acquisitionZONE Products for the week of November 18, 2002
Gyroscopes are used to measure the rate at which objects rotate. The information provided by gyros can be used to trigger automobile airbags during rollover, improve the accuracy and reliability of GPS/navigation systems, and stabilize moving platforms such as automobiles, airplanes, robots, antennas, and industrial equipment.
"Analog Devices' ADXRS gyro has a very low noise output which enables us to use it in many applications to augment GPS receivers where critical location information is required during temporary disruptions of GPS signals," said Michael S. Perlmutter, Executive Vice President, Fibersense Technology, one of the world's leaders in advanced inertial navigation and stabilization technology based in Canton, MA. "In our product designs, reliability is critical and immunity to shock and vibration is a must. Competitive solutions were not able to match Analog Devices in terms of size, performance or reputation for MEMS reliability." For greater reliability assurance, the new device is also the only commercially available gyro with a full mechanical and electronic self-test feature that can operate while the sensor is still active.
By developing the new gyro on the iMEMS process, Analog Devices leverages over 10 years of MEMS manufacturing experience. "Analog Devices is the industry leader in the MEMS accelerometer market today due in large part to the world class quality and reliability delivered by the iMEMS process," said Franklin Weigold, vice president and general manager of the Micromachined Products Division, Analog Devices. "Using the iMEMS process, we can offer our customers a roadmap that will provide high quality, high performance gyros for as little as $10."
Patented design techniques that leverage the iMEMs process capabilities
have resulted in a MEMS sensor that provides extraordinary immunity to shock
and vibration. The ADXRS gyro delivers stable output in the presence of
mechanical noise up to 2000 g over a wide frequency range. This unique capability
is important in applications such as automotive rollover detection, where
a rollover event must be accurately detected while the vehicle is experiencing
shocks and vibrations induced by collisions with other objects and off-pavement
surfaces.
analogZONE Says . . .
Lee Goldberg
I'm not sure which part of ADI's announcement is more interesting, the integrated gyro itself, or the technology that produced it. The ability to put signal processing electronics (and eventually perhaps more) on the same chip as a MEMS structure opens up new possibilities for products that would have been too bulky, too costly, or even impossible to produce otherwise. While I have neither the time nor the expertise to expand on the possible applications of integrated MEMS/Electronics chips at the moment, I will speculate that they could spawn new generations of medical products, improve the state of the art of industrial controls, and perhaps have applications in smart homes, energy management, and other areas I cannot foresee.
For the moment then, I'll concentrate on the chip itself, and its potential applications. For some more insights into the actual workings of the device, see the following section which contains analysis from analogZONE's sensor specialist, Paul McGoldrick. I agree with ADI that its highest volume market (at least for now) will be in the automotive industry, doing relatively mundane things like supplementing the navigation information in GPS systems and detecting rollovers to deploy airbags in the correct sequence, but lots of other really interesting applications lie on the horizon.
In the automotive market, these little devices will probably be used in souped-up anti-skid system known in the industry as dynamic vehicle control. My friend Andy has something like this in his Corvette, and it will deliberately alter the braking on selected wheels to keep the car optimally glued to the road under even the most ridiculous conditions. Once the gyro's cost comes down to the point where auto makers can use them in quantity, the accuracy and multi-axis capability that the unit delivers should enable designers to create systems that sense not only lateral acceleration, but also body roll, yaw around the CG, and maybe even vertical motion from wheel hop or bumps! I imagine driving a car with a fully active suspension and drive-train might be unnerving to the uninitiated, somewhat akin to riding a living creature, as the car slithers, squirms, twitches, and holds the road with a mind of its own, to keep the vehicle traveling where you point the steering wheel.
Then, of course, there are the unusual lower-volume applications that ADI is not concentrating on, that I find ever-so-much-more fascinating. Being a pilot, the obvious things that come to mind are aviation-related. While GPS is great for telling you where you are and where you are headed, it is really poor at telling you much about where you are pointed. And unless you happen to be flying a multi-million-dollar aircraft equipped with a ring laser gyro, or other inertial navigation system, you are most likely relying on a mechanical gyroscopic compass to know which way you are pointed, and a gyro-equipped artificial horizon to know where your nose and wings are relative to the ground. Being a high-precision device whose basic technology dates from the 40s or earlier, it is no wonder that gyroscopic flight instruments are not terribly reliable, very susceptible to vibration and aging, and always rather expensive.
