of pecking order predominates among those who flash matchbox-sized handsets in public; whomever can squeeze his phone into the smallest pocket wins bragging rights to being right on the cusp of the leading edge.

Thanks to smaller, more potent batteries, as well as enormous advances in chip, filter and resonator technology, mobile phones have shrunk to but a mere fraction of the first-wave Wall Street bricks that emerged in the late 80's.

And yet, it still doesn't seem to be enough. At least for a handful of engineers and developers who are working on the next generation of supra-mini phones so miniscule they could fit snugly into the user's ear canal. A handful of companies, among them Discera and Agilent, are applying micro-electro mechanical systems, or MEMS, technology to mobile phones, in the hope that a phone stripped of most of its traditional internal elements can be cut down to size.

Wearable wireless

"With MEMS resonators, your cell phone could be worn around your wrist, or like a ring around your finger," says Clark Nguyen, an associate professor in the department of engineering and computer science at the University of Michigan, who is one of a handful of engineers that are developing MEMS technology for wireless use. "A cell phone could become something that you wear all the time, but don't really think about until you need it. Like a watch."

Much of what takes up space in a mobile phone is designed to allow you to receive calls intended for your phone, as opposed to someone else's. The filters and resonators beneath the keypad are passive components that filter out unwanted radio, TV, and other frequencies, and funnel in the desired ones unimpeded. These components tend to be constructed from ceramic compounds, and for that reason, they provide an intractable obstacle to shrinking devices beyond the current threshold.

"Any high frequency electronic device, in general, requires filtering," says William Mueller, a MEMS research and development manager with Agilent technologies. "If you're moving from a high to a low frequency, that frequency has to go through a mixer, and that creates all kinds of signals that can jam your own signal. But any two-way electronic device needs filtering."

In addition, mobile phones do not have integrated circuit boards - all of the components are separate, which takes up lots of space. "Passive components, like crystal filters, cannot be integrated with the transistor electronics on a single board," says Nguyen. 'We can't have everything on a single chip."

The burden of bulk

Thus, even the smallest phones are hindered by unnecessary bulk. Engineers now want to replace those cumbersome filtering components with MEMS filters, which would not only lead to the massive shrinkage of phones -potentially to the size of Mr. Nguyen's pinkie ring - but would also provide far greater efficiency for outgoing and incoming calls. For anyone who has cursed one dropped call too many, or thrown up their hands trying to make a local call without interference, MEMS components could be the Godsend they've been looking for.

There are essentially three basic filters used in phones: An electrical resonator, which is usually made from ceramic, a surface acoustic wave resonator, or SAW, and a bulk acoustic resonator. Frequency size directly affects the performance of these filters - the higher the frequency, the more difficult it is to process and still provide a strong signal.

Currently, the filters used to weed out unwanted frequencies on most phones are sorted with surface acoustic wave devices, or SAWs. When the desired frequency meets up with the SAW, the device will vibrate contentedly. But SAWs are vexingly non-discriminatory, and are prone to vibrate when any number of rogue frequencies hit them. This leads to those moments when we want to throw our mobiles across the room. By contrast, MEMS are more finely honed instruments, and will only vibrate when a specific frequency hits it - kind of like a finely calibrated tuning fork.

"I liken high-frequency resonators to guitar strings," says Nguyen, the founder of a company called Discera, whose mandate is to design functional MEMS resonators that work not only for phones, but also across all wireless applications. "But in fact, MEMS resonators are only 10 microns long, so they vibrate at a much higher frequency than guitar strings - they can vibrate at frequencies up to 100 megahertz."

At such extraordinarily high frequencies, mobile phones do not need nearly as much power to operate - in fact, hardly any power is needed. That has huge implications for all kinds of wireless applications. Because SAWs are woefully inefficient and tend to produce weak signals, power amplification must be used to compensate. That leads to cumbersome handheld units with heavy-duty battery power. Better selectivity can lead to leaner, meaner, more energy efficient phones.

For his part, Nguyen sees MEMS resonators ushering a brave new wireless world. "In cities, you could have wireless sensors above you every ten feet or so, on buildings," he says. "They could monitor temperature, air pressure, any number of things. It could also be used for agriculture. You could have crops spread out over a large area, and wireless sensors could tell you which crops are good, and which are bad, which could lead to more efficient farming. It could also help usher in wireless offices, as well."

There's always a hitch

The MEMS components themselves are exceedingly small - a clutch of filters can amount to mere fractions of a centimeter. But MEMS components are fragile and delicate; so attuned are they to the slightest variation in movement that something as prosaic as air molecules or moisture can interfere with the internal device's vibration selectivity. Which means that an internal vacuum must reside in the covering over the MEMS devices, and that has created complications for Discera as it tries to bring MEMS-enabled mobile phones to market.

"I would say that vacuum encapsulation is the biggest problem we're facing right now," says Clark Nguyen. "In the mobile phone business, cost is everything, and we have to figure out a way to hermetically contain the resonators cheaply."

In order to skirt that vexing issue, other companies are avoiding tuning elements in favor of more stable membranes that reside within something called a film bulk acoustic resonator. Agilent has one such device, called an FBAR resonator. Although it has already been used in PDA module cards, Agilent wants FBAR to become standard for mobile phones that would still be considerably larger than phones using MEMS filters, but 90 percent smaller than SAW devices. "What (Discera) is doing with MEMS is different from what we're doing," says Agilent's William Mueller. "Their filter is a relatively complex mechanical structure, whereas what we have is a self-pulsing membrane similar to an accordion."

The FBAR is already being used in various wireless devices. Samsung has integrated the resonator in a wristwatch phone that will possibly hit the market next year, and a company called AirPrime is using the FBAR for a wireless module that works for Handspring PDA's.

Regardless of what technology is being used, the holy grail of phone miniaturization remains integrating the filters and resonators with the electronics onto a single microchip, thereby achieving what Seinfeld's George Castanza once called "significant shrinkage" - a process that might be as distant as a decade away from becoming a reality.

"Our substrate is silicon, which is a lot easier to scale than, say, crystal," says Agilent's Mueller. "But the economics aren't right at the moment. What happens is, if you put everything on a single chip, the yield on everything has to be perfect. Besides, SAWs have a huge base right now. It'll take time."

Discera's Nguyen, however, says that the big mobile phone companies are closely tracking the progress of what he and his five-man R&D team are doing with MEMS resonators, and they like what they're seeing so far. "They've been responsive to it," says Nguyen of the big telecom monoliths. "Interest is peaking right now. They're waking up to it, and asking us, 'when can we start to put this on our road map?'"

If Nguyen gets his way, it will be sooner rather than later.

Marc Weingarten is an LA-based writer whose work appears in Business 2.0, The Los Angeles Times, Smart Business, Entertainment Weekly, The Village Voice, Vibe and San Francisco magazine.