TPS65988: Sinking and sourcing on both USB-C ports

Hi Ben,

We will assign this thread to one of our engineers and they will respond within the week.

Thanks and Regards,

Chris L

Hi Chris, 

Thanks for your reply. 

Could you please be more specific rather than "refer to datasheet?"
If I decide to use the BQ24292i charger with one VBUS input only with TPS65988 instead of BQ25792, it is possible, as long as I use the two PFETs to ensure both chargers dont supply VBUS sink to the charger? 

What part would you suggest that can do the same as the TPS65988 series then?

Regards

Ben

Hello,

Benjamin Todