Patent Application
20090002579
Not Applicable.
The present application relates generally to contrast compensation for liquid crystal displays, and in particular, to contrast compensation of liquid crystal displays used in high light flux projections systems.
Liquid-crystal displays (LCDs) are widely used in projection displays for large screen televisions and monitors. One particularly successful LCD-based projection system is a WGP-based LCoS microdisplay system, which uses both wire grid polarizers (WGPs) and liquid crystal on silicon (LCOS) panels. This projection system, which has been proven to exhibit both high resolution and high image contrast when compared to other microdisplay technologies such as transmissive liquid crystal (xLCD), digital light processor (DLP), and direct-view LCD, typically uses three or more microdisplay panels (e.g., for the red, green and blue colour bands) to improve on-screen brightness. In addition, in order to enhance the on-versus off-state sequential image contrast of the projection system, a moderately low magnitude linear retarder is typically placed before each microdisplay panel to compensate for residual birefringence of the microdisplay panel in the off-state. As is well known in the art, this residual off-state birefringence typically leads to off-state leakage, which manifests as a bright dark-state that is very obvious when displaying dark video content, and which significantly lowers the on-state/off-state contrast ratio. The use of a moderately low magnitude linear retarder to compensate for the even lower magnitude residual off-state birefringence of the display panels provides contrast compensation has been shown to significantly improve the contrast ratio.
For example, consider the conventional 3-panel WGP-based LCoS microdisplay projection system shown in
Each WGP 15, 14, and 13 is a polarizer/analyzer formed from a plurality of parallel micro-wires that transmits light having a polarization orthogonal to the direction of the parallel micro-wires and reflects light having a polarization parallel to the direction of the wires (e.g., if the polarizers are designed to pass horizontal or P-polarized light, as illustrated in
As discussed above, the moderately low magnitude linear retarders 21a, 21b, and 21c, are compensating elements used to improve the contrast performance level of the projection system, which is otherwise limited by the residual birefringence of the LCoS panels in the dark (e.g., off) state. For example, in the absence of the moderately low magnitude linear retarders 21a-c, the P-polarized polarized light that illuminates each microdisplay panel in the off-state is slightly elliptically polarized upon reflection due to the residual birefringence of the LCoS panels 20a-c. When the elliptically polarized light, which contains both a P- and an S-component, is transmitted to the corresponding WGP 15, 14, 13, the S component is reflected to the X-cube thus allowing dark state light leakage onto the large screen and limiting the contrast of the projection system.
Use of the moderately low magnitude linear retarders 21a-c improves the contrast level by providing in-plane retardance that compensates for the retardance resulting from the residual birefringence in the LCoS panels 20a-c. In particular, each moderately low magnitude linear retarder 21a, 21b, and 21c introduces a phase retardance that cancels the linear retardance resulting from the inherent birefringence of the corresponding LCoS panel. In general, the term “in-plane retardance” refers to the difference between two orthogonal in-plane indices of refraction (at a predetermined wavelength) times the physical thickness of the optical element. Since each low magnitude linear retarder 21a, 21b, and 21c is required to provide a predetermined amount of in-plane retardance, they are often configured as A-plates (i.e., an optical retardation element having its extraordinary axis oriented parallel to the plane of the plate). For a vertically aligned nematic (VAN) LCoS panel the linear retardance resulting from the inherent birefringence in the off-state is approximately 2 to 5 μm across the entire visible band. Accordingly, the moderately low magnitude linear retarders 21a, 21b, and 21c are typically required to exhibit approximately 10 nm to 20 nm A-plate retardance. Since the moderately low magnitude linear retarders 21a, 21b, and 21c are used to provide this relatively low magnitude linear retardance they often termed trim retarders.
Notably, these trim retarders 21a-c are typically oriented such that their slow axes are configured at approximately orthogonal azimuthal alignment to the slow axes of the LCoS panels 20a-c (i.e., termed “crossed axes” configuration), while their fast axes are configured at approximately orthogonal azimuthal alignment to the fast axes of the LCoS panels 20a-c. The terms slow axis (SA) and fast axis (FA), as used herein, refer to the two orthogonal birefringent axes when the linear retardance is measured at normal incidence. Notably, the SA and FA locations change with off-axis illumination as well as reversing the SA/FA roles for a negative out-of-plane retardance component at a large angle of incidence.
