BQ25792: Query on Autonomous Operation of TPS25751 + BQ25792 Charger Configuration

Part Number: BQ25792
Other Parts Discussed in Thread: TPS25751, BQ25798, BQ25731

Dear Team,

We are using the PTPS25751DREFR for USB-PD source/sink management and the BQ25792 as the battery charger in our design. The schematic is attached for your reference.

We would like to understand how much of this system can operate autonomously, without any host-side (I²C) intervention or firmware configuration.

Our requirement is to obtain approximately 7.4 V / 5 A at the VSYS output of the BQ25792, with VBUS/PoE input around 12 V, while ensuring that the battery continues to charge at a reasonable rate. The system will be primarily wall-powered, switching to battery power only when the external input is unavailable.

Could you please review the schematic and help clarify the following points:

  1. Autonomous operation: To what extent can the PTPS25751DREFR and BQ25792 function independently without I²C programming? Are the current strap configurations sufficient to meet the above power requirements?

  2. Minimal configuration path: If I²C configuration is necessary, can the MPU (powered from the default VSYS output) subsequently configure both ICs at runtime to enable full performance?

  3. Programming and tools: Can these devices be programmed using any standard I²C host (e.g., Arduino, STM32, or other generic MCU), or is a dedicated interface or GUI tool required? Since the Aardvark adapter is quite expensive for early prototype builds, we would appreciate any low-cost alternatives or evaluation tools you can suggest.

  4. Design optimization: Any recommendations to further reduce software dependency or make the charger + PD controller combination more self-sufficient would be extremely helpful.

  5. Additionally, could you advise on thermal management considerations when drawing up to 5 A from VSYS, and any layout or component recommendations to ensure reliable operation at that load.

Thank you for your time and support. We look forward to your feedback on our schematic and guidance to optimize this implementation.TI_power_review.pdf 

  • Hi Venkatesh,

    Regarding 1, once loaded, the TPS25751 firmware functions and controls the BQ25792 without I2C interaction.  As long as the TPS25751 is connected to USB-C port, the BQ25792 is enabled.  However, if there is no USB-C port and different input source is used, the TPS25751 firmware disables the charger.  The firmware must be modified to accommodate a non-USB-C power source.

    Regarding 2, BQ25792 can be re-configured via I2C using the TPS25751 I2C passthrough. The TPS25751 team can answer if MCU can make changes to operation via I2C after the firmware is installed.  

    Regarding 3, any I2C host can be used.

    Regarding 4, see 1.

    Regarding 5, the BQ25792 package does not have a thermal pad and so gets quite hot above 4.5A.  I recommend fsw=750kHz and 2.2uH inductor for highest efficiency.  Dropping VBUS from 12V to 9V for smaller in to out differential would also improve efficiency.  Lastly, connecting the VIN, PMID, GND, SYS and BAT pins to as much copper (big planes/pours on multiple layers) for heat sinking will also help keep the die temp below thermal regulation threshold.

    On the schematic, the 10uF caps between the charger' input MUX FETs are too large. They will likely prevent the 

    Also, I see there is an option for 1S battery?  The charger's MINSYS voltage cannot be set above the VREG voltage so for 1S battery, MINSYS cannot be set higher than 4.1V not 7.4V.

    Regards,

    Jeff

  • Dear Jeff,

    Thanks for your response. I have a few follow-up points for clarification:

    1. I hope you were able to review the schematic. We have implemented a dual-input design at the supply input — both USB-C and PoE. The charger is expected to work with either source. Could you please advise on the correct configuration for this setup in the battery charger?
      Additionally, once configured, will there be any runtime impact when switching between sources? For example, if the configuration is done for a non–USB-C input (PoE) and we later supply power through USB-C, would that affect operation? If so, what is the recommended way to mitigate such issues?

    2. Both the battery charger and PD controller can be configured using the Aardvark tool, correct? Since they share the same I²C path, I wanted to confirm if that’s the correct approach. I noticed dual I²C connectors on the EVM boards and wanted to verify this setup.

    3.  Is there any package variant available with a thermal pad for improved thermal performance?

      Could you suggest any alternative solution that supports higher current capability, meets our requirements, and can operate in standalone mode (without I²C control)?

  • Hi Venkatesh, 

    Regarding 2, BQ25792 can be re-configured via I2C using the TPS25751 I2C passthrough. The TPS25751 team can answer if MCU can make changes to operation via I2C after the firmware is installed.  

    You can use an MCU to command TPS25751 to execute I2C writes/reads to BQ25798 with commands "I2Cw" and "I2Cr". 

    Thanks and Regards,

    Raymond Lin

  • Hi Venkatesh,

    Not sure how this went for 21 days without response but to answer your questions above.  There is a not another charger with integrated FETs with same power level in a better package with thermal pad.  To improve BQ25798 efficiency and thermal perform, you can change to 750kHz switching frequency and use a 2.2uH inductor.

    For higher current capability you would have to move to a charge controller with external FETs like BQ25731 in https://www.ti.com/tool/PMP41013.

    Regards,

    Jeff