Google

Saturday, February 16, 2008

Inventors unveil robot to fill car gas tank

Robotic arm fitted with sensors carefully opens car's flap, unscrews cap

Image: A car-fuelling robot

Dutch inventors on Monday unveiled a car-fuelling robot able to fill a vehicle's tank with gas without the driver needing to leave

EMMELOORD, Netherlands - Motorists nostalgic for the time they could sit tight while attendants braved the elements to fill their tanks may yet see those days return — compliments of a Dutch robot.

Dutch inventors unveiled on Monday a car-fuelling robot they say is the first of its kind, working by registering the car on arrival at the filling station and matching it to a database of fuel cap designs and fuel types.

A robotic arm fitted with multiple sensors extends from a regular gas pump, carefully opens the car's flap, unscrews the cap, picks up the fuel nozzle and directs it towards the tank opening, much as a human arm would, and as efficiently.

"I was on a farm and I saw a robotic arm milking a cow. If a robot can do that then why can't it fill a car tank, I thought," said developer and petrol station operator Nico van Staveren. "Drivers needn't get dirty hands or smell of petrol again."

He hopes to introduce the "Tankpitstop" robot in a handful of Dutch stations by the end of the year. It works for any car whose tank can be opened without a key, and whose contours and dimensions have been recorded to avoid scratching.

Asked whether he would trust his car to a robotic garage attendant, Jelger De Kroon, filling his black Alfa Romeo at a nearby gas station, said: "Why not? I guess I could keep my hands free and clean, but I'd hope they have good insurance."

New double-A battery outlasts competition

Panasonic’s Evolta earns certification from Guinness World Records

Image: Panasonic batteries

TOKYO - Japan's Panasonic is introducing a double-A household battery that will keep gadgets running 20 percent longer than rivals do, so long that Guinness World Records has dubbed it the world's longest-lasting alkaline battery.

The new Evolta — whose name is derived from "evolution" and "voltage" — has proved its mettle against products from Duracell and Energizer, as well as its own upscale Oxyride batteries, according to Matsushita Electric Industrial Co., which makes Panasonic brand products.

Evolta's longer life results from more material being packed inside it, the new materials it's made with and its sturdier seal than predecessors had.

Guinness certified it in Tokyo Tuesday as "the longest-lasting AA alkaline battery cell," based on testing under guidelines set by the industry's International Electrotechnical Commission.

Evolta's 10-year shelf life — up to 60 percent longer than rivals' — also makes it more attractive to store in disaster preparedness kits. Other batteries last five to seven years on the shelf, Panasonic officials said.

Evolta goes on sale in April in Japan, and is planned for overseas markets later this year, Matsushita said. A pack of four will cost about $5.40 in Japan, about 15 percent more than regular batteries and 3 percent more than Oxyride. Prices outside Japan aren't set.

Multi-core chips make computers run faster

Parallelism allows industry to keep up with Moore’s Law


PCs will continue to get faster, in accordance with Moore's Law, but it won't be like the old days when vendors brought out successive generations of microprocessor chips that ran at faster and faster speeds, to the delight of the users.

"They've hit the wall because of heat," Tom Halfhill, senior analyst for the Microprocessor Report newsletter in San Jose, Calif., told LiveScience.

Currently, chip speeds have topped out at a little under 4 gigahertz (4 billion cycles per second) because they get hotter as they run faster, and at higher speeds they fry themselves.

Chip-makers respond
The industry's answer: adding more "cores" (i.e. copies of the processor circuitry itself) that run at today's speeds. Vendors are offering two, three and four processors per chip, while eight-core versions are expected.

The cores run in parallel, Halfhill explained, and the machine's operating system (often modified Windows or Linux) knows how to divide up its tasks between the cores, to enhance overall performance, and, of course, keep up with Moore's Law.

Formulated decades ago by Intel co-founder Gordon Moore, Moore's Law states that the power of computer chips — and therefore computers — can be expected to double every other year.

The problem, Halfhill added, is that typical end-user applications, such as word processors, spreadsheets or browsers have no facilities for using multiple cores.

"Most application programs were written to run on one processor, and adding parallelism is not trivial," Halfhill noted. "It is happening slowly, but not all desktop programs can run in parallel — and if you are just typing with a word processor, what difference does the extra speed make?"

