WO1996019900A1

WO1996019900A1 - Liquid crystal video projector

Info

Publication number: WO1996019900A1
Application number: PCT/FR1994/001516
Authority: WO
Inventors: Christophe Nicolas; Jean-Yves Eouzan
Original assignee: Thomson-Csf
Priority date: 1994-12-22
Filing date: 1994-12-22
Publication date: 1996-06-27

Abstract

Liquid crystal video projector comprising at least a light source (S) emitting a light beam; a polarising separator (SP) receiving the light beam and providing two orthogonally polarised sub-beams; two liquid crystal imagers (LCDA, LCDB), each placed on the path of a sub-beam; an objective (L3) receiving the two sub-beams (F4, F5) after going through the two imagers; the video projector being characterized in that it has also a combination mirror (M3) situated probably at the vicinity of the input lens of the objective (L3), the imagers (LCDA, LCDB) being substantially symmetrical to each other with respect to the combination mirror (M3); and in that the two sub-beams modulated by the imagers converge towards the objective (L3) according to two mean directions forming a non-nule angle between each other and produce two images of the source (S) which are substantially distinct in the plane of the input lens (P) of the objective (L3). Application to stereovision video projector.

Description

LIQUID CRYSTAL VIDEO PROJECTOR

The invention relates to a double liquid crystal video projector and more particularly to a video projector which is convertible between stereovision and ordinary vision with, in the latter case, an increased brightness compared to the state of the art. The number of liquid crystal cells is doubled compared to the number strictly necessary to obtain a conventional (non-stereoscopic) image, but the device has an optimized energy efficiency.

By stereovision is meant relief vision of images projected on a passive screen. Among different stereovision methods, the one that concerns us consists in projecting two images, A1 and A2, on a passive screen, the light beams used to project these two images being polarized. Their polarization directions are perpendicular to each other, ie Dl and D2. The observer wears glasses with two polarizing filters. The polarization direction of each filter coincides with that of one of the two projected images. So, for example, the right eye sees one of the two projected images, A1, and the left eye sees the other, A2. Of course, the screen must allow the observation of polarized images: while an ordinary diffusing screen "depolarizes" the light during the reflection of the rays, a metallic screen makes it possible to preserve the polarization properties. These screens exist and are used in cinemas (at the Futuroscope in Poitiers, for example).

One method for producing the two polarized images is to use two separate video projectors. Each of them is associated with a particular video signal and a polarizer is interposed on the path of the light rays coming from each of these projectors. The two polarization directions are perpendicular to each other. The projection on the same screen of the two images thus produced, combined with the use of polarized glasses gives the observer the illusion of a three-dimensional image.

In order to obtain the projection of the two polarized images, two cathode ray tube projectors can be used by placing a polarizer at their output. The disadvantages are:

- problem of superposition of images and risk of destabilization of this superposition due to thermal drift over time;

- more than half of the light produced by each projector is lost due to polarizers; - in an ordinary cathode ray tube projector, there are three projection objectives. It is therefore difficult to perform a zoom function. This is all the more difficult if two combined projectors are used.

It is interesting to use projectors with liquid crystal imagers (or LCD) for the following two reasons:

- the light modulated by these imagers is, in principle, polarized at the output of each LCD imager;

- It is known to combine together the light rays modulated by different imagers, for example when each modulates a primary color, and to use a single projection objective common to all the imagers. This objective can therefore include a "zoom" function;

- possibility of averaging faults due to LCD imagers.

It is understood that a video projector with liquid crystal imagers comprises at least one LCD imager. In the general case, it comprises: - either a single LCD imager; for a color image, each pixel is provided with a red, green or blue colored filter;

- either three LCD imagers, each illuminated by a beam of monochrome red, green or blue light.

FIG. 1 describes a conventional optical architecture of a three-color video projector. Dl, D2, D3, D4 are dichroic mirrors. Ml and M2 of ordinary mirrors. Generally the light emitted by the lamp is not polarized. The imagers LCDy, LCD3, LCDR each include a polarizer which absorbs half of the flux, the other half being modulated by the liquid crystal cell. In general, for a rectangular LCD imager (4: 3 or 16: 9 format for video), the input polarizer of the LCD is oriented at 45 ° to the sides of the rectangle, this for reasons of homogeneity of the contrast throughout the surface. Likewise, the LCD output polarizer is oriented at 45 °. For the most common case of positive LCD imagers, that is to say transparent under a zero control voltage, the input and output polarizers are oriented at 90 ° from one another. The orientation of the polarizers of an LCD imager at 45 ° to the sides of the image is at the origin of a characteristic of liquid crystal projectors which will prove to be important for the invention which will be described below: 'off-axis' illumination (Figure 2). Indeed, so that the contrast is uniform over the entire image and that it is maximum, it is necessary to illuminate each imager at a non-zero average angle (which can reach 6 °). One of the main disadvantages of liquid crystal video projectors is their low light efficiency, typically between 1 and 2%. One of the main reasons for this low light efficiency is that the LCD imager includes a polarizer. As a result, more than half of the light is lost upstream of the imagers.

