Connect with us

Idea for Stroboscopic Flat Panel Display

Discussion in 'Electronic Design' started by IsaacKuo, Mar 17, 2007.

Scroll to continue with content
  1. IsaacKuo

    IsaacKuo Guest

    I have this idea for creating a large flat panel display, and wonder
    if it's feasible.

    The basic idea is for the panel to be a thin slab chamber of
    water or gel. The edges are coated with a reflective mirror
    surface, so total light reflection keeps light inside the panel.
    One side is perforated with tiny holes, so the water's surface
    tension keeps it flat.

    Three stroboscope LEDs feed red, green, and blue light
    in turn into the edge of the slab. These need to be able to
    provide extremely short duration flashes at precisely
    60hz (or greater).

    The display is shaped using a row of ultrasonic transducers
    along one edge of the display. These act as a phased array,
    concentrating sound waves onto individual pixels. The
    computer electronics sum up the waveforms of every displayed
    pixel. As a result, when the strobe flashes, the "active" pixels
    are places where the sound is concentrated. In these places,
    water pressure causes the water surface at the perforations
    to be bumpy. These bumps spoil total internal reflection,
    and as a result light can escape at the active pixels.

    Note that these sound waves only concentrate onto the desired
    pixels momentarily. This is why a strobe is required. The
    light "captures" the image at the precise point in time when
    the sound waves are concentrated as desired. The rest of
    the time, these sound waves interfere in arbitrary ways. If
    light were applied continuously, imaging may still be vaguely
    possible but it will be extremely blurry--each active pixel
    would have an hourglass shaped halo around it.

    The duration of the strobe pulses need to be maybe one
    microsecond or less, but the brightness needs to be able to
    illumate the entire panel, if necessary (for a pure white image).
    Is that feasible? Would multiple LEDs be necessary or better?

    I'm not sure what requirements are for the linear transducer
    array. Assuming a pixel size of around 1mm, the transducers
    need to operate at at most a 1mm wavelength and need to
    be at most 1mm wide.

    I think this concept could be suitable for inexpensive flat
    panel displays. Most of the display is simply a water
    chamber surrounded by bulk glass or plastic. One
    edge has the ultrasonic transducer array, which I
    imagine would be the most expensive component. Is
    there a way to make this component less expensive?

    The display itself is naturally transparent. For use as a
    TV or computer display, you'd want it to be backed by
    a black coating, of course. For an artistic display, the
    transparency may be considered a feature rather than
    a flaw. This panel may be cut to any convex shape
    (as long as the transducer array has an unobstructed
    view to all pixels in the active display area).

    Ideas? Criticisms?

    Isaac Kuo
     
  2. I am not 100% sure how your system works, _one_ array of
    transducers would have problems creating _one_ pixel
    or one row of pixels in the other dimension.
    The idea to use fluid is not new, look up 'Eidophore'.
    In the Eidophore a scanning electron beam was used to manipulate
    the surface of an oil film.
    In a vacuum and electron beams can go anywhere fast.
    So how do you addres coordinate x,y?
     
  3. IsaacKuo

    IsaacKuo Guest

    My idea was to build up an image in time, so that the waves
    constructively reinforced to create the desired image at the
    instant the strobe light was flashed.

    However, I have since realized that I had the approach all
    backwards. This approach requires very precise control
    of the transducers in a highly complex way.

    The transducers were doing a very complex job, while
    the LEDs were doing a very simple repeating job (just
    strobing periodically).

    Instead, the greatly superior approach is to have the
    transducers do a relatively simple repeating task, while
    the LEDs do a relatively simple modulated task. The
    transducers produce sound waves to constructively
    reinforce on a simple raster scanning pixel, while the
    LEDs create the image by modulating their brightness
    levels.

    My new refined idea has a different construction. From
    back to front:

    1. Black rear coating.

    2. Clear reflection coating.

    3. Plastic optical chamber.

    4. Perforated clear mask.

    5. Water chamber.

    6. Clear front wall of water chamber.

    The plastic optical chamber is a slab of plastic with a
    mirrored coating around the edge. It's made of a
    hydrophobic material. Red, greed, and blue LEDs
    feed this chamber with light from the edge, where
    total internal reflection contains the light.

    The perforated clear mask is also made of a
    hydrophobic material. The surface of the water
    has bumps which poke into the perforations, but
    these bumps don't touch the optical chamber
    unless sound vibrations provide sufficient pressure.

    The water chamber has maybe a dozen blade
    shaped sonic actuators spaced across the bottom
    edge. These form sound waves which constructively
    reinforce to a raster scanning pixel. Within the
    currently active pixel, pressure is sufficient to
    make the water bumps touch the optical chamber.
    Light leaks out from the chamber at these contact
    points, and reflects off the curved bump surfaces
    to deflect the light into a forward cone.

    The waveform required for creating the raster
    scanned pixels is calculated in a straightforward
    manner. For each pixel, there's a particular time
    "t" when it should be activated. The required
    waveform is a spike at time t-d/s, where d is
    the distance between this pixel and the sonic
    actuator, and s is the speed of sound in water.
    By summing up all of the spikes, you get a
    periodic waveform for each sonic actuator.

    This approach vastly simplifies the electronics
    design task. The Red, Green, and Blue LEDs
    are simply fed analog amplified signals straight
    from the VGA cable. The challenge is calculating
    the transducer waveforms and syncing them
    up to the incoming hsync and vsync signals.

    Isaac Kuo
     

  4. Perhaps you could make a small prototype to see if any
    problem some up.
    A prototype with only a few pixels, and one color.
     
  5. No idea ... but an interesting variation would be to feed audio into the
    transducers and shine powerful LED or laser light from below (for
    transmission) or above (for reflection) such that the "picture" gets
    projected onto a ceiling or wall. You could even put a small amount of oil
    on top for a thin-film effect.

    Hmm, might have to go and try this ...

    [...]
     
  6. IsaacKuo

    IsaacKuo Guest

    Yes, I think this refined design is getting simple
    enough that I can try it out. My original concept was
    just excessively complex.

    I've currently refined the concept so there's no need
    for messing with water. There are now only two
    layers:

    1. Rigid optical sheet of clear plastic edged with
    mirror coatings and three LEDs.

    and

    2. Flexible vibrating sheet of with a frosted inner
    surface. The edge is supported at 12+ points
    by vibration actuators.

    Instead of sound pressure waves, this sheet is
    vibrated transversely, like a rubber sheet.
    The frosted inner surface contacts the optical
    sheet at a single pixel at a time (raster scanning).

    Compared to my earlier concept using water,
    this is a lot less "fiddly". Unfortunately, the frosted
    surface will reflect ambient light so the black
    levels are no better than a front projector. This
    may be mitigated by using a dark coating, but
    of course this cuts down on the display's own
    light also.

    Isaac Kuo
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-