Since the slow axes of the trim retarders 21a-c and the slow axes of the LCoS panels 20a-c are configured at orthogonal azimuthal orientations, the role of the fast/slow axes switches from the trim retarder 21a-c to the LCoS panel 20a-c for normal incidence light. In other words, light having a specific polarization is alternately delayed more then less, or vice-versa, in the trim retarder 21a-c and the LCoS panel 20a-c, respectively. If the linear retardance of each trim retarder 21a-c matches the linear retardance of the corresponding LCoS panel 20a-c in the off-state, the net effect is zero relative delay for the incoming polarization, and as a result, an unchanged polarization (i.e., the output light is not elliptically polarized). The corresponding WGP 15, 14, 13 and/or optional clean-up polarizer then rejects the output light so that the dark-state panel leakage does not appear on the screen. Since the trim retarders 21a-c do not alter significantly the throughput of the panel on-state, the resulting sequential contrast (full on/full off) is excellent.
While each trim retarder 21a-c should, in theory, provide a linear retardance that matches the linear retardance of the corresponding LCoS panel 20a-c in the off-state, in practice, the linear retardance of both the LCoS panels 20a-c and the trim retarders 21a-c tends to vary within each component due to manufacturing tolerances in device thickness and material birefringence control, as well as operational drifts (temperature, mechanical stress etc). As a result, it is more common to provide a trim retarder that exhibits a higher linear retardance than the residual off-state retardance exhibited by the corresponding LCoS panel to ensure adequate compensation. For example, a trim retarder with a linear retardance of 5 nm (at λ=550 nm) could be provided to compensate for a vertical aligned nematic (VAN) LCoS exhibiting 2 nm of linear retardance at normal incidence (at λ=550 nm).
As is known to those skilled in the art, this mismatch in linear retardance requires offsetting of the optic axis of the trim retarder 21a-c, relative to the nominal crossed axes configuration described above. In other words, the trim retarder is mechanically ‘clocked-in’ by rotating its azimuth orientation away from the crossed-axes configuration until an increase in the contrast ratio is experimentally observed. This practical assembly is shown in
Various technologies have been used to fabricate trim retarders. For example, some examples of materials used to form trim retarders include uniaxially stretched polymer films such as polyvinylalcohol (PVA) or polycarbonate (PC) films, uniaxially aligned films of liquid crystal polymer (LCP) material, non-tilted biaxial organic foils such as cellulose acetate, molecularly birefringent inorganic crystals, and inorganic thin films.
In much of the prior art, trim retarders are fabricated as true zero-order trim retarders. For example, trim retarders are often fabricated from polymer films that have been stretched to provide a relatively low magnitude retardance. However, for materials having a relatively high birefringence, such as some inorganic crystals and/or LCP materials, forming a true zero-order retarder is challenging. For example, in order to fabricate a zero-order trim retarder having about 10 nm of linear retardance in the visible region, a quartz wave plate (configured as an A-plate, and having a birefringence's of 0.009 at 550 nm) would need to be approximately 1.1 microns thick. Even a similar quartz plate configured as a zero-order quarter-wave plate would require a thickness of 10-20 microns. I practice, it is very difficult to polish birefringent crystal plates to physical thicknesses less than about 100 microns (e.g., they are too thin for easy fabrication and handling).
One approach to fabricating a moderately low retardance trim retarder with materials having a high birefringence is to use a dual-layer configuration. For example, in one embodiment two birefringent crystal plates having different magnitudes of linear retardance are oriented in a crossed-axes configuration to form a pseudo zero-order retarder. In fact, commercial quarter-wave plates are often fabricated by laminating two quartz plates having their slow axes oriented substantially orthogonal to each other, wherein the difference in thickness of the two plates provides zero-order quarter-wave retardance. In other embodiment, two birefringent crystal plates having the same retardance (e.g., two half-wave plates) are oriented at a non-90 degree relative azimuthal angle offset to form a pseudo zero-order retarder. Unfortunately, both of these embodiments require an increased number of components and thus, are associated with increased manufacturing costs. In addition, there is also increased cost related to the required relative alignment.
