Successful flip chip assembly of large die with many bumps for applications such as image sensors depends upon the substrate and the die that is being placed upon it having parallel, coplanar surfaces when they are brought in contact. Any deviation from coplanarity can cause open or poor electrical connections and may even misalign the die by imparting a sideways sliding motion during placement pressure.

Two conventional approaches for establishing and checking coplanarity rely on optical or laser methods. An optical autocollimator projects an image onto the mirrored surface of the substrate, measuring the relative angle by the deviation of the reflected image from the projected one. Mechanical controls adjust the angle of the substrate until the images are superimposed at coplanarity.

Laser-based coplanarity adjustment depends upon a comparison of laser beam reflections from shiny targets at several locations on the substrate. Again, the substrate angle may be mechanically adjusted to achieve coplanarity.

Many flip chip bonders do not have either of these expensive add-ons. Establishing or verifying coplanarity then becomes a cumbersome trial-and-error procedure of making sample assemblies that can be removed and microscopically checked for coplanarity. This eliminates the expensive optical or laser equipment, but at the cost of considerable time consumed in repeated samples and adjustments. Verification that coplanarity is maintained through a production run means more samples

Sensor Products Inc. offers a pressure-sensitive film as a simpler, quicker, less costly method to establish the coplanarity of the two surfaces. The sensors are thin-film elements that measure contact pressure and distribution between any two mating surfaces. Placing pressure-sensitive film between the surfaces and bringing them into contact at the normal assembly pressure provides a permanent color-coded “footprint” of the pressure applied at every point.

The colors show the magnitude of the pressure. The uniformity of pressure across the contact area verifies coplanarity. The location and magnitude of any non-coplanarity is readily seen, for easy readjustment. Re-verifying during a production run simply means inserting another piece of film. 

About George A. Riley

George A. Riley has more than 20 years experience in microelectronics packaging, including 10 years developing and manufacturing flip chip assemblies for more than 50 customers. He is an industry consultant, educator, and web entrepreneur who has authored more than 90 papers and articles. He can be reached at (508) 753-3572, by emailing griley@flipchips.com or by visiting his website www.flipchips.com

By Jason Lomberg, Technical Editor

The Advanced Precision Kill Weapon System (APKWS), a laser-guidance platform from BAE Systems, scored four hits in four shots during the final phase of testing. APKWS adds precision laser guidance capabilities to 2.75 inch rockets—specifically, those found in attack helicopters.

The APKWS uses a form of the semi-active laser seeker known as DASALS (distributed aperture semi-active laser seeker). Installed between the warhead and the Hydra 70 rocket motor, the seeker aperture is divided into four elements, and each element is placed on the four wings of the guidance section. 

Little modification is required for the APKWS. According to the company, “the system requires no platform integration or aircraft modifications, and the mid-body design of its guidance section enables use of existing warheads, fuzes, and rocket motors.” APKWS can be fired from any helicopter that can launch 2.75-inch rockets, including the AH-1 Cobra, UH-1 Huey, OH-58 Kiowa Warrior, and AH-64 Apache.

Michelle McBride, co-project manager for the Navy's Airborne Rockets group, mentioned that, “The Navy is very pleased that APKWS has entered the final phase of testing and that we are nearing a Milestone C decision.” The Milestone C decision means that the system is ready for low-rate initial production. APKWS is the only laser-guided 2.75-inch rocket undergoing this full U.S. government evaluation.

APKWS couldn’t be timelier. Today’s enemy doesn’t play by our rules—insurgents use human shields, hide among the populace, and launch attacks from urban centers. In this vein, General Stanley McChrystal issued a highly-controversial memo banning close air support (CAS) except under “very limited and prescribed conditions.” APKWS probably won’t change policy, but it could go a long way towards limiting collateral damage. 

Is the search giant really making its own mobile device? Does it matter?

(Slate) - For months, tech blogs have been salivating over the possibility that Google will soon release a cell phone of its own. This discussion has always been a little strange. Google launched its mobile operating system Android two years ago, and we've seen the release of several Android devices since then, including the much-acclaimed Motorola Droid. So aren't all Android phones really "Google Phones?"

Not according to the rumor mill, which insists that Google—forced to deal with outside cell manufacturers and mobile carriers—has never had the chance to build the phone it really wants to build. The "real" Google Phone would be designed from top-to-bottom by Google, and the company would sell it directly to customers without any interference from cell carriers. "Like the iPhone for Apple," TechCrunch's Michael Arrington has written, "this phone will be Google's pure vision of what a phone should be."