The ADI gyro provides an excellent alternative to a mechanical gyro that might address all the above shortcomings - including price. Analog Devices accelerometers have already been incorporated into at least one solid-state gyro, manufactured by Crossbow Systems. In fact, analogZONE even reported on a low-cost digital cockpit for home-built aircraft that employed one of Crossbow's MEMS-based Attitude and Heading Reference Systems. While it worked rather well and did not exhibit some of the nasty habits of traditional units, the unit was rather costly ($2000 to over $4000, depending on options), which gave it little or no cost advantage over a civil aviation-grade gyro.
Much of that expense lies in the fact that Crossbow had to add the signal processing electronics to the raw MEMS sensors. Now, with integrated electronics, and some other improvements for better accuracy and calibration, this new part should cut the cost of an aviation-grade gyro down to $750-$1000, or even much lower, if sufficient volumes are produced. This opens up lots of possibilities including replacing the traditional "ball and needle" attitude and direction instruments with much more reliable and precise products at very competitive prices. Such a gyro system could also help to greatly improve the accuracy and quality of the information from GPS-based navigation systems, and perhaps form the basis of a simple, low-cost autopilot system. I look forward to the day that I'll be able to fly my next homebuilt with a panel that borrows heavily from F-16s and 747s.
Of course, a cheap, reliable gyro platform would also lend itself to autonomously-piloted vehicles too - both winged and wheeled. While it won't involve the huge unit volumes of the automotive industry I imagine that the military will be buying lots of these devices to support its current love affair with remotely-piloted vehicles and "smart" weapons. One can take some consolation that they will improve the accuracy and reduce the cost of the devices we are using to deliver pinpoint judgment upon our enemies.
But lest we be overly morbid, there are lots of commercial applications for these sensors as well. While smaller in volume on a per-product basis, Dan Krakauer, ADI' s program manger for the iMEMs program speculates that there will be thousands of lower-volume applications that have not even been conceived of yet. One area he finds promising is that of stabilization platforms, which might be used to steady cameras, antennas, or even work areas against the pitching and rolling of a moving plane, land vehicle, or ship. There are probably some other exciting areas where the gyros could be used to generate navigation information for small, remotely-operated vehicles that are used to inspect and repair pipelines, search for survivors in collapsed buildings, and explore the ocean - all places that GPS signals cannot penetrate.
With an initial price of $30 each in 1000-piece lots, and a projected target of under $10 in five years, it's most likely we'll be seeing lots of these devices popping up in expected, and unexpected places.
|
Comments From Paul McGoldrick…
At my last count of the listings there were 148 organizations developing commercial MEMS technology. While the majority of those that have successful commercial models believe that the largest profits will be in the non-military sector, there is a huge military interest in the technology for radar, radio, steerable antennas and inertial measurement units. There are also extreme uses such as guidance on a Howitzer shell, which requires the ability to withstand forces up to 75,000 G. The materials that have been, and are being, looked at in the MEMS arena include plastic, aluminum, composites, and piezoceramic. However, the use of semiconductor processes - commercially launched by Analog Devices - offers advantages that are unobtainable from other materials: The ability to include the processing and control circuitry on the same die.
Talking with Craig Core (ADI's Wafer Fab Manager in the Micromachined Products Division) was an interesting experience in understanding how all the manufacturing processes you have learned about CMOS are turned on their heads for the purpose of producing mechanical objects with fixed fingers, moving fingers (with supporting beams), tethers and anchors - looking like multiple sets of tuning forks intertwined. And yet, on the same die, there are conventional circuits looking after the mechanical area's needs and taking its outputs for processing. The detectable capacitance change in ADI's MEMS process is 12 zF (10e-21 F), with a 30 µG resolution and a 0.00016 angstrom detectable detection. And yet it can take 3000 G at 15 kHz!