Another approach to providing a moderately low retardance is to a use multiple-order trim retarder. For example, a tenth-order quarter-wave retarder (e.g., 5.25 waves) should behave similarly to a zero-order quarter-wave retarder (e.g., 0.25 wave). Although calculations have shown that clocking characteristics of multiple order retarders may be similar to their zero-order counterparts, they are not generally ideal for trim retarder applications due to their high dispersion. For example, consider the theoretical linear retardance (at normal incidence) as a function of wavelength of a 0.25 waves quartz retarder and a 5.25 waves quartz retarder, illustrated in
In U.S. Pat. No. 5,576,854 to Schmidt et al, contrast compensation is provided with an approximately quarter-wave retarder (e.g., 0.27 waves). More specifically, quarter-wave retardance (e.g., 0.25) is used to compensate for skew ray depolarization of the MacNeille polarization beam-splitter (PBS), while the additional retardance above a quarter-wave (e.g., 0.02 wave) is used to compensate for birefringence in the LCD panel. Unfortunately, since compensator requirements for WGP-based polarization beam-splitting devices can differ significantly from those based on a MacNeille PBS, this approach has not provided a successful solution to contrast compensation in WGP-based LCoS microdisplay projection systems. In fact, in U.S. Pat. No. 6,909,473, it is stated that performance results indicate that the use of the approximately one quarter-wave plate compensator can even degrade contrast ratio in WGP-based LCoS microdisplay systems.
In WO 01/79921 A2, Candee et al also propose using a quarter-wave plate to provide skew ray compensation of the MacNeille PBS prisms. In addition, Candee et al propose two different embodiments for compensating residual off-state birefringence in the reflective panel. In the first embodiment, the above-mentioned quarter-wave plate is slightly misaligned. In a second embodiment, an additional quarter-wave plate or an additional half-wave plate is misaligned. More specifically, the orientation of the second quarter-wave plate or half-wave plate is slightly rotated from the principal coordinate plane of the imager panel (also S- and P-plane of the optical system). Notably, this approach is also not expected to provide a successful solution to contrast compensation in WGP-based LCoS microdisplay systems. For example, as discussed above, the use of a quarter-wave plate is associated with poor performance in WGP-based LCoS microdisplay systems, whereas the use of a half-wave plate is expected to cause the panel on-state brightness to decrease such that the resulting sequential contrast (full on/full off) is negatively affected and system throughput degraded. In addition, aligning the second half-wave plate to approximately half the angle offset of a second quarter-wave plate from the S- or P-axis does not work.
In would be advantageous to provide an improved trim retarder for WGP-based LCoS microdisplay systems.
The instant invention relates to contrast compensation in liquid crystal display (LCD) projector systems, where the LCD exhibits small magnitude residual in-plane retardance in the off-state. The contrast compensation is provided with a near zero-order half-wave retarder. Advantageously, the near half-wave retarder delivers optimal dark-state crossed polarization output without appreciably degrading the on-state, in WGP-based LCoS projection systems. Furthermore, the near half-wave retarder is readily fabricated using a single-layer birefringent crystal with a manageable thickness tolerance. In addition, the near half-wave retarder exhibits an angular sensitivity comparable to prior art small magnitude trim retarders.
In accordance with one aspect of the instant invention there is provided a liquid crystal display projection system comprising: a reflective liquid crystal display panel having residual off-state birefringence at a predetermined wavelength; and a trim retarder for compensating for the residual off-state birefringence of the reflective liquid crystal display panel and for increasing an on-state/off-state contrast ratio of the liquid crystal display projection system, wherein the trim retarder includes a single-layer retarder element having an in-plane retardance for compensating for an in-plane component of the residual off-state birefringence, the in-plane retardance shifted from a half-wave at the predetermined wavelength by a predetermined amount, the predetermined amount less than about 0.15 wave at the predetermined wavelength.
In accordance with one aspect of the instant invention there is provided a method of improving contrast ratio in a liquid crystal display projection system, the method comprising: providing a trim retarder for compensating for residual off-state birefringence of a reflective liquid crystal display panel in the liquid crystal display projection system, the trim retarder including a single-layer retarder element having an in-plane retardance for compensating for an in-plane component of the residual off-state birefringence, the in-plane retardance shifted from a half-wave at the predetermined wavelength by a predetermined amount, the predetermined amount less than about 0.15 wave at the predetermined wavelength.
In accordance with another aspect of the instant invention there is provided a method of improving contrast ratio in a liquid crystal display projection system, the method comprising: determining a residual off-state retardance of a reflective liquid crystal display panel in the liquid crystal display projection system; determining a first in-plane retardance for compensating for the residual off-state retardance and for increasing an on-state/off-state contrast ratio of the liquid crystal display projection system; and positioning a trim retarder in the liquid crystal display projection system, the trim retarder including a single-layer retarder element having a second in-plane retardance, the second in-plane retardance substantially equal to one of a half-wave plus the first in-plane retardance and a half-wave minus the first in-plane retardance, the first and second in-plane retardances determined at a same wavelength in a visible region of the electromagnetic spectrum.
