• State
  • Locked Locked
  • Replies 11 replies
  • Subscribers 33 subscribers
  • Views 182 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TPS25750: GPIO4 & GPIO5 Load Capacitance

Sam Bruner
Prodigy 50 points
Part Number: TPS25750

GPIO4 and GPIO5 can be used for USB D+ and D- charging current identification using the BC1.2 spec. Since these pins will be additional capacitive loads on the USB data lines that can affect signal integrity, I wanted to know the load capacitance of these pins in all relevant states.  Mostly, I'm curious to know what the load capacitance will be once BC1.2 handshaking has completed and USB data is being sent between the SoC/MCU/uProc and a PC/host/etc.  I could not find this information in the datasheet and I have only been told that these pins will become "high impedance" once BC1.2 handshaking is complete.

  • 0
    TI__Mastermind 25820 points

    Sam,

    I will have to review some internal data to see if this information has been captured for this device.  

    I am aware of a few customers who have used the BC1.2 negotiation of the TPS25750D with a FULL SPEED USB device on a MCU, but I am not aware of anyone who has used it with a HIGH SPEED (240MBPS) USB link.  The signal integrity of the routing stub will be difficult to contend with even if the capacitance is very low.

    Regards,

    Chuck

  • 0 in reply to Chuck Branch
    Prodigy 50 points

    Thanks for the feedback, Chuck. That was my concern also. I'd like to do USB high speed (480Mbps) and wondered about the stubs, even if the capacitive load could somehow be reduced after BC1.2 handshaking was complete. One thing I had considered was a USB switch that bypasses the path to the TPS25750 after the identification has been completed. Does that seem feasible?

  • 0 in reply to Sam Bruner
    TI__Mastermind 25820 points

    Sam,

    The switches will cause the part to disconnect and reconnect, so it is not a good solution.

    Does you Hub/MCU support BC1.2 charging?  It might be best to just bypass the TPS25750D.

    Regards,

    Chuck

  • 0 in reply to Chuck Branch
    Prodigy 50 points

    Unfortunately, our MCU doesn't have this BC1.2 charger detection built into it. Detecting through the TPS25750 seems to be our only option, excluding a standalone identification chip. I'm a bit surprised the TPS25750 is constantly checking the connection status. I thought once the handshaking was complete, the chip wouldn't check again unless it detected power was lost and regained. Otherwise, I would think regular handshaking would interfere with real USB communication that could occur after handshaking (assuming a CDP). 

  • 0 in reply to Sam Bruner
    TI__Mastermind 25820 points

    Sam,

    I have learned from others on the apps team that we have had multiple customers successfully implement 480MBPS USB with route through on the USB lines of the TPS25750.

    I am going to attempt to get an example layout for you.

    Regards,

    Chuck

  • 0 in reply to Chuck Branch
    Prodigy 50 points

    Chuck,

    Thanks for the update! Our system requirements have been shifting a bit over time. We'll need to prioritize support for Quick Charge 2.0 and/or 3.0 functionality and BC1.2 support isn't as important. This is due to total power consumption needs of our system compared to available power from BC1.2 (7.5W max). We are also able to manually check for D+/D- short without the IC and that captures the most important use case of BC1.2 for our needs (an AC adapter with no USB data connection). It is still interesting for me to know what the pass through implementation would be and/or the parasitic capacitance of those pins in pass through mode, but it seems QC2.0 or 3.0 is not supported natively by the TPS25750 so we'd probably need another IC or another solution. 

    Sam

  • 0 in reply to Sam Bruner
    TI__Mastermind 25820 points

    Sam,

    The TPS25750D does not support QC2.0 or QC3.0, so if these are considered critical, then you will need to derive that functionality elsewhere.

    I also want to warn you that from a compliance standpoint, QC3.0 is expressly banned by the USB IF.

    You can support these modes electrically, but you will not pass a USB compliance test with QC3.0 enabled.

    Regards,

    Chuck

  • 0 in reply to Chuck Branch
    Prodigy 50 points

    I appreciate the information, Chuck. I'd still be curious to know the parasitic capacitance of these pins if used in pass through mode, even though I doubt we will use them in this way. If that's not worth your time as an unlikely use case, I understand, but I'd love to have the data if someone has already measured it and it just takes some time to find it.

  • 0 in reply to Sam Bruner
    TI__Mastermind 25820 points

    Sam,

    I will see what I can do, but in order to even work, it has to be less than 1pF in the disable mode, or it would reduce the USB Bandwidth too much to build a functional system.

    Regards,

    Chuck

  • 0 in reply to Chuck Branch
    Prodigy 50 points

    Chuck,

    My main question was what the capacitive loading would be if we passed the USB data through the part all the time. In other systems, we've been able to pass a high speed eye diagram with close to 10pF of parasitic capacitance from a pass through part.

    Sam

  • 0 in reply to Sam Bruner
    TI__Mastermind 25820 points

    Sam,

    We would have 2 different loads:

    1.  The load when we are doing BC1.2 detection or advertisement

    2.  The load when we tristate the bus when we are done with detection.

    The load when we are disabled will be much lower than the load during detection.  I am working to get the answer for the first load.