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.

DS90UB921-Q1: Flickering issue with the 921-948 pair

Part Number: DS90UB921-Q1

Hello,

Can a missing or incorrectly generated DE input signal to 921 serializer cause loss of LOCK on the 948 deserializer?

How does the 948 PLL acquire its LOCK status (i.e. what criteria need to be met to indicate LOCK and UNLOCK)?

Is there any relationship or dependency between LOCK status and PASS status with the 921-948 pair?

Regards,
Alan Ocampo
Analog Field Applications
Texas Instruments Inc. 

  • Hello Alan,

    DE is not considered in the lock between 921->948. There are some systems which do not use DE at all. Also please note that 948 does not have PASS functionality in normal mode. This is reserved for 940 and will be removed from the 948 datasheet in the next revision. PASS pin on 948 is only used for BIST. 

    Best Regards,

    Casey 

  • Additional observations on the display flicker / loss of lock behavior from my customer:

    When they set an RGB value of say less than some value (still testing), then the flicker goes away. If they set the RGB at max, they can get the LOCK condition to toggle and screen to flicker.  

    >> What explains the LOCK sensitivity to the RGB value?

    Also, they mistakenly wrote to register address 0x23 of the 921 with some arbitrary values, and they were able to get the flicker / unlock PLL condition to stop.  

    >> Can you share the functionality of reserved register 0x23?

    Thanks,

    Alan

  • Hello Alan,

    RGB values don't have any correlation to LOCK. Can you please probe the input PCLK to 921 and check the jitter according to the scope settings on datasheet table 6.8? Also what is the PCLK value for this video and cable type? Can you verify it matches the table 1 configuration for the FSEL strapping?

    0x23 is reserved because it should not be used normally to resolve a system issue. Bit 6 is probably the only bit or relevance and it should affect the back channel speed but it has not been tested. 

    Best Regards,
    Casey 

  • After more debug, the customer thinks it’s related to RGB timing.  Their RGB signals failed when they checked it directly on a scope. And flicker went away when they inverted the PCLK on at least one unit. They will continue testing, but we can close this for now. Thanks for the suggestions.

    Alan