High Efficiency Transparent Optical Scanning Element

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

High Efficiency Transparent Optical Scanning Element

Postby Stephen » Wed Feb 14, 2007 6:07 am

Imagine a transparent optical scanning disc or drum with translucent pits embossed or drilled part way through one surface as a substitute for apertures. Imagine a linear modulated light source, such as a linear bank of LEDs, illuminating a portion of one edge of the transparent scanning element to transmit a linear light beam through the disc. As the transparent scanning element rotates about its axis, the linear light beam sequentially illuminates each translucent pit as it passes within the boundary of the beam, forming one scanning line.

All the light from the LEDs is concentrated into a single line, rather than spread out over the entire area of the display. Thus, in the case of a 32 × 48 pixel display, each pixel is 1/48 th of the brightness of the light source instead of 1/( 32 × 48 ) th of the brightness that a conventional opaque scanning element with apertures would provide.

A photosensor with a linear field of view could substitute for the light source for camera use. The photosensor would have an improvement in signal to noise ratio of as much as 32 times, since the area of view is limited by this amount compared to a photosensor with a rectangular field of view with a conventional opaque scanning element.

I have posted a description and illustration of this concept at http://www.taswegian.com/NBTV/images/TOSE.pdf .
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

High Efficiency Transparent Optical Scanning Element

Postby Stephen » Wed Feb 14, 2007 7:43 am

It is possible to stack several of the transparent scanning elements on a single drive spindle for a three dimensional display. In this case each transparent scanning element has its own linear light source. A displacement signal, separate from the modulating signal, sequentially triggers each linear light source to reproduce three dimensional content without any special viewing devices. John Logie Baird describes the generation and use of such a displacement signal in his British Patent GB373196 filed 18 February 1931.
Last edited by Stephen on Thu Feb 15, 2007 2:30 am, edited 1 time in total.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Re: High Efficiency Transparent Optical Scanning Element

Postby Andrew Davie » Wed Feb 14, 2007 9:36 am

Stephen wrote:All the light from the LEDs is concentrated into a single line, rather than spread out over the entire area of the display. Thus, in the case of a 32 × 48 pixel display, each pixel is 1/48 th of the brightness of the light source instead of 1/( 32 × 48 ) th of the brightness that a conventional opaque scanning element with apertures would provide.


Couldn't one place an appropriately curved mirror where the LED array sits, and have a single LED/photosensor operating instead of a line of them? This would give a further 48-fold improvement.

Cheers
A
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: High Efficiency Transparent Optical Scanning Element

Postby Stephen » Wed Feb 14, 2007 10:35 am

Andrew Davie wrote:
Couldn't one place an appropriately curved mirror where the LED array sits, and have a single LED/photosensor operating instead of a line of them? This would give a further 48-fold improvement.

Cheers
A


It would be possible to have such a curved mirror as you suggest that captured the light from a single LED. The problem is that a stationary mirror would still have to form a line of light so that each translucent pit receives illumination from the bottom to the top of the picture. Thus it would have to disperse the energy of the single point LED over an entire line so that it would offer no improvement. For that matter, the light source could comprise a single high power LED behind a couple of transversely mounted cylindrical lenses with appropriate focal lengths to approximate a line source.

The only way that I could conceive of directing all of the light from a point source to each translucent pit is with a separate mirror for each pit mounted along the edge of the disc that would receive light from the light source and direct it toward the pit. However, this gets into a complex design wherein we would need 32 accurately fabricated and aligned mirrors along the perimeter of the disc that are all spinning at high speed. I suspect it would be less trouble to turn to a Weiller mirror drum.

Nevertheless, this is probably the highest efficiency "Nipkow" disc arrangement possible. Interestingly, although the back side of the transparent disc should have a black baffle for best contrast, the back of the disc could be open if the display has a powerful light source. The display image would be viewable from the back as well as from the front and appear to float in mid air like a hologram.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

High Efficiency Transparent Optical Scanning Element

Postby Stephen » Wed Feb 14, 2007 12:57 pm

Of course, the linear light source could shine upward toward the translucent pits instead of sideways toward them. This way the light transmission to each translucent pit is reduced by a degree of about 1/32 instead of 1/48. This would be a slight improvement.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Postby Andrew Davie » Wed Feb 14, 2007 2:58 pm

How about placing a very light and small mirror on the Nipkow disk itself, one for each 'aperture' point. The mirror would be placed at 45 degrees to the viewer, so that it points to the light source off the edge of the disc.

An alternate, the mirror being so attached that it can pivot. Now include a magnetic substance on the bottom-end of the mirror and a very powerful magnet at the base of the (single) LED light source. The idea is that the magnet would be attracted to point the mirror exactly at the (single) LED light source, so no matter where it was in the 'scanline', the mirror would point correctly.

