CurtPalme.com Home Theater Forum

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

Oh, and let me add that the illustration that I used, uses the same LED chips that that Marquee claims to use. Its the same as those used in some RPTVs.

garyfritz · Joined: 08 Apr 2006 Posts: 12088 Location: Fort Collins, CO

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

**AnalogRocks** · Posted: Sun Jan 03, 2010 9:53 pm Post subject:

(pulls up a desk at the back of the class room)

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

Nashou66 · Joined: 12 Jan 2007 Posts: 16171 Location: West Seneca NY

**AnalogRocks** · Posted: Sun Jan 03, 2010 10:30 pm Post subject:

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

stefuel · Posted: Sun Jan 03, 2010 11:52 pm Post subject:

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

Alright, but I'm not going to make it easy. There will have to be participation. So you guys can understand whats happens HERE, you'll have to answer questions for me to continue.

First, lets go back before LEDs and look at a DLP that is bulb lit. The bulb is white light so how do you think it is able to make colors on the screen?

Spanky Ham · Joined: 22 Mar 2006 Posts: 5643 Location: Comedy Central

Mac,
I am interested in how the LED pjs work along with how well the current RPTVs are holding up.

As for VDCs engine, I am not sure if it is laid out like in Mac's post. I do know that Scott said they aren't doing anything different than the other LED pjs.

dturco · Posted: Mon Jan 04, 2010 12:34 am Post subject:

The spinning color wheel is timed to send the individual r,g,b light through?

Also I am VERY interested in the LED projectors, for the 3 year cycle that I see before GOZER needs replacement with a new comparable technology OR tubes.

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

**AnalogRocks** · Posted: Mon Jan 04, 2010 1:37 am Post subject:

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

So if I was to say incorrect? Very Happy

**AnalogRocks** · Posted: Mon Jan 04, 2010 1:45 am Post subject:

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

It's not NECESSARY, though. Each DMD in a 3 chip system gets fed separate red, green, and blue light anyway. Why mix and unmix from RGB to white then back to RGB? You lose lumens by doing that as no optical system is 100 percent efficient.

I don't quite understand youir comment "and although 3 chip is possible I haven't heard of any yet". Certainly you know that 3 chip DLP engines were the first type of DLP engines to be developed and all the highest output projectors use 3 chip engines.

The Phlatlight is also available in a single RGB module (driven sequentially, R, then G, then B, typically) for use with one chip engines, and they're available in individual color modules as well.

The single RGB module, driven sequentially, allows you to create a single chip DLP projector that has NO color wheel, too.
This by itself is a nice development because we know that the color wheel motor is a weak point in the design. Particularly
in unsealed optical engines.

I think that the ideal configuration would be to dedicate one Phlatlight color module to its own DMD optical system, lighting it up right at the DMD with appropriate light guides to ensure even DMD illumination and minimal losses. The system output would be combined conventionally into a single lens.

CJ

garyfritz · Joined: 08 Apr 2006 Posts: 12088 Location: Fort Collins, CO

Answering Mac's question: DLPs take the white light from a bulb and split it into 3 colors. Some split it simultaneously, sending the 3 colors out to 3 DLP chips; others split it temporally, with a color wheel that passes sequential R/G/B/etc colors.

Ahh. I got it, Mac. The light from the 3 lasers follows the path in your diagram, but NOT AT THE SAME TIME. It never mixes into white light (except in your visual perception system) because only one of the 3 LEDs is on at once. Duh. Smile

macgyver655 · Joined: 22 Aug 2007 Posts: 8508

Ok then, lets just say that it's partially correct. The color wheel is not timed to allow the light to pass through the colors. The color wheel just spins, usually anywhere from 8 or 9 thousand to 15 or more thousand RPMs. Also note that there are different color wheels, designated by number of segments. A 4 segment is usually RGB and white, a 7 segment adds cyan, magenta, yellow. There may be other varieties but you get the basic idea.

So you have a bulb that is at a constant brightness and a color wheel that is just spinning away. As you can see, the colors are never mixed in this sequence, they each pass at their own individual time.

So as the varying colors of light come off the color wheel they are then reflected onto the DMD (Digital Micromirror Device). The DMD is a device that has several hundred thousand microscopic mirrors that can be turned to either reflect or not reflect light. Any light that is reflected is then directed to the lens which projects the image.

So how does this happen.

Lets slow down the color wheel to observe 1 revolution and we will use a 4 segment wheel. As the red part of the wheel is in front of the light tunnel (a mirrored tunnel that the light from the bulb passes through) the red light beam is reflected to the DMD. Now the processor will determine which mirrors to turn to reflect the red beam through the lens. Which mirrors are turned it based off the video signal so the DMD will have a reflected image.

As the wheel turns to the green part, the process now begins as to which mirrors are turned to reflect the green part of the image...... and so on.

There is a marker on the color wheel so the processor knows the location of the wheel at all times.

The DMD must reflect each color multiple times to produce 1 frame of image and then 60 frames per second if that is the selected refresh rate.

The intensity of a color in various location of the image is determine by how fast a mirror is turned on. The stronger a color is needed, the faster a mirror is cycled on.

ANY QUESTIONS ??

Edited for correction.

garyfritz · Joined: 08 Apr 2006 Posts: 12088 Location: Fort Collins, CO

Um, minor quibble. The brightness isn't determined by the NUMBER of mirrors, but by HOW MUCH TIME each pixel's mirror is flipped to the "full on brightness" position. Those little suckers flap like a mofo to regulate the level of brightness -- more "on" time (during a single frame) translates to a brighter pixel during that frame. See e.g. bullet 2 on this page.