New "Telescopic Pixel" Displays Could Outperform LCD and Plasma
Liquid crystal displays (LCDs) have become the overwhelming choice for both desktop and mobile computing because they offer the best combination of image quality, price, and power efficiency of the current display technologies. But LCDs still have a lot of room for improvement, as they only transmit 5 to 10 percent of the total backlight to the user, and can account for up to 30 percent of the total power consumption of a laptop. In this week's Nature Photonics, researchers from Microsoft and the University of Washington report a new display technology called "telescopic pixel" that transmits 36 percent of backlight radiation.
The new pixel design is based on a tried-and-true technology: the optical telescope. Each pixel consists of two opposing mirrors where the primary mirror can change shape under an applied voltage. When the pixel is off, the primary and secondary mirrors are parallel and reflect all of the incoming light back into the light source. When the pixel is on, the primary mirror deforms into a parabolic shape that focuses light onto the secondary mirror. The secondary mirror then reflects the light through a hole in the primary mirror and onto the display screen.
Each pixel is produced in two halves by standard photolithography and etch techniques. The secondary mirror is simply a lithographically patterned array of aluminum islands on glass, but making the primary mirror is a bit more complicated. First, an indium tin oxide (ITO) electrode is deposited on a glass substrate and coated with polyimide. The polyimide acts as a support and electrical insulator for the primary mirror. Aluminum is then sputtered onto the polyimide and photolithography is used to pattern 20 µm diameter holes, forming a two-dimensional array that will eventually line up with the secondary mirrors.
The last step in the primary mirror fabrication is a dry etch, which preferentially removes polyimide from under the holes in the aluminum layer, resulting in sections of aluminum that are suspended in free space. These free-hanging sections of aluminum can be deformed by applying a voltage between the metal and the ITO layer. Once assembled, each pixel is 100 µm in diameter. This fabrication method is low-cost and compatible with the infrastructure currently used for LCDs.
Performance tests on arrays of telescopic pixels suggest they hold substantial promise for future displays. As mentioned above, backlight transmission was measured at 36 percent, and simulations indicated that this could reach 56 percent with design improvements. In a modern laptop with a five-hour battery life, this increase in efficiency could lead to almost 45 extra minutes of battery time without reducing screen brightness. Pixel response time was 0.625 ms—a dramatic improvement over LCDs, which have 2 to 10 ms response times. These response times may also be fast enough to allow sequential color processing where colors are displayed as rapid pulses of red, blue, and green from each pixel, streamlining fabrication and device design. Also, the intensity of each pixel can be smoothly varied from zero to one hundred percent for more realistic gray scales and color shades.
The worst and possibly crippling property of the displays was contrast. Experimental measurements conducted with non-collimated light showed a contrast of 20:1. Simulations indicate that ratios of up to 800:1 may be possible, which would put these displays on par with LCDs. Several easily correctable experimental factors led to the low contrast numbers, and I would expect to see much better numbers in future prototypes.
Given the substantial performance gains, amenability to current fabrication methods, and Microsoft's involvement, this report could signal the beginning of a new display technology. These displays have the potential to be faster than LCDs, more scalable than plasma, and cheaper and more energy efficient than both. It's a long, arduous path from the university lab bench to the laptop on your desk, but I wouldn't wager too much against seeing telescopic pixels in some capacity in the future.
By Adam Stevenson, Ars Technica
