Announcement

Collapse
No announcement yet.

USB 3 failures with extension cables using BurninTest 8.1 1022 and USB3.0 test

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • USB 3 failures with extension cables using BurninTest 8.1 1022 and USB3.0 test

    Hello,

    The hardware is detailed in the attached BurnInTest logs.

    We are trying to determine root cause on errors that are being reported by both PassMark BurnInTest v8.1 Pro (1022) and with USB3.0Test V2.0 (1009) when a USB 3.0 extension cable is used.

    We are using firmware 2.5 only with our loop-back devices and have issues on systems that are idle using 1-4 loopback devices at a time. We seem to have software issues (BurnInTest 8.1, USB3.0Test), USB 3.0 loopback issues and USB 3.0 extension cable (from Amphenol with P/N USB3FTV2SA03NASTR) issues with multiple cables and devices.

    We find that errors require using specific PassMark USB 3.0 loop-back plugs with a specific extension cable. I have no issues when testing same extension cable using USB 3.1 SSD and other tools (such as HD_Speed). I also note the same throughput to the USB 3.1 SSD as is observed with the PassMark USB 3.0 loopback plugs in benchmark mode.

    Failures under BurnInTest appear to coincide with the re-enumeration happening with the changes to the test mode (loopback/benchmark).

    Numerous failures also noted with some of our brand-new devices and the connection at the plug itself which may later require using RTV to address. We have not inspected cable and plugs for contamination yet with a black light.

    Troubleshooting is very challenging when using more than one loopback as the software is randomly picking the wrong interface as the failing one; despite it being obvious by the count and the state of the loopback device per the LCD. I note that enabling verbose logging with USB3,.0Test drops the error rate by > 75%.

    Thank you,
    Josiah

    Attached Files

  • #2
    The version of BurnIntest you are using is over 3 years ago and we no longer support that release. See
    https://www.passmark.com/products/bu...st/history.php

    How long are these extension cables that you are using (Amphenol spec seems to indicate 30cm)? As this is an extension cable I assume you also have some other normal USB cable in use as well (maybe the one we supplied?). What's the total length of the cabling?

    may later require using RTV to address
    We don't know what RTV is in this context sorry.

    The errors do seem to indicate a hardware failure, but it could just as easily be a bug in the USB host controller's device driver.

    There is no magic solution to these type of errors. It's needs detailed investigation and some trial and error.
    e.g.
    - Does it all work, without error, when the extension cables are not used.
    - Does it all work, without error, when a different PC is used (e.g. a standard desktop machine)
    - Does using BurnInTest V9 make any difference
    - Is the USB host controller's device driver up to date
    - Do you have a USB protocol analyser available (Beagle 5000 for example)
    - Have you tried using some other (short) off the shelf USB extension cable.





    Comment


    • #3
      When reviewing this issue I note that we ended up having to use a hardware-based tester to pre-screen both the Amphenol and the PassMark USB 3.1 cables.There were defects found in both.

      Comment


      • #4
        Hard to comment without knowing the details of the testing.

        Comment

        Working...
        X