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.

TUSB8020B: TUSB8020B for power negotiation (900mA) of a HS USB device

Part Number: TUSB8020B
Other Parts Discussed in Thread: TIDA-00287

Hello,

I am currently designing an USB-powered device where I communicate with the host through HS data lines but need more than 500mA to operate in "full-featured" mode. As most of the target hosts will have USB3 capabilities I thought about using a bus-powered USB3 hub IC as front-end for my device and let it negotiate the desired 900mA with the host, while I do not impact the rest of the system (communication will still occur through the HS lines). Communication through HS lines is a hard requirement, as there is no SuperSpeed capable MCU in the market for my application (I need to act as a device with plenty of interrupt and isochron EPs and there is no usable USB3.0 audio driver to support my application).

For clarification: the hub will be embedded in a device and only the upstream connector will be accessible to the user. One downstream connector will be left unconnected and the other just connected to the HS lines of the rest of the device.

I based my assumption on this thread: e2e.ti.com/.../getting-900ma-for-usb-2-0-using-tusb8020

I have the following questions:

1- Will the insertion of the TUSB8020B hub suffice to get the 900mA from the host? I wanted to just populate one of the downstream ports with the rest of my device (HS) and let the hub do the power negotiation by itself. Would that work?
2- I have seen that there is a reference design with this IC acting as a bus-powered hub, the TIDA-00287. It is though not very clear to me how the IC should be configured to appear and act as a bus-powered device. How do the internal registers and pins have to be configured to achieve such behavior?
3- I wanted to be able to detect when the host is NOT USB3-capable, in order to adjust my power consumption to the maximum 500mA of USB 2.0 (non "full-featured" mode). I assume that the PWRCTLX pins will be asserted when the hub has completed a successful power negotiation with the host but, how can it be differentiated between the cases in which the power negotiation occurred through the SS lines (900mA available) or through the HS lines (500mA available)?

The device I am currently designing would require certification and the following thread caused me some concerns: e2e.ti.com/.../tusb8020b-can-power-be-provided-from-the-upstream-usb-connection

There are two issues with the hub:

4- If the hub allegedly always reports as self-powered, how does the use-case implemented in the TIDA-00287 reference design work? In the documentation it is explicitly stated that the hub operates as a bus-powered device right in the description on the first page of its design guide (www.ti.com/.../tidu428)
5- I assume that the main blocker could be the fact that the hub exceeds the power budget before enumeration. By looking at the table in the section 7.7 of the datasheet it is not clear to me that this will be the case with a USB 3.0 host and a single USB 2.0 downstream device. Can this information be confirmed? In this case, what is the point of the above mentioned reference design, if it is not spec-compliant?

Thank you very much in advance for your time,
Alberto

  • Hi,
    Your issue is being assigned to the appropriate engineer. Due to the US Holiday there may be a delay in response to your post. We apologize for any inconvenience.

  • Hi Alberto,

    USB host ports do not typically limit power draw on ports based on whether a connection is USB 2.0 or USB 3.0.  Power is available on the port and a USB 2.0 host port is required to supply at least 500 mA and a USB 3.0 host port is required to supply at least 900 mA.  Many host ports can supply more.  The USB specs have requirements around how much power a USB 2.0 or USB 3.0 bus powered device can draw in both active modes and suspend modes to prevent interoperability issues, but the requirement is for the bus-powered device to limit the power consumption, the host does not typically do that.

    So the TIDA-00287 is a bus powered application, but it is not a compliant one.  The TUSB8020B will always report as a self powered hub in its descriptors and it exceeds the suspend mode power consumption required by a bus powered device.  Functionally the reference design works well since most USB host ports provide extra power.  Please note that a compliant bus powered hub application would necessarily limit the power available to the attached downstream devices to prevent the power consumption from the host port from exceeding 900 mA / 500 mA total for both the hub and the devices combined.

    I don't think the TUSB8020B is a good fit for this application.  It is not a compliant bus powered hub and it would add in considerable power consumption to your application if bus powered.

    Regards,

    JMMN

  • My main concern is not to be able to make a difference between the scenarios when the host is USB2.0 capable (I can draw max 500mA) or USB3.0 capable (I can draw max 900mA).
    Probably it would be acceptable to violate the 100/150mA pre-enumeration current but not to draw 900mA from an USB2.0 host.

    I was hoping the hub would allow me to detect when the power negotiation was performed over SS lines and thus 900mA are available, as my device will be blind regarding to the communication over those lines. Is there a way to retrieve such information from the hub controller?

  • Hi Alberto,

    The TUSB8020B does provide upstream status outputs (HS_UP and SS_UP on pins 35/36) that maybe useful determining the upstream signal connection speed.  Please note that if you are bus powering the application, the current draw in USB 2.0 mode from the hub alone would be 250 mA, so that would make the max draw of a bus powered device connected to the bus powered hub less than 250 mA.

    Regards,

    JMMN

  • Thank you again very much for your helpful reply. Those signals seem to be exactly what I need for my application. I have overseen them while reading the datasheet Slight smile

    My initial assumption was that the hub controller connected to an USB3.0 capable host and with a single HS downstream port active would consume significantly less than 250mA @5V. By seeing the section 7.7 of the datasheet I would think that my use case would need to allocate a similar amount of current for the hub controller as the 2.0 host / 1 HS device active scenario. This would mean, with 90% efficiency of the DC/DC regulators, around 55mA @5V. I could still live with twice that figure. Is this assumption very unrealistic?

    Communication over the SS lines would only happen during enumeration (in that case I could spend the 250mA without problem, as my system will not yet be up and running) and then it would remain inactive from the application viewpoint.

  • Hi Albert,

    Sorry about that, I was looking at the wrong entry!  Yes, the hub would consume 48 mA @ 3.3V and 71 mA @ 1.1V in USB 2.0 mode with a single downstream device.  With good efficiency regulators, that would be 55 mA at 5V (or with less efficient regulators closer to120 mA) .  When there is a USB 3.0 connection made between the host and the hub, power consumption would increase until the USB host puts the USB 3.0 portion of the hub into low power mode.

    Regards,

    JMMN

  • That is great news! I looks, after all, that TUSB8020B could be a good fit for my design.
    Thank you very much for your clarifications.