TI中文支持网您好！
我北京闻亭泰科技术发展有限公司的光学工程师：吴丽萍。
我仔细读了2022/07/11TI出的文档：
Optical_Design_Guidelines_For DLP_Products_revA.pptx。此文档对于光学工程师帮助非常大，里面有大量的光学知识。对于文档中的知识还有一些疑惑，请帮忙解答，具体如下：
1.如下图,文档中是表述复眼/积分棒出射光斑到DMD后会变大3X左右，这个拉长是不是DMD的翻转角度引起的。比如：DMD翻转角度是17°，入射到DMD角度为34°，这个34倾斜导致光斑的拉长，拉长倍率计算如下：1/cos（34°）=1.2X。不知道我这个理解是否正确?
2.如下图，请确认TIR/RTIR棱镜空气间隙，以前的经验是TIR棱镜空气隙厚度<0.2mm，RTIR棱镜空气隙厚度为7-10um，与您写的文档正好相反，还请核实，谢谢！
3.如下图，红框部分是说入射到积分棒的角度最佳是在±26°~30°之间吗？设计积分棒的时候是否就按照这个角度设计，匀化效果最佳？我这样理解是否可以？
4.如下图，能详细说明一下zemax软件怎么实现的吗？主要是为了实现什么功能？
因最近正在设计新的光学引擎系统，您的解答对我帮助巨大。感谢您耐心的看完问题，期待您的回复，非常感谢！
敬候佳音！
Fortune Xu：
您好，感谢您对TI DLP技术的关注，您的问题已收到，会尽快给您回复，谢谢。
,
Jeremy_Wu：
请参考光学工程师回复：
1. The anamorphism we show is a 30% stretching not 3x. Generally yes, the anamorphism is caused by the oblique illumination angle onto the DMD. Your example is correct that for a TRP the tilt is 17deg and illumination angle is 34 degree although sometimes customers may increase the illumination slightly and tradeoff efficiency for contrast. Primarily the anamorphism is contributed by the light rays travelling through the prism. There is a difference in optical path length on one side of the illumination bundle than the other. For example, in the diagram showing a right angle prism (RTIR), you can imagine light rays on the left side would travel through less glass than the right side therefore causing a difference in optical path length which stretches the image. These anamorphism values are estimates and vary by architecture, design of the prism (i.e. refractive index, angles, etc.), DMD tilt angle, and illumination angle.
2.The slide is correct. The TIR prism on the left side of the slide shows an airgap in the projection path. This needs to be very small otherwise introduces significant astigmatism aberration that will be difficult to correct in the projection lens design which will affect image quality MTF. We recommend <10um but suggest customers simulate this in their design to see the effect on MTF. The RTIR prism has the airgap in the illumination path only. This can be larger (<200um) since we do not care about imaging quality of the illumination bundle on the DMD.
3.Yes, as it states in the text box, we recommend ±26-30° cone angle light input into the tunnel. This should provide reasonable homogenization of the light bundle at a reasonable tunnel length. The tunnel should be simulated to find a length where the uniformity meets the customer’s target. Of course if the cone angle into the tunnel is much smaller, then the tunnel would need to be much longer which grows the projector length/size. The tunnel needs several reflections (at least 4-5 in my experience) to properly homogenize the illumination.
4. The slide explains how to implement it in Zemax already. The table you see is from Zemax. This slide is mainly intended for optical designers, they should be familiar with it. The geometric image analysis on the right side will follow same instructions as the previous slide. The main purpose is to analyze the pupil of the lens so see if there is an F/# mismatch between the illumination and the projection. For example, if you illumination is designed for F/2.4 and projection lens is F/2.4, you want to make sure light is being properly captured and if there is excess light that the light can be apertured in the illumination to prevent this light from entering the lens since it doesn’t make its way to the image. It will be blocked by the projection lens aperture but could cause scattering. The reason the illumination pupil may not accurately match is again due to the anamorphism of the illumination which not only stretches the illumination on the DMD but also stretches the pupil if not properly maintained. We recommend blocking this light in the illumination using a ‘cat’s eye’ (elliptical) aperture for example.
,
WU LIPNG：
Thanks for your detailed replying.
I have understand the questions.
Best wishes.
