CA1243644A

CA1243644A - Illumination system reflector

Info

Publication number: CA1243644A
Application number: CA000486832A
Authority: CA
Inventors: Ralph J. Gagnon
Original assignee: Hughes Aircraft Co
Current assignee: Raytheon Co
Priority date: 1984-07-16
Filing date: 1985-07-15
Publication date: 1988-10-25
Also published as: IL75787A; JPS61502711A; IL75787A0; AU4546085A; AU569404B2; ES8609665A1; ES545193A0; WO1986000685A1; NO860827L; KR860700283A; EP0188541A1

Abstract

ILLUMINATION SYSTEM REFLECTOR

ABSTRACT OF THE DISCLOSURE
An illumination system reflector (18) which provides maximum energy to an aperture (25) with high uniformity and at high efficiencies, the reflector (18) being particularly adapted for use in liquid crystal light valve projection systems. A discharge lamp (34) is utilized in conjunction with an elliptical reflector (18), the axis of the reflector between the primary and secondary focus being tilted at an angle (.theta.) to the lamp axis. The axis of the ellipse (22) is displaced from the lamp axis a predetermined amount (DH), the combination of tilt (.theta.) and displacement (DH) being effective to increase illumination uniformity and efficiency at the aperture (25). Other system parameters, such as throw distance (CTHRO) and cone angle (.alpha.) also affect uniformity and efficiency for a typical design in which the parameters are selected to provide an optimized system being disclosed. A computer program simulation may be utilized for parameter selection. The vertical displacement (DH) parameter is shown to be related to the lamp envelope dimensions.

Description

~36~

ILLUMINATION SYSTEM REFLECTOR

1. Field of the Invention The present invention relates to an improved reflector for use in illumination systems and, in particularl the use of the reflector in liquid crystal light valve projectors.

2. Description of the Prior Art The development of the liquid crystal light valve has opened the door to substantial progress in the state of the art of high quality large screen projector systems. The details of the operation and use of such project~rs are described for example in U.S. Patent No. 3,824,002, entitled ALTERNATING CURRENT LIQUID CRYSTAL
LIGHT VALVE, issued on July 16, 1974 to T. D. Beard and assigned to the assignee of the present invention.
The system uses a powerful light source such as a xenon arc lamp to illuminate a liquid crystal light valve through collimating and polarizing optics. The xenon arc lamp is used with an elliptical reflector to illuminate the system light valvesO
Although the reflector utilized in the aforemen-tioned projection system provides light of satisfactory uniformity, the demand for projection systems with higher image quality, including increased brightness and contrast, has resulted in further efforts to provide improved projection light sourcesO

.
' ~43~

1 One prior art approach to the problem of uniform illumination of an image plane is illustrated in U.S.
Patent No. 1,275,120 wherein a surface of revolution is produced from an ellipse section with its major axis displaced to produce a ring of second foci. However, in order to achieve reasonable uniformity of intensity across the film or aperture, it is necessary to effectively defocus by placing the film or aperture at a distance from the second focus of the optical system, an unacceptable requirement for modern day projector systems.
Wilkinson U.S. Patent No. 3,720,460 improves upon the system disclosed in the aforementioned patent by using a surface of revolution constituting a single elliptical segment reflector with its axis displaced to ; a predetermined angle ~ related to: 1) the interfocal distance of the reflector, and 2) the aperture size to produce predetermined intensity across the illuminated aperture. In essence, the objectives are achieved by the rotation of the segment of an ellipse constituting a reflector surface about the generating axis of the ellipse passing through the first focus and at such an angle that a ring of foci circumscribes the aperture illuminated by the ligh~ source and reflector. Where the li~ht source is non-uniform in radiance along its length as is typical of arc discharge lamps wherein the intensity of the arc is greater at one of the electrodes, the arc immediately adjacent to that electrode is imaged at the aperture extremity. Although the Wilkinson lamp reflector provides a prede~ermined intensity distribution across an illuminated aperture, such as the film gate aperture of a movie projection system, the uniformity and efficiency levels required for certain applicatiQns, such as the liquid crystal light valve projection system described hereinabove, is no~ achieved by the Wilkinson .~

; ' ::

.

