• State Resolved
  • Locked Locked
  • Replies 19 replies
  • Answers 5 answers
  • Subscribers 28 subscribers
  • Views 3197 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DLP470TP: Single RGBW LED illumination results from a simulator. Are these values and this setup recommended and acceptable?

Mike Sh.
Expert 1405 points

Part Number: DLP470TP

For a head mounted display we decided the best option, if it will perform well, will be to use a single LED containing the red, green and blue emitters to save space and weight.

We found one candidate for the job, the Luminus Color Mix LED which can do at least 60 lumen output.

We ran simulations with it and th results are here:

At the very corner edges the DMD is 79.8% bright compared to middle. Anywhere closer than those tips it is over 80%, as can be seen in the simulation result images. Is this an acceptable value?

Here is how it was set up:

Second question I guess is are there any recommendation to what TIR or RTIR prism to use? My optical engineer used a PMMA prism pair for beam redirection due to the relative low brightness and no risk of plastic optics flexing or melting. We are a startup and haven't built a DLP optical engine before. Would appreciate answer to these two questions to know if we are good to go or whether we should change the prism and/or LED.

  • 0
    TI__Mastermind 23070 points
    Hi Mike,
    Thank you for provide detailed information. I will review this with our team and get back to you before end of this week.

    regards,
    Vivek
  • 0 in reply to Vivek Thakur
    TI__Mastermind 23070 points
    Hi Mike,
    Thank you for your patience.
    Illumination – The simulation results seems to be good. The intensity variation from center to corner is within the range of typical illumination design for DLP system. You may want to consider 10-15% overfill to get better uniformity across DMD.
    Your question:
    Second question I guess is are there any recommendation to what TIR or RTIR prism to use? My optical engineer used a PMMA prism pair for beam redirection due to the relative low brightness and no risk of plastic optics flexing or melting. We are a startup and haven't built a DLP optical engine before. Would appreciate answer to these two questions to know if we are good to go or whether we should change the prism and/or LED.

    Vivek – RTIR prism may be preferred because of lower cost. Typically TIR prism requires tightly controlled air gap which may increase manufacturing cost. The prism material may require refractive index around 1.73 at f/1.7 optics, which could be difficult in PMMA.

    LED – You decision about using all (R,G&B) in one package is good and it will simplify illumination optics. I am not aware of a projection product using this LED. I do not know about efficiency of this device. I suggest that you contact Osram and Luminus to find best LED options for you.
    I hope this answers your question.

    Regards,
    Vivek
  • 0 in reply to Vivek Thakur
    Expert 1405 points

    Thank you. I will ask my optical engineer to make the changes.

    But can you explain why high refractive index is needed for f/1.7 optics or why f/1.7 optics are assumed as default?


    Are RTIR prisms you mentioned available as off the shelf optical components or are they custom manufactured parts?

    There are some but I don't know if they are suited for this size DMD 

    The LED we simulated is Luminus Color Mix SBM-40-LC

    We could not find anything similar. If you know anywhere else we can look for a similar LED product please let me know.

    It may not be as bright for pico projection with only 133 lumens for the red channel but for a near eye display it seems perfect. I can contribute the project file if TI will find it useful.

    ''I am not aware of a projection product using this LED. I do not know about efficiency of this device.''

    What do you mean by efficiency? Power consumption or something else/more?

  • 0 in reply to Mike Sh.
    TI__Mastermind 23070 points
    Hi Mike,
    Thank you for your patience. Please find response to your questions:

    Your question:
    But can you explain why high refractive index is needed for f/1.7 optics or why f/1.7 optics are assumed as default?

    Vivek> F/1.7 optics is the fastest optics we can use with TRP pixel due to Etendue match for DMD @ 17 degree mirror tilt angle. The ON-state light out from DMD creates 17 degree light cone which requires a high index material for internal reflection in prism

    Your question:
    Are RTIR prisms you mentioned available as off the shelf optical components or are they custom manufactured parts?
    There are some but I don't know if they are suited for this size DMD

    Vivek> Mostly custom built RTIR prism. You need to work with your optical team to explore if you could you off the shelf prism.

    Your question:
    We could not find anything similar. If you know anywhere else we can look for a similar LED product please let me know.

    Vivek> I am not aware such LED in production. You may want to inquire if they could provide you a custom built or a prototype unit. The other option would be to get packaged bare LED die as per your requirement.

    Your question:
    What do you mean by efficiency? Power consumption or something else/more?

    Vivek > light output for a given power. At this stage power may not a concern for you.

    Regards,
    Vivek
  • 0 in reply to Vivek Thakur
    Expert 1405 points
    Thank you.

    "The ON-state light out from DMD creates 17 degree light cone which requires a high index material for internal reflection in prism"

    I assume this is not the case with laser illumination when the laser beam is perfectly collimated and the size of the DMD, correct?
  • 0 in reply to Mike Sh.
    TI__Mastermind 23070 points
    It applies to all light source because the basic principle of DLP based projection is Light source is imaged on DMD. The DMD image is either projected through projection lens and viewed using Eyepiece .

    regards,
    VIvek
  • 0 in reply to Vivek Thakur
    Expert 1405 points
    DMD is just a mirror, no reason why it should inherently create a light cone. There must be a specific reason as even the optical engineer I asked is puzzled. Please explain what it is.
  • 0 in reply to Mike Sh.
    TI__Mastermind 23070 points
    Hi Mike,
    Thank you for your patience.
    You are right, DMD is just a mirror and it reflects the light. In a typical design, the DMD is illuminated with a light which has angle/cone. As a result the reflected light has a cone.

