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.

TUSB1064: 10G-EXPANSION-EVM sink board configuration as shipped

Part Number: TUSB1064
10G-EXPANSION-EVM sink board configuration as shipped
I am trying to evaluate the TPS65987EVM as a UFP, and I have the 10G-EXPANSION-EVM sink board connected.
I cannot get USB 3.0 through the demo setup (USB 3.1 devices enumerate as USB 2.0). Using a TotalPhase Beagle 5000 v2, it appears that LFPS polling is failing to establish a link.
I am struggling to verify pin-strapping options of the TUSB1064 mux on the sink board.
The six resistors for configuring SNK_DPEQ[1:0], SNK_SSEQ[1:0], SNK_EQ[1:0] are all installed. Without removing these to measure them, I am assuming they match the schematic (PSIL068A(001)_Sch 20.PDF) and are therefore all 1k pull-ups/-downs to SNK_VCC and GND.
Despite the internal pull-ups/-downs, this puts every EQ pin at very nearly 1.65V, which leaves the EQ settings very susceptible to noise/power rail behaviour at startup, unless I am missing something.
What is the default or intended configuration of this part? Could misconfigured EQ settings be responsible for the failing of LFPS polling?
Regards
  • Hello Akira, 

    I've assigned this issue to the relevant expert who will get back to you shortly

  • Hi Akira,

    If the LFPS signal is polling, it could mean that the USB 3.0 handshake is not being established. Are you able to see the USB communication?

    For the pin-strapping, if you look at the TUSB1064 datasheet section 8.3.3 it also mentions that there are internal resistors and combined with the external resistors they can be configured to the different levels. Could you probe the voltage on your EVM EQ pins to verify what level it could be at? I have also tagged a TUSB1064 expert to help.

     do you know what could be going on for this issue and how to resolve? 

    Thank you,

    Hari

  • Hari,

    What do you mean by being able to see USB communication? I have a TotalPhase Beagle USB 5000 v2 analyzer to view USB communication.

    Regarding the pin-straps:

    The datasheet says the internal pullup is 45k, internal pulldown is 95k. Combined resistances would be pullup = 978R, pulldown = 989R. This gives a calculated voltage of 1.659V, a mere 9mV above the R/Float threshold in the datasheet. Moreover, the datasheet says the internal pulls are isolated after startup, which I am assuming means that I cannot accurately measure the voltage that actually corresponds to the latched state. Nevertheless, the config pins measure 1.62V while the board is powered up, which is close enough to 1.65V that I cannot infer the state of these pins.

    Regards,

    --Akira

  • Akira

    The equalization does not apply to the LFPS since the LFPS is a low frequency signal. Can you please probe both sides of the TUSB1064 and see if the LFPS is being sent from both the host and the device?

    If you are not seeing LFPS on both sides, then I would check the host/sink/TUSB1064 RX termination to make sure the termination is properly enabled.

    Once the TUSB1064 TX sees the host/device RX termination, then it will enable its own RX termination on the other side. If the termination is not there, then either the TUSB1064 lane configuration is wrong or host/sink not enabling its RX termination.

     To exit LFPS.Polling 

    1. At least 16 consecutive Polling.LFPS bursts meeting the Polling.LFPS specification defined in Section 6.9 are sent.
    2. Two consecutive Polling.LFPS bursts are received.
    3. Four consecutive Polling.LFPS bursts are sent after receiving one Polling.LFPS burst.

    Thanks

    David