~4~6~

1 reflector. Further, additional optical components, including a relay lens, may be required in the Wilkinson system to project the image onto the aperture, thus reducing overall system efficiency.
Thus, what is desired is to provide an improved illumination system reflector in which the uniformity of the illumination distribution across an aperture is substantially increased over prior art systems and wherein the overall collector efficiency of the system in which the reflector is utilized is also increased.

SUMMARY O THE INVENTION
The present invention relates to an illumination system reflector which provides increased energy to an aperture with high uniformity and at high efficiencies, the reflector being particularly adapted for use in liquid crystal light valve projection systems. A
discharge lamp is utilized in conjunction with an elliptical reflector, the axis of the reflector between the primary and secondary focus being tilted at an angle to the lamp axis. The axis of the ellipse is displaced from (hereinafter also called "vertically displaced" for ease of description relative to the drawings) the lamp axis a predetermined amount, the 2~ combination of tilt and vertical displacement being effective to increase illumination uniformity and efficiency at the system aperture. Other system parameters, such as throw distance, cone angle, and lamp dimensions also affect uniformity and efficiency, a typical design in which parameters are selected to provide an optimized system being set forth. A computer program simulation may be utilized for parameter selection.

, .

: ;

~2~36~
3a An aspec-t of the invention is as follows:
In an illumina-tlon sys-tem for producing substantial-ly uniform light distribution across an aperture incluc~-ing a light source consti-tuting a lamp having a pair o~
spaced electrodes producing a region of light there-between having a brightness gradlen-t varying from a maximum brightness adjacent to one electrode and a lesser brightness adjacent to the second elec-trode, said spaced elec-trodes and the region of light therebetween generally falling on a line defining the system axis of said illumintion system, the improvement comprising: a reflector constituting an elliptical surface of revolution about a longitudinal system axis coinciding with a line be-tween said electrodes through said region of light, the axis of said elliptical surface being displaced a predetermined distance from and tilted a-t an angle relative to the system axis prior to rotation thereof, to an ex-tent tha-t a relatively uniform light distribution occurs across the aperture.

, , ~' ' , ~4~6~L

.
For a better understanding of the invention as well as further features and objects thereof, reEerence is made to the following description which is to b0 read in conjunction with the accompanying drawing wherein:
FIG. 1 illustrates the prior art Wilkinson system, FIG. 2 shows a cross-section of the reflector constructed in accordance with teachings of the present invention;
FIG. 3 illustrates a negatively decentered ellipse segment;
FIG. 4 shows the curve segment of FIG. 3;
FIG. ~ shows a cross-section of a reflector formed in accordance with the teachings of the present invention, FIG. 6 illustrates the geometry for computing the source image displacement provided by a shaped lamp envelope; and FIGS. 7a and 7b illustrate a computer simulation of a full reflector cross-section and an arc lamp utilized in conjunction therewith.

DETAILED DESCRIPTION OF THE INVENTION
._ FIG~ 1 illustrates the prior art Wilkinson projection light source and optical system. In particular, a reflector 10 nearly ellipsoidal in shape, has a light source 13 comprising the arc from a gas discharge lamp having a significant length L and located at the first focus fl as illustrated. The reflector 10 actually is a surface of revolution of the line 5 around the axis A of the reflector. The line S is a segment of an ellipse with its major axis M passing through the first focus fl and intersecting ~he axis A at a deviation , , .

