Microsoft Surface Book Teardown Reveals Muscle Wire Mechanism

November 4, 2015 by Richard Baguley 27 Comments

It’s hard to resist the temptation to tear apart a shiny new gadget, but fortunately, iFixIt often does it for us. This helps to keep our credit cards safe, and reveal the inner workings of new stuff. That is definitely the case with the Microsoft Surface Book teardown that they have just published. Apart from revealing that it is pretty much impossible to repair yourself, the teardown reveals the mechanism for the innovative hinge and lock mechanism. The lock that keeps the tablet part in place when in laptop mode is held in place by a spring, with the mechanism being unlocked by a piece of muscle wire.

We are no strangers to muscle wire (AKA Nitinol wire or Shape Metal Alloy, as it is sometimes called) here: we have posted on its use in making strange robots, robotic worms and walls that breathe. Whatever you call it, it is fun stuff. It is normally a flexible wire, but when you apply a voltage, it heats up and contracts, much like the muscles in your body. Remove the voltage, and the wire cools and reverts to its former shape. In the Microsoft Surface Book, a single loop of this wire is used to retract the lock mechanism, releasing the tablet part.

Unfortunately, the teardown doesn’t go into much detail on how the impressive hinge of the Surface Book works. We would like to see more detail on how Microsoft engineered this into the small space that it occupies. The Verge offered some details in a post at launch, but not much in the way of specifics beyond calling it an “articulated hinge”.

UPDATE: This post was edited to clarify the way that muscle wire works. 11/4/15.

E-Ink Display Moves On-Screen Controls Off-Screen

August 26, 2015 by James Hobson 37 Comments

Tablet computers have come a long way, long way. It finally seems like they’ve found their niche in the market, and now maybe they can catch up to more traditional computers. The Microsoft Applied Sciences division came up with a cool prototype design for a new tablet, one with a secondary e-ink input display.

The tablet interface makes use of e-ink strip above the keyboard. While it might not seem like much, this frees up a bunch of screen real estate, allowing you to have various icons and shortcuts off screen. It makes a ton of sense for digital artists as they can draw on the screen, but also have their toolkit open right below them — almost like real painting/drawing.

One of the other great uses for something like this is a signature pad — with everything going digital, when is the last time you had to print, sign, and scan a document back to someone? They even developed a dedicated email app you can use solely on the e-ink screen, allowing you to maximize the use of your main screen for something like a video chat.

The demo is pretty cool, and we often wonder why there aren’t more phones with e-ink displays integrated into them — is this just the beginning?

Continue reading “E-Ink Display Moves On-Screen Controls Off-Screen” →

Drivers For 3D Printers And Why We Need Them

August 20, 2015 by Brian Benchoff 88 Comments

Manufacturers of 3D printers have a lot to do before they catch up with makers of the cheapest 2D, paper-based printers. If you’ve ever taken an inkjet apart, you’ll most likely find some sort of closed-loop control on at least one of the axes. The 2D printer will tell you when you’re out of ink, when a 3D printer will go merrily along, printing in air without filament. File formats? Everything is Gcode on a 3D printer, and there are dozens, if not hundreds of page description languages for 2D printers.

The solution to some of these problems are drivers – software for a 3D printer that slowly consumes the slicing of an object, printer settings, and placing an object on the bed. It’s coming, and the people who are responsible for making your 2D printer work with your computer are busy at work messing up the toolchain for your 3D printer.

The latest version of CUPS (C Unix Printing System) adds support for 3D printers. This addition is based on meetings, white papers, and discussions in the Printer Working Group (PWG). There has already been a lot of talk about what is wrong with the current state of 3D printer toolchains, what can be improved, and what should be completely ignored. Let’s take a look at what all of this has accomplished.

