STUSB4500 I2C hot swap

[danielweberdlc]

We're using this controller in our project and have the VCC pin tied to VBUS so the device does not pull power from the battery when USB is not connected (it was pulling ~125uA which is too much for our battery life as the main power is not turned off).

We've had a few cases of I2C bus hangs and I'd like to confirm this controller will not disrupt the other bus devices in the powered off state.

[TDK]

If the device is not powered, it will pull down SDA and SCL which will certainly disrupt the bus.

[danielweberdlc]

An unpleasant revelation, given I just built a bunch of boards. irritatingly the note you reference is not in the version of the datasheet I have - I see the note was added in revision 8 in late 2022 after I designed the boards. All of my test boards built and ran fine - guess I should have checked earlier.

I didn't power the device since I have 2 USB-C ports and each one requires a controller and adding the static power draw drains the battery too quickly.

[TDK]

In general when working with chips, it's common to have this due to protection diodes that go from pins to the power rails. Most I2C chips are going to behave this way. STM32 chips can accept I2C signals while not powered but it's the exception to the rule, not the norm.

[danielweberdlc]

Regarding  POWER_OK2, POWER_OK3 and VBUS_EN_SNK pins, what is their behavior when there is no power on the chip?  Are they also pulled down?  Can I simply use the I2C to program the NVM and then disable the I2C interface, as the only needed purpose is programming?