BQ25611D: Cannot keep VSYS when entering HIZ mode with adapter present

Part Number: BQ25611D


Tool/software:

Hi

I'm currently trying to comply with USB suspend current by entering this IC into HIZ mode, while keeping the system powered from battery.

However, after I set EN_HIZ bit to 1 in REG00 - voltage on VSYS collapses, unpowering the system completely.

I've also tried to set BATFET_RST_EN and BATFET_DIS bits to 0 before entering HIZ mode - but it does not help and charger IC still disables power to the system completely.

Is there some configuration I'm missing to make it work properly - if it is possible at all?

For more context -  the system behind VSYS consumes around 4.2mA in this state, so with 2.3mA added by charger IC it's not an option to keep it all powered from VBUS and hope to comply with 2.5mA USB suspend current.

  • Hello,

    I have attached a 3.4V battery with 5V input with 340mA ICHG and 0.5A ISYS. When I hit EN_HIZ, the SYS drops to the battery voltage and the battery supports the load. It does not crash. What are you using to test for the battery? Do you have some bulk capacitance on BAT?

    Regards,

    Mike Emanuel

  • I'm using a Li-Poly battery charged to 4.1V. Yes, there is some bulk capacitance attached to battery line.

    CE_CHG also goes high before entering HIZ and we configure ICHG as 1380mA. Watchdog is also disabled.

    The circuit in question:

  • Hello,

    Please list all register settings before and after the load is enabled. Please detail all steps taken to replicate the issue, including the system load. Are you using a real battery or battery simulator? If battery simulator, how much output capacitance is present?

    Regards,

    Mike Emanuel

  • It's not a simulated battery, it's this pack https://www.fullymaxbattery.com/china-1500mah_3_7v_lithium_ion_battery_electronics_lithium_ion_polymer_battery-35945874.html

    Register dump right before HIZ state is entered (pairs are reg:value):

    00h:0x0E, 01h:0x1A, 02h:0x97, 03h:0x01, 04h:0x40, 05h:0x8E, 06h:0x66, 07h:0x48, 08h:0x44, 09h:0x00, 0Ah:0x80, 0Bh:0x54, 0Ch:0x05

    Measuring VSYS shows this rail slowly dissipating voltage from around 0.7V as if it's floating with some capacitors discharging, while VBUS is still at 5V.

    The board is connected to a USB hub, receives enumeration and then USB hub is disabled on PC to force USB suspend event on the bus (alternatively PC can be sent into sleep mode). Upon receiving this signal firmware does some housekeeping (disable LEDs and some other peripherals it does not need) and sets EN_HIZ bit in REG00.

  • I think I've found the issue.

    Our system also reacts to VBUS disconnected by entering the charger into shipping mode to conserve power.

    However VBUS state itself system detects when it polls charger REG0A VBUS_GD - which switches to 0 once we request charger to enter HIZ mode.

    DS does not mention explicitly that entering HIZ mode also updates VBUS_GD state, but since poor source detection is based on applying a certain load - it seems understandable that disconnecting charger from VBUS completely includes resetting poor source detection state.

    Can you confirm this is the behavior we are observing?

    Register dump seems to be confirming this:

  • Hello,

    We are currently at a conference and will try to respond by end of Monday. However, I believe that in HIZ that VBUS_GD is low. I would design the firmware to not disable the BATFET immediately when losing the input. If you have a system load and lose the input, you do not want to immediately cut off the BATFET as the system will crash.

    Regards,

    Mike Emanuel