Of course this is offset by the problem of the force involved with the disc rotating at 750rpm or so... would it be possible to have a magnet strong enough to overcome any mispositioning by centrifugal (centripetal?) force. In any case, the mirrors would be mostly pointing the right direction in each rotation -- the magnetic attraction would just have to provide very minor adjustments to angle. I would expect, though, that it's not possible to find a magnet strong enough.

In place of a single magnet at the LED source, one could have a series of magnets behind the Nipkow disc, in a line from the LED source to the closest edge of the image scanned area. The angle wouldn't be quite so right, but for such a small image size would it make a difference?

But anyway, having mirrors on each aperture would provide effectively the same system, no?

Cheers
A
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Postby Stephen » Thu Feb 15, 2007 12:53 am

The underlying problem would be to have a simple system for capturing the entire beam of the LED in a single pit throughout its arc of revolution in the viewing area. If the beam of the LED is wide enough that it reaches all of the little mirrors for each pit or aperture, it is already too wide for each little mirror to capture more than a small portion of it. If the beam of the LED was on the order of the width of the pits themselves, then each mirror would have to intercept the narrow beam and direct it to its corresponding pit.

That is why I suggested, for instance, 32 small mirrors along the edge of the disc that would so direct the narrow beam to each pit. If they were fixed in angle, there would be some light drop off at the top and bottom of the display. If as you suggest these mirrors were rotatable, this would alleviate this problem.

My feeling is that the complexity involved for this ultimate efficiency would not be worth it. What fascinates me about the transparent disc is that it is extremely easy to fabricate and conceivably have a brightness level similar to a mirror screw scanning element, which is a significant improvement. Making the disc may be as simple as placing a paper template with the pit positions marked under a blank disc and then punching or drilling pits over these positions.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

High Efficiency Transparent Optical Scanning Element

Postby Stephen » Thu Feb 15, 2007 4:47 am

If one were really interested in maximum efficiency, the easiest way might be to leave the transparent scanning disc alone and modify the light source arrangement. Consider the linear light source as a linear array of LEDs. There is no reason that they must all be modulated together simultaneously. They may each sequentially turn on and off as each translucent pit moves within each one's respective beam path. I am not necessarily suggesting a linear array of 48 LEDs, but lets say eight of them, so once a pit passes through the six pixel width of the first LED in the array, the first LED switches off and the second LED in the array switches on for the next group of six pixels, and so forth.

The switching mechanism for each LED could be a timing signal or reflected light from the pit picked up by a bank of photosensors. Switching eight LEDs in this fashion would allow light transfer to a degree of 1/6 instead of 1/48. Again, I am not sure if it is really worth the complexity since it matters little whether all the LEDs are on simultaneously as long as the drive amplifier has sufficient power.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Transparent scanning element.

Postby Stephen » Sun Mar 25, 2007 1:39 am

If the transparent scanning element is in the form of a ring instead of a disc, the inner perimeter may have the shape of a 32-sided polygon, one side aligned with each pit. The inner perimeter would then receive a reflective coating, such as aluminium paint. If the polygon has the right diameter, the reflective surface of each side of the polygon would reflect light back from a single modulated LED mounted along the outer perimeter of the transparent element toward its corresponding pit as it rotates through its corresponding scan line, thereby keeping each pit illuminated for its entire scanning line with the beam of a single modulted LED.

In this way, each pit receives nearly all the light from the single modulated LED, increasing efficiency to an extent that might approach the efficiency of a Weiller mirror drum display. The 32-sided surface of the inner polygon need not have any degree of optical precision, since it would only serve to generally reflect light that passes through the transparent element back toward toward the pits. With this arrangement, each pit receives its illumination from reflected light for most of its scan. It only receives illumination directly from the LED for the fraction of its scan when it is directly in line with the beam of the LED through the transparent element.

A support disc or inner rim could mount the transparent ring-shaped element on a drive motor shaft.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Transparent scanning element.

Postby Stephen » Thu Mar 29, 2007 10:01 am

Today I tested the principle of the proposed transparent scanning element with an old clock crystal upon which I secured several translucent "pits" along a portion of the edge. Using a laser pointer with the beam aimed through the edge of the clock crystal from the opposite side, each pit lit up brightly when I rotated the clock crystal to let the beam hit it. As expected, the clock crystal remained completely transparent to the beam so that each pit appeared as a moving point of bright red light in mid-air whenever it passed through the beam.

Perhaps after I receive and assemble the MUTR televisor that I have on order I can use it as a test platform for a real 32 line transparent scanning element. I would replace the supplied scanning element with a transparent one and the backlit LED board with an edgelit one.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am


Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 26 guests