DLP470TP: Refractive Index of RTIR

Hi, Vivek. Cherng works with me and I asked him to clarify this info. He has worked with AR and imaging optics before.
But the provided info still seems incomplete:

1) You have mentioned to Cherng the DMD needs a 17 degree converging illumination beam and the beam coming out of it is 17 degree diverging light cone, but it is still not clear to us why and how it is linked to etendue of the LED. (Yes, I forwarded the illustration you sent me to him).


2) Same is the case with the refractive index. He ran some simulations in Lightroom and Zemax and 1.51 RI seemed adequate for total internal reflection. Sadly there isn't a reference Zemax design for bottom illuminated 4710 and only side illuminated 3010. So he designed his own TIR prism.
We also contacted several TIR prism manufacturers who do prisms specifically for DLP and none seemed to provide such high refractive index or even ready 4710 prisms and prism designs and all asked to provide custom design and custom order?


3) To be clear by slower optics you mean smaller aperture?


Please let us know if any of our provided info is lacking and if so what else we can provide to help you help us.

I'll ask Cherng to follow and participate in this conversation since he is the optical engineer.

Thank you Vivek.

1. Is this the case with direct angled illumination or also via a TIR prism? Can you please remind us why there are 3 states instead of 2 (one to direct light to absorber and another to direct light to the projection lens or eyepiece, and 3rd state for what?...)?

2. "The largest collection angle at TRP Pixel DMD could be 17 degree" - this doesn't seem clearly worded. You mean the illumination beam going into the DMD can be converging at 17 degrees or lower? And this is because the projection lens has to be placed at a certain distance from the DMD due to off-state beam separation and/or TIR prism so for proper contrast and brightness the lens aperture size is limited as can't be too big as will collect off-state beam as well but can't be too small as will simply be too dim? is that all?

If so, it seems like slower optics could be used with smaller collection angle with LEDs as well and it would only sacrifice brightness?

3. Okay, now it makes sense.

And again by slower optics you only mean smaller aperture right?

"I will try to find a reference design with bottom illumination share off-line with you."

Please do. You can PM it to Cherng or me.
Thanks.

"The incoming light to DMD needs to converge at 17 degree OR LOWER."

Okay, this is what I wanted to hear. So the DMD is on the focal point and a beam converging (at most, but optionally lower than) 17 degrees on the DMD will diverge from the DMD 17 degrees. Right?

This is all I believe but I'll discuss with Cherng if we have everything sorted out and also forward him the reference design. Thank you.

Cherng, I believe the misunderstanding came from the idea that the TIR prism refractive index has to be >1.7. It seems we need >1.7 if we want the illumination beam to converge 17 degrees. But we might not need a LED beam focused that much as we are not building a pico projector where we want as much lumens as we can get from a given LED. I believe in your design we managed to use a PMMA (1.5 RI) TIR prism as the beam wasn't converging onto the DMD at such "extreme" angle . Is my understanding correct?

Yes, the previous design the beam is not converging and we just use PMMA for the TIR prism. 

As Vivek mention, some converging angle to DMD will help improve the contrast. Thus, I think this is good input to improve the illumination optics design.  

Yes of course, it makes sense now.

I think since the DMD is on the focal plane and image plane, the illumination beam absolutely has to converge with LED illumination vs laser. I don't think with LEDs its easy to have a perfectly parallel beam to have a focal point both at the DMD and projection screen or eyepiece lens.

However, you can have any converging angle below 17 as long as brightness requirements are met for the specific use case. More converging angle just allows to focus a brighter LED with a bigger emitter onto the DMD.

But having a converging angle is not for improving contrast as far as I understood but just to get the illumination beam in an image plane on the DMD. I think contrast is determined by ensuring the 1) on, off, flat beams don't overlap and 2) the projection optics have a smaller aperture (slow optics).
But please do correct me if I'm wrong, Vivek.