DLP2010EVM-LC: Trigger not precisely correct with 1D 8 bit patterns.

Prodigy 100 points

Replies: 10

Views: 119

Part Number: DLP2010EVM-LC

In the document https://e2e.ti.com/support/dlp/f/94/t/729968# it states "...Note 2: There is currently a bug in the GUI limiting the maximum exposure time for 8-bit patterns to 10464us. This is planned to be fixed in the next GUI update. "

Please clarify, has this been fixed in the current Feb 2019 release?

In any case I have the following problem with trigger out 2, and it may be that I am taking the wrong signal ground. 

I have created a 1D 8bit  Vertical Pattern set using just one image in the entire set. 

Pre-Exp dark time is 2000 uS, Exposure Time is 10400 uS and Post-Exp Dark is 200 uS. 

Trigger Out 2 is Enabled and not inverted, all else is disabled. 

I am triggering a camera and wiring that I used to trigger successfully from a 4500, from Trig_Out2 on J4 Pin 4.

The documents don't appear to explicitly state where to take the trigger out signal ground from (only describes four of the pins on J4), so using a DMM I identified pin1 as ground.  I took my ground for the Trigger Out 2 Signal signal from Pin 1 on J4. 

Perhaps this is wrong, perhaps it is the cause of my problem by picking up PCB digital noise etc. 

What I am finding is that when I use the trigger out 2 signal to trigger a camera capture, the results are different every time.

This same approach used to work precisely right with the 4500. 

Numbers I see on my board are DLPx010EVM-LC and DLP017A

I would be most grateful for any solutions you could propose. 

 

10 Replies

  • Expert 3115 points
    Hi Tony,

    The bug was only for splash patterns and it has been fixed. We'll remove the note.

    What type of illumination are you projecting the pattern with? Mono or RGB?

    Have you checked whether the timings you're using are supported? Please refer to this page: e2e.ti.com/.../2878269

    Regards
    Azad
  • In reply to Azad:

    HI,

    I am projecting blue. I am not certain I understand the info in your link. Please evaluate whether you consider my settings below to be "supported", as I believe they are:

    Pre-Exp dark time is 2000 uS, Exposure Time is 7500 uS and Post-Exp Dark is 200 uS.

    Do we know for sure if it is proper to take Trig Out 2 signal ground from J4 pin 1?

    Where should trigger signal grounds be taken from?

    Regards,
    Anthony

  • In reply to Tony Higgens:

    Please let me know if this is being looked in to. 

    Is there a particular person it should be assigned to? Surely there must be an expert who knows basic issue like trigger ground pin?

  • Expert 3115 points

    In reply to Tony Higgens:

    Hi Tony,

    Pin 1 of J4 is indeed GND, so that may not be the cause of the issue.

    Can you explain what you mean by "the results are different every time"? Attaching a couple of images might be helpful.

    For 8-bit single color internal patterns, the supported exposure time range is 3464µs - 7600µs. Please confirm that you're exposure time falls within this range.

    Regards
    Azad
  • In reply to Azad:

    Whilst Pin 1 of J4 is of course a ground,   it may be a noisy digital ground - are we absolutely certain I am not supposed to take trigger ground from somewhere else?

    I confirm that I still have the issue at an exposure of 7500uS with my single color 1D 8 bit pattern.

    My pattern is supposed to be a sine wave, and indeed to the eye it looks like one when the system is on continuous repeat.

    However, what I capture is like the image in different stages of being built up.

    In the attached image, upper middle, shows the "Source Image" as created for upload via the TI software.

    Lower left and lower right are both captures of the resulting projection. The images are a little dark, but if you look at the red marks added along the bottom of the image it will give emphasis to how I get "banding" of differing structures at differing times. Camera exposure have been varied from below 7500 to many multiples of it with short dark times, problem persists.  

  • Expert 3115 points

    In reply to Tony Higgens:

    Hi Tony

    I'll investigate this issue on my side and get back to you.

    Would it be possible for you to share the input sine wave source image?

    Regards
    Azad
  • In reply to Azad:

    I have "inserted" the 1D source image below. Being only one pixel high, you could easily miss it!

    One other thought I had is whether there is a certain physical length for the trigger wires, above which they might be subject to clock skew (due to the capacitance of the wiring)? Perhaps the threshold is lower for this projector compared to the 4500? 

    I have one other related query that I would appreciate an urgent answer for. Does the issue of "supported" and "unsupported" timings also affect the DLP4500EVM? I mean, is it possible to enter a value that the software will accept, but which the 4500 doesn't technically support? 

    Thanks

  • Expert 3115 points

    In reply to Tony Higgens:

    Thanks Tony. Let me look into this.

    DLPLCR4500EVM is based on an older system architecture and doesn't have the concept of "supported" and "unsupported" timings. This is specific to the new DLPC347x controller based EVMs.

    Regards
    Azad
  • In reply to Azad:

    Hi, 

    I may have homed in closer to the problem, but I would need further info to solve. 

    With a scope, I noticed that the trigger 2 output waveform is 3.3V if it is an open connection. When I connect the camera (the trigger is an optocoupler onboard the camera) the peak voltage of the waveform drops too low. This was never an issue with the 4500. 

    Does the 2010 have a weaker output capability compared to the 4500? Do I need an op-amp voltage follower / buffer to avoid loading the 2010 trigger 2 output pin for example?

    Thanks, 

     

  • Expert 3115 points

    In reply to Tony Higgens:

    Hi Tony,

    The EVM already has buffers on the trigger signals to prevent this. Please refer to the EVM schematics for more information on this.

    Have you installed a jumper on connector J11 of the main board to set the trigger voltage level? Please refer to section 7.3 of the EVM User Guide for this.

    Regards
    Azad