Reconfigurable PCs
Beyond multiplying, the cores are also eventually expected to specialize, and even let PCs reconfigure themselves on the fly, predicts analyst Rob Enderle, head of the Enderle Group in San Jose, Calif. If the user is running a video game, some of the cores would convert themselves into graphics processors to handle the workload. If the PC is handling computation-heavy jobs like encryption, they would convert into plain-vanilla processors.

Antivirus software could move into one core, wall itself off from infection and from there monitor the rest of the system, he added.

"All in all, we should be able to look forward to more intelligence in our applications, especially from the parts that operate in the background, such as spell-checkers or virus-checkers," Enderle said. "The more intelligence you can build in, the more likely it is to do something amazing."

But software will take serious rewriting if the chips are to have more than eight cores, said Halfhill, or many of the cores are likely to remain idle.

However, processor chips with multiple, fully employed cores have become the norm in other fields, he added. Chips that handle Internet traffic, cell phone tower traffic and cryptography typically have scores if not hundreds of cores, he noted, adding that each core is typically less complex than a PC processor.

'Green' robot self-propels through sea

Glider has crisscrossed Virgin Islands Basin more than 20 times

WASHINGTON - A seagoing glider that uses heat energy from the ocean to propel itself is the first "green" robot to explore the undersea environment, U.S. researchers said on Thursday.

They said the glidler had crisscrossed the 13,000-feet-(4,000-meter-)deep Virgin Islands Basin between St. Thomas and St. Croix more than 20 times since it was launched in December.

And it could keep going on its own for another six months, the team at the Woods Hole Oceanographic Institution and Webb Research Corporation in Falmouth, Massachusetts, predicted.

"Gliders can be put to work on tasks that humans wouldn't want to do or cannot do because of time and cost concerns," Dave Fratantoni of Woods Hole said. "They can work around the clock in all weather conditions."

Such robots can carry sensors to measure temperature, salinity and biological productivity.

They usually surface from time to time to fix their positions using the Global Positioning System and to communicate via Iridium satellite to a laboratory.

Most gliders rely on battery-powered motors and mechanical pumps, the researchers said. This one draws its energy from the differences in temperature between warm surface waters and the colder, deeper layers of the ocean.

"We are tapping a virtually unlimited energy source for propulsion," Fratantoni said.

He said data collected by the glider would help researchers understand how eddies in the region affect ocean circulation and move around the larvae of fish as well as pollutants.

Vision of the future seen in bionic contact lens

Prototype includes light-emitting diodes, basic wiring and a tiny antenna

Thumper has seen the future.

Researchers at the University of Washington have created the prototype for a bionic contact lens — recently tested on rabbits — that includes light-emitting diodes, basic wiring for electronic circuits and even a tiny antenna. Future versions, the scientists believe, could serve as a flexible plastic platform for applications such as surfing the Internet on a virtual screen, immersing gamers in virtual worlds and monitoring patients’ medical conditions.

Babak Parviz, an assistant professor of electrical engineering at the University of Washington, said he and his collaborators began by thinking about contact lenses and their normal purpose of correcting vision. What if his group’s collective expertise in nanotechnology and microfabrication could transform the lenses into something else entirely?

Adding displays directly onto the lenses, visible to the wearers but no one else, could project critical information onto windshields for drivers or pilots or superimpose computer images onto real-world objects for training exercises. And with a wireless connection to the Internet, the lenses could allow bus or train riders to surf the Web on virtual screens suspended in midair or pave the way for gaming enthusiasts to immerse themselves in virtual worlds with no restrictions on their range of motion (although perhaps adding a further nuisance for fellow commuters).

A boon for mobile devices
If successful, the bionic lens could prove a huge boon for mobile-device manufacturers.

“One of the problems is that we can make the electronics smaller and smaller, and then the user wants to interface,” Parviz said. “A really tiny display is not useful.” But putting that display directly onto the user’s contact lenses would effectively solve the problem of size and allow personal electronics to continue shrinking.

Whether a future iPod will come equipped with a bionic contact remains to be seen — literally — but a lens with a basic display could be ready in the near future.

Parviz said the health care field also might benefit from the technology. “How do we constantly monitor someone’s health?” he said. “It turns out that a lot of indicators that tell if a person is healthy or not show up on the surface of the eye.”

A biosensor-equipped lens could provide a non-invasive way of gleaning that information and sending it on to a database or serving as a relay station for data or power from retinal implants designed to correct vision problems.

