Linked by Hadrien Grasland on Sun 4th Sep 2011 12:39 UTC
Graphics, User Interfaces "Okay, it's been about one week and there are not much answers anymore, so it's time to thank everyone who participated, close this survey, and publish the results, along with some interpretation." Everything is released under Creative Commons CC0 license, so anyone interested, please help yourself.
Thread beginning with comment 488724
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[3]: LED
by Neolander on Wed 7th Sep 2011 07:14 UTC in reply to "RE[2]: LED"
Neolander
Member since:
2010-03-08

60% ? Impressive, I wonder how they achieve that. If they start from white light, isn't color filtering supposed to eat its share anyway, even if it's done sequentially ?

I mean, assuming that the Red, Green, Blue was a partition of the optical spectrum with equivalent energy to each color...
-When displaying red, green and blue have to be filtered out, so there go two thirds of the energy.
-When displaying green, red and blue have to be filtered out
-When displaying blue, red and green have to be filtered out

Or isn't it so ?

EDIT : Okay, according to this https://secure.wikimedia.org/wikipedia/en/wiki/Time-multiplexed_opti... , they don't start from white light, but have a light source that can intermittently emit red, green, and blue light instead.

How can such a source work ?

Edited 2011-09-07 07:21 UTC

Reply Parent Score: 1

RE[4]: LED
by transputer_guy on Wed 7th Sep 2011 14:40 in reply to "RE[3]: LED"
transputer_guy Member since:
2005-07-08

The TMOS cell is a very elegant but also simple design in principle (maybe not so in production). It uses the effect of two glass surfaces coming into contact, light bleeds easily and efficiently from one to the other, a technique that is also used by spies to tap into fiber optic cables, (in hidden closets perhaps).

The single light cell per pixel uses a capacitive MEMs device that pushes a clear top surface against the back plane surface or light substrate. Each MEMs cell actually has hundreds of redundant point contacts so no dead pixels. The back substrate has R,G,B LEDs along the side edges pulsed in time order. The color cycling can be done at very high frame rates so would be fully integrated by the eye. Since each pixel only uses 1 Fet, it can also be much smaller than the tri cell allowing very high pixel resolutions.

Unipixel ended up selling its substrate optical backplane patents to Rambus, this can also be used for flat LED lighting applications that are likely to be huge soon. Unipixel may have removed most of the interesting material from their website though.

Some of the ideas of TMOS are quite similar to the TI DLP. While TMOS is planar, DLP has to magnify & project the image on a CMOS MEMs chip to a screen or wall, just as efficient but not as desirable image quality and needs spinning prisms, color wheels, lamps, lenses, diffusers, Fresnel lenses etc. The images are still processed the same at the pixel level by pulse width modulation of the on off light switchs.

Funnily enough, Sony has a HD LED/LCD TV based on R,G,B,W cell that should be nearly twice as efficient as plain RGB. The idea here is that a 4th LCD cell in each pixel lets the majority of the White part of the pixel light through unfiltered. So if the RGB pixel value is 0x5080A0, then reduce that to 0x003050 and add 0x50 to white.

Reply Parent Score: 2

RE[5]: LED
by Neolander on Thu 8th Sep 2011 09:08 in reply to "RE[4]: LED"
Neolander Member since:
2010-03-08

I've just realized that I've forgotten a big something in my previous calculation of theoretical LCD efficiency. Color filters may have something like 30% efficiency, but behind them there is another big power-waster : the back polarizing filter, which eats 50% of the incoming light's energy if it is absolutely perfect. This makes the theoretical LCD efficiency, given that perfectly lossless components of equal refraction index are used, 15%. With this in mind, 5% for a real-world device is not even that bad...

About your explanation of TMOS now : interesting use of frustrated TIR ! I'm still amazed that a mechanical device can be fast and efficient enough to achieve good display refresh rates with low power consumption, even though I shouldn't be since I've seen the IMOD/Mirasol demos...

Wonder for how long these capacitive MEMs they use in the FTIR layer can retain their state in real-world use. If it can withstand low refresh rates when displaying a still image, this tech could be used to build some nice transflective screens.

Also, I didn't know that it's transistors that are the biggest component which ends up limiting DPI in screens, so thanks for the information.

Big thing which I don't like in this screen concept is that it's yet another display device where displaying black and other dark colors is a huge power waste. Guess only OLED can do something about this...

Another question is that of vision angles. If you look sideways at a TMOS screen, you see light that has traveled a longer path inside of the FTIR layer, and which will thus be darker.

Edited 2011-09-08 09:24 UTC

Reply Parent Score: 1