Video Projector on a Chip?

Stile 65 writes "Cornell researchers have made a 0.2mm-squared mirror mounted on carbon fibers that can oscillate at 2.5KHz, 'caus[ing] a laser beam to scan across a range of up to 180 degrees.' These can be mounted on a chip, and in combination with lasers, arrays of such mirrors on a chip can be made into a video projector. From the article: ''"It would be an incredibly cheap display," [Cornell grad student Shahyaan] Desai said. And the entire device would be small enough to build into a cell phone to project an image on a wall."' This display is made possible because of the innovative use of carbon fiber instead of silicon in MEMS. Unlike a standard DMD, this type of device would have one mirror per scanline, not one mirror per pixel, allowing the chip to be much smaller."

Comments

Re:A Step Up (down in size) from this

[DanMc]

Agreed! Although Light Blue Optics Ltd's devices aren't available on the market yet, and they claim to have no mirrors, prisms, or moving parts. I poked around in Feb when LBO announced their laser projector tech. I couldn't believe no one had tried to make a cheap laser/mirror scanning projector. I found some patents on the technology from the 70's that appear to be used by companies doing the "Pink Floyd laser light show" type devices. They just don't understand what they have could be used to kill off the multi-million dollar LCD projector market, home theater, and even win the LCD/Plasma/OLED/etc TV wars.

Prediction:

[The MAZZTer]

Virus that causes porn movies to randomly play through this display on cell phones. Man that would be embarrassing in a public place like a mall or something. Which is exactly why someone is going to make it.

At least CRTs had phosphor "memory"

[Jeff DeMaagd]

What does this have? Phosphors hold their brightness a little bit, down a reducing curve. This sort of display would have the scan line refresh issue of CRTs without the benefit of the fade curve, the light disappears immediately, so then it's just retina response time. I would expect that this would have to have a pretty high refresh rate to not be annoying. Will this allow three-chip operation? Consumer DLPs have a "rainbow effect" because only one chip flashes out the red, green and blue parts of the image. This doesn't bother everyone but I suspect that this system will have similar laments.

Re:At least CRTs had phosphor "memory"

[Veldcath]

If you read the article, they are talking about having one mirror per scan-line... and talking about oscillating the mirrors at a rate of 2.5 kHz. So it would be drawing 2,500 times a second. And they're also saying it would be one mirror per line, so you'd be seeing the entire display refresh 2,500 times a second. Considering that movies are projected at twenty four frames per second and our persistance of vision handles that with relative ease, I suspect a scan rate of 2.5 kHz would be more than adequate to create a very solid-looking image.

-V

Re:manufacturing efficiency of carbon fibre?

[kfg]

The largest silicon chips approach a billion devices at a cost of $0.0001 cent per device.

And that's why LCD displays cost about a buck.

What is the manufacturing efficiency of carbon fibre?

The carbon fiber in a thirty dollar fishing pole is measured in kilometers. In this device the carbon fiber elements are measured in microns. Only one device per scanline is needed.

As per my first sentence generally manufacturing costs swamp materials costs when building per unit, but as per my fourth sentence a 1024x768 display would require 768 elements, rather than 786,432.

KFG

KFG

No ^ 3

[Anonymous Coward]

Yes you need a lot of power, but you do not need all that power in a single coherent beam.

Since the image has a fibre per scanline you can use lots of low power laser diodes.

Doesn't sound too good to me...

[GreyPoopon]

From the article...

"You need something incredibly stiff to oscillate at a resonant frequency of 60,000 times a second (the line-scanning rate of most video displays)..."

Desai first showed that micrometer-scale carbon fibers can bend like tiny fishing rods by more than 90 degrees and can be made to vibrate billions of times without breaking down.

So, even at 500 billion times, that would be a lifetime of only 2314 hours?? No thanks. Please post again when they get it up to 10,000.

"Carbon is normally a brittle material," Desai said, "but in the fiber form it resists breakage. We have some data implying that if it lasts three and a half days it's going to last forever."

There's science for you...

Re:How powerful...

[budgenator]

About the same as a common laser pointer 5mW is a lot of laser light at least in red, not sure how much more you'd need for a decent green or blue laser so let's guestimate 25mW total for indoor usage, a 1/2 W would probably handle outdoors in almost direct daylight I'd guess. These figures are very reason considering what a plasma display or a LCD projector would consume.

Re:and society marches backward...

[Jeremi]

And the entire device would be small enough to build into a cell phone to project an image on a wall." This is just what we need.

Actually, it is just what we need, if the goal is to replace personal computers with cell phones. Imagine 10 or 20 years from now, ugly beige boxes have gone the way of the VCR and everybody just carries their "PC" with them in their pocket wherever they go. Wireless Internet access is available everywhere, of course, and while you can still use the small screen on the train, you can also sit down at any desk and use your phone the same way you use a PC now: with a full-size projected display, keyboard, and mouse. So now we've got the display part solved (in theory), the next step is to figure out how to fit a mouse and keyboard into a cell phone.... :^)