• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Subscribers 27 subscribers
  • Views 1556 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

8-bit Gray scale images issue when using modulated light

user4566018
Intellectual 630 points
Other Parts Discussed in Thread: DLP7000, DLP9000

Hi,

My application is on time-of-flight depth measurements. I am using the DLP300 with LED removed and using a modulated laser diode instead. When projecting binary patterns (1-bit) such as checkerboard, all goes fine.

However, when using 8-bit Gray patterns the displayed pattern is flickering. This does not happen when I use continuous (non-modulated) light. So the problem is the modulation of the laser together with the DMD. 

I assume that the modulation of the light interferes/overlaps with the bit-planes arrangement and the timings on the DMD.

Any ideas for solving this issue? it is very important

Thanks a lot,

Pepe

  • 0
    Intellectual 630 points
    I manage to solve the problem by increasing the time the laser in ON. This was the DMD setting time is less than my laser. All good
  • 0 in reply to user4566018
    TI__Intellectual 2305 points
    Pepe,

    Thanks for let us know that your solution with longer on-time works.

    Regards,
    A214529
  • 0 in reply to A214529
    Intellectual 630 points

    Hi A214529,

    I am coming back with this problem. Due to application constrains, I need to reduce the integration (exposure) time. The problem now come back: I see flickering when projecting 8-bit gray scale patterns. In fact, increasing the exposure time only reduced the problem, it did not solve it.

    Has anyone encounter the same problem when using modulated light and the DLP Lightcrafter (or any other MDM-based projections)?

    Can you help me overcoming the problem? is it even possible?

    Thanks in adavance, Pepe

  • 0 in reply to user4566018
    TI__Mastermind 30600 points
    Pepe,

    Can you describe in a little more detail how you're modulating the laser? What is your ideal system operation? You want to sync the mirrors with the laser pulses?

    Thanks,
    Paul
  • 0 in reply to Paul Goedeke
    Intellectual 630 points

    Hi Paul,

    My application is time of flight. In my application, the modulated light is a square wave modulated at 20 MHz, that is only active 1.25 ms each 10 ms. You can see this in the duty cycle of my application. That timing is repeated over time. In the case where Exposure time equals 10 ms, that is the light is continuously active, I see no flickering. As the exposure time increases, the flickering is less. My problem is that I need short exposure times (i.e, more flickering). May syncing the mirrors with the pulses is one solution, can you  think in others? how can I sync the mirrors with laser?

    Thanks a lot, Pepe

  • 0 in reply to user4566018
    TI__Mastermind 30600 points

    Hi Pepe,

    Thanks for your clarification.I believe one reason you're seeing the flickering has to do with how the DLP chip displays the 8-bit patterns. Each of the 8-bit patterns gets broken into 1-bit "bit-planes." I recommend that you read Chapter 4 of the user's guide for an understanding of how this works.

    How many 8-bit patterns are you trying to project? I speculate that perhaps you are running up against the speed limitations for the Lightcrafter (120Hz for 8bit patterns) and that your 8.75ms "dark time" as shown above is causing the flickering that you're seeing.

    The Lightcrafter does have input and output triggers you could use to possible sync your laser, but I'm not sure if that will help your issue.

    Thank you,

    Paul

  • 0 in reply to Paul Goedeke
    Intellectual 630 points

    Hi Paul,

    Thanks for your explanation. I am using static image mode and loading the picture below.  I think, but do not know how to calculate it, that in my exposure time do not fit an integer number of periods of the PWM controlling the mirrors. Can you help me in calculating it? I think if I chose my exposure time to be an integer number of those, it will work out fine. Thanks for your help!

  • 0 in reply to user4566018
    TI__Mastermind 30600 points
    Pepe,

    I did the math and it works out that what you're trying to do is exceeding the capabilities of the LCr3000. If you're trying to squeeze an 8 bit pattern into 1.25ms, that's approximately 800Hz frame rate. Unfortunately for 8bit patterns, the limit is 120Hz.

    This platform is not capable of those speeds. When you run a 10ms period like you were doing, that's 100Hz which is possible and that's why you're not seeing any flickering. I assume that you're using the LED enable signals to control your laser.

    Take a look at the DLP7000 or DLP9000 - those are our high speed devices and might be able to help you accomplish your goals.

    Thanks,
    Paul
  • 0 in reply to Paul Goedeke
    Intellectual 630 points

    Hi Paul,

    The limit of 120 Hz, does it also hold for static images? I think this is a crucial question, because after reading the documentation the pattern frame rate limitation of 120 Hz is for pattern sequence mode, right?

    I am NOT changing my pattern every 1.25 ms as you may have understood. It is a static image not changing over time, it is not sequenced. I am in static image mode. I edited the picture and changed the name to 'Active light time: 1.25 ms', so you are not confused with exposure time (which is also a term we use in Time-of-flight to describe a different concept than in the pattern sequence mode of Lcr). 

    Thanks for answering the questions, Pepe

  • 0 in reply to user4566018
    Intellectual 630 points
    OK I think I got it. It seems that ALSO in static image mode there is the 120 Hz limitation, which translates into 0.0083 seconds. If I set my active light time to that value the Flickering completely disappears. There is no way out of this limitation.

    Thanks a lot for your help, Pepe