• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 29 subscribers
  • Views 414 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.

DLPC3479: Confirm that the DLPA3000 current output has stabilized when the DLP4710 begins to output LSB

Daniel Shen34
Prodigy 70 points
Part Number: DLPC3479
Other Parts Discussed in Thread: DLPA3000, , DLP4710, DLP4500

Hi!

How can I confirm that the DLPA3000 current output reaches steady state in trigger-in 1-D pattern mode before the LSB PWM of the DMD starts flipping?

I have tried adjusting the pre-dark exposure, and the experimental result refers to the time difference between the trigger-in and the start of charging for the DLPA3000 output.

Due to the difference in the steady state duration of DLPA3000 output between 1A and 6A, we want to ensure that the LSB does not start flipping within the current rise time (approximately 80us) of the slower 6A current, so as not to affect the gamma value.

  • 0
    TI__Genius 10305 points

    Hi Daniel,

    Do you own a custom board with DLPC3479 + DLPA3000 combination?

    What DLPA3000 current output are you looking at? Do you have scope captures that can be shared? 

    Regards,

    Lori 

  • 0 in reply to Lori Wang
    Prodigy 70 points

    Hi Lori,

    Yes, we have our custom board, DLPC3479 + DLPA3000 with Red Channel only firmware

    This is a waveform output of DPLA3000 with 5A, and I think there should be no problem with the rise time of 80us. However, we need to confirm if this rise time may conflict with the DMD flip start time of DLP4710, causing unstable LSB of gray scale.

    Thanks!

  • 0 in reply to Daniel Shen34
    TI__Genius 12785 points

    Hi Daniel,

    Thank you for sharing this plot. We are discussing this among the team.

    Can you say if there is any poor performance you are experiencing while asking about the rise time? We understand that you do not want to alter the gamma value, but this platform goes through a validation process before release.

    Regards,

    Austin

  • 0 in reply to Austin Snyder
    Prodigy 70 points

    Hi Austin,

    Our product is still in the verification stage, and our optical verification department has some doubts about the light output and DMD flipping timing. As we often experiment with LEDs with different characteristics, the rise time of the current naturally varies slightly. However, for my EE department, we cannot provide a simple measurement experiment to prove to the optical RD that the exposure grayscale of the LSB is completely correct.

    If we could set one of the trigger outs as the signal for the start of LSB flipping, we could easily use an oscilloscope to confirm the moment when the DLPA3000's output current has reached stability.

    Thank You!

    Daniel

  • 0 in reply to Austin Snyder
    Prodigy 70 points

    Alternatively, if you could provide a parameter setting that is different from pre-dark exposure, specifically for adjusting the delay time for the start of DMD flipping, it would be better for us to experiment and confirm that the stable current output time fully covers the DMD exposure time.

  • 0 in reply to Daniel Shen34
    TI__Genius 10305 points

    Hello Daniel,

    The DMD mirrors are controlled by a sequence of load and reset instructions. For a 8-bit pattern display, the sequence starts with MSB first not LSB, so the non-zero rising time of the LED via DLPA3000 should not affect the LSB grey scale. 

    I hope this helps. 

    Regards,

    Lori 

  • 0 in reply to Lori Wang
    Prodigy 70 points

    Oh, sorry, I didn't know since when the established impression (maybe DLP4500?) was believed to be LSB First. It would be better if it is actually MSB First.

    However, the rise time of the LED current may affect the weight of the MSB. I still hope to have a verification method. Can you provide it?

    Thank you!

    Daniel

  • +1 in reply to Daniel Shen34
    TI__Genius 12785 points

    Hi Daniel,

    If the primary concern is validating the gamma curve, I would recommend testing a greyscale ramp, followed by purely red, green, and blue ramps. This can be done either through the video source feeding into the DLPC3479 or through the internal test pattern generator (TPG). However, do note that the TPG method will limit you to the 60Hz frame rate. Section 3.2.5 of the DLPC3479 Software Programmer's Guide provides details on how to set an internal test pattern. These commands are also available in the advanced menu of the DLP EVM GUI.

    I suggest the ramp as it is useful in observing the linearity of light output. Another option is to have the start and end values match for the ramp and increment slowly from 0 to 255. This will allow you and your team to have the entire projection increment. Placing a photo-detector attached to an oscilloscope will show the light output.

    Regards,
    Austin