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.

USB C Data line signaling, Battery Charging application

Other Parts Discussed in Thread: TUSB320, TPS2511, TPS2561A, TPS2543, TUSB320EVM

Hello.  I am interested in using the TUSB320 from TI and configuring as a UFP.  The chip would be in a mobile battery pack, for example, and would communicate to whatever usb c host there is, through the 24-pin USB C connector.  I really don't care about the data lines.  I just want to use the power from the host (and also use some of the spare data lines to provide MORE power for faster charging, and would use a 1:2 switch to switch the lines between two different circuits).

I know the TUSB320 would present a pull down resistance on the CC pins, since it is acting as a UFP.

I have looked at a schematic in the TUSB320 datasheet (figure 14) where the TUSB320 is in UFP in I 2C Mode Schematic.  I would want to use it in this way, basically, but the schematic does not show what is done with the USB 2.0 and superspeed diff pair data lines.  I want to know if there is a chip like the TPS2511, but for the data lines on a USB C connector.  That is, is there some sort of chip that I can put down on my board and connect the data lines to so that i still get power from the host but have all other data lines (besides the CC pins of course.  Maybe i need those to work between the host and my TUSB320)?

My concern is that if I just leave all of the data lines unconnected/open on my board, then the host will give freak out and decide not to provide any power at all.  Is there any reason to believe that the host would quit supplying power all together if the all of the data lines (except CC pins!) are just left open?  Because that's probably the simplest solution in my mind.  Or, I can just short them together.  I'd do that if I know that's ok.  That's easy to.  I want my battery pack to work universally, with any kind of usb c host.

Also, I am looking for a schematic for the TUSB320 Evaluation modul (EVM).  I can't find a schematic on the product's TI web page.  But, I wonder if the answer lies there.  That is, if the designers of that board already know about this, they designed the module to provide power even though all of the data lines (besides CC pins) were just shorted together or left open.

Looking forward to a response!  Sometimes I don't have luck with forums..

Matt

  • Hello Matt,

    The TUSB230 will present a pull down resistor on the CC pins. This will generate a detection on the Host side and therefore your device will have access to VBUS. At this point, the device should be able to draw up to 3A from VBUS (depending on the advertised current mode; USB3 standard, 1.5A or 3A). Please let me know if you want to support legacy battery charging (BC 1.2), in order to research how has to be done.

    Best regards,
    Diego.
  • Thanks, Diego!

    1. I would like to have the ability to support legacy battery charging (according to the BC 1.2 spec.).

    2. Ok, so it seems that as long as the CC pins present the proper pull-down resistance to the host, then I can leave all other lines open (except all ground and power/5V lines) because my device would be getting power from the host. Great. In this case, I will just leave all of the data lines open. That's all right?

    3. Another question. Looking at the USB-C connector, there are 4 V_BUS lines.
    Let's say I want to preserve the reversibility of the connector, but that I want to re-purpose 2 of the VBUS pins, let's say. That is, I will take the 5V from 2 of the VBUS pins and NOT connect them to the other 2 VBUS pins on my board. Instead, I will use the 2 re-purposed V_BUS pins to power a "stand-alone" circuit, and these 2 re-purposed V_BUS pins are their own net in my schematic, say "VCC_5_2" while the 2 original V_BUS pins that I won't re-purpose might make up a "VCC_5_1" net in my schematic.
    In this scenario, would their ever be a host that finds out about this, is unhappy, and decides to not provide power to any of the 4 v_BUS pins, for any reason?
    In this case where I would just be using 2 V_BUS pins (2 that are not re-purposed) to power most of the electronics on my board, i suppose i limit myself to 1.5A as opposed to 3 A, which you pointed out, Diego.

    4. I want to understand what you mean by advertised current mode. If my battery charger with its TUSB320 is acting as a UFP, what would make the TUSB320 advertise "3A" and what would make the TUSB320 advertise "1.5A" or "1.0A"?

    5. Is there a schematic available for the TUSB320 Evaluation modul (EVM)?

    Thanks!!
    Matt
  • Hi Matt,

    To comply with BC 1.2 spec you will need to use TUSB320 + a device like TPS2543 or TPS2561A, using this configuration you don´t leave data lines open since data lines are connected with TPS device.
    Please refer to BC1.2 spec on this link: www.usb.org/.../devclass_docs

    To answer your third question, yes you can have two VBUS nets just be careful not to draw more current than the host can provide.
    All 4 VBUS pins are coming from the same source, consider this for your design.
    The Host won't shut down VBUS unless it has reached it's limit current draw.

    The TUSB320 does not advertise the current draw in UFP mode, in fact it detects the amount of current the host is able to provide, by means of CCpins, so, it's the host the advertiser.
    Please send me an e-mail to request TUSB320EVM schematic to JoseCarlos.Gil@TI.com

    Regards.
    JC
  • Thanks for the good information, JC.  

    I didn't know about the TPS2543 or TPS2561A.  My impression is that these chips are mostly used in a "DFP" role.  Perhaps they can be used in an "UFP" role.  It is not entirely clear to me.  Nonetheless, what I'm really looking for is something like the TPS2543, acting in a UFP role, and would present the proper electrical signatures on the data lines on all differential pairs.  That is, I am looking for a TI chip that, on detection of "legacy" devices using the USB 2.0 protocol, would present the proper electrical signatures on A6/A7 or B6/B7 (USB 2.0 D+/D- on USB-C connector) of the USB-C connector.  But, the TI chip would also have 4 pins which detect a device using the RX superspeed data lines, and then output the proper electrical signatures on the TX superspeed data lines.  

    So, I'm trying to imagine something like the TPS2543 that would be the USB-C version (but also support legacy devices with USB 2.0).  The chip would have 4 of its pins dedicated to both superspeed diff pairs.  The chip would detect a certain kind of host and then output the proper signatures on the other 2 of 4 superspeed data lines.

    Thanks,

    -Matt

  • Let's say I use the TPS2561A in a UFP role.  I connect the data lines from the USB-C connector to the "IN" side of the switch.  But then what do I connect the data lines on the "OUT" side to?  A simple resistor to ground?

    I am not using any of the data lines for anything.  So, I just want to tie them off and forget about them, so the host can work with the TUSB320 and provide power to my board via a USB-C connector.

    Thanks,

    Matt

  • I still need to know what to do about the data lines. That is, should they be left open, shorted together, pulled up/down, or if there is such a chip that does it. It doesn't look like the TPS2543 is meant to go with USB-C connectors, and the TPS2561A is a switch .. I would still need to know what goes on the non-connector side of the switch to properly load the data lines (the data lines that I'm not using..)

    Thanks,
    Matt Overlin
  • Hi Matt,

    I'm reviewing this with apps team, as soon as I get something I'll get back to you.

    Regards.
    JC
  • Hi Matt,

    If your design is used to receive charge only through Type-C port, you can left open data lines since the control is made with CC1 & CC2 pins.

    Regards.
    JC

  • Ok, great.  So, I would plan to leave the SSTX pair, SSRX pair, old DM/DP pair, as well as the SBU lines open.

    Can the SBU lines be left open as well?

    Thanks,

    Matt

  • Find out an answer by looking at the USB 3.1 spec. It recommends leaving the SBU pins open. Or, the SBU pin can be pulled down to ground with a resistor with resistance > 950 kOhm .. which the spec. says is the functional equivalent of an open anyway.

    I think I have all the information I need now.
    Thanks.
    -Matt Overlin
  • Hi Matt,

    Yes SBU lines can be left open.

    Please let me know if anything else is needed.

    Regards.
    JC