miércoles, 10 de febrero de 2010

Coils, filters, switches and capacitors manufactured with micromechanics for RF


Ericsson Hopes for
Mechanical Mobiles

Coils, filters, switches and capacitors manufactured  with micromechanics for RF sector provide more
compact and energy-efficient designs.

When mechanics in micro-format can replace traditional
electronics, mobile phones can be made both smaller
and more energy-efficient. These are the expectations
of both researchers and industry when it comes to the
future of micromechanics in the RF (radio frequency) sector.


MEMS, or micro-electro-mechanical systems, have
now found applications as blood pressure meters in the
medical industry, mirrors in projectors and as a
ccelerometers in the automotive industry. But small
mechanical structures can also be used as building
blocks in a radio. In research labs around the world,
there are now MEMS components that can replace
electronic switches, radio filters, coils and capacitors
in transmitters and receivers. But the big break-through,
as well as the introduction of commercial products,
will have to wait for the time being.

"It's not just a matter of developing a component; it must
also be mass-produced," says Shu-Ang Zhou, who
coordinates Ericsson's RF MEMS efforts.
He and Ericsson are not interested in just a few
components; for RF MEMS to be of interest, it must
be possible to develop components with good
performance at competitive prices and in runs that
suffice for mass-produced mobile phones. Ericsson
is now looking both at RF switches for smart antennas
and filters, coils and capacitors in MEMS structures.

"I believe that MEMS will initially be in the terminals
because the technology makes it possible to reduce
both energy consumption and size. In base stations,
energy consumption and size are not as important."

The reason for Shu-Ang Zhou, as well as many others,
putting so much hope in RF MEMS is that constructions
can be made for which energy loss is insignificant in
comparison with today's electronics solutions.
An example is a diode or transistor switch that is
replaced by a mechanical switch in an antenna.
While the diode constantly consumes power, the
mechanical switch functions like a common relay.
Moreover, MEMS components have good linearity
and high bandwidth, which is important for future
broadband radio communications systems.


The disadvantage is that mechanics are slower than
electronics, but for switching antenna elements in a
smart antenna, mechanics are fast enough.

It is not just the power benefits that make Shu-Ang Zhou
believe that the telephone will be the first application for

"There is also the uncertainty of how reliable the
components are that will result in them being used in
the terminals first. One might have a terminal for two
or three years; a base station should last considerably

Edvard Kälvesten, president of the Swedish MEMS
company Silex, believes, however, that the base station
can become the first application for which RF MEMS
comes into use.

"It will be difficult, price-wise, to replace existing
components in mobile telephones. However in the
base station, it's acceptable if it costs more."

His company is conducting a project with RF MEMS
together with an unnamed customer. They expect, just
as Ericsson, that products will be introduced within
three years. "Of our products, RF MEMS is the one
that demands the longest take-off run," Kälvesten explains.

According to Shu-Ang Zhou, Ericsson does not plan
to manufacture its own RF MEMS components. Instead,
collaboration is underway with likely subcontractors,
for example, Philips Semiconductors. This is the procedure
the company employs with most of its components, but in
the MEMS area, collaboration can demand more.
"It's a little different because it's an entire system. This
requires that the system engineer works together with the
component manufacturer."


Collaboration with Philips is conducted, in among other
ways, through the newly started EU project Mems2Tune,
which shall run for a period of three years and target the
development of switches and adjustable capacitors.
Joost van Beek from Philips is leading the project, and
precisely as his Ericsson colleagues, is convinced that the
future is in mass-produced MEMS for mobile telephones.

"Philips is most interested in the applications for
mass-produced electronics, and the market for RF MEMS
is there if you can just make a sufficiently inexpensive and
reliable product. I don't believe that that a switch in
RF MEMS needs to be more expensive than a
corresponding diode switch."

For Philips, the focus is on development in producing
MEMS products that can be manufactured in a process
that is as close as possible to the standard processes that
they already use.

"We feel that only small modifications are needed to be
able to use MEMS and we are exerting ourselves to change
as little as possible," says Joost van Beek.

    Aside from the adaptation of production for RF MEMS,
the major challenge facing Philips and the other manufacturers
is in finding packaging that is sufficiently inexpensive and
substantial to protect the mechanical parts.

"At present, there is no satisfactory standard packaging
for MEMS," says Joost van Beek.
On the Mems2Tune project, it is the research institute
Umec that is working with packaging. But Philips also has
a project underway on which attempts are being made to
construct packing directly on the silicon chip at the same
time as the MEMS circuit is manufactured.


The problem with the small switches is that dampness
and oxygen, for example, can cause the switches to
corrode and perhaps fasten in one position. But even
for other structures such as resonators, the packaging
issue is a problem.

Because of the resonators' smallness, the surrounding
air influences how well they oscillate. For this reason,
efforts are being directed towards including a vacuum
in the packaging.

Joost van Beek believes that there can be products that
employ RF MEMS technology on the market as early
as this year, but not from Philips. Which company that
will be first, he does not wish to reveal, but he believes
that the first products will be switches.

   Another company that has had projects with
RF MEMS is the antenna manufacturer Allgon.
But there, development has been on hold as of late.

"The market is more focused on the immediate future
and is today less inclined to take risks when we've had
a year without growth behind us. Functioning MEMS
at a reasonable price is further away and we are less
active just now when it comes to RF MEMS," says
Björn Berndtsson of Allgon Mobile Communications.

Motorola, which previously conducted MEMS
projects with Allgon, has developed a switch that is
now undergoing durability testing in their labs.

"We've demonstrated switches that manage ten billio
n switchings without failure," says Regina Cirmonova,
press spokeswoman at MotorolaSemiconductors'
product department in Geneva.

For the future, the dream among researchers is to be
able integrate MEMS and electronics on the same plate.

"It's hard to get it to work; each part of the production
chain weakens it," says Peter Enoksson, a professor in
microsystems at the Chalmers University of Technology
in Gothenburg, Sweden.

"The components today are often a combination, perhap
s in the same package. But we are looking at which ways
are best for integrating them, for example, as flip chips or
multi-chip models."

Página: www. radantmems.com
Realizado: Franco A Rivera C.
Asignatura: CRF

No hay comentarios:

Publicar un comentario