Draper Labs pursued the MEMS gyro concept in about 1990 (but without integration) although Bendix tried an integrated approach about ten years earlier. ADI started development of an integrated MEMS gyro in 1994 and showed a sample off at a MEMS conference in 1998.Customer samples went out in 2000 and now the first two products, the ADXRS150 (150º/s) and ADXRS300 (300º/s), are shipping; they are functional up to 10,000º/s. A later product will offer 75º/s performance.
The devices use the accelerometer performance that has made MEMS so marketable for ADI and add Coriolis* acceleration detection to provide a signal proportional to the rate of rotation. The key difference between an accelerometer and a gyroscope is that in the latter proof mass must be in motion to generate a Coriolis signal. Although they are 2.5-V devices (with a 5-V rail) the products have a voltage multiplier on-chip to provide the 12 V needed for the gyro body to obtain a good noise floor.
Separate electrodes are used for drive sense and feedback while the proof mass electrode is shared with all three. The mechanical element (dither frame) in the sense mode is used as a noise-shaping one-bit sigma-delta filter. Resonance of the dither frames is achieved electrostatically and the Coriolis effect displaces the sensing structures The drive direction and velocity are amplified and shaped and used to switch the demodulator in the amplified Coriolis acceleration signal path. The latter is then further amplified to the output. Force feedback is applied back to the Y/sense dynamics. Regulators, bias, and a temperature sensor (to allow for temperature coefficient calibration) are all included on-chip as is trim and test for BIST.
ADI, according to David Krakauer (Gyro Program Manager), is focusing on vehicle rollover detection systems, enhanced GPS performance in poor reception locations, robotics (including toys), avionics and antenna stabilization. Other markets, not spelled out, are undoubtedly personal safety systems for people who are physically at risk from falls, camera stabilization, security systems and similar platforms.
The devices are in a 32-ball BGA with a 30 mW power consumption. Obviously, compared to any other gyro in the market the size and the performance-for-size are an incredible jump in the history of the gyroscope.
The basic design of the gyro appears to originate from work at UC Berkeley under DARPA F30602-97-2-0266 (Professors Bernhard E Boser & Roger T Howe) where two designs of X-axis rate gyros have been implemented. Devices in the second generation (with Coriolis effects along the Z-axis) were fabricated by Sandia National Labs and Berkeley Microlab before being submitted to ADI for fabrication in MEMS.
Data Sheet ADXRS150
Data Sheet ADXRS300
*The Coriolis Effect, often incorrectly described as a force, was postulated by Gustav-Gaspard Coriolis, a French mathematician/engineer, in 1835. He showed that if Newton's laws are to be applied to rotating objects there would be apparent trajectory errors in objects not on the Earth's surface. This would be most noticeable in an object leaving, airborne, one of the Earth's Poles with a destination on the equator (the fastest moving line of points on the earth's surface.) By the time the projectile reached the equator the earth would have moved Eastwards and it would miss its target. The path taken was a straight line but to an Earth observer it would appear that the projectile had veered to the right in the Northern Hemisphere and to the left in the Southern Hemisphere. (Similar affects can be attributed to the Earth's other two axes of movement: Around our Sun and by our movement through the Milky Way.)
The right/left rule applies to anywhere on the respective Hemispheres and the effect can be clearly seen in the movement of clouds and other weather patterns, as well as in oceanic waters. The angular speed of the earth at 360º/sidereal day (approx. 72.7 microradian/s) is insufficient for the effect to be noticed on small objects (like water running out of a bathtub.)
Abstracts From UCB Work:
http://buffy.eecs.berkeley.edu/IRO/Summary/01abstracts/aseshia.1.html
http://buffy.eecs.berkeley.edu/IRO/Summary/01abstracts/dubravka.2.html
http://buffy.eecs.berkeley.edu/IRO/Summary/01abstracts/elliot.1.html
http://buffy.eecs.berkeley.edu/IRO/Summary/01abstracts/sunil.1.html
See also "A Monolithic Surface Micromachined Z-Axis Gyroscope with Digital Output," by Jiang, Seeger, Kraft and Boser.