In order to double the luminous flux of the projected image, one solution consists in using several projectors arranged in parallel, all projecting the same image.

This solution has the drawback of multiplying the electrical consumption and of posing the same problems of image convergence as cathode ray tubes (with less ease in solving these problems).

Similarly, to project images in stereoscopy, two liquid crystal projectors can be used in parallel. As the light is polarized at the output of each LCD imager, after passing over the dichroic mirrors, it becomes elliptical but the ellipse is greatly elongated and the loss of light due to the use of polarized glasses should not be too great. However, you must turn the LCD imagers of one of the projectors relative to the other in order to obtain two polarization images perpendicular to each other.

In order to circumvent the problem of low efficiency, a solution as shown in FIG. 3 consists in doubling the number of LCD imagers compared to the number strictly necessary in a video projector comprising a single projection objective (see Patent EP 372,905) . In the basic configuration there are two LCD imagers, each being illuminated by polarized light.

A collimated beam of white light is divided into two complementary components polarized in two directions perpendicular to each other, by the action of a polarization splitter. Each component illuminates a projector LCD1, LCD2, the two components modulated by the imagers are then recombined by a mixing cube sensitive to polarization. On one of the channels, after separation, you can rotate the polarization direction by 90 ° so that it is identical to that of the other channel.

In a more elaborate configuration, there are up to six LCD imagers, in order to obtain a color image. Each LCD imager modulates light in one of the three primary colors and one of the two complementary polarizations. In both cases, all the light emitted by the lamp is used.

This system could be used in stereoscopy. It would suffice to address each LCD with the proper video signal. However, this system turns out to be impractical when evaluating some numerical quantities. In fact, the separating cube (or mixer) sensitive to polarization is, by construction, very sensitive to the incidence of light rays. It is characterized by a typical angular acceptance of ± 3 °. In other words, it must present beams of almost collimated rays. However, in a video projector, after modulation of the beams by the LCD imagers, it is necessary to make the rays converge in the aperture of the objective; this is the role of the field lens.

If we can find an optical system which makes it possible to adapt the rays modulated by each LCD imager and a cube mixing polarizations, a simple calculation of order of magnitude of the geometric expanses shows that the dimension of this cube is prohibitive.

The geometric extent of the LCD imager is defined by its surface, S> LCD _> and an opening number: N (see conference report: Japan Display 92, p. 121-124). This number of apertures can represent the aperture of the projection objective, or a solid angle taken on the isocontrast curve (angle of view) or even the solid angle defined by the image of the light arc ( of the arc lamp) and the image focal length of the field lens. It is advisable to choose, among these three solid angles, the one which effectively limits the luminous flux. The formula for the geometric extent of the LCD imager is: ELCE> DCL ⁼ ^X ^(an opening GTE solid)

E_QD: geometric extent

^ LCD ^: LCD imager surface

ELCD = S _LCD . π / (4.N ² )

N being the number of apertures of the objective. For a cube cube of side C, of angular acceptance ± α, the geometric extent is defined as being the product of the surface of the face of the cube by the solid angle defined by the angular acceptance:

Ec = C ² .2π [l - cos (α)]

If Eç> EJ_ _), there is no loss of flow for geometric reasons. On the other hand, if Eç <E] _çγ) we can define a geometric efficiency of mixing, βg which is a factor of flux transmission:

_ C ² .8. [L - cos (α)]. N ^; ^C g ^{_ C} C '^ CD ^_ _ç

^Û LCD

Typical digital application

N = 4.5, S _L CD = ^{84 cm2} (5.5 "diagonal LCD in 16: 9 format): E _ç_) = 3 cm ² .sr (326 mm .sr)

Minimum dimension of the mixing cube to adapt the geometric ranges:

ELCD = E _C -> C = 19.5 cm If the cube is smaller, the geometric mixing efficiency is: βg ≈ 2.6.10 "3.C ² (C expressed in cm)

For example for the standard dimension C = 5 cm, we find eg = 0.07.