    I want to clarify that normally our customers do not illuminate DMD with collimated laser beam for the following reasons:
    1. It is difficult to create large enough spot to cover complete DMD with overfill (10-15%). The efficiency will be low.
    2. Uniformity could be big issue.
    3. Such system also have higher speckle because angular diversity is less.
    4. Diffraction efficiency could be poor due to a very slow optics (f#).
    I hope this explains.

    regards,
    Vivek
  • 0 in reply to Vivek Thakur
    Expert 1405 points

    Thank you.

    We have a reference system where a multimode laser beam has its speckle reduced with a vibration flyeye lens acting as a speckle reducer or "deskeckler". However that is beyond this topic or my current project. I was just trying to understand what you are saying.


    I understand what we are talking about now however I don't understand when you say it has to be 17 degrees. Neither does my optical engineer, and we don't want to continue developing an optical engine fearing we may have missed something about these DMDs. Can you explain where the 17 degree cone comes from specifically please?

  • 0 in reply to Mike Sh.
    TI__Mastermind 23070 points
    Mike,
    The 17 degree was in reference to using LED to illuminate TRP DMD . In order to maximize the light out , the fastest optics one can use is F/1.7. The F/1.7 will have a +/- 17 degree light cone.

    In case of laser illumination, you will able to use a slower optics which will result in a smaller light cone.

    regards,
    Vivek
  • 0 in reply to Vivek Thakur
    Expert 1405 points
    I will pass this info to my optical engineer.
  • 0 in reply to Vivek Thakur
    Expert 1405 points

    Hello, Vivek. This is the response I got:

    "Could you please ask them how they define F(/1.7) for the illumination optics."

    To add to what he said, I can say what I personally suspect you mean. Please tell me if it is so, as I'm not an optical engineer myself:

    I assume fast optics is for collecting most light focused onto a DMD which we may not need due to low brightness requirements. Because the more you focus LED light to a smaller spot (such as DMD), the wider the focused beam cone is going to the DMD and the same going out, and since there's a gap between the DMD and projection lens due to the prisms the lens aperture would need to be larger for a larger LED beam cone going into it to fit it. I guess that's what fast optics mean for projection lenses?

    If not please explain by answering the question I forwarded from the optical engineer.

    Thank you.

  • 0 in reply to Mike Sh.
    TI__Mastermind 23070 points
    Hi Mike,
    The light collection angles at LED and at DMD are different and that may be causing confusion. By F/1.7 illumination optics, I am referring to angle at the DMD.

    To maximize light collection and efficiency, the Etendue of light source should match Etendue of the DMD. The Etendue of DMD is defined by active array area of DMD and light cone. The largest light cone angle for TRP DMD is 17 degree without compromise in image quality. The F/1.7 illumination optics is referring to light cone at DMD.

    On LED side, the collection light cone for a LED could be anywhere from 60 degree to 90 degree. For efficient collection, the LED Etendue defined by LED emitting area and angle should be smaller or equal to that of DMD (area & 17 degree).

    In case of laser illumination, the emitting area is relative very small and one can use smaller cone angle (slower optics) at DMD to match Etendue and efficiently collect all the light.

    I hope this answer the question.
    Regards,
    Vivek
  • 0 in reply to Vivek Thakur
    Expert 1405 points

    Thank you, but I'm sorry you're still not being very clear.

    > The largest light cone angle for TRP DMD is 17 degree without compromise in image quality.

    Why 17, and what do you mean by image quality and efficiency? Are we talking about how bright we can achieve with a given LED? OR more such as vignetting of the projection? Please specify what quality and efficiency means here.

    Does the 17 here have anything to do with the TRP tilt angle? Becuse if yes, it is again not clear how the tilt angle matters. My optical engineer said PMMA RTIR prisms appear to provide the necessary retroreflection for the LED and the DMD in Zemax and LightTools and your previous message which I forwarded to him is not clear and it is still not clear why high refractive index prisms are needed.

  • 0 in reply to Mike Sh.
    TI__Mastermind 23070 points

    DMDEtendue.pdfHi MIke,

    The 17 degree and F/1.7 is related to TRP DMD tilt angle.   I have attached a set  slide which may explain this subject. The slide example  is for 12 degree  VSP pixel architecture.  This are also applicable for 17 degree TRP pixel.

    regards,

    Vivek

  • 0 in reply to Vivek Thakur
    Expert 1405 points
    Thank you very much, I will pass on this info, hopefully it makes it clear. I have passed on your previous info already to two independent optical engineers I work with, one experienced in LCoS AR optics and another in camera imaging optics and both had hard time understanding the explanation.
    Hopefully this will clear things up.


    Thanks for the continuing help.