Further features and advantages of the embodiments of the instant invention will become apparent from the following detailed description, taken in combination with the appended drawings, in which:
a is a plan view of a trim retarder used for contrast compensation in a WGP-based LCoS microdisplay system in accordance with one embodiment of the instant invention; and
b is a perspective view of
It will be noted that throughout the appended drawings, like features are identified by like reference numerals.
In order to provide an improved trim retarder for a WGP-based LCoS microdisplay system it is necessary to look at some of the preferred characteristics of trim retarders. Ideally, a trim retarder should be able to (a) deliver extremely low crossed-polarization leakage in the light off-state; (b) deliver nearly unchanged crossed-polarization output in the light on-state of the compensated panel versus uncompensated panel; (c) exhibit good mechanical clocking sensitivity for the initial alignment (i.e., when clocking is required) and for long-term alignment drift; (d) provide a high contrast ratio of on-state intensity versus off-state intensity over a given red, green, blue band or the entire visible wavelength band; and (e) exhibit good retardation magnitude and orientation uniformity.
Moreover, in addition to providing in-plane retardance, it may be advantageous for the trim retarder to provide out-of-plane retardance. While in-plane retardance is typically provided with an A-plate (i.e., an optical retardation element having its extraordinary axis oriented parallel to the plane of the plate), out-of-plane retardance is typically provided with a C-plate (i.e., an optical retardation element having its extraordinary axis oriented perpendicular to the plane of the plate). While a C-plate does not provide any net retardation for normal-incident rays (i.e., normal incident light is unaffected by the birefringence), rays incident off-axis (i.e., at an angle to the extraordinary axis) experience a net retardation that is proportional to the incident angle. Accordingly, out-of-plane retardance is typically provided to increase the field of view of LCoS panels. A C-plate is considered to be positive if the retardance increases with angle of incidence and negative if the retardance decreases with angle of incidence. Alternatively, a C-plate is considered to be negative if the retardance product Δnd is negative (e.g., if ne-no is negative). Since vertically aligned nematic (VAN)-mode LCoS panels typically function as +C-plates, it is common for the corresponding trim retarders to include both an A-plate component for compensating for the residual off-state in-plane retardance (i.e., A-plate retardance) and a −C-plate component for compensating for negative out-of plane retardance (i.e., −C-plate retardance). The resulting full-function trim retarders are conveniently termed A/−C-plate trim retarders.
As discussed above, there are a few materials that can be used to provide an A-plate retardance suitable for contrast compensation of WGP-based LCoS microdisplay systems, and which largely meet the above requirements. For example, trim retarders providing 10 to 30 nm retardance within the visible wavelength band have been fabricated using various deposition methods. These thin-layer structures have been shown to provide high contrast results over wideband while maintaining good azimuthal insensitivity to clocking. Another material that shows very high potential for fabricating trim retarders is molecularly birefringent inorganic crystal. The use of inorganic birefringent crystal in high light flux projector applications, such as digital cinema projection, is advantageous due to its high durability and/or stable birefringence when exposed to high light flux conditions. Unfortunately, since current grinding and polishing techniques are incompatible with providing birefringent crystals with low to moderate zero-order retardances, unless used in a dual-layer configuration, their use as trim retarders has not been established.
In accordance with one embodiment of the instant invention, a trim retarder fabricated from a relatively high birefringent material (e.g., a birefringent inorganic crystal or LCP layer) configured as an approximately half-wave plate (HWP) is used for contrast compensation of a WGP-based LCoS microdisplay system. Since the trim retarder is designed to provide approximately half-wave retardance in the visible region of the electromagnetic spectrum, the trim retarder is easier to fabricate and/or handle. In addition, since the trim retarder provides approximately zero-order half-wave retardance (i.e., is not a multi-order retarder) it is not highly dispersive across the red, green, and/or blue bands (i.e., it is not associated with a large dispersion in the visible region).
