• State
  • Locked Locked
  • Replies 11 replies
  • Subscribers 33 subscribers
  • Views 127 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:

TUSB2046B daisy chain together

TUSB2046B: Chained TUSB2046B not going into suspend mode

W A
Prodigy 30 points
Part Number: TUSB2046B

Tool/software:

I have previously gotten a TUSB2046B to work as a self-powered hub using reference designs, and peripheral devices successfully connected to a host. I am iterating on the design using a second TUSB2046B to connect to the first, expanding from 4 connected devices to 7, with one port of the upstream hub being used to connect the downstream hub. On each of the external device ports, I am using the recommended capacitors, resistors, and TVS diodes, along the lines of Figure 11 from the datasheet.  

When evaluating the chained design, I am seeing the upstream hub suspend with no host/devices connected as expected, but the downstream port does not. I realized that I did not use the capacitors and resistors between the hubs. There is no difference in implementation between the two hubs, other than this one connection, so I assume leaving these components out (pull down resistors in particular) is the cause of the downstream hub not suspending - does this seem reasonable?

Thanks!

  • 0
    TI__Guru*** 145435 points

    HI WA:

      With 1st hub upstream port connected, did 2nd hub got detected?

    How did you know 1st hub downstream port not  suspend?

    Best

    Brian

  • 0 in reply to Brian Zhou
    Prodigy 30 points

    Sorry, I had a typo - I meant to say upstream (1st) TUSB2046B hub suspends, but downstream (2nd) hub does not. I am looking at the suspend pins of each hub via test points. When no device is connected, the upstream hub is suspended (logic high) and the downstream is not (logic low). The same is true if I connect a device to the downstream hub, and the device is not visible to the host. If I connect the same device to the upstream hub, both hubs are not suspended and the device connects to the host. It looks like sources describe pull ups and pull downs on the line controlling the connection, so I think the lack of pull downs is causing a connection issue. I'm guessing adding the capacitors and terminating resistors won't hurt either.

  • 0 in reply to W A
    TI__Guru*** 145435 points

    I see, can you send  schematic for review?

    Best

    Brian

  • 0 in reply to Brian Zhou
    Prodigy 30 points

    Sent via private message.

  • 0 in reply to W A
    TI__Guru*** 145435 points

    Hi W A:

    Do you have 1.5k pullup on DP0 of 2nd hub?

    Do you have 15k pulldown for 1st hub D1+/D1-?

    also need to remove1M pulldown on reset pin for both hub.

    Regards

    Brian

  • 0 in reply to Brian Zhou
    Prodigy 30 points

    No, mentioned above that I'm fairly confident that leaving out the pullup and pulldown resistors is causing the issue. I will remove the 1M pulldown.

  • 0 in reply to W A
    TI__Guru*** 145435 points

    ok, pleas update once you have test result.

    Best

    Brian

  • 0 in reply to Brian Zhou
    Prodigy 30 points

    Our manufacturer had finished the pullup and pulldown test additions on the DP and DM lines before I could ask for them to also remove the RC resistor, so I went ahead and tested just those data line changes. Interestingly, the downstream hub does not suspend when booted and reset from external power, but works perfectly once it is connected to the host (the VBUS from the upstream host is not used on this board, all components including resets and external connectors are powered from the external supply). The second tier hub connected to host both allows devices to connect to host and suspends when nothing is connected, and stays suspended when disconnected from the computer. This feels like something of a startup race condition, given that the hub starts to work only with upstream connection,, without an additional reset signal. I will work on removing the RC reset pulldown, but are there any other suggestions?

  • 0 in reply to W A
    Prodigy 30 points

    Just tested removing the RC pulldown resistor, with no luck. The downstream hub still doesn't suspend on startup. I also tried increasing the reset capacitor to 220 nF on just the downstream hub, to delay the reset behavior. I don't think it's the reset circuit, though, as the downstream hub works once connected to host without further reset, indicating that startup is working correctly. I also don't think it's downstream external ports, as the hub suspends immediately when connected to the computer, without the downstream ports changing. My only thought is to add the other hub-to-hub connection components, the terminating resistors, the capacitors, and potentially additional wire resistance - any suggestions?

  • 0 in reply to W A
    Prodigy 30 points

    Tested two of my old prototypes, which each contain 1 hub with all connection components in place. If the two are powered on simultaneously, or the downstream hub is powered while already connected via usb to the upstream hub, the downstream does not suspend. If they are powered on without usb connection, then connected, the downstream hub switches from suspended to not suspended upon connection. In all cases, the downstream hub suspends when the upstream hub is connected to the host, and stays suspended when the upstream hub is disconnected from the host. Seems like it's something to do with the connection event between the hubs.

  • 0 in reply to W A
    TI__Guru*** 145435 points

    Hi:

      this is very old hub, let me discuss with our system guy   to see if they have any suggestions.

    Regards

    Brian