The size of the mixing cube and, consequently, its mass, must therefore be large in order to take advantage of the efficiency of the polarized mixture.

In order to increase the luminous flux, it has been proposed to combine several video projectors which would project the same video image. The rays are combined thanks to the use of flat mirrors upstream of a single projection object (see document Japan Display'92 - 117). This solution has the following advantages: no adjustment of image convergence, compatibility with stereoscopy. On the other hand, it has the disadvantage of requiring as many sources of illumination (arc lamp) as there are projectors. This results in high energy consumption.

The object of the invention is to remedy these drawbacks. The invention therefore mainly relates to a liquid crystal video projector comprising at least:

- a light source emitting a light beam;

- a polarization splitter receiving the light beam and providing two orthogonally polarized sub-beams; - two liquid crystal imagers each placed on the path of a sub-beam;

- an objective receiving the two sub-beams after crossing of the two imagers; characterized in that it also includes a combination mirror preferably located in the vicinity of the entrance pupil of the objective, the imagers being substantially symmetrical to each other with respect to the combination mirror; and in that the two sub-beams modulated by the imagers converge towards the objective in two mean directions forming a non-zero angle between them and form two substantially distinct images of the source in the plane of the entrance pupil of the goal. The various objects and characteristics of the invention will appear more clearly in the following description given by way of nonlimiting example and in the appended figures which represent:

- Figures 1 to 3, architectures of three-color liquid crystal video projectors of known types;

- Figure 4, an embodiment of a video projector according to the invention

- Figures 5 and 6, explanatory illustrations of the projector of the figure

- Figure 7, an alternative embodiment of the system of Figure 4;

- Figures 8a and 8b, an application of the invention to a chromatic system;

- Figure 9, a diagram for illustrating a calculation of geometric ranges of the projector elements according to the invention. Referring to Figure 4, we will therefore describe a basic architecture of a projector according to the invention. This architecture essentially includes:

- a light source S;

- a polarization splitter SP;

- two sets of independent liquid crystal light modulators or imagers LCDA, LCDB;

- at least one mirror combining the modulated beams M3;

- a single projection lens L3.

The source S (arc lamp) and its parabolic reflector provide a beam of collimated white light FI. The separator cube SP gives two sub-beams F2, F3 whose directions of polarization are perpendicular to each other. Plates λ / 2 Ql, Q2 arranged on the paths of the sub-beams make it possible to adapt the direction of polarization of the sub-beams to the direction of each polarizer of the imagers LCDA, LCDB. The field lenses (L1, L2) arranged upstream of each imager cause the rays to converge so that they form images Tl, T2 of the arc of the source S in the vicinity of the pupil P of the objective Preferably , these two images T1, T2 of the arc of the source are juxtaposed.

According to FIG. 4, the mirrors M1, M2, M3 make it possible to bring the two sub-beams in substantially the same direction towards the objective L3. The mirrors M1, M2 transmit to the imagers LCDA, LCDB of the beams F4, F5 slightly inclined with respect to the normal to the imagers (for example 3 to 6 degrees). This tilting optimizes the contrast of the imagers between their on and off operation. The normal to the imagers are such that after reflection on the mirror M3, they are substantially collinear with the optical axis of the objective L3. The sub-beams after reflection of F5 on the mirror M3 therefore form an angle between. They form in the plane of the pupil of the lens L3, two images substantially distinct from the axis of the lamp S.

In an ordinary video projector, each field lens, placed upstream of each LCD imager converges the light rays by forming an image of the arc of the source in the plane of the entrance pupil of the L3 objective. By associating two imagers which share the same objective, it is necessary to ensure that each of the two arc images can be transmitted by the pupil P of the objective L3. The general idea of the invention consists in forming the two corresponding arc images close to one another, in the plane of the pupil of the common objective, and in using a plane mirror (or several) so that the rays coming from the two imagers all enter the pupil P of the lens L3.

The arc images must therefore be juxtaposed in the plane of the pupil P of the lens L3 while, of course, the images of the imagers must be superimposed in the plane of the projection screen. This is guaranteed by the collinearity of the axes normal to LCD imagers. The tilted illumination property of LCD imagers in video projectors is used for the juxtaposition of arc images, as shown in Figure 5.