Continue reading “Drivers For 3D Printers And Why We Need Them” →

The Machine That Japed: Microsoft’s Humor-Emulating AI

August 17, 2015 by Kristina Panos 18 Comments

Ten years ago, highbrow culture magazine The New Yorker started a contest. Each week, a cartoon with no caption is published in the back of the magazine. Readers are encouraged to submit an apt and hilarious caption that captures the magazine’s infamous wit. Editors select the top three entries to vie for reader votes and the prestige of having captioned a New Yorker cartoon.

The magazine receives about 5,000 submissions each week, which are scrutinized by cartoon editor [Bob Mankoff] and a parade of assistants that burn out after a year or two. But soon, [Mankoff]’s assistants may have their own assistant thanks to Microsoft researcher [Dafna Shahaf].

[Dafna Shahaf] heard [Mankoff] give a speech about the New Yorker cartoon archive a year or so ago, and it got her thinking about the possibilities of the vast collection with regard to artificial intelligence. The intricate nuances of humor and wordplay have long presented a special challenge to creators. [Shahaf] wondered, could computers begin to learn what makes a caption funny, given a big enough canon?

[Shahaf] threw ninety years worth of wry, one-panel humor at the system. Given this knowledge base, she trained it to choose funny captions for cartoons based on the jokes of similar cartoons. But in order to help [Mankoff] and his assistants choose among the entries, the AI must be able to rank the comedic value of jokes. And since computer vision software is made to decipher photos and not drawings, [Shahaf] and her team faced another task: assigning keywords to each cartoon. The team described each one in terms of its contextual anchors and subsequently its situational anomalies. For example, in the image above, the context keywords could be car dealership, car, customer, and salesman. Anomalies might include claws, fangs, and zoomorphic automobile.

The result is about the best that could be hoped for, if one was being realistic. All of the cartoon editors’ chosen winners showed up among the AI’s top 55.8%, which means the AI could ultimately help [Mankoff and Co.] weed out just under half of the truly bad entries. While [Mankoff] sees the study’s results as a positive thing, he’ll continue to hire assistants for the foreseeable future.

Humor-enabled AI may still be in its infancy, but the implications of the advancement are already great. To give personal assistants like Siri and Cortana a funny bone is to make them that much more human. But is that necessarily a good thing?

[via /.]

Massive Microsoft Machinations For Makers

April 30, 2015 by Brian Benchoff 89 Comments

If you’re not stuck in the tech news filter bubble, you may not have heard the Microsoft Build Developers Conference is going on right now. Among the topics covered in the keynotes are a new Office API and a goal to have Windows 10 running on a Billion devices in a few years.

There are, however, some interesting things coming out of the Build conference. Windows 10 is designed for hackers, with everything from virtual Arduino shields running on phones, Windows 10 running on Raspberry Pis, and Visual Code Studio running on OS X and Linux.

This is not the first time in recent memory Microsoft has courted the maker market. Microsoft begrudgingly supported the hardware dev scene with the PC version of the Microsoft Kinect, and a year or two ago, Microsoft rolled out drivers for 3D printers that were much more capable than the usual serial interface (read: the ability for printer manufacturers to add DRM). To the true, tie-die wearing, rollerblade-skating, acoustic coupler-sporting, Superman III-watching hackers out there, these efforts appear laughable – the product of managers completely out of touch with their audience.

Depending on your perspective, the new releases for the Arduino, Raspberry Pi, and other ‘maker-themed’ hardware could go one way or the other.

As far as educational efforts go, the Windows Remote Arduino and Windows Virtual Shields for Arduino are especially interesting. Instead of filling a computer lab up with dozens of Arduinos and the related shields, the WVSA uses the sensors on a Windows 10 smartphone with an Arduino. Windows Remote Arduino allows makers to control an Arduino not through the standard USB port, but a Bluetooth module.

If Arduinos aren’t your thing, the Windows 10 IoT preview for the Raspberry Pi 2 and Minnowboard Max is out now. The Win10 IoT distribution does not yet have working WiFi or Bluetooth, making it the single most useless operating system for Internet of Things devices. It was, however, released at the Build conference.

