TPS65988: Debug adapter issues

Hi Aya,

I have an issue with the custom board, that is why I would like to use the debug probe. It would be much easier to use the existing Application Customization Tool instead of implementing the same thing in the MCU.

Maybe you can help me with that other issue instead.

The port controller works well if I power it externally. The device can negotiate the required 12V with the charger and switch to the USB power source for operation. This is what I see with the PD analyzer:

VBUS_UP
SOURCE_CAPABILITIES
GOODCRC
REQUEST
GOODCRC
ACCEPT
GOODCRC
PS_RDY
GOODCRC

If I try to connect the charger in a dead battery state, the board will get caught in a boot loop. Using a PD analyzer, I can see the following happening in an infinite loop

VBUS_UP
SOURCE_CAPABILITIES
SOURCE_CAPABILITIES
SOURCE_CAPABILITIES
VBUS_DN

What can be the reason for this?

Best regards,

Adam

Hi Adam, 

Is the port negotiating the 12V sourcing, sinking, or DRP? From your description, it sounds like the port is intended to behave as a source (providing power to devices downstream), is that correct?

In a dead-battery state (i.e. powered from VBUS), the device would not be able to source power, until the dead-battery flag is cleared through I2C 4CC command "DBfg"

Best Regards, 

Aya Khedr 

Hi,

Sorry, if I wasn't clear there. Here is some more context:

The device in question is a docking station for smartphones. It is supposed to have 2 options for powering: DC plug and USB Type-C. If you plug in both at the same time the DC plug will be disconnected automatically. The TPS65988 has one port strictly as a source (connected to the phone) and the other strictly as a sink (possibly connected to a phone charger).

The system works perfectly if I power it from the DC plug, it is able to charge the phone the TPS65988 acting as a source.

The system also works if I start with a DC plug, then plug in the USB charger and then remove the DC plug. The TPS65988 in this case will sink 12V and once stable the board will switch over to draw power from USB instead of the DC plug.

The issue happens when I try to power the board using the USB charger. There is no phone connected at this point. The TPS65988 is acting as a sink. I can see that the MCU and TPS65988 can power up in legacy 5V mode. Once the MCU configures the TPS65988 to negotiate the 12V the whole system resets. Voltage drops to 0V then back to 5V and the cycle starts again.

Best regards,

Adam

Hi Adam, 

Thank you for clarifying. 

Adam Kisari said:

The issue happens when I try to power the board using the USB charger. There is no phone connected at this point. The TPS65988 is acting as a sink. I can see that the MCU and TPS65988 can power up in legacy 5V mode. Once the MCU configures the TPS65988 to negotiate the 12V the whole system resets. Voltage drops to 0V then back to 5V and the cycle starts again.

Is the USB charger PD capable? What is the power rating of the adapter? 

Also, have you tested with other Type-C PD adapters? Is the behavior the same? 

Best Regards, 

Aya Khedr 

Hi,

Yes, the charger is PD capable. It is rated for 12V 1.5A max. I have tested with different adapters too.

The PD negotiation happens without any issues when not in a dead battery state so the charger must be compatible.

Best regards,

Adam

Hi Adam, 

What is the dead-battery configuration set to? And how is the configuration being loaded to the PD controller (MCU or SPI flash?)

Best Regards, 

Aya Khedr 

Hi Aya,

There is no flash on the current board, but I already added one to the next revision in case it might be necessary to achieve what we want. The goal would be to do this without flash if possible.

I have an MCU that would load the "Low Region" bin file generated by the Application Customization Tool when the system is powered up.

I tried multiple bootstrap configurations.

In the meantime I found the following post. It seems to describe my exact situation/problem, but it has no public solution.

e2e.ti.com/.../tps65988-seems-stuck-in-legacy-mode

Best regards,

Adam

Hi,

Please expect delay in response due to Thanksgiving holiday.

Regards

Hi Adam, 

Thank you for your patience. 

Is there a step in the process where you could be resetting the PD controller? (maybe through the 4CC command "GAID")

Best Regards, 

Aya Khedr 

Hi,

These are the commands I execute for the fw update

PTCq
PTCr
PTCs
PTCd (in a loop)
PTCc
PTCq
Gaid (Warm reset, I also tried without this)

Best regards,
Adam

Hi Adam, 

I will review the sequence provided and get back to you with feedback by end of week. 

Best Regards, 

Aya Khedr

Hi Aya,

Do you have any news for me?

Best regards,
Adam

Hi Adam, 

Thank you for your patience. 

I don't see any obvious issue with the flow described above. The MCU seems to be loading the configuration correctly since you are not observing issues when in a non dead-battery condition. 

1- Does the MCU write to any of the PD controller's registers after the configuration is loaded to the PD? There are some registers such as Global System Configuration (27h) and Port Configuration (28h) that if modified, will cause a port disconnect and reconnect. If a port disconnect occurs, VBUS will go low, the MCU will reconfigure the PD controller, and the cycle would repeat. 

2- The dead battery configuration should be set to one of the following when loading the configuration by MCU. 

Best Regards, 

Aya Khedr