0
\$\begingroup\$

I'm designing a board based on a USB Hub HS 2.0 that has five permanent attached ports, two of them with a high current consumption (180 mA typical, 250mA max. per device).

I would like the board to work just using an USB Type C port as the power source.

I've chosen a USB2517 from microchip, and I'm having some issues with the design (and my knowledge about the USB standard):

  1. If I use a bus powered USB 2.0 design, as my devices will probably draw more than 500mA they won't work as expected, right?

  2. Can I use 5k1 CC resistors and power the USB Hub as it was a self-powered hub rather than a bus-powered one? Will this work besides having a 2.0 USB Hub?

  3. I'm having doubts about the power distribution to the downstream ports. I've chosen a ganged configuration but I'm not fully sure. If my devices are permanent, should I use individual current sense and power control or using a ganged ports control is correct?

Thanks in advance!

\$\endgroup\$
7
  • \$\begingroup\$ I believe this isn't allowed as devices downstream of a self-powered hub are only allowed to use 100mA - I think. But if you are willing to bend the rules, you could make it work anyway. What do you mean permanent attached ports? \$\endgroup\$ Commented Feb 24, 2023 at 23:19
  • \$\begingroup\$ I'm using a USB hub to connect several bridge devices such as ethernet-to-usb bridge or usb-uart brige. All of them are routed within the same PCB, that's what I mean with permanente attached. As far as I know, the 100mA limit would just apply for the bus-powered devices, right? \$\endgroup\$
    – fapps
    Commented Feb 25, 2023 at 8:02
  • \$\begingroup\$ ah well then I assume the rules about power sharing for USB hubs do not count, because this is not a USB hub, it's a device. I thought I read somewhere a specification about devices with built-in hubs, but I can't find it now. \$\endgroup\$ Commented Feb 25, 2023 at 8:10
  • \$\begingroup\$ You are talking about the power going alongside the USB ports, but you won't be routing the power through the hub chip, so don't worry about that, but I would worry about how the computer will see the power management, and I don't know how to do that properly. Maybe the hub should be labeled as self-powered, or maybe there's another way to tell it to ignore power management for this device, or something. \$\endgroup\$ Commented Feb 25, 2023 at 8:12
  • \$\begingroup\$ well, I just learned that hubs do actually control power, so maybe I'm wrong about the power not going through the hub chip \$\endgroup\$ Commented Feb 25, 2023 at 8:21

1 Answer 1

0
\$\begingroup\$

can I be sure that placing the 5k1 CC resistors will ensure 1.5A of power always?

No, you cannot be sure about this. Your job is to put 5.1k resistors to ground, and then to measure the voltage level(s). 5.1k is a must, you have no other choice.

The host ports, however, can signal their particular capability as 500/900 mA, or 1500 mA, or 3000 mA, depending on pull-up they provide (and are designed for).

Some Type-C ports might provide only basic capability, so drawing boldly 1.5A might not work. Your job is to detect the port capability and act accordingly.

You can find more details here

https://electronics.stackexchange.com/a/327946/117785 https://electronics.stackexchange.com/a/442152/117785 https://electronics.stackexchange.com/a/517087/117785

\$\endgroup\$
1
  • \$\begingroup\$ I understand this should be the way to go in a fully compatible USB device. However I'm looking for a simple device. If the host is an USB 3.1, the provided current should be at least 1.5A in most cases right? \$\endgroup\$
    – fapps
    Commented Mar 5, 2023 at 9:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.