Also announced was a partnership with a fabulous hardware project hosting site, Hackster.io. Microsoft and Hackster.io will be collaborating with hackathons and other events focused on Windows technology. I get why they wouldn’t want another, vastly more popular project hosting site doing this, but I’m a little confused at why Instructables wasn’t the top Microsoft pick.

As always, you may express your infinite derision in the comments below. Spelling Microsoft with a dollar sign will result in a ban.

Keystroke Sniffer Hides As A Wall Wart, Is Scary

January 14, 2015 by Bryan Cockfield 79 Comments

For those of us who worry about the security of our wireless devices, every now and then something comes along that scares even the already-paranoid. The latest is a device from [Samy] that is able to log the keystrokes from Microsoft keyboards by sniffing and decrypting the RF signals used in the keyboard’s wireless protocol. Oh, and the entire device is camouflaged as a USB wall wart-style power adapter.

The device is made possible by an Arduino or Teensy hooked up to an NRF24L01+ 2.4GHz RF chip that does the sniffing. Once the firmware for the Arduino is loaded, the two chips plus a USB charging circuit (for charging USB devices and maintaining the camouflage) are stuffed with a lithium battery into a plastic shell from a larger USB charger. The options for retrieving the sniffed data are either an SPI Serial Flash chip or a GSM module for sending the data automatically via SMS.

The scary thing here isn’t so much that this device exists, but that encryption for Microsoft keyboards was less than stellar and provides little more than a false sense of security. This also serves as a wake-up call that the things we don’t even give a passing glance at might be exactly where a less-honorable person might look to exploit whatever information they can get their hands on. Continue past the break for a video of this device in action, and be sure to check out the project in more detail, including source code and schematics, on [Samy]’s webpage.

Thanks to [Juddy] for the tip!

Continue reading “Keystroke Sniffer Hides As A Wall Wart, Is Scary” →

FTDI Screws Up, Backs Down

October 24, 2014 by Brian Benchoff 252 Comments

A few days ago we learned chip maker FTDI was doing some rather shady things with a new driver released on Windows Update. The new driver worked perfectly for real FTDI chips, but for counterfeit chips – and there are a lot of them – the USB PID was set to 0, rendering them inoperable with any computer. Now, a few days later, we know exactly what happened, and FTDI is backing down; the driver has been removed from Windows Update, and an updated driver will be released next week. A PC won’t be able to communicate with a counterfeit chip with the new driver, but at least it won’t soft-brick the chip.

Microsoft has since released a statement and rolled back two versions of the FTDI driver to prevent counterfeit chips from being bricked. The affected versions of the FTDI driver are 2.11.0 and 2.12.0, released on August 26, 2014. The latest version of the driver that does not have this chip bricking functionality is 2.10.0.0, released on January 27th. If you’re affected by the latest driver, rolling back the driver through the Device Manager to 2.10.0.0 will prevent counterfeit chips from being bricked. You might want to find a copy of the 2.10.0 driver; this will likely be the last version of the FTDI driver to work with counterfeit chips.

Thanks to the efforts of [marcan] over on the EEVblog forums, we know exactly how the earlier FTDI driver worked to brick counterfeit devices:

[marcan] disassembled the FTDI driver and found the source of the brick and some clever coding. The coding exploits differences found in the silicon of counterfeit chips compared to the legit ones. In the small snippet of code decompiled by [marcan], the FTDI driver does nothing for legit chips, but writes 0 and value to make the EEPROM checksum match to counterfeit chips. It’s an extremely clever bit of code, but also clear evidence FTDI is intentionally bricking counterfeit devices.

A new FTDI driver, presumably one that will tell you a chip is fake without bricking it, will be released next week. While not an ideal outcome for everyone, at least the problem of drivers intentionally bricking devices is behind us.