Over the weekend, we got a hint that this mythic Google Phone might be for real. At the company's "all-hands" meeting on Friday, Google gave employees a slick new phone to try out. In no time, descriptions of the device began to leak out on Twitter. Later, the Wall Street Journal added more details: The phone is called the Nexus One, and though it will be manufactured by the Taiwanese company HTC, it will carry only Google's logo and will be sold online directly to consumers, not through a carrier. Thanks to Engadget, which managed to get a gallery of pictures, we also know that the phone is a real looker.

Google's only on-the-record statement is vague—on the company blog, it says that it gave employees the device so they can test out "new mobile features and capabilities." So is the Nexus One the true Google Phone, a device set apart from every other Android phone—or is it just the next incarnation of Android, a device meant to show off features that will soon be available on all phones that run Google's OS?

If Google really does plan to sell a phone that carries exclusive features—if it is really producing software that it won't share with other device manufacturers—the move would mark a huge rift in the wireless industry. Indeed, that's precisely why I'd argue that the Nexus One isn't anything special. It's entirely possible that Google will experiment with selling a phone directly to consumers, but I'd be shocked if the device did anything that other Android phones can't do.

Click Here for the rest of the article.

Editor's Note: One day we'll see a data commodities exchange, where all assets and bandwidth from all providers is sold to the highest bidder. We need to be careful that this does not become a venue for speculation.

(NetworkWorld) - Amazon's cloud computing division is unveiling an eBay-style auction service that will let users bid on unused virtual server capacity, potentially allowing customers to lower the cost of running applications on Amazon's Elastic Compute Cloud. Known as "Spot Instances," the price of this cloud-based server capacity changes based upon supply and demand, unlike Amazon's usual fixed prices for server instances.

"With Spot Instances, customers bid on unused Amazon EC2 capacity and run those instances for as long as their bid exceeds the current Spot Price," Amazon says. "The maximum price you specify is not necessarily the price that you will pay. For example, if you specify 50 cents as your maximum price and the Spot Price is 30 cents for the period, you will pay only 30 cents. If the Spot Price increases, you will pay the new price (until it exceeds your maximum, at which time your instances will be terminated)."

The fixed pricing scheme, which customers can still access instead of the auction-style pricing, ranges from about 8 cents per hour to $3 per hour.

Since Spot Instances can be terminated without warning, once a customer is outbid, they shouldn't be the only source of capacity allocated to enterprise applications that need 24/7 uptime.

Spot Instances come in the same sizes and software types as fixed-price virtual servers. Amazon says it developed the auction-style system in response to customers who said they wanted additional capacity at lower cost, and were willing to be flexible about when they run their applications.

Click Here for the rest of the article.

By Jason Lomberg, Technical Editor

Collateral damage has become synonymous with modern urban warfare. In fact, the concept of “total war,” where there is “less (or no) differentiation between combatants and non-combatants (civilians),” has existed since The Peloponnesian War. As far back as 431 BC, civilian casualties were considered an acceptable (and often desirable) outcome of total warfare. Today, the DOD defines collateral damage as lawful, “so long as it is not excessive in light of the overall military advantage anticipated from the attack.” This unfortunate trend has prompted development of the Autonomous Rotorcraft Sniper System (ARSS), essentially an unmanned surgical instrument of death.

 
The ARSS takes the RND Edge 2000 sniper rifle (replete with .338 Lapua Magnum rounds), mounts the rifle onto a highly-stabilized turret, and attaches the system to a Vigilante unmanned helicopter. The nexus of the ARSS, the Precisions Weapons Platform (PWP) lightweight turret, was designed by Utah State University’s Space Dynamics Laboratory (SDL). According to SDL, the PWP is “intended to integrate a sniper system onto a rotary–winged unmanned aircraft that will allow sniper capability on a platform that can achieve line of sight to a target by going above or around obstacles. This system could provide an accurate delivery of fire with little or no collateral damage, which is particularly useful in urban settings.”

The title Autonomous Rotorcraft Sniper System (emphasis mine) is a misnomer since the system is, at best, semi-autonomous. Auto-pilot handles the flying, while software assists the aiming. But it’s a human operator, using a modified Xbox 360 controller, who “pulls the trigger.” SDL describes their “operator’s station” as follows: “similar to a video game with video imagery from the cameras displayed on a flat–panel monitor. A gaming controller is used for aiming, arming safing, and firing tasks.” The ARSS doesn’t make critical decisions on its own—an important prerequisite of true autonomy. But with its surgical accuracy, the ARSS could greatly reduce collateral damage.

The RND Edge 2000 can fire up to 10 aimed shots per minute, so the ARSS can handle multiple targets in quick succession. Better still is the cost per trigger pull-- $4. By comparison, Hellfire missiles run about $68,000 each. The non-existent blast radius of the RND Edge limits collateral damage, but this comes with a tradeoff—the AGM-114 Hellfire II has a max range of 8,749 yd vs. 1,914 for the .338 round. ARSS will require more of an “up close and personal” approach.
Inevitably, the ARSS will draw the ire of groups who feel that video game violence desensitizes our youth. The inclusion of an Xbox controller will raise red flags. I don’t buy this argument (i.e. “When war is a videogame, it's much easier to pull the trigger.”). Similarly, I don’t believe robots make war more desirable. To paraphrase Patrick Lin, George Bekey, and Keith Abney, authors of Autonomous Military Robotics: Risk, Ethics, and Design, “these objections seem to imply that we should not make any improvements in the way we prosecute a war and, indeed, should return to more brutal methods (e.g., bayonets).” 

By Jason Lomberg, Technical Editor

Back at CES ’09, I had the chance to preview Hillcrest Labs’ MEMs Accelerometer-based motion-control technology. They’ve since marketed it as a consumer device, the Loop Pointer. Essentially an air mouse on steroids, the Loop Pointer is one of the coolest gadgets I’ve seen in a while.

The Loop Pointer is a disc-shaped device applicable in a wide range of media apps. There’s a left click (‘select’), a right click (‘back’), and a scroll wheel. But that merely describes a tethered mouse. The Loop Pointer’s bread and butter is its motion-sensing technology, powered by MEMs Accelerometers. Navigating your PC feels odd at first, but the “air surfing” quickly becomes second nature. Unlike the WiiMote and various other motion control devices, the Loop Pointer suffers no line-of-sight restrictions. Simple flicks of the wrist send the pointer across the screen.

As for the tech specs, the Loop Pointer is 122 mm x 30.5 mm (4.8” x 1.2”) and 4.9 oz, with a frequency of 2.4 GHz (2401 MHz to 2482 MHz) and a range of up to 10 m (33'). Hillcrest describes the Loop Pointer as a “mouse for your TV.” It “works on your TV when you use the TV as the monitor for a PC or Mac.” Lacking this set-up, I tested the Loop Pointer as a standard PC mouse and a PS3 remote.

The plug and play functionality is simple—connect the USB RF transceiver, put batteries in the Loop Pointer, turn it on, and you’re good to go. At first, navigating the Web feels odd without a tethered mouse. But you get used to it quickly. The Loop Pointer also serves as a makeshift PS3 DVD remote. Navigating through the PS3’s menus is a breeze. While playing a movie, flicking the Loop Pointer left or right will rewind or fast-forward, respectively. But the Loop Pointer is so sensitive that this is hard to control. It’s better simply to lock the device during playback.

I would highly recommend the Loop Pointer. And at $99, it’s a steal. You’ll never look at your mouse the same way again.

By Jason Lomberg, Technical Editor

The Air Force has confirmed the existence of a new Stealth Unmanned Aerial Vehicle (UAV). Designated the RQ-170 Sentinel, “The Beast of Kandahar" has been spotted over Afghanistan (see photo) as early as 2007. The military plans to deploy it for reconnaissance and surveillance missions.

Sources speculate that the RQ-170 will be roughly the same size and have similar payload capability as a Predator. This means it could lug from 500-1,000 lbs of surveillance equipment. The “Beast” was developed by Lockheed Martin's Advanced Development Programs.

Some critics feel the stealthy “Beast” is unnecessary in Afghanistan. Jamey D. Jacob, an associate professor of mechanical and aerospace engineering at Oklahoma State University, asks, “Why does the U.S. need to have a super secret stealth UAV in Afghanistan? The Taliban and Al Qaeda don't have radar seeking missiles we know of, so Predators and Global Hawks should work fine.” Dr. Jacob speculates that the RQ-170 may be useful against adversaries (like Iran) who do possess radar based ground-to-air missiles.

But Dr. Jacob is forgetting something. “Close air support” has become a political hot potato. On 6 July, 2009, General Stanley McChrystal issued a memo banning airstrikes (in Afghanistan) except, “under very limited and prescribed conditions.” If surveillance is the primary mission, and the “Beast” performs this role better, then the RQ-170 will take over recon duty.