In order to further understand the use of the approximately half-wave retarder as a trim retarder, consider the general retarder solutions space illustrated in
In order to evaluate the off-state and on-state characteristics of a near half-wave trim retarder, the electro-optic (EO) curve is used. A VAN-mode LCoS panel was driven to a range of voltages (i.e., the on-state LCoS voltage was over 5V and the off-state voltage was 1.2V) and the normalized reflectance was converted to effective LCoS in-plane retardance. The measurement was performed for the Green band (e.g., 510 to 570 nm) with a f/2.4 cone of light. As an approximation, the effective in-plane retardance was inferred at λ=550 nm at normal incidence using
where Δnd is the single-pass retardance of the VAN-LCoS panel at a given voltage, λ is the illumination wavelength, and φp is the orientation of the slow-axis relative to the P-polarization (i.e., φp=45 degrees). The inferred panel in-plane retardance is plotted in
Referring to
The on-state reflectance of the uncompensated panel reaches a maximum value at about 5.2V voltage driving. With a BCE or QWP compensated panel system, the required voltage to reach the maximum reflectance is increased slightly (e.g., 5.35V). This can be optimized by Gamma correction found in typical panel operation. The use of either a HWP-minus or HWP-plus retarder compensator results in a slightly lower on-state maximum reflectance. In the case of a 30 nm offset from the HWP condition, the throughput reduction is about 4.5% (e.g., reaches a normalized reflectance of about 95.5%). This lost of brightness is due to birefringence interaction of retarder/panel in the on-state and does not include the insertion loss of absorption and reflection due to the additional optical component.
Since mechanical clocking of the trim retarder slow-axis versus the system ‘S’ and ‘P’ axes (and hence the panel slow-axis) is typically implemented for commercial LCD light engine assembly, one important characteristic of the trim retarder is the tuning range (e.g., the ideal trim retarder will have a relatively broad tuning range, or in other words, will exhibit good mechanical clocking insensitivity).
Referring to the top plot, the BCE and QWP-compensated panels show a relatively flat and symmetric response to the clocking of the retarder slow-axis over ±3 degrees, in the on state. Referring to the middle plot, the reflection of the QWP-compensated panel appears to change more with angular tuning, in the off-state. The resulting contrast tuning curves indicate that the BCE compensated panel provides about 1.7 degrees of FWHM (i.e., full-width half-maximum or 50% contrast bandwidth), whereas the QWP compensated panel only delivers about ⅓ as much contrast bandwidth (e.g., approximately 0.57 degree). In other words, the QWP-compensated light engine system is calculated to be 3× as sensitive to angular drift of the retarder element versus a BCE-compensated LCD system. This is a serious drawback. If fact, it is believed that one of the reasons that the prior art failed to provide quality contrast compensation with the approximately one quarter-wave retarder is the low mechanical angle tuning tolerance of the quarter-wave plate. In other words, a quarter-wave plate or near quarter-wave plate is extremely sensitive in its clocking behaviour.
In comparison to the BCE and the QWP, both the HWP-minus and HWP-plus retarder compensators are calculated to be slightly asymmetric in their response to the angular tuning in the light-on state. On the other hand, the off-state panel reflection for both the HWP-minus and the HWP-plus retarder compensators is nearly identically to the BCE/panel reflection. In fact, the associated contrast bandwidth for these two large magnitude retarders appears to be almost the same as the BCE-compensated LCD systems (e.g., at about 1.65 degree FWHM). In other words, these large magnitude retarders exhibit almost the same tuning sensitivity of a small magnitude BCE. Since the on-state light throughput of these large magnitude retarders is only a few percent worse than a corresponding BCE or QWP retarder compensated LCD system, it appears that it would be advantageous to use a HWP-minus or a HWP™-plus retarder as a compensator for LCD panel rather than a QWP retarder, where the relative angular clocking of the retarder/panel is de-tuned by, for example, thermal drift of the optical assembly.
It is noted that as the compensator retardance exceeds the QWP magnitude and approaches that of a HWP, the optimal retarder axes (fast and slow) of the compensator begin to deviate from the S- and P-axes even more. This is contrary to what is taught in WO 01/79921 A2, wherein the HWP axes are closer to the S- and P-axes than the QWP retarder axes. In the reflective LCoS projection system described herein, the QWP retarder compensator is double-passed, yielding a half wave net retardance upon reflection. Consequently, this QWP retarder has to be aligned with a small angular offset from either the P- or S-axis. This yields a small fraction of the half wave retardance upon double-pass transmission as the effective retardance to compensate for the small magnitude panel residual retardance. When the retarder compensator is slightly higher magnitude than a QWP in single pass, the double-pass retardance is larger than half-wave. In order to produce the same effective retardance for panel compensation, the deviation angle of the retarder compensator, having a retardance higher than a QWP but less than a HWP in single pass, from the P- or S-axis has to be increased.