~Z~36~

1 angle ~. The net result is that the second focus of the ellipsoidal shape reflector 10 is not a single point but a ring of foci f2M lying in the plane P. The angle Q is selected to be equal to an angle whose tangent substantially equals one half the maximum transverse dimension of an associated aperture 15 over the interfocal distance of the ellipse. Typically, deviation angle ranges from 1/2 to 3.
Given the parameters of the system including the light source radiance gradient, arc length L and rotation angle ~, the system provides light energy through the aperture 15 at a relatively uniform intensity.
The present invention improves upon the Wilkinson system by increasing both collection efficiency and the uniformity of energy distribution at an aperture in the manner hereinafter described.
Before going into the particular details of the present invention, some general considerations are set forth. In essence there is an inherent conflict between opposing requirements and the design of an elliptical or quasi elliptical reflector for an illumination system. Ideally, it is hoped to collect 100~ of the light from a lamp bulb with 100% uniformity inside a spaced apart prescribed aperture. However, it is clear that this means the illumination pattern should ideally be in the form of a step function which is not physically possible. The taper at the edge of the actual pattern may either be allowed to take place outside the collection aperture which gives uniformity at the expense of

3~ efficiency or inside the aperture which gives efficiency at the expense of uniformity. Obviously the best way to obtain high efficiency would to place a point source at the first focus of the reflector which is an ellipse of revolutionO All light will be collscted and delivered to the second focus. Efficiency will be 100% but , ~2~

1 uniformity will be 0% because all the light is concen-trated at the center of the aperture. To improve uniformity, the collection aperture can be moved away from the second focus. Generally, the desired improvement S is best achieved by moving the aperture towards the first focus. As one does so, the image of the point source becomes defocused and eventually the blur circle covers the whole collection aperture. At this point uniformity is better with the light covering the whole aperture but it is not acceptable for many applications.
If one tries to improve uniformity further by additional movement of the aperture, the light will spill across the boundary and efficiency will begin to decline.
The present invention has been provided to overcome the above challenges and to improve upon the aforemen-tioned prior art techniques. As will be explained with reference to FIG. 2, computer simulations have shown that a number of parameters affect the overall efficiency and uniformity of an illumination reflector system in addition to tho~e noted by Wilkinson. Efficiency ~and uniformity), for example, depends to a large extent on cone angle, lamp size tsince the lamp source is not a point source but occupies a defini~e volume), and the throw distance. But other system parameters including the radius of the aperture distance between ~he foci of the ellipse, the eccentricity of the ellipse, the lamp illumination cut-off angle, the offset of the source from the first ellipse focus, etc., also affect efficiency and uniformity~ Some of these parame~ers are fixed by the constraints imposed upon the illumination reflector system, i.e., the aperture size may be pre-determined as well as the throw distance by the physical confîguration of the system in which ~he reflector is utilizedO Although a computer program for simulating ~L2436~

l the reflector design can be utilized to optimize the variable parameters to provide, based on certain given input parameters, a re~lector which optimizes c~llection efficiency and illumination uniformity, it has been S found that appropriate values for one of the variable parameters, the vertical displacement of the ellipse axis from the lamp, or optical axis, will significantly increase efficiency and uniformity notwithstanding the values chosen for the other parameters (within pre-l~ determined limits). In addition, appropriate selectionof the variable tilt parameter will also enhance system performance.
FIG. 2 shows a cross-section of a portion of a reflector 18 and a curve 20 which is a segment of an ellipse. Th geometry utilized in constructing the reflector 18 in accordance with the teachings of the present invention is also shown. The intersection point 21 of the ellipse axis 22 and the reflector portion 20 is raised ~decentered) by distance DH above the system optical axis (i.e., the major ellipse is displaced an amount DH from optical axis A). Then the ellipse portion 20 is rotated about the optical axis A
at an angle ~ about the left side axial intercep~. The curve 20 above the tilted ellipse axis 22 is then a partial cross-section of the solid reflector. The surface of revolution thus is a section of an ellipse that has been tilted, decentered and then rotated about the optical axis.
Since the equation for an ellipse with a coordinate system origin at its center is (a ) (b ) ~l) - ' ~;~3~

