OPEN SLATE
PROJECT
| Back to Open Slate Home Page | Date: 2003.01.20 (2007.11.21) |
I am not an engineer. In Internet jargon that would be rendered as IANAE, but who would understand? The significance of my IANAE status is that I do not feel qualified to suggest a hardware design, and I fear those who are engineers might not take us seriously. What is needed, in my opinion, is participation by genuine engineers. Having said that, I offer this attempt at defining a slate.
The design presented here is intentionally simple. This is due to the goal of Open Slate computers being self-made. Every student in a fully functional Open Slate enabled school should be able to assemble and maintain one of these basic designs. Specialists will certainly development -- slate modders -- and will find it lucrative to build jazzy customized slates for a fee. This in turn will stimulate industrial design and give students opportunities to develope and implement creative ideas.A slate is comprised of the following components:
The most important part of a slate is the frame. The design will be simple, clear, universal, and easily produced. The frame contributes strength, both literally and figuratively. It must be thick enough to minimize flex and twist, even when a slate is dropped. This stiffness must be achieved with minimum weight -- a heavy slate will be left behind.
The frame will include a standard pattern of holes for mounting other components, such as the system board and LCD panel. The process of mounting components may involve unique adaptors, such as has been standard practice in mounting hard drives. The early adoption and commitment to a limited number of frame designs (perhaps even just one) will contribute to the success of the project during the pioneering phase.
The frame should be made of aluminum. An alloy commonly found in aircraft construction, stiff, light, easily bent on a hand operated brake, and resistant to corrosion. I have seen two alloys highlighted in advanced bicycle component design, 6061-T6 and 7075-T6, which might be an example of what I am talking about. (IANAE, remember?) Another possiblity popular in light aircraft is 2024-T3. These alloys are popular for parts machined from a solid block of metal, and for tubing used in bicycle frames, stems, and handlebars -- they may not be the best choice for fabrication by bending. I offer these not as a recommended alloy, but as an example of the kind of quality I am looking for.
Making a frame should just be a matter of cutting out a pattern from flat sheet metal, bending it on a brake, and securing the corners. Ideally this last step should be done with TIG or MIG welding, but that might be beyond the reach of the self-made crowd. An alternate, bolt-up version should be available.
The frame can be cut from a single flat piece of sheet metal. The tricky work involves cutting away the material at each corner that will make room for the bends. That done, each edge will be folded twice at ninety degrees, resulting in a squared-off "C" cross-section with the open side facing inward.
Another design that we should consider would be to use a thicker piece of metal without any bending.
The shell is a light-load bearing skin that protects the innards and gives the slate its unique character. Material can be anything, from wood to titanium, but the preferred material will be plastic. Ambitious slate makers and commercial manufacturers will probably use current injection-molded fabrication methods. The more highly-prized slates will use methods more appropriate to hand fabrication, such as hand laid-up fiberglass. Advanced model airplane design, such as used in F3-B sailplanes, may be a source of inspiration.
Standardization of the frame design creates the opportunity for diversification in shell design. Slate-modders (as I propose they be called) can invest in designs, jigs and molds knowing that the shells they produce will be a bolt-on fit. Compare this to lighting fixtures. The fact that light bulbs come in a few common sizes allows lamp designers to design lamps knowing that consumers will not have a problem finding replacement bulbs.
There will be two special challenges for shell designers, hinges and buttons. There is no requirement that a slate have hinges, but the shell design should provide for protection of the LCD panel. The finest solution to this problem I have ever seen is that used on the Apple Newton, starting with the 110. A much more pragmatic solution is used in newer, smaller, and much less expensive PDAs, such as the Visor Handspring. An example of a design I would not recommend is the Visor Treo -- I think the fold-up lid is too fragile.
It seems as though the slate will require at least one power mode switch. I would prefer that this all be done via the touch-screen, but that would require leaving it powered up full time, which is sure to degrade battery life. Designing even one button that is functional, reliable, and that does not provide an easy entry for dirt is a challenge.
The chording keyboard is covered in a separate section below.
There is no standard design for the system board other than to specify inputs and outputs. This is done in the spirit of Formula One race car design rules, to foster innovation. Designers may prefer a mother-daughter board design, and some features can be made available through PC-CARD devices. Power management, while not specified, is expected.
The system board must be able to accommodate the following inputs and outputs:
The choice of ICs (chip sets) should promote battery life over speed. High performance slates may be capable of processing video, at least DVD playback but possibly video editing, but the expectation is that to do so would require running on A/C power. All slates should be able to display less demanding video formats typically used for web-based streaming video. (There are serious issues involving video at this time, due to the lack of a widely used open-source video format.)
The system board will typically mount to the frame with the aid of a board-unique adapter, although it is hoped that eventually the popularity of the Open Slate design will result in boards designed specifically for the frame.
