• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 33 subscribers
  • Views 684 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.

Original question:

TUSB8041-Q1: Hub loses connection

TUSB8041: Reset/Loss of Ports USB 2.0

Jason Wilmoth
Prodigy 10 points
Part Number: TUSB8041

Hi All,

I'm using the TUSB8041RGCT in a PCA to talk between a host and multiple controllers.  All coms are USB 2.0 only, and do not have connections for 3.0.  Intermittently, the TUSB8041 seems to reset (all com ports in device manager disappear momentarily), then will come back online.  I was passed this design from a previous engineer so am not 100% up to speed on the TUSB8041, but have read the datasheet extensively to try to understand what could be causing this issue.  

Originally, I figured it may be my 3.3 and 1.1 supply timing.  However, both regulators are the same series (Semtech SC189).  In an attempt to ensure this wasn't my issue, I tried changing the GRSTz capacitor from 1uF to 4.7 uF and did not see an improvement.

To aid my question, this is the schematic for the upstream hub.  The downstream hub is set up identical and can be seen coming into the upstream hub at DN1 (pins 1 and 2).

Has anyone had this issue, or can recommend a possible cause and fix?  Alternatively, is there an issue with the design overall?

Thanks,

Jason

  • 0
    TI__Guru 60870 points

    Hi Jason,

    Can you send a clearer version of the schematic?  If you don't want to post on the forum, accept my friend request and send it directly to me.

    Does the hub have a good thermal pad connection with adequate solder paste?

    Regards,

    JMMN

  • 0 in reply to JMMN
    Prodigy 10 points

    Hi JMMN,

    Thanks for the quick reply.  I've attached a pdf version of the schematic.

    I've also attached screenshots of the top paste layer with the silkscreen as well as an additional screenshot with the top layer.

    1665.USB Hub.pdf

    Let me know if there's anything else you'd like to see in more detail.

    Thanks,

    Jason

  • 0 in reply to Jason Wilmoth
    Prodigy 10 points

    As a quick follow on, the host device for this system is an Intel NUC, and I updated the serial port settings from selective suspend enabled (5 second timout), to selective suspend enabled (3600 second timeout).  I haven't been able to reproduce the disconnect issue since I made this change but it seems like I'm bandaiding the issue.

  • 0 in reply to Jason Wilmoth
    TI__Guru 60870 points

    Hi Jason,

    I'm not seeing anything obvious in the design.  Can you put a scope on the 3.3V / 1.1V rails and check if there are any dips that could explain it?  Do just the downstream devices disappear / reappear or the hubs too?

    Regards,

    JMMN

  • 0 in reply to JMMN
    Prodigy 10 points

    I scoped the power on timing of the 1.1 (1.2 in my case) and 3.3 rails.  See screenshot below.  Its hard to decipher exact timing, but 1.1 is stable approximately 240 usec before 3.3.  All I could really find in the datasheet is that 1.1 should be stable a minimum of 10 usec before 3.3.

    Working on your second question.  I believe its just the downstream devices that disappear momentarily, but trying to reproduce failure.  Now that I want to see a failure, the system seems to be holding strong though.  Will update with further info.

  • 0 in reply to Jason Wilmoth
    TI__Guru 60870 points

    Hi Jason,

    I think your power on reset timing is ok.  You would see an issue at first connection if there was a reset timing issue, not after actively running for awhile.  I was more interested to see if there was noise on the 1.1V and 3.3V rails that could be causing the devices to drop off the bus.  Power consumption can vary quite a bit during active use so if there isn't enough bulk capacitance on the rail it could get noisy.

    Regards,

    JMMN