1 ~he surface of revolution of the ellipse described in FIG. 2 is obtained by substituting r =~x2 + y2 (2) for y. By coordinate transformation the new reflector equation is:

a~
[(r - DH)cos~ - Zsin~]2 = 1 (3) In order to implement ~he design of the reflector, it is necessary to simplify equation (3) to generate the reflector's surface of revolution. The simplified equation of the form Az2 + Brz + Cr2 + Dz ~ Er + E = 0 (3A) can be derived. Without going into details of the calculation, A = cog20 + sin~; B = sin~cos~

C = sin2~ + cosi~ (3B) D = -BDH- 2cos0; E = -2CDH- 2sin0; F = CDH2 + 2DH sin0 (3C) a a a ~ii9~!4 1 Although FIG. 2 indicates that decentration occurs above the optical axis A, the ellipse is preferably placed below the optical axis (negative decentration) to bring illumination source 26 closer to the curve of the ellipse. This arrangement is shown in FIG. 3, points 1 trl=DH) and 2 denoting where the curve crosses the r axis (z=0). Points 3 and 4 denote where the curve crosses the z-axis (r=0), the curve between those two points being used to calculate the reflector design.
FIG. 4 shows the relevant part of the curve used to generate the reflector surface (the termination point of the curve 20 being set at ~ = 45 to illustrate the typical illumination cutoff angle provided by a selected illumination bulb). Rotation of this curve about the z-axis generates the desired reflector surface.
FIGo 5 shows, for illustrative purposes, a drawing (not completely to scale) of a reflector cross-section fabricated in accordance with the teachings of the present invention with typical dimensions included (rear opening dimension for insertion of an arc lamp 26 is included FIG. 7a).
After reflection off the elliptical surface, the light rays are directed to the aperture plane 24 as illustrated in FIGS. 2 and 5. A measure of system performance is obtained by evaluating each ray at the aperture according to whether or not it lies within the aperture 25 of diameter d. Efficiency is then defined as the ratio of the number of rays which pass through the aperture 25 to the total number of rays generated by the arc lamp 26. Uniformity (or non-uniformity) in turn is defined as the energy distribution over the aperture or 1 - Ener~ (Max) Energy (Min) , .

6~

1 Thus when e~uation t4) approaches zero, the uniformity is close to 100~. ~oth uniformity and efficiency are a function of the previously mentioned system parameters.
It should be noted that the reflector designed in accordance with the teachings of the present invention can operate directly into the aperture 25 without the need for an intervening illumination relay lensD The advantage is that the expense and light loss associated with the use of the relay lens are avoided. Since acceptable uniformity levels at relatively long throw distance can be achieved, a highly efficient design is thus provided.
FIG. 2 also illustrates an idealized version of illumination source 26 positioned on the optical axis A, a distance zoff from the first ellipse focus 28.
The illumination source 26 is actually a simplified version of the arc 29 struck between the cathade 30 and anode 32 of the arc lamp 34 utilized ~FIG. 7a) in the system reflector.
It has been discovered that the envelope shape 40 surrounding the arc 29 will, in fact, have an effecct similar to negative decentration~ In particular, it has been discovered that it is possible (in certain circumstances) to design the lamp envelope 40 to effectively provide the required negative decentration of the reflector without actual displacement of the reflector.
A calculation for estimating equivalent decentration caused by a real lamp envelope follows. If the lamp is in the shape of concentric spheres centered at the source, then light will be transmitted through without any deviation and there will be no source displacement.
But if the envelope is ovoid shape, so that ~he center of curvature of the surfaces is displaced from the arc 29, then the image of the arc can be displaced from .

