TCPP02-M18 - Spurious over-current protection fault occurs.

[Shane Kent]

Hello - 

My team and I are utilizing the TCPP02-M18 in combination with the TI TPS65987 USB-C Controller in a USB Power Provider application.  

Right at the moment when a transition from the nominal 5V to a higher negotiated voltage occurs the TCPP02-M18 triggers an over-current fault and removes power from the Sink device. This is confirmed by reading the I2C registers of the TCPP02-M18 where Register 0x02 reports 0x82. When I clear the fault via the I2C writes the same cycle occurs.  

If I short over the sense resistor, R21 in our schematic, the over-current fault doesn't occur. Page 9 of the UM2973 for the X-NUCLEO-SRC1M1 suggests that the drain-to-gate capacitance for the MOSFET must be greater than 20pF.  I'm using the Vishay SQJQ904E-T1_GE3 as the dual N-Channel MOSFET in the provider path which has a gate capacitance of ~0.5nF to 1.5nF, so an additional capacitor should not be necessary.

The Sink device that I'm using in these tests on consumes ~20mA on the VBUS net, so we should be very far from any real overcurrent events. I do not see any transition events on the IANA output of the TCPP02-M18 that would indicate an overcurrent is actually occurring.

Below is a screenshot of the TCPP02-M18 portion of our schematic.

Do you have any guidance or suggestions on how to continue to debug this?

- Shane

[siddhipandare]

Hi Shane, 

Did you find an answer to this? or resolved this? We are facing the same issue. Thanks !

[Shane Kent]

Hello -

No, unfortunately we did not get an answer to this question and to resolve
it we designed the TCPP02-M18 from our solution. It's a shame that ST does
not seem to prioritize support for these types of issues.

- Shane

[codex174]

ya bro it’s not real overcurrent just spike when vbus jump
ur sense resistor catching that noise so chip trips
try tiny cap across sense resistor or small gate resistor on mosfet
that smooths edge and stops false fault

[siddhipandare]

Hi, we did not observe any spike on the scope. When I changed the PD contract from 15V 3A to 15V 2.95A (non-standard, just for testing) the OCP did not occur. This suggests it may not be a spike or inrush phenomenon. We are still investigating but I hope STM responds soon.

[siddhipandare]

Hi Shane,

Here’s what we found:

The overcurrent events were indeed valid OCPs, confirmed by measuring the current across the shunt and with a current probe. Previously, we were monitoring the IANA pin output, but it did not capture the 200 us surge. I’d recommend checking the output using a current probe instead of relying on the IANA pin signal.

It is not some spike but the device (laptop) actually draws a surge for 200us.