To juxtapose the two arc images from the two imagers, the plane mirror M3 is used in the vicinity of which the arc images are formed. The axes x _a and x ( _j (in solid lines) are the median axes normal to LCD imagers. In order for the image of the LCDB imager reflected by the mirror M3 to be superimposed on that of the LCDA imager, the mirror M3 is placed along the bisector of x _a and xj-. Thus after reflection of the rays which have been modulated by the LCDB modulator, the axes x _a and xb merge into a single axis: x. This axis x is that of the projection objective L3. Thanks to the off-axis illuminations, along z _a and zj _j (hatched lines) and by shifting the mirror M3 relative to the point of intersection of the two axes x _a and xj- ,, it is possible to juxtapose the images of arc, ImA and ImB, formed by the two projectors. Thus no energy is lost in the combination of the beams. For this, it is necessary that the angles of illumination are of opposite signs between the of them spotlights. In FIG. 5, the letters "b" and "h" associated with the imagers respectively represent the bottom and the top of a video image.

Preferably, the two images of the source arc are in the plane of the pupil of the objective. It is advantageous to form the arc images in the immediate vicinity of the mirror, so as to make them almost contiguous, which minimizes the necessary opening of the objective and reduces the overall size.

In the foregoing, the images produced by one imager must therefore be superimposed on the images produced by the other imager. On the other hand, if a projection circuit, that of the LCDB imager for example is characterized by an illumination off-axis directed downwards (for example), the projection circuit of the LCDB imager must be characterized by an up-axis off-axis illumination.

As shown in Figure 6, this affects the polarization of the projected images.

Indeed, the direction of inclination of the illumination of the LCDA imager or

LCDB up or down depends on the diagonal according to which the input polarizer is oriented. The same goes for the direction of the mean inclination of the modulated beam relative to the output polarizer of the imager. In Figure 6, the

LCDA and LCDB are assumed to each belong to a projection circuit. To meet the following two conditions:

- tilting off axis in opposite directions; - superimposition of the images coming from the two imagers LCDA and LCDB, the polarizations of the projected images P ^ and Pg are necessarily perpendicular to each other. According to a preferred embodiment, the polarization directions are oriented at 45 ° from the video image. In particular in the case of a chromatic system, the separation using chromatic mirrors has led to elliptical polarizations, P ^ and Pg being the major axes of the ellipses which makes this orientation of the polarizations advantageous.

FIG. 7 represents an alternative embodiment of the invention according to which the two imagers LCDA and LCDB, instead of being along perpendicular axes, are along parallel or collinear sensitive axes. The two sub-beams F4 and F5 coming from the two projection circuits illuminating the imagers LCDA and LCDB are each reflected by mirrors M4 and M3 respectively so as to image the source in ImA and in ImB in the plane of the pupil of the objective L3.

The invention also includes a variant for which each of the two projection circuits comprises several imagers. In particular we can consider the association of two light modulator assemblies of conventional structure (described in FIG. 1), which would share the same light source and a common objective.

FIG. 8a represents the association of two projection circuits A and B. Each projection circuit comprises three LCD imagers as well as the conventional dichroic mirrors. The white source S (arc lamp + reflector) illuminates a polarization splitter cube SP which produces two sub-beams F2, F3 of white light whose main axes are shown in hatched lines. The sub-beams F2 and F3 are transmitted to assemblies respectively called projector A and projector B. Each projector A and B is as shown in FIG. 8b for example, and functions like the projector of FIG. 1. Each imager LCDN, LCDB , LCDR therefore makes it possible to modulate a beam of a primary color. The FS ^ beams. FSg of the two projectors A and B are recombined together using a mirror M3 (see Figure 4) or two mirrors M3, M4 (Figure 7) so as to form in the plane of the pupil of the lens L3 , two separate images of source S. The dichroic mirrors separate each of these two beams into beams of red, green or blue colored light.

Unlike the case of an ordinary projector, these colored beams are polarized. However, their polarization direction is not necessarily identical to that of the LCD polarizer. A λ / 2 plate adapted to the color concerned, placed just before each LCD makes it possible to make these polarization directions identical.

In FIG. 8a, it is assumed that all the LCDs of projector A undergo an off-axis illumination downwards, while all the LCDs of projector B undergo an off-axis illumination upwards. Thus, a combination mirror placed near the pupil plane of the objective makes it possible to orient the rays modulated by the projector B into the opening of the objective, while the rays modulated by the projector A pass under this combination mirror. .

An order of magnitude calculation makes it possible to verify that the solution presented above is realistic.

The 150 W metal halide arc lamps are characterized by an arc length of 5 mm, they are standard components in video projection. Their geometric extent is of the order of 150 mm ² .sr. This is determined by the size of the arc and the angular radiation pattern. The required dimension of the mixing mirror is calculated by considering that an image of the arc must be able to be formed on its surface. The value of the geometric extent of the lamp is equal to that of the geometric extent defined by the field lens and the image of the arc in the pupillary plane (see Figure 9). Let f be the focal length and D the diameter of the field lens. D is also the diagonal of LCD. Let d be the diameter of the image of the arc.

_ field lens area x arc image area

(focal distance)

-_ π ² d ² .D ² E = - x

16 f ²

For example, for E = 150 mm ² .sr, D ≈ 5.5 ", f = 250 mm, we find as the diameter of the image of the arc in the plane of the pupil of the objective L3: d ≈ 28 mm

Assuming that the mixing mirror M3 of the beams is inclined at 45 °, its minimum dimension, φ is therefore: φ = d / cos (45 °) = 40 mm

Considering that two arc images must be contained in the pupil, we deduce the number of apertures of the lens L3: N = f / 2 (2.d) N = 4.5 Illumination angle of the LCD : Arctg (i) = d / (2.f) -> i = 3 °

If the LCD imagers are illuminated at a higher angle, this imposes a larger lens aperture.

These numerical values are quite realistic.

The size of the separator cube must be sufficient so that its angular acceptance does not become a limiting factor of flow. By taking the formula of the geometric extent of the cube, and adapting it to the numerical value corresponding to that of a lamp, we deduce the required dimension of the separator cube, ie its side C:

E = C ² . 2π [l - cos (oc)] = 150 mm .sr -> C = 13 cm

The cube is therefore relatively large, but this is much less of a problem than when it is used as a mixer. Indeed, this time it is not in the path of the imaging rays. For the rays of illumination, the influence of aberrations is therefore much less important.

The system of the invention has the following advantages: - increased brightness, double at most, for a non-stereoscopic video projector;

- compatibility with the projection of stereoscopic images when the adequate video signal is supplied to each of the two projectors; - a common objective for all polarizations and primary colors which facilitates adjustments.

Claims

1. Liquid crystal video projector comprising at least:

- a light source (S) emitting a light beam; - a polarization splitter (SP) receiving the light beam and providing two orthogonally polarized sub-beams;

- two liquid crystal imagers (LCDA, LCDB) each placed on the path of a sub-beam;

- an objective (L3) receiving the two sub-beams (F4, F5) after crossing of the two imagers; characterized in that it also comprises a combination mirror (M3) preferably located in the vicinity of the entrance pupil of the objective (L3), the imagers (LCDA, LCDB) being substantially symmetrical one of the other with respect to the combination mirror (M3); and in that the two sub-beams modulated by the imagers converge towards the objective (L3) in two mean directions forming a non-zero angle between them and form two images of the source (S) substantially distinct in the plane of the pupil input (P) of the lens (L3).

2. Projector according to claim 1, characterized in that the polarization splitter provides two sub-beams whose axes form an angle of substantially 90 ° and in that it also comprises a first and a second deflection mirrors (Ml, M2) allowing the two sub-beams to be brought back to a common path, a third mirror (M3) being placed near the optical axis of the objective to reflect one of the sub-beams towards the pupil (P) of the objective (L3) while it lets pass the other sub-beam towards the pupil of the objective.

3. Projector according to claim 1, characterized in that the two sub-beams are recombined towards the objective (L3) using two mirrors (M3, M4).

4. Projector according to claim 1, characterized in that each imager (LCDA, LCDB) is associated with a lens (L1, L2) imagining the source in the plane of the pupil of the objective (L3).

5. Projector according to claim 4, characterized in that the images of the source are juxtaposed in the plane of the entrance pupil (P).

6. Projector according to claim 1, characterized in that it comprises a half-wave plate placed between the polarization splitter (SP) and each imager (LCDA, LCDB) to transmit to each imager 45 ° polarized light.

7. Projector according to claim 2, characterized in that the first and second mirrors (Ml, M2) transmit beams to the imagers (LCDA, LCDB) (F4, F5) making an angle with the normal to the imagers and the normals to the imagers are such that after having been brought in the same direction by the third mirror, they are substantially collinear with the optical axis of the objective (L3).

8. Projector according to claim 1, characterized in that it comprises an imager by primary chromatic range, placed between the separator and the objective; in each optical circuit located between the separator and the objective, dichroic separators separate the light of different wavelengths and each transmits it to an imager, the light coming from each imager being transmitted to the pupil (P) of the objective (L3).