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.

MMWCAS-RF-EVM: Large skew in differential pairs of the MIPI CSI routing

Part Number: MMWCAS-RF-EVM
Other Parts Discussed in Thread: AWR2243, , MMWCAS-DSP-EVM, TIDEP-01012, TIDEP-01017, TDA2

We are debugging an issue we have with multiple "MMWCAS-RF-EVM + MMWCAS-DSP-EVM" setups. We often lost 1 out of 4 AWR2243's data after the system running for a while.

We noticed that on the MMWCAS-RF-EVM board, the MIPI CSI traces from the AWR2243 to the connectors have pretty large skew in the diff pairs.

Usually the requirement is like 5 mil, any they are often of 30-40 mil. (by our measurement on the AD layout file)

First question is why? and second could this contribute to our unstable AWR2243 data issue?

Thanks,

Li

  • Hi,

                 Can you please elaborate a more on the unstable data issue? I believe this is your own board(based on mail exchange).

    regards

  • Hi Abhed,

    Thanks for your reply.

    The unstable issue actually only happened on TI-Dev board..... we have similar issues with 4 out of 5 samples we have (TIDEP-01012+TIDEP-01017). 

    It haven't happened on our custom designed radar prototype yet.... But we really want to understand the issue before our next revision/massproduction.

    1. Sometimes when we first start the board, there will be one channel of EVE data missing, and one of the Lattice FPGA shows CDONE low (i.e. the corresponding status LED for that FPGA is off). We tried to reset that particular FPGA and nothing happens. We see channel 2 & 4 has this issue from time to time.

    2. Other times, after our code on TI-DEV running on TI_DEV for a while, we lost one channel data out of no reason, after we reset that FGPA of the failing channel, it would become normal again. 

    We are not sure if these 2 issues are related and neither did we know if it's related to the dramatically mismatched CSI diff pair issue on the TIDEP-01012.

    We're hoping you guys have more insight.... 

    Thanks!

    Li

  • Hi Li,

             Couple of points:

    1. Did you observe the data drop or missing on multiple RF EVMs and multiple TDA2 boards?

    2. Has the capture board same for all the RF EVM's?

    3. Have you tried reflashing the FPGA's of TDA2 board?

    4. Do you use TDA2 board for your custom RF board data capture also? has the issue been observe don that too?

    Regards

  • 1. Did you observe the data drop or missing on multiple RF EVMs and multiple TDA2 boards?

    Yes 4 out of 5 sets of RF EVM + DSP EVM have issues.

    2. Has the capture board same for all the RF EVM's?

    we have 5 sets of RF EVM + DSP EVM (in 3 locations of the world...)

    3. Have you tried reflashing the FPGA's of TDA2 board?

    It's interesting question, we never thought that would help (since they mostly works...) but we will try this.

    4. Do you use TDA2 board for your custom RF board data capture also? has the issue been observe don that too?

    our custom board has complete 2243+TDA2, we didn't use the TDA2 reference board. we haven't seen issue there so far (mentioned in the post)

  • Hi,

          In your custom board which host do you use? In point 4 you mentioned TDA2 does that mean you have Lattice FPGA for CSI2 bridge on your custom board too?

    Regards

  • That's correct we have 4 Lattice and TDA2 on our custom board, just like those on the reference design

  • Hi Liu,

                   Do let us know the update when you have tried the re-flashing of the TDA2 board. Or advice if we can close this ticket.

    Regards