Question about collimated display systems.

[Salire]

A 6' radius is the sweet spot for my project. I'm glad you guys are focusing on this. I'm also interested in the options for vacuums. I'd thought of the variable speed and am interested in your results. From my reading, I'm under the impression that very little vacuum pressure is actually required. Have you thought about things like bilge blowers?

[Salire]

Or maybe a kitchen ventilation blower?

[super2277]

Hi!

how much is needed for half circle curved screen for 737ng with shell ? wide and lendght

[wledzian]

See post #299
http://www.mycockpit.org/forums/show...l=1#post132173

[Dinther]

This is the most fascinating thread I ever read. Very exited about the results achieved by wledzian and geneb also surprised why it has gone so quiet around your efforts.

Anyway, to give your results some reproducible backing. For that I want to produce a Sketchup model with mathematically correct dimensions. The reason I use Sketchup is because I have a copy of Shaderlight 2 which enables me to do raytracing and thus virtually experiment with variations to your setup.

I like to see a much more accurate description of what exactly the guideline dimensions are. You speak of a spherical section meaning a piece of a perfect hemisphere. If so, which piece? Can you define the upper and lower part of it in degrees?

Same question with regard to the projection surface. I believe you said both hemispheres have their center at the same point. Is this correct or was there some tweaking. After all, the viewers eye position must be important here too.

What exactly is the viewers eye position anyway?

At the start of the thread people correctly speak of a parabolic mirror. A parabola is not spherical. The characteristic of a parabolic mirror is collimation, a spherical section does not do that. Interestingly when a chain is suspended between two points it naturally forms a parabola. The same is true for a piece of mylar under vacuum I suppose.

Would it be right to conclude that your mirror is spherical in a horizontal axis and parabolic vertically?

collimated screen.jpg
Image of sketchup model. The model uses a parabolic mirror. (Not a spherical section because I simply could not get that to work)

collimated-raytrace.jpg
Image of shaderlight 2 raytrace. In this raytrace you can see how the office environment is reflected in the mirror and also a piece of the screen with test pattern. Obviously things should fall in line when the viepoint moves to where the pilot would sit. I have not achieved this yet as I don't have any real solid data to model with. I hope those details will be made available so that we can have a solid record in once place how this screen is done right.

Sketchup model can be downloaded here: Collimated display study by dinther - 3D Warehouse

[Mike.Powell]

The mirror surface in a commercial collimated display system is spherical, not parabolic. With the appropriate frame, a mylar sheet stretched by differential pressure will form a very nearly spherical surface. The shape of a hanging chain is a response to the unidirectional gravity field, while the mylar shape is the response to non-unidirectional pressure which is normal to the curved mylar surface.

Spherical optics introduces geometric image distortion. This is compensated by warping the image source. To a first approximation the image is placed on a spherical surface having a radius half that of the mirror and sharing the center of curvature. Further optimization moves the image surface closer to the mirror with the result that the image surface is technically a toroid.

Google's US patent search is a good source of information on this technology. Search on "infinity displays".

[Joe Lavery]

Hey Mike, you lost me at Commercial.. I think I have the skill to build one of these and the tools to do it, but I'd need at idiot's guide to the technical side of things.

You should write a book on the construction, with plans and dimensions I'm sure there's a sale for it.

Joe.

[Mike.Powell]

Hey Mike, you lost me at Commercial.. I think I have the skill to build one of these and the tools to do it, but I'd need at idiot's guide to the technical side of things.

You should write a book on the construction, with plans and dimensions I'm sure there's a sale for it.

Joe.

I think Gene is working on that though I don't know what the status is.

[wledzian]

I've been handling the optics design, and Gene has been taking it from there, designing the structure. We still can't sell full kits or plans, as one critical element is still under active patent. We can get you pretty close, and there's enough information available here for you to take it the rest of the way.

Dinther, the mirror is spherical, not paraboloid. True, the collimation is not perfect, but it doesn't have to be. It just has to place the image far enough away that stereopsis is no longer the dominant depth cue. Motion parallax and perspective take it from there. Trust me when I say the effect is stunning.

[castle]

I think Gene is working on that though I don't know what the status is.

yes, Gene is just about done with the design for a 737 setup so i won't steal his thunder.

But a couple of things to keep in mind.
1) you will need a large space around the cockpit. it is a large structure. The projection screen alone has a radius greater than 5 feet and the mirror structure has a diameter over 16 feet and you will need some room behind the structure for walk-around space. having a 12 foot ceiling helps. you might get by with 10 feet. I will be building mine in a 14x24x12.5 foot garage stall.


2) a CNC router is almost mandatory. precision is a must and the projection screen has 63 ribs forming the surface.

3) would recommend a mylar width of 116 inches. The design calls for a minimum of 107" with margin of 3" on each edge. It is a huge piece over 30 feet in length; coming up a little short in gluing the mylar to the frame is a no-no, you can trim the excess, but you can't add a patch segment.

Wayne and Gene have done a fabulous job and will share the design sans the patent with anyone crazy enough to accept the challenge. But it can be worked out by studying the patent, crunching the numbers and observing and viewing the pics.

Gene will be writing a "howto" doc on the design and instructions for building and I will also be adding my two cents along the way.

Jack