Power management will be a key technology. In normal use a slate will rarely be placed in a completely powered off state. In terms of today's laptop designs, the slate that has been "turn off" will be in suspend mode. Given suitable mass storage, the equivalent of hibernation mode could be an additional option. In this case the power management software should be capable of automatically entering hibernation mode, even from suspend mode. Like the Newton, a slate should able to detect a low battery condition and refrain from entering full power mode. It appears that to do power management right will require a special BIOS. Perhaps we could collaborate with one of the open BIOS projects.
One of the features our slate design must have is the ability to change batteries on the fly. There are two ways to accomplish this, by the use of a backup battery, and through the hibernate option.
The backup battery prevents data loss when changing the main battery and provides long-term protection when the main battery becomes discharged. The Newton works this way; primary power comes from four AA batteries, and the backup is a watch battery.
Hibernation requires use of suitable mass storage, most likely a hard drive. The good thing about hibernation is that it is permanent; power can be completely removed. On the minus side, it takes time. There is a very real danger that if hibernation is invoked due to a low battery condition, the battery could run down too low to complete dumping the memory image to disk.
Battery life will be a critical aspect of slate design. With current technology, few laptops have sufficient battery life to provide for a full day's use. Increasing capacity means increasing weight, and a heavy slate will be left behind.
One way to extend working time is to provide for easy battery replacement. Carrying an extra battery in a backpack is preferable to a heavy slate.
What we want to avoid are situations where slate users need to be plugged into A/C power in environments intended to be mobile. Today, classrooms and meeting rooms designed for laptop use offer A/C power -- along with Ethernet jacks -- at every seat. The Open Slate concept is to eliminate this kind of anchorage.
Think of this as the battery compartment. Some shell designers will prefer to put a cover over the battery, but the battery design should make this unnecessary. I can think of two examples of what I have in mind, the way batteries attach to power tools such as a Makita power drill, and the way the Sony VAIO SR series laptop battery attaches. The goal is to make battery swapping simple and fast.
A sensing mechanism should be incorporated that can respond appropriately when a battery is replaced. As the battery is removed, the system should go into suspend mode (assuming it was in normal operation). Conversely, when a fresh battery power is attached, the system should come out of suspend mode. This might be possible using micro-switches in the battery latches, but it might be easier to make it a part of the power management system. This would require a back-up battery with enough capacity to do whatever it is that happens when suspend mode is entered, which should be possible.
A nice-to-have feature would be a battery charging station incorporated into the A/C adaptor. Some camcorders use this design. This allows a second battery to be charging while the slate is running off battery power. Current laptop designs require that a battery be charged while in the computer, so charging takes the "mobile" out of mobile computing.
This piece is rather self-explanatory, yet it has proven to be as much a hurdle as any other. Unlike desktop systems, there does not appear to be a readily available source for the bits and pieces to build laptop systems, much less a slate.
A slate requires a digitizer, a method for sensing pen location and movement. There are two common methods for making a digitizer, resistive and active. The resistive method senses the pressure of a stylus -- or any object -- touching the surface. PDAs, including the Newton, work this way. In the active method the digitizer transmits a low power alternating magnetic signal, which is picked up by a circuit in the stylus. The induced signal from the stylus is sensed by the digitizer. This design allows the digitizer to sense the stylus when it is close to, but not necessarily touching, the surface of the digitizer. Most graphics tablets, including those made by Wacom, use this method. One problem with the active design is that it only works with a matching stylus.
Another approach that should be investigated would be to separate the digitizer from the screen. Many laptops use a resistive digitizer as a fingertip actuated mouse ("touchpad"), and people quickly adapt to the separation of image and input surface. Indeed, the traditional mouse requires the same abstraction. As intriguing as it may seem to be able to interact directly with the image, there are good reasons not to go that route:
The most compelling reason not to use a separate digitizer is size -- such a design will make the slate much larger than the display. Another challenge will be accommodation of lefties, but since a slate is supposed to be built by its owner this is solvable.
This part is straightforward. Standard hardware should be used. The connectors should all be on the side opposite the user; connections on the sides will interfere with the hands.
There is no requirement that a slate have a keyboard. In fact, laptop-style QWERTY keyboards are discouraged except as plug-in accessories. What is encouraged are chording keyboards built into the bottom of the slate, along the left and right sides. Details of its design will be the subject of another article. For now, just know that this will involve something like six or eight small buttons on each side which are pressed by the four fingers while holding the slate at the edge with the thumb on top. It is called a chording keyboard because the 100-odd codes necessary to enter ASCII text (and internationalization must be taken into account) are created by pressing multiple keys at once. If this sound weird, consider the fact that you already do it with today's QWERTY keyboard when you alter the meaning of a key with the shift, alt, or ctrl keys on a PC, or with the equivalents on a Mac. Millions of kids use tricky key combinations when they play video games, and millions more when they play a music instrument. Whatever the design ends up being, making the keys strong, comfortable, and reliable will be a challenge.