0
\$\begingroup\$

I have upgraded a design with a USB2412 to a design with a USB2514Bi chip to have more USB ports available.

During testing we discovered the device works like expected when we connect it to a USB A port. When the device is connected to a USB C port it does not work until we disconnect and immediately reconnect the device. The only difference between both ports is of course the USB PD protocol.

We have measured and confirmed the USB PD configuration is OK (5.1k pulldown resistors in place on both CC1 and CC2). +5V is correctly supplied to the device. When we look at the current draw, something strange happens. On the first connection over USB C 38mA is drawn. After reconnecting 78-82mA is drawn, the same as when it is connected to a USB A port.

When we look at the device manager, the generic USB hub (so the USB2515Bi chip) does not show up at the first connection. After reconnecting it does show up (and so do the downstream connected USB devices). It seems like the USB2515Bi is simply not starting at the first power-up. Maybe because the USB PD power negotiation takes some more time than the normal USB powering does?

On the old version of the board with the USB2412 chip, this problem does not occur.

I looked at the reset circuit in both the USB2514B and USB2412 Hardware Design Checklists. In the USB2412 document (figure 7-2) the R=10kΩ and C=0.1μF. In the USB2514 document (figure 7-2) these values are R=10kΩ and C=1μF. On the PCB we used R=100kΩ and C=0.1μF. The description in the USB2514 document says [The values for the “R” resistor and “C” capacitor are not critical and may be adjusted per individual system needs or preferences.]. As the timing on both the USB2514 and USB2412 chips are equal (see USB2514 datasheet figure 5-1 and USB2412 datasheet figure 3.1), we did not change the R/C values on the new board. But as we run into this issue now that may be the solution.

Do you have suggestions on this? Maybe I have to delay the reset trigger? Do you have recommended values for R/C in the reset circuit for use with USB PD powering?

\$\endgroup\$
8
  • \$\begingroup\$ I'd suggest following the manufacturers suggested values and seeing what happens. \$\endgroup\$
    – Finbarr
    Commented May 4, 2023 at 8:44
  • \$\begingroup\$ @Finbarr, suggested value for which parts? \$\endgroup\$
    – user220456
    Commented May 4, 2023 at 8:51
  • \$\begingroup\$ If you already have recommended values from manufacturer, and are not using them, why not just use the recommended values? Also, have you done any debugging if the values even are the issue or not? It might be somewhere else. \$\endgroup\$
    – Justme
    Commented May 4, 2023 at 9:46
  • \$\begingroup\$ @Justme, just a clarification, Maybe because the USB PD power negotiation takes some more time than the normal USB powering does? \$\endgroup\$
    – user220456
    Commented May 4, 2023 at 10:02
  • \$\begingroup\$ @Newbie but you are not using USB-PD. \$\endgroup\$
    – Justme
    Commented May 4, 2023 at 10:13

0