36~gL

1 the arc itself (displacement of the arc image source from the optical axis A is equivalent to physically displacing the reflector from ~he optical axis).
Consider two circles of radlus R and R ~ T, (FIG. 6) where the thickness T between the circles is the thickness of the envelope 40. The illumination or arc source 26 is placed at a distance r from the inner surface of the envelope 40, so that the distance from the sour~e to the common centers (center of curvature 0 27 ) iS R-r. A horizontal axis is drawn through the source 26 and the upper part of Figure 6 is rotated about this axis. This generates an envelope which is similar to the actual arc lamp bulb envelope shown in FIG. 7a.
As shown in FIG. 6, the refraction of the source rays throu~h the envelope 40 generates an image 46 of the source which is displaced from the source 26A The image displacement is caused by the curved lamp envelope 40. The displacement occurs in a given cross-section and the direction depends on the particular orientation of the cross-section. The magnitude of this displacement D is determined as follows:
From the law of sines, sin~ = sint~ 0) (5) R-r R
Assuming the angles are small, it is concluded ~hat:

~ R-r Also note that the angle incre~ent relative to the center of the circles as the ray passes through the thickness T is equal to:
= T~ (7) ntR+T) . ,.

l Hence the ray exits the envelope at the new coordinate defined by the angle:
= r + 1 T
R-r n R+T
s When the ray passes through the thickness T it arrives at the outer surface with incidence angle 0 ~ ~
and therefore leaves with incidence angle e - na~r which is equal to:
~ l _ T (9) R+T

Equations (8) and t9) give the position and direction of the ray as it leaves the outer envelope 40.
This may be used to project the ray backwards to the point where it intercepts the vertical axis. This is the position of the image of the source. By trigonometry the distance from the center o~ the two circles to the image may be computed:
R (lO) Image distance = r + R + l T
R-r R+T n R*T
As to be expected, whether we choose n = 1.0 or T = 0.0, we get an image distance equal to R-r, the same as the source distance.
To get the displacement of the image 46 from the source 26, the source distance R-r is subtracted from the image distance. An expression with a complicated denominator results; approximating the multiplier of (R-r) in the denominator by 1.0 gives the following approximate expression for the source displacement D:
D - n-1 T (l _ ~)2 (li) , ~2~369~

1 This vanishes, as it should, when n = 1I T = 0, or r = R.
The way this would be used is as follows. Assume that the maximum lamp inner radius is equal to r, the thickness is T, and the index is n~ Then, if it is desired to achieve a displacement D, the radius of curvature of the inner envelope surface should be chosen equal to:
R = r (12) 1 - / nD
~(n-l)T
For example, suppose that the bulb diameter is 38 mm.
Then the outer radius is 19 mm. If the thickness is 3 mm, then the inner radius r is 16 mmO Suppose the index of refraction of the envelope is 1.46 and the desired displacement D is 0.2 mm. Then the radius of curvature of the inner surface should be equal to R = 16 = 30 mm (13)

4 6 x o .
~ 0.~6 x 3 while the radius of the outer surface would be 33 mm.
This geometry roughly is what is shown in F~G. 7a and will displace a source by 0.2 mm. At small required reflector decentrations which occur at small throw distances CTHRO, the source displacement caused by the lamp envelope may be used in lieu of an actual physical displacement. For relatively long throw dis~ances (i.e., 28" or greater~ the required decen-tration, typically about 0.78 mm, is large enough ~5 ~243~