The calculated linear retardance of the four compensators (e.g., BCE, QWP, HWP-minus, and HWP-plus) in the green channel are shown in
The calculated contrast ratio of a VAN-mode LCoS panel compensated with an optimally clocked retarder (i.e., for maxima contrast illustrated in
Notably, while the large magnitude retarder compensators (e.g., HWP-minus and/or HWP-plus) do provide somewhat reduced contrast compensation relative to the BCE0 and QWP0 compensators (e.g., see
Advantageously, the large magnitude retarders (e.g., HWP-minus and/or HWP-plus) are readily fabricated as a zero-order retarders using materials having high birefringence. Accordingly, the approximately half-wave retarders can be fabricated from inorganic birefringent crystals, such as quartz, which is known to be durable and stable in high light flux conditions. In addition, current grinding and polishing techniques can be used to fabricate the approximately half-wave crystal plate, as a zero-order retarder, with reasonable thickness tolerance.
Notably, the thickness tolerance of a zero-order approximately half-wave quartz retarder is much higher than a small magnitude retarder BCE0 fabricated with a single-layer quartz structure. For example, by assuming a quartz layer thickness tolerance of ±3% (including ±3σ range), the required thickness variation lies within ±0.1 μm for the small magnitude BCE0 quartz retarder. On the other hand, the Γ=245 nm HWP-minus and Γ=305 nm HWP-plus single-layer large magnitude quartz retarder will have approximately ±0.8 cm and ±1 μm thickness tolerance (assuming the same ±3% thickness tolerance and nominal targeting). Hence, it can be expected that the single-layer large retarder tolerance is some 8× to 10× better than the small magnitude zeroth order BCE retarder.
Further advantageously, the large magnitude HWP-minus and HWP-plus retarders provide good tuning angle sensitivity. In particular, the angular tuning characteristics are comparable to the BCE0. For example, consider the modeled contrast spectra of a compensated VAN-mode LCoS panel illustrated in
Clearly, while a small magnitude retarder such as BCE0 is ideal for compensating the residual off-state panel retardance of an LCoS panel in terms of contrast and azimuthal angle sensitivity, the near half-wave retarders (e.g., HWP-minus and/or HWP-plus) offer a reasonable compromise between contrast, azimuthal angle sensitivity, and suitable materials/fabrication techniques. In particular, the near half-wave retarder substantially maintains the contrast tuning insensitivity of the BCE and allows for a large thickness tolerance with an acceptable on-state throughput loss of a few percent. For a high flux panel system, where grinding and polishing of solid birefringent crystals is utilized to fabricate the retarder compensators, the near HWP retarder is the most cost-effective and delivers the required high contrast performance similar to the BCE0. In addition, by utilizing a single retarder layer, the overall system contrast is not impaired by the presence of circular retardance, which is typically found in multi-layer angularly-offset retarder compensator.
Notably, in these calculations, the simulation used single-crystal quartz material dispersion models for (ne, no) indices for the trim retarder and a typical LC model for the LCD panel. Of course, practical trim retarders may be implemented with a variety of technologies having a variety of dispersive properties.
Referring to
Referring to
Referring again to
The trim retarder 140 includes a first retarder element 142 that has approximately half-wave retardance. The difference in retardance between a true zero-order half-wave plate and the first retarder element 142 is selected to provide a retardance magnitude suitable for contrast compensation of the LCoS panel 130. In general, the difference in retardance will be between about 0.005 wave and 0.15 wave of the wavelength of interest, which if the trim retarder is used at 550 nm, corresponds to a linear retardance between 2 nm and 82 nm. More typically, the trim retarder will be required to provide between 10 and 40 nm of in-plane linear retardation (i.e., which, at 550 nm, corresponds to a retardance of about 0.02 wave and 0.07 wave). For example, in one embodiment the difference in retardance is approximately 0.055 waves, which at 550 nm corresponds to a retardance of approximately 30 nm. This retardance value is highly suitable for providing contrast compensation of a VAN-mode LCoS panel that exhibits an off-state panel retardance of approximately 2 nm at λ=550 nm. Note that when the difference in retardance from: HWP is approximately 0.055 wave, the first retarder element 142 will have a retardance of either 0.455 or 0.555 wave.