Microfabrication technique
One of the first big obstacles for the team was resolving the fundamental incompatibility between the fabrication process for microchips and light-emitting diodes and the types of polymers used for contact lenses. To get around the issue, the researchers first constructed electronic circuits from ultra-thin metal layers — each only one-thousandth the width of a human hair — and fashioned diodes so small that nearly 100 could fit within an inch.

On the lens itself, the researchers created multiple receptor sites that each attracted a separate component by exploiting the same capillary forces that push water up through a plant’s roots. This microfabrication technique allowed the tiny parts to self-assemble on the surface of the lens and bind themselves together to form the different devices. Although circuitry covers much of the current prototype’s surface, Parviz said there should be enough room on the periphery of the lens to ensure that future nano-gadgetry wouldn’t obstruct a person’s view.

For the prototype, the group successfully integrated an antenna, tiny metal wires for an electronic circuit, and red light-emitting diodes onto the lens surface. Harvey Ho, a former graduate student in Parviz’s lab, presented the research Thursday at the Institute of Electrical and Electronics Engineers’ international conference on Micro Electro Mechanical Systems in Tucson, Ariz.

Focusing on image quality
Some scientists have been less gung ho. Daniel Palanker, a retinal implant expert at Stanford University, questioned the ability of a display generated by the contact lens to produce a sharp image on the retina of its wearer’s eye, noting that the normal focal distance for seeing objects clearly is about 25 centimeters in front of a person’s eye.

But Glenn Chapman, a professor in the School of Engineering Science at Simon Fraser University in Burnaby, British Columbia, said researchers could overcome that obstacle by precisely adjusting the angle of incoming light emitted by diodes on the contact lens.

Assuming the light beam is high-quality, he said, correcting the beam's incoming angle could make up for the cornea's lack of focusing ability and instead allow the transparent crystalline lens behind the eye's iris to focus the image onto the wearer's retina. Of course, contact-wearing rabbits won't be able to tell researchers when they've hit upon the right angle to produce a crisp image, Chapman said, but an artificial eye overlaid with the lens could do the trick.

Parviz said his team also will try to pair microlenses to each pixel in a display created by the contact lens, hopefully manipulating the image and changing its perceived location in such a way that the viewer would have the feeling of seeing an in-focus picture suspended in midair. But he agreed that the challenge will be a complicated one. "It's unprecedented," he said. "No one has ever tried to form an image right on the surface on the cornea."

As for the microfabrication process, Chapman said he was impressed with how the self-assembly technique allowed the University of Washington researchers to basically float into position metal pieces that normally don't adhere well to plastic surfaces. The technique had been developed previously by Parviz and others, he said, "but it's a nice application of it."

And with further advances in the microfabrication field, Chapman said, the potential is "very high" for wires that are essentially invisible to the human eye and for even tinier organic light-emitting diodes.

'Like a normal contact lens'
If size isn’t necessarily a limitation, adequate power could be. Parviz said his group is now working on the issue of how to run displays or biosensors without the need for awkward batteries. So far, the prototype’s lens-mounted antenna has shown promise in collecting radio frequency waves and turning them into useful energy.

If all goes well, putting in or taking out the bionic lens should be as easy as popping a regular one in or out, he said. “It should feel like a normal contact lens. It should be completely smooth against the surface of the eye.”

Which isn’t to say the lens is inconspicuous. “If you look into the rabbits’ eyes, you would notice that something is going on,” Parviz said.

Nevertheless, the rabbits tolerated the bionic lenses well during their 20-minute fittings, though none of the systems have yet been switched on. The group has yet to seek permission for the necessary safety trials in humans.

If safety and engineering issues are addressed, future iterations could perhaps be engineered to camouflage the circuitry, thus sparing bionic lens-wearing commuters the stares of passersby swearing they’d just seen the Terminator or a visor-less Geordi La Forge from “Star Trek: The Next Generation.”

A 'revolution' in batteries


Nature Nanotechnology / Stanford
Photomicrographs show silicon nanowires before and
after charging (left and right, respectively).

If you've ever rushed to save your files before your laptop battery gave out, or scrambled to recharge your iPod, or wished out loud for the resurrection of the electric car ... relief is in sight.

Yet another battery breakthrough is on its way to market, taking its place alongside improved hybrid-electric vehicles, the promise of ultracapacitor systems and even better AA power cells. Next-generation batteries could well last several times as long as current power packs, thanks to nanotechnology.

"This idea will have a really high impact on battery technology," said Stanford chemist Yi Cui, who is the lead researcher behind a study appearing in this month's issue of Nature Nanotechnology. "This is really revolutionary."