1 to overshadow the image displacement caused by thelamp envelope, thus physical displacement is still required. In this latter case, the contribution to source displacement by the lamp envelope need not be taken into account. In most situations in which liquid crystal light valve projectors are utilized, the throw distance CTHRO is sufficiently long to overshadow bulb caused image displacement.
As set forth hereinabove, a computer program has ~een devised which responds to various input parameters, including tilt ~ and decentration DH, to provide an optimized reflector simulation such that light generated by the source (aperture lamp) is reflected to the aperture substantially uniform in energy distribution, the collection efficiency at the aperture being relatively high. Although the computer program per se is not considered part of the present invention, some of the results of the computer simulation are set forth hereinafter to illustrate the multiplicity of additional reflector parameters which effect uniformity and efficiency and which should be considered in the reflector design. Typical input parameters used in the program with respect to the ellipse (see FIGS. 2-4) and which affect uniformity and efficiency are: eccentricity (c/a~; throw distance (CTHRO); cone angle (~); ellipse diameter (2a); the previously noted decentration (DH);
and tilt (~); aperture offset (ZAP); source offset (ZOFF); aperture size (d); and focus-to~focus separation (distance between first ellipse focus 28 and second ellipse focus 31). Input parameters related to the source 26 (see FIG. 4) include lamp diameter; and lamp illumination cutoff angle (XLU - XLU2)o In ~he compu~er optimization, the illumination system was redesigned for various cone angles, throw distances, lamp sizes, ., ;

~lZ~3644 1 eccentricities, focus-to-focus separations, decentrations and tilts. Measures of uniformity and efficiency were provided, the computed efficiency being multiplied by a nominal value (typical}y 0.7) to account for energy which is generated outside of the computational model.
Among other observations, it has been ascertained that as the gap size (distance between electrodes 30 and 32, FIG. 7a) increases, collection efficiency decreases due to the increase in arc size; as lamp envelope size increases, efficiency decreases since the lamp envelope will block some of the ligh~ from the reflector to the aper~ure; ~hat efficiency increases with larger cone angles (the cone angle being limited by the design of the projector lens provided, for example, in the liquid crystal valve projector system) and that efficiency and uniformity decrease with very long or short throw distances CTHRO.
Design data for the reflector shown in FIG. 5 which has been fabricated in accordance therewith follows:

Ellipse: Data Ran~es Eccentricity 0.958 0.947 to 0.962 Throw distance 35 in ---Cone angle 5 ---Diameter 7.942 in ---Decentration -0.788 mm-0.888 to -0.688 Tilt 0.474 0.574~ to 0~374 Aperture offset -6,06 in ---Source offset -0O280 mm ---Aperture 2.00 in ---Performance:
_ Efficiency 0.54 ~~~
Uniformity 0.08 ---364g~

1 The above data was generated by initially assuming a 35 inch throw, a 5 cone angle and utilization of a 2.2 KW lamp generating 80,000 lumens. The collection efficiency of 0.7~ x 0.7 - 0.54 and uniformity of 0.06 is considered to be an excellent reflector design.
Typical ranges for these parameters for providing a highly efficient illumination system with high uniformity at the aperture are set forth in the right hand column. Ranges in performance factors are also set forthr The variations and parameters in the above ranges will provide an illumination system reflector which meets the requirements for liquid crystal light valve projector systems~ The tilt and dec~ntration concepts described hereinabove and the selection of the other reflector parameters c~n also be utilized to design reflectors for use in other projector systems, such as motion picture projectors.
FIG. 7a illustrates a more detailed view of the reflector design shown in FIG. 5, arc lamp 34 being 2~ centered on the optical axis A. Reflection of the rays 50, 52...88 emanating from arc 29 by reflector 18 is also illustrated.
FIG. 7b is an enlargemen~ of the cathode region and shows the illumination arc 29 represented by a "ball of fire", or source 26, for computer simulation purposes and represents the maximum light brightness produced by the arc tthe arc actually comprises light having a brightness gradient varying from the maximum adjacent cathode 30 to a lesser brightness adjacent anode 32). The position of one of the rays 50 tangle ~) and its direction is also shown in FIG 7b. The computer simulation assumes that the lamp illumination cutoff angles are XLU and XLU2~ rays emanating ~rom source 26 between these two angles being reflected from the inner surface of reflector 18.

, ., -: LZ~36~

1 The present invention thus provides a technique for improving the efficiency and uniformity characteris-tics of illumination system arc reflectors used in liquid crystal light valve projector systems and other types S of projection systems. In essence, the shape o the reflector is determined to provide maximum efficiency and uniformity given certain predetermined parameters, such as the physical configuration of the lamp and the throw distance required in the reflector system. Among the parameters which can be varied given the fixed, pre-determined parameters, the ~ilt of the ellipse axis from the optical axis, and the offset or decentration of the ellipse from the optical axis, contributes significantly to increased efficiency and uniformities; the latter to a greater degree than the former. Other parameters can be adjusted to obtain the combination of parameters which provide the optimum performanceO In this regard, although cone angle, lamp size and throw distance have been determined to significantly affect efficiency and uniformity, they are less likely to be variable since they are generally determined by other system constraints.
One important feature of the present invention is the discovery that lamp envelope size can influence the decentration parameter and with a proper design for smaller throw distances, physical displacement of the ellipse surface to fulfill the decentration requirement may be eliminated.
While the invention has been described with ref-erence to its preferred embodiment, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the true spirit and scope of the invention. In addition, many modifications may be made to adapt to a particular situation without departing from its essential teachings.

Claims

What is Claimed is:

1. In an illumination system for producing substantially uniform light distribution across an aperture (25) including a light source (26) constituting a lamp (34) having a pair of spaced electrodes (30, 32) producing a region of light (29) therebetween having a brightness gradient varying from a maximum brightness adjacent to one electrode and a lesser brightness adjacent to the second electrode, said spaced electrodes (30, 32) and the region of light (29) therebetween generally falling on a line defining the system axis of said illumination system, the improvement comprising: a reflector (18) constituting an elliptical surface of revolution about a longitudinal system axis (A) coincid-ing with a line between said electrodes (30, 32) through said region of light (29), the axis (22) of said ellip-tical surface being displaced a predetermined distance (DH) from and tilted at an angle (.theta.) relative to the system axis (A) prior to rotation thereof, to an extent that a relatively uniform light distribution (50 through 88) occurs across the aperture (25).

2. The system of Claim 1 wherein the axis (22) of said elliptical surface (18) is displaced from (D) and tilted to a predetermined angle (.theta.) from the system axis (A) prior to rotation about said system axis (A) to increase the amount of the light distribution across said aperture (25).

3. The system of Claim 2 wherein said elliptical surface (18) is configured based upon throw distance (CTHRO), cone angle (.alpha.), eccentricity (c/a),ellipse diameter (2a), aperture offset (ZAP), source offset (ZOFF) and aperture size (d).

4. The system of Claim 2 wherein the shape of said spaced electrodes (30, 32) are confined within an envelope (40) which is such that the image (46) of the light (26) adjacent to said electrode is displaced therefrom.

5. The system of Claim 1 wherein said region of maximum brightness of said light source (26) is displaced from the first focus (28) of said elliptical surface by a predetermined amount (ZOFF) to increase the uniformity of the light distribution across said aperture (25).

6. The system of Claim 1 wherein said aperture (25) and the first focus (28) of said elliptical surface is spaced apart prior to rotation about said longitudinal axis a selected distance (DH+(a-c)sin.theta.) to increase the uniformity of the light distribution across said aperture (25).

7. The system of Claim 1 wherein the angle (.theta.) between a line extending from the extremity of said aperture (25) to the extremity of said reflector (18) and the axis of said system (A) is selected prior to rotation about said longitudinal axis to increase the uniformity of the light distribution across said aperture (25).

8. The system of Claim 1 wherein said displaced predetermined distance (DH) is in the range from about 0.688 to about 0.888 millimeters.

9. The system of Claim 2 wherein said angle (.theta.) is in the range from about 0.374 to about 0.574 degrees.

10. The system of Claim 1 wherein the displacement (DH) increases the light collection efficiency at said aperture (25).

11. The system of Claim 6 wherein said selected distance is such that the image displacement produced by the envelope of said lamp is insufficient by itself to provide the required displacement (DH) of said intersection point.

12. The system of Claim 1 wherein said illumination system is utilized in a liquid crystal light valve projector system.