The first retarder element 142 is typically formed as a single-layer retarder element using a relatively high birefringence material, such as molecularly birefringent inorganic crystal or LCP, which can function as an A-plate. For example, in one embodiment the first retarder element 142 is fabricated as a near half-wave quartz retarder. In this embodiment, the quartz layer will be self-supporting or will be supported with a transparent substrate. In each case, the quartz layer will be configured as an A-plate such that its optic axis (i.e., which is also the slow axis for this uniaxial material) lies in the plane of the quartz layer.
The trim retarder 140 also includes a second retarder element 144 to increase the field of view of the LCoS panel 130. Accordingly, the second retarder element 144 will function typically as a C-plate. For example, in this embodiment, the second retarder element 144 is shown to include two form-birefringent anti-reflection (FBAR) stacks 144, each of which functions as a −C-plate and is coupled to a different side of the first retarder element 142. Each FBAR stack 144 is a periodic stack typically formed from alternating layers of contrasting refractive index materials. For example in one embodiment, each FBAR stack includes alternating layers of high and low refractive index materials. In another embodiment, each FBAR stack includes alternating layers of high, medium, and low refractive index materials. In each case, the thickness of each layer that contributes to the form-birefringence is limited to a fraction of the operating wavelength (e.g. a fraction of λ=550 nm). As is known in the art, a periodic stack of alternating index layers having thicknesses much less than the wavelength of light can be designed to form a zeroth order sub-wavelength grating (ZOG) that functions as a −C-plate retarder. Since the −C-plate retardance of these diffractive elements arises from the structure (form) of the alternating layers rather than from molecular birefringence, the alternating layers may be formed from normally isotropic materials. For example, some examples of suitable materials for the alternating layers include organic and inorganic dielectrics such as silica (SiO2, n=1.46), tantala (Ta2O5, n=2.20), alumina (Al2O3, n=1.63), hafnia (HfO2, n=1.85), titania (TiO2, n=2.37), niobia (Nb2O5, n=2.19), and magnesium fluoride (MgF2, n=1.38). FBAR coatings are discussed in further detail, for example, in U.S. Pat. No. 7,170,574, which is hereby incorporated by reference.
Advantageously, since the trim retarder 140 can be fabricated entirely from inorganic materials (e.g., if the first retarder element 142 is formed from quartz while the second retarder element 144 is formed from thin film inorganic dielectric layers) a very stable and durable optical retarder that is ideal for use in high light flux conditions is provided. In addition, since the FBAR stacks include an anti-reflection function, reflections from the first retarder element 142 are reduced without needing to provide an additional anti-reflection coating. In fact, this full-function A/−C plate trim retarder provides an excellent balance between simplicity, durability, and low manufacturing costs.
Further advantageously, since the A-plate retardance of the trim retarder 140 is provided with a near half-wave retarder 142, the trim retarder 140 will function in an azimuthal angle insensitive manner, and can be formed using micro-fabrication techniques that require relatively loose physical thickness tolerances. In addition, since the near half-wave retarder 142 provides a moderately high through-put in the light-on state, the contrast compensation is acceptable.
Note that a true half-wave plate (i.e., which provides exactly half-wave retardance or 0.5 wave) is not suitable for use as the retarder element 142. In particular, a true half-wave retarder is expected to provide a low throughput in the light-on state, and thus a reduced contrast ratio. In fact, the instant invention is distinguished from the prior art (e.g., WO 01/79921 A2 to Candee et a1) in that the retardance is selected to be shifted from exact half-wave retardance by a small amount. Since the retardance is shifted from true half-wave retardance, the trim retarder can be fabricated to provide a higher in-plane linear retardance than the residual in-plane off-state linear retardance of the VAN-mode panel, and such that the trim retarder can be clocked-in during assembly of the projection system.
Of course, the above embodiments have been provided as examples only. It will be appreciated by those of ordinary skill in the art that various modifications, alternate configurations, and/or equivalents will be employed without departing from the spirit and scope of the invention. For example, while the embodiment described with reference to
This application claims priority from U.S. Provisional Application No. 60/947,156, filed Jun. 29, 2007, which is hereby incorporated by reference.
| Number | Date | Country | |
|---|---|---|---|
| 60947156 | Jun 2007 | US |