The key innovation involves using silicon nanowires instead of the usual carbon to store energy in a lithium-ion battery's anode.

Silicon has more than 10 times as much charge capacity as carbon. If commercial batteries could live up to that performance level, you could theoretically be running your laptop for 20 to 40 hours straight rather than the typical two to four hours. An electric car could go 400 miles on a charge rather than 40 miles.

Of course, the reality is more complex than the theory. But more about that later. The first question is whether this technology is actually for real. If silicon is that good at storing electrical energy, why isn't it being used already?

That's where nanotechnology makes the difference: For years, engineers have been trying to harness silicon electrodes for battery applications. But the problem with silicon is that its volume bulks up by a factor of four when you add the lithium - and then shrinks by the same factor when power is extracted. That quickly pulverizes an electrode made of silicon film or particles, rendering the battery useless.

Cui and his colleagues took a different approach: They grew nanowires of silicon directly on a stainless-steel plate. Each wire was about 90 nanometers wide, or a thousandth of the width of the typical human hair. When the filaments were filled with lithium-ion power, they thickened up and lengthened into curls, like tiny spongeworms - but they retained their resiliency through dozens of power cycles.

"This idea really made these silicon materials possible to be used in battery technology," Cui said.

Challenges still lie ahead: First of all, Cui's team focused on retooling the anode, which is just one of the electrodes in a battery. To get the full tenfold improvement, Cui told me, "you would need to improve also the other electrode ... but with one electrode improvement, you can improve a lot already." For example, you could make the anode smaller, leaving more space for a bigger cathode.

Cui's team also found that there was a one-time capacity drain after the first charge. But that's no biggie. The nanowires' storage capacity was still about eight times higher than carbon, Cui said. "This won't prevent this technology from going forward," he said.

On the plus side, silicon-nanowire batteries wouldn't have to look like the battery bricks that are typically used in laptops or cell phones. "It's a fundamentally different structure from the current technology," Cui said. And that could result in batteries that are better-shaped to conform to the available space.

Cui said a patent application has been filed for the technology, and he's considering starting up a company to commercialize the concept. So when might silicon-nanowire batteries hit the market? "I'm thinking in the next three to five years," Cui said.

Some companies are already knocking on the lab door. Cui acknowledged that Tesla Motors, the company working on an all-electric sports car, is just one of the outfits expressing interest. "There are lots," Cui told me, "but it's better not to mention their names now."

Sunday, November 11, 2007

Analog companion IC melds power, audio functions for wireless handhelds

The AT73C206 analog companion power supply and audio interface IC from Atmel Corporation integrates a high-quality audio interface and all the necessary power management functions for handheld wireless communication/multimedia devices on a single chip.

Power functions include nine LDOs, one high-current DC/DC converter, a standalone continuous battery charger with integrated pass transistor, a four-channel white LED driver and a self-contained RTC.

The AT73C206's voice/audio analog interface comprises a 16-bit voice codec, 16-bit stereo audio DAC with headset driver and a 300-mW speaker driver. Packaging is in a 9 x 9-mm, 96-ball ultra-thin BGA.

Pricing and availability: The AT73C206 is priced at $2.90 in quantities of 100K units. A reference design and samples are available now.

Atmel Corporation, 408-441-0311, www.atmel.com

Product literature:
AT73C206 data sheet (summary) (PDF)
AT73C206 data sheet (complete) (PDF)

Single-supply multiplexers save board space, cost

Milpitas, Calif. — Intersil Corp. has released two single-supply, triple 4:1 buffered multiplexers for use in high-end consumer video products. The ISL59451 offers an integrated DC-restore feature and the ISL59452 is designed for AC-coupled systems.

Intersil said the ISL59451 and ISL59452 devices are the first single-supply, buffered triple 4:1 multiplexers in the industry. This feature set is important for consumer customers designing systems where video input is the only feature requiring a negative rail, said the company. By eliminating the need for a negative rail, these ICs will help simplify the design of products such as projectors, high-definition TVs and broadcast equipment, as well as save board space and cost for original device manufacturers (ODMs).

The ISL59451's integrated DC-restore feature lifts the video signal above ground and into the operating range of the multiplexer, ensuring the video is in the correct voltage range for processing. For systems using an AC-coupled design, the ISL59452 is offered without the integrated DC-restore feature and at a slightly lower price.

Using these parts, customers can choose between AC or DC coupling video over the 0-V to 5-V range, while maintaining excellent video characteristics from the multiplexer that are usually associated with ±5-V performance, said Intersil. These parts can also be used as an input stage, where multiple channels of video are coming into a device (such as a projector) but only one channel needs to be selected at any one time.

Pricing: The ISL59451 is available in 5 x 5 QFN packaging in quantities of 100-999 for $6.49. The ISL59452, offered in the same packaging and quantities, is priced at $4.88
Product information: ISL59451 and ISL59452

Intersil Corp., 1-408-432-8888, www.intersil.com

Toshiba expands motor controller line

San Jose, Calif.—Toshiba America Electronic Components Inc. (TAEC) added three products to its line of motor drivers: a stepping motor driver (TB6560AHQ/AFG) and two H-bridge DC motor drivers (TB6559FG and TB6593FNG).

The TB6560AHQ/AFG is designed to control the two-phase bipolar stepping motors used in printers, where the driver's two-phase operation can provide smoother transition between motor speeds to prevent paper jams.

The TB6560AHQ/AFG is a pulse width modulation (PWM) controlled, chopper-type, sinusoidal, micro-step bipolar stepping motor driver. It supports the excitation modes 2-phase, 1-2-phase, 2W1-2-phase and 4W1-2-phase, as well as forward/reverse mode. The TB6560AHQ/AFG provides a low vibration, high-performance drive for two-phase bipolar type stepping motors using only one clock signal. The driver was developed in Toshiba's BiCD (bipolar/CMOS/DMOS) process.

The TB6559FG and TB6593FNG DC motor drivers target applications such as vending machines, gaming machines, printers and small appliances.

The TB6559FG full-bridge DC motor driver IC offers support for two motor control-operating modes, either internal constant current PWM (for Vref between 0V and 3V) or direct PWM (for Vref between 4.5V and VREG). The TB6593FNG motor driver IC has a very low R(on) of 0.35 ohms, which allows the device to provide high performance while drawing power at levels similar to a motor driver intended for battery-powered applications. The TB6559FG operates with an output current of up to 2.5 amps (peak) and supply voltages of up to 30 volts. The TB6593FNG operates with an output current of up to 3.2 amps (peak) and supply voltages of up to 13.5 volts. These devices were also developed in Toshiba's BiCD manufacturing process.

Pricing: The evaluation price in 10,000 unit quantities for the TB6560AHQ/AFG is $3.30 each, $0.62 for the TB6559FG and $0.66 each for the TB6593FNG.
Availability: Engineering samples of the TB6560AHQ/AFG are scheduled for availability in the fourth quarter 2007, with mass production scheduled to begin by end of year. Both the TB6559FG and TB6593FNG are currently in mass production.
Product brief: click here.

Toshiba America Electronic Components Inc., 1-949-623-2900, www.toshiba.com/taec

Software helps verify A/D convertors


Jivaro tools apply user-controllable reduction techniques upon parasitic data files obtained from layout extraction tools

EdXact's Jivaro netlist reduction tool was chosen by e2v, a developer and manufacturer of specialised components and subsystems. e2v is using Jivaro to help verify its A/D convertors, developed using Jazz Semiconductor's SiGe BiCMOS 0.18-micron process. The new convertor family includes quad 8bit devices capable of delivering conversion rates from 1.25 to 5GSps from a single chip at 1W per channel.

EdXact provides software solutions that help to accelerate the overall backend IC physical verification cycle.

Jivaro tools apply user-controllable reduction techniques upon parasitic data files obtained from layout extraction tools.

This helps simulation tools to cope with huge data files while still maintaining accuracy.

Jivaro was chosen to help accelerate postlayout verifications on the A/D convertors' design On that A/D mixed signal design, Jivaro-enabled simulations proved to execute three to 10 times faster, with the same level of accuracy, allowing more verifications to be executed and larger blocks to be verified, while still meeting tape-out delivery schedules.

Multiple simulation runs increase verification confidence, which is crucially important during the tape-out process.

Moreover, Jivaro's ease of use allows quick adoption of the tool by designers.

'Thanks to Jivaro, we were able to meet our just in time constraints and to increase our verification targets as well' according to D Boisgontier, manager of e2v's Broadband Data Convertor Development Team.

Edxact: contact details and other news
Email this article to a colleague
Register for the free Electronicstalk email newsletter
Send a comment on this recent news story to the Editor
Electronicstalk Home Page