Other Parts Discussed in Thread: AM5728, , TPS2052
I'll paste an image of my USB circuit below. It may look a little wonky - sorry.
We use two TPS2052B chips to switch power to three USB jacks. The switches are controlled by a TUSB4041, which in turn is connected to an AM5728'S USB2 port. We see a strange intermittent problem: sometimes the switch doesn't connect 5V to the VBUS, and we measure about 1V, +/- .2V. I get this result even if I disconnect the load from the jack.
I suspect that the TVS is slowing the discharge of the 5V, which means the cameras connected to the USB jacks are pretty high-impedance as well. Could having 1V or so on the output of the TPS2052B, while the 5V supply is at zero, create some kind of sneak circuit that locks up the switch somehow?
Alternate theory: I measured the enable signals with an ohmmeter and it read 0V, which seems lower than normal logic would read. Could it be open somehow?
I also considered TPS2052B's thermal shutdown, but the chips aren't getting very warm. OCn signals are high, so overcurrent doesn't seem to be the issue.
Circuit diagram:
