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.

DS90UB962-Q1: margin analysis with BCC_EN_ENH_ERROR=1

Part Number: DS90UB962-Q1
Other Parts Discussed in Thread: ALP

Dear expert,

In customer board, we met very strange phenomenon with margin analysis. (ALP is from ti.com)

Customer system is UB935 with UB962. For most channel, margin analysis can run smoothly. But for one particular channel in one particular board, margin analysis failed. Both good capture and bad capture are attached as below. The phenomenon can be reduplicated every time. 

Then we compared registers between good channel and bad channel and find difference is 0x47,BCC_RESP_ERR=1 for bad channel. 

Then we set 0x46,BCC_EN_ENH_ERROR =0 to disable Enhanced Error checking in BCC for bad channel. Then margin analysis can run smoothly as good one every time.

BTW, the "bad" channel shows video as good as other channels in customer system. 

And we did further test for the bad channel and find out with BCC_EN_ENH_ERROR =1, margin analysis fail only when SP >11 or EQ<6.  I attached Pic3 as "Good Margin result with bad channel BCC_EN_ENH_ERROR =1, with SP<12". Pic 4 is  "Good Margin result with bad channel BCC_EN_ENH_ERROR =1, with EQ>6"

Need your help to explain why BCC_EN_ENH_ERROR impact margin analysis result so seriously. (Once BCC_RESP_ERR is set once, margin analysis will return RED result forever. Even we rerun margin analysis , it will return all RED)  If we read 0x47 to clear BCC_RESP_ERR, margin analysis can rerun normally at the beginning and fail like Pic 1 below again. 

If there is really serious physical problem for the bad channel, why margin analysis doesn't show any abnormal when BCC_EN_ENH_ERROR =0? And we also didn't find any abnormal for the video traffic in this bad channel in system level.

Looking forward your help.

Pic1: Good margin capture:

Pic2: Bad margin capture:

Pic 3: Good Margin result with bad channel BCC_EN_ENH_ERROR =1, with SP<12

Pic 4: Good Margin result with bad channel BCC_EN_ENH_ERROR =1, with EQ>6

  • Hello,

    It sounds to me like what is happening is you have a poor performing channel on one of the boards and disabling the bit in question causes many of the backchannel errors that the MAP tool is looking for to not be reported.  So by disabling BCC_EN_ENH_ERROR you are removing the ability for the MAP tool to check for all errors, which results in a much better performance. 

    Regards,

    Nick

  • Nicholas,

    If the channel is really bad, after BCC_EN_ENH_ERROR disabled, margin analysis can run and its result should detect some abnormal.  But the result is same good as other channels. This is very hard to explain.

  • Hi Ryan,

    The MAP tool checks the following bits.

    So run that channel with BCC_EN_ENH_ERROR set to 0 and one and read register 0x47 (BCC_STATUS).  Let me know the results.

    Regards,

    Nick

  • Nick,

    Pls see below two register dump attached. 

    Register dump with BCC_EN_ENH_ERROR=0; 0x47 value is 0x0; 0x4E value is 0x02

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/BCC_5F00_disBCC_5F00_EN_5F00_ENH_5F00_ERROR_5F00_TEST.nrd

    Register dump with BCC_EN_ENH_ERROR=1; 0x47 value is 0x21; 0x4E value is 0x7B

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/BCC_5F00_ENBCC_5F00_EN_5F00_ENH_5F00_ERROR_5F00_TEST.nrd

  • Hi Ryan,

    When BCC_EN_ENH_ERROR is set then the errors show up in 0x47 which cause the MAP tool to read errors and give poor results.  My thinking is that if this is only occurring on one particular channel on one particular board then it is probably a hardware related issue.  Could try to swap our device out with a different one to see if the error on that board still exists to know for sure?

    Regards,

    Nick