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.

DS90UB914A-Q1: Back-Channel Enable and Disable During Normal Operation

Part Number: DS90UB914A-Q1

[ DS90UB913A & DS90UB914A-Q1 ] Back-Channel Enable and Disable During Normal Operation

Hi,

In order to improve the receiver eye integrity, my customer tries to implement the sequence which turn-on and off the backchannel communication with 0x02[2].
i.e. the back channel is enabled only when it's required.

Would you please let me know the minimum duration which requires for UB913A and UB914A to communicate and configure correctly?
Again, no guarantee required...

Thanks,
Ken

 

 

===========================================================================================

Initial Response from TI

===========================================================================================

Not sure why they would want to switch on and off unless they are multiplexing two cameras or inputs.
This is not a validated use case so we wouldn't have timing info to provide. By the way, the register should be 0x01[2] for BC enable/disable.

  • Hi Pal,

    Thank you for your support.

    We are facing the image flicker issuer. From a debug test, when we checked the CMLOUT waveform, by disabling the back channel communication, we could have more mask margin on CMLOUT waveform.

    As you may aware, yes, there are some points on customer board need to be optimized to improve eye integrity, however it's not really close to the phase of launch production. That's why they cannot implement hardware modification at this moment.

    Anyhow, in order to minimize the risk of image flicker, the customer now consider whether back-channel can be enabled only when it's required.
    Any ideas to help customer?


    BTW, the debug was done together with your local FAE. 
    Also, yes, you are right, 0x01[2] to enable/disable bc.

    Thanks,
    Ken

  • Ken,
    I hope they are not using CMLOUT during normal operation as this is meant for only debug. Can you provide additional information on the CMLOUT performance differences with and without the backchannel and the customer board issues that they've already identified?
  • Hi Pal,

    No, we just use CMLOUT only for debugging. I can share the whole details at off-line discussion.
    Can you send an e-mail to me? tanaka-k(at)macnica.co.jp

    What I can say here is that the eye of BC enabled from CMLOUT was more degraded or less margin comparing with BC disabled eye.

    We have performed several measurements such as S11, TDR and CMLOUT measurement with several different register settings.
    Also, we compared the result with UB914A-EVM as a reference. From these measurements, the characteristic impedance of customer board was not optimized.
    (Not exact within +/- 5%, nor +/- 10% of 50ohm.)

    Thanks,
    Ken

  • Hi Pal,

    How can we support customer?

    Thanks,
    Ken

  • Hello Ken,

    The forward channel will still function normally when the back channel on the 914A is disable.
    If customer does not need communication over the back channel, there should be no problem to turn it off.

    As Pal mentioned, ideally the customer would fix their design. It would be prudent to wait a minimum of 100ms after enabling the back channel before attempting any I2C communication from the deserializer to the serializer. As disabling back channel is non-standard use case it is recommended customer to perform validation at the system level.

    Regards,