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.

Artifacts with Discovery 4000 Developer Board

Other Parts Discussed in Thread: ALP

Hello,

our Institute purchased a Discovery 4000 Developer board in 2009. Since some time, we are seeing artifacts on the DLP chip. Vertical lines appear on the patterns displayed. The artifacts change a little over time and are not necessarily effecting a whole column on the chip.

After contacting the local distributor, we were told that the board has a production error and was replaced for free in 2009. Unfortunately, we were not informed of this opportunity. Is it still possible to receive a free replacement?

Regards

Steffen Matthias

  • Hello Steffen,

    Welcome to the DLP section of the TI-E2E Community.

    The production error only affected the DDR2 address space.  It should not affect the image.

    What you are seeing suggests a poor connection of one or more of the LVDS data lines to the DMD.  If it or they are making poor contact it may have significant capacitive effects on the line.

    Are the lines equally spaced or have a pattern period of 32 (XGA) or 64 (1080p)?  For the 1080p this would likely be on ony one half of the device.

    Please try reseating and tightening both ends of the flex cable(s) connections.  Also reseat and tighten the c-spring interposer under the DMD.  You may want to check the DMD and DMD board pads where the interposer makes contact to make sure there is no debris on these surfaces.  Also, with a microscope (preferaby a stereo unit) inspect the c-springs in the interposer to make sure none of them are damaged.

    I hope this helps.

    Fizix

  • Hello Fizix,


    thank you for your reply! The DLP chip is XGA. The patterns appear to be somewhat periodic, but sometimes only affect parts of a column. Unfortunately, cleaning the contacts and tightening the connection did not help.


    We are using the ALP-API to generate the patterns on the chip. With the "basic" version of the API binary patterns are displayed without artifacts. The artifacts only appear when using the "high-speed" version, which is using the DDR2-RAM.

  • Hello Steffen,

    If you have access to the Discovery section of MySecureSoftware (formerly called the Discovery Extranet) please reference the Discovery D4000 Technical Advisory called, "Discovery™ 4000 DDR2 SO-DIMM Technical Advisory" dated 02/09/2009. 

    This advisory shows that the several P/N pairs were swapped that limited the address space of the DDR2 SO-DIMM to 1/2 the actual max size intended.  i.e. 2G vs 4GB.  It should not affect the image in any way.

    Please contact ViALUX regarding the actuall D_CLK rate of the ALP "high-speed" on the D4000 platform.

    Does the ALP report the version number of the DLP Controller (i.e. 1 - 4)?  If it does, can you provide that information?

    Fizix

  • Hello Fizix,

    thank you for your reply.

    ViALUX wrote me that the D4000 board is no longer supported due to the production error. Can you tell me where I can find the signal on the board? I could try and check the clock with an oscilloscope.


    The ALP-API reports the following information:

    Serial Number: 4184

    DMD Type: .55 XGA Type X (D4000)

    Driver Version: 0.2.5.7

    USB Controller Firmware: 3 Jul 2008

    FPGA Configuration Date: 29 Apr 2009

    As the artifacts seem only to appear when using the DDR2-RAM, we tried to exchange the module. Unfortunately, the same errors were still present with the new memory module.

  • Hello Steffen,

    Please request access to the "MySecureSoftware" DLP Discovery pages from ViALUX, if you do not already have access.  On these pages you will have access to D4000 schematics, layout files (Gerber) and the Technical Advisories that I referenced.

    Did this issue exist from the very beginning (when the board was first bought) or did it apear later on?

    Also please ask ViALUX if they know what DMD data clock rate is used int "high speed" mode.

    Fizix

     

  • Hallo Fizix,


    I can only find the schematics for the D3000 and D4100 in the MySecureSoftware section. Where are the D4000 files located?

    I am not sure when the artifacts appeared, I started working with the development board 6 month ago. Then they were already present.

    ViALUX is not providing technical support for the D4000 anymore unfortunately.

  • Steffan,

    You are correct.  I had forgotten that D4000 information was removed when we moved to MySecureSoftware.

    Although they are not supporting it anymore, I will ask them to provide a the D-clock frequency information.

    Fizix

  • Hello Steffan,

    The data clock is only running at 200 MHz and with a minimum of 4 clock cycles between commands, so this is not the issue. 

    This sound very much like poor Ohmic contact somewhere.  Typically this can be caused by a damaged "C"-spring in the interposer to the DMD.  You will need to inspect the interposer under a microscope to be sure.

    In the final analysis, we strongly encourage you to consider moving the Discovery 4100 based platform.

    Fizix

  • Hello Fizix,


    thank you very much for your reply. Is it realistic, that the interposer is damaged, even though it was never detached from the DMD?

  • Hello Steffen,

    Probably not, but it is possible one of the contacts was marginal at the beginning so that over time it has become worse (slight oxidation, thermal expansion/contraction cycles).

    Otherwise I am at a loss, unless there a cold solder joint in the system whose ohmic contact has slowly degraded over time.

    Fizix