• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 475 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.

TPS2546: Mode Change from Divider1 to Divider2 (DCP_AUTO)

Alan R Olson
Prodigy 40 points
Part Number: TPS2546
We are using this USB Charge Port Controller to manage a USB type A receptacle.  The ILIM_HI setting is 1.2A for a 1A DCP.
We wish to support portable devices that look for Apple signatures on D+/D- as well as shorted D+/D- per USB BC 1.2 DCP.
When we connect a device (iPhone or static load) we observe the chip change modes from DIVIDER1 to DIVIDER2 with a 2 second pause (VBUS is discharged).
The datasheet explains that the mode change occurs if the load device draws more than 750mA.
The port DCP is configured for 1A so changing to DIVIDER2 mode (2.1A) is not desirable and may confuse the attached portable device.
Can you help us understand why the mode change is happening and why 750mA is the threshold?
Thank you!
  • 0
    TI__Genius 9630 points

    Hi Alan,

    Have you seen any functional issues from the device being in divider 2? If the device is still outputting the correct current, the phone may be able to recognize divider 2 but still output the desired current. 

    If you need to be in divider 1, you can use CTL1, CTL2, CTL3, and ILIM_SEL to force the device into DCP divider 1 as shown in the table below. 

    Thanks,

    Emma

  • 0 in reply to Emma Hoying
    Prodigy 40 points

    The functional issue with the behavior is that the MFI certification lab flags the 2 second reset and mode switch as a failure with their automated test system (ATS).

    Configuring the device for exclusive divider1 mode precludes the use of DCP_Shorted for non-Apple devices.

    I am curious why 750mA was selected as a threshold for trying divider2 mode when the load is already less than the 1A current advertised by divider1 mode.

    Thanks!

  • 0 in reply to Alan R Olson
    TI__Genius 9630 points

    Hi Alan,

    The designer for this part is currently out of office until early November so I will not be able to get back to you on the reasoning behind 750 mA until then. 

    Also, the TPS2546 can support charging schemes tested during MFI compliance. However, the TPS2546 has not been tested or certified for MFI compliance so if you are seeing a failure during MFI compliance testing, it may be valid. The device can still properly support these charging schemes, though. 

    Thanks!

    Emma

  • 0 in reply to Emma Hoying
    Prodigy 40 points

    Thank you for following up with the chip designer.  I hope to read their comments the first week in November.

    Cheers!

  • 0 in reply to Alan R Olson
    TI__Expert 6280 points

    Hi Alan,

    The TPS2546 will first attempt to do Divider 1 which is the first generation of the proprietary charging. When current goes above the 750mA the TPS2546 will attempt to try Divider2 so see if the device connected support the higher current. If it is determined that the device supports divider2 it will remain in that state, if not it will go back to divider1. 

    This allows a device that supports divider2 to charge at the highest current possible and still allow devices supporting the first generation divider1 to charge at the rated current. 

    The reason VBUS is discharged is to adhere to the USB Spec and BC1.2 spec, where a device determines the attach through the presence of VBUS. VBUS is discharged to simulate the disconnect and restart the timing for each of the detection. In a BC1.2 only, Divider1, or Divider2 only charger, the VBUS will never toggle in a normal working condition since the advertisement does not change.

    When you use the DCP auto feature it will have to cycle through all the modes but will have to remove/apply VBUS as it switches.

    When running compliance for a specific mode you should set the control bits according to the test you want to run to not have the DCP conflict.

    Jacob

  • 0 in reply to Jacob Ontiveros
    Prodigy 40 points

    Thank you for this reply.