USB-C Preventing Reverse Flow

[ACharnley]

Say you have a battery connected to a buck connected to a USB-C power-bank which works in either input/output mode. To declare the buck as an output resistors are used on the USB-C CC lines. Normally they are connected to VBUS, but if the buck is turned off then these are likely going to be closer to ground, in which case they now function as a potential sink indicator.

The power-bank on the other-hand might be monitoring the CC line to see whether it has a voltage at all (so it's connected to an output), whether when it provides a voltage the read is the same (so it's not connected). These in conjunction with whether there is a voltage on VBUS.

All this because potentially the power-bank can reverse flow through the diode of the upper FET on a buck and cause it to fail if the current draw is great. It's a common failure in a few products I've seen where the CC resistors aren't handled properly either at input/output end.

I'm mostly thinking to connect the CC resistors to a permanent voltage source rather than VBUS in hope that the power-bank will know it's connected to a power source but the power source is not necessarily functioning right now.

Be good to know if others have came across this situation!

 

[technofreak]

Hi, If you observe powerbank has different ports for charging the powerbank and connecting other devices for charging from powerbank.

If you can post schematic will be more easy to understand.

 

[ACharnley]

This is not always true, many power-banks now have input/output USB-C port support and the crux is how detection is implemented. Many bucks use a 400ohm discharge resistor when off and if USB-CC pins are tied by resistor to the VBUS line then they now become almost tied to ground, which advertises the output as an input.

And potentially you now have reverse flow through the upper FET diode of the buck. This may not matter, but if there's a current consumer upstream of the buck then the diode is going to warm up, and...

 

[technofreak]

This is not always true, many power-banks now have input/output USB-C port support and the crux is how detection is implemented. Many bucks use a 400ohm discharge resistor when off and if USB-CC pins are tied by resistor to the VBUS line then they now become almost tied to ground, which advertises the output as an input.

And potentially you now have reverse flow through the upper FET diode of the buck. This may not matter, but if there's a current consumer upstream of the buck then the diode is going to warm up, and...

Thank You, I have seen these in laptops, did not see in powerbanks.

came across this article from Digikey while searching..

Designing In USB Type-C and Using Power Delivery for Rapid Charging

 

[ACharnley]

I think there's a little more to it and not all power-banks are conformal. My thinking is it needs a CPU to first check that a) there's VBUS and b) there's CC pull-up and then configure to source current, otherwise if there's no VBUS and the CC pull-down is correct then configure as an output.

* I said before the CC resistors would configure as a pull-down if they're tied to VBUS, this is true but the values would make the CC voltage read 'illegal', so that should, in theory, stop the downstream pumping current out. Unfortunately I've seen one instance where it ignores this logic.