• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 6 answers
  • Subscribers 35 subscribers
  • Views 2383 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

SN65LVPE502: USB3.0 Compliance test

Toshihiro Watanabe
Genius 10235 points

Part Number: SN65LVPE502

Hello,

We applied USB compliance test which SN65LVPE502 is on the board and the result is failed as below items.

TD.1.1 No LFPS output

TD.1.2 No LFPS output, there is a waveform like CP0

TD.1.3 passed

TD.1.4 passed

TD.1.5 No loopback mode

Questions

 1. Do we have a similar report which compliance test doesn't pass by using SN65LVPE502?

 2. Do TI applied compliance test on SN65LVPE502? If it has done, how was the result?

I think there are some more questions after this but please let me ask this first.

Best regards,

Toshihiro Watanabe

  • 0
    TI__Guru 64326 points
    Toshihiro,

    Please see link below, there is a compliance report linked on the device page of SN65LVPE502.
    www.ti.com/.../SN65LVPE502
  • 0 in reply to Malik Barton57
    TI__Genius 10235 points
    Hello Malik-san,
    It seems TD.1.1 and TD.1.2 are not tested and TD.1.5 are information.
    What does this information mean?
    Does SN65LVPE502 relate with no LFPS output? or is this host side issue?
    Does SN65LVPE502 relate with not become loop-back mode?
    SN65LVPE502 set in Compliance mode while compliance test. Is this mode relate with no LFPS output and loopback?
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Guru 64326 points
    Toshihiro,

    You not need to put LVPE502 into compliance mode, compliance will run even in normal mode.
  • 0 in reply to Malik Barton57
    TI__Genius 10235 points
    Hello Malik-san,
    Why TD.1.1, TD.1.2 and TD.1.5 can't pass in the customer case?
    Compliance report of LVPE502 doesn't have information about TD.1.1, TD.1.2 and TD.1.5 is only information. These items are matched with the customer failed items.

    Compliance mode seems not relating this issue. The customer did both cases.
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Guru 64326 points

    Toshihiro,

    Can you see the LFPS signal coming from the host? Please probe in between Host and SN65LVPE502 to confirm.

  • 0 in reply to Malik Barton57
    TI__Genius 10235 points
    Hello Malik-san,
    Is this LFPS measurement ok if we use USB3.0 devices? We may not be available USB3.0 compliance test equipment.
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Guru 64326 points
    Toshihiro,

    LFPS can be seen with USB 3.0 device as well. It is the first thing that happens in the SuperSpeed Handshake to enable SuperSpeed communication. I would try looking for LFPS from the host with USB 3.0 device connected downstream. If found then probe between SN65LVPE502 and USB device to see if LFPS is passed through.

    In TD 1.1 and TD1.2 are you seeing any CP patterns? If so this mean that LFPS did pass through the device, at minimum.
  • 0 in reply to Malik Barton57
    TI__Genius 10235 points
    Hello Malik-san,
    We measured the waveform by using USB3.0 device. It seems LFPS doesn't pass in normal mode and pass in compliance mode.
    I will send the waveform and schematic offline. Please check it.
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Genius 10235 points
    Hello,
    Any update?
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Guru 64326 points
    Toshihiro,

    SN65LVPE502 in normal mode does not turn on internal termination as it does in compliance mode. If SN65LVPE502 does not see termination it will not output the LFPS. Is their proper 50 ohm termination via oscilloscope or USB 3.0 device during this test? I could not find anything in the schematic that would cause this issue.
  • 0 in reply to Malik Barton57
    TI__Genius 10235 points

    Hello Malik-san,

    Yes, this measurement is done by USB3.0 device terminated in both cases.

    Is LFPS waveform in compliance mode proper?

    If compliance mode waveform is correct, why the compliance test doesn't start?

    Why does LFPS not output in normal mode even though USB3.0 device connected?

    Best regards,

    Toshihiro Watanabe

  • 0 in reply to Toshihiro Watanabe
    TI__Genius 12770 points

    Watanabe-san,

    I have a couple questions and comments to clarify.  I don't understand how USB3 compliance testing is being performed with a differential probe, this is not how the testing is defined or performed.

    To perform the USB3 compliance testing, the USB test fixture is used to connect the Tx differential pair of the DUT directly to a scope via SMA cables (The scope provides 50 ohm termination), using a differential probe will not provide the proper termination (Depending on the probe, it could also affect the termination detection when used to probe a live signal).  Once the DUT (Host, Device or HUB) detects termination, it will begin to send LFPS for 360ms then timeout and enter the compliance mode state.  At this point the DUT will be sending the CP0 compliance pattern (And sequence through the various compliance patterns when receiving a Ping.LFPS).

    When speaking about Compliance Mode with the SN65LVPE502, there are two aspects:

    1. Compliance mode of the SN65LVPE502 selected by the CM pin.  This simply enables the termination of the Rx pins of the SN65LVPE502.

    2. System Compliance mode where the DUT is in Compliance mode outputting the compliance patterns for Compliance testing.

    Can you detail the test set-up, are you changing the CM pin of the LVPE502(This is not necessary for Compliance testing)?

    Can you perform the test with test fixtures using the scope termination?

    We do not see any issue with outputting LFPS when performing the compliance testing as defined (Using test fixtures with SMA cables and scope). 

    As a debug alternative only to observe LFPS, you may be able to use a high-impedance single-ended probe.

  • 0 in reply to Undrea Fields
    TI__Genius 10235 points
    Hello Malik-san,
    If I can get your point correctly, you are doubting the load is not terminated correctly by 50ohm. Is my understanding correct?
    Does scope mean oscilloscope 50ohm termination or different measurement equipment?
    If it means oscilloscope, is single-ended 50ohm termination no issue? I think one side is connected to GND.
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Genius 12770 points

    Watanabe-san,

    It appears that the termination is not correctly detected when testing with a differential probe and USB device.  In order to capture the LFPS signal for compliance testing, the outputs of the SN65LVPE502 should be connected directly to a scope (50 ohm terminated) via SMA cables.  To clarify the test set-up please refer to the links below from USB.org describing the USB3 electrical testing and test fixtures needed to perform the compliance testing.  

    www.usb.org/.../EnhancedSuperSpeedPHYComplianceTestSpec.pdf

    http://www.usb.org/developers/estoreinfo/SuperSpeedTestTopologies.pdf

    Can you confirm how the testing is being performed?  

    -Undrea

  • 0 in reply to Undrea Fields
    TI__Genius 10235 points
    Hello Undrea-san,
    I can get the customer measuring by USB3.0 device. I will send it to you offline. Please check it.
    Best regards,
    Toshihiro Watanabe