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.

DS90UB913A-Q1: Suggested Power-Up Sequencing

Part Number: DS90UB913A-Q1

Hi Team,

I have a question regarding Figure 8-1. Suggested Power-Up Sequencing.

It is mentioned that a bit error may occur when the t3 Max condition of 16ms is not met.

If PDB is asserted after the t3 provision, will the bit error be automatically cleared to the default?

When the t3 provision cannot be observed, only the 0x27 register is not initialized and needs to be written manually?

If the 0x27 register cannot be cleared, does the data output start to DES with the error still in the register?

In the case of no register write, I would like to know the problem with t3 > 16ms.

The following post I found.

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/971550/ds90ub913a-q1-parity-error-counted-up-by-powerdownpll-0x27

”Having 1 or 2 parity error at power up is something possible due to device not LOCKing immediately.”

Best Regards,

Kenji

  • Hello Kenji,

    If PDB is asserted after the t3 provision, will the bit error be automatically cleared to the default?

    The errors are caused by writing to register 0x27 because this resets the forward channel. You will not be able to write to any registers if you have not yet asserted PDB. To clear the bit errors you will need to reset the error counter resisters on the deserializer.

    When the t3 provision cannot be observed, only the 0x27 register is not initialized and needs to be written manually?

    When t3 cannot be met, the writes to register 0x27 are used to reset the forward channel. All registers on the 913a should automatically be properly initialized.

    If the 0x27 register cannot be cleared, does the data output start to DES with the error still in the register?

    If register 0x27 is not cleared, the forward channel will not be able to function as this register is used to power down the forward channel PLL and NCLK.

    In the case of no register write, I would like to know the problem with t3 > 16ms.

    if t3 > 16ms the 0x27 register write is a requirement for correct operation of the device otherwise start-up issues may be experienced.

    Best regards,

    Joshua 

  • Hi Joshua-san,

    Thank you for your advice.

    My user is designing before the t3 regulation change and cannot meet t3 < 16ms.

    We cannot allow startup problems to occur.

    Do you have any suggestions to solve this problem other than writing to the 0x27 register?

    Best Regards,

    Kenji

  • Hi Kenji,

    I am sorry to say that we do not have another solution, other than the suggested writes to register 0x27, if t3 < 16ms cannot be met.

    Best regards,

    Joshua

  • Hi Joshua-san,

    I have not had any problems to date.

    If I have a startup problem, is it likely that turning the power off again will clear up the problem?

    Regards,

    Kenji

  • Hello Kenji,

    We cannot guarantee that the problem will be fixed with a power cycle, although it may fix the issue until the next power cycle.

    We recommend the register writes to 0x27 to ensure that the startup issues never occur if t3 > 16ms.

    You can try to not use the register writes if you have not experienced problems yet, but if you do begin to experience issues on startup please use the register writes.

    Best regards,

    Joshua

  • Hi Joshua-san,

    If a bit stands in the 0x27 register, does this mean that Power Down and Normal Operation will be repeated?

    Therefore, is it correct to say that the serializer signal is not output?

    Our customer is designing before PCN issuance.

    The rate of occurrence may be quite small, but does this mean that the device is faulty?

    Best Regards,

    Kenji

  • Hi Kenji,

    0x27 is the workaround should the required timing sequence not be fulfilled. 

    It was identified that only a small sub-set of units have issue under these certain power-up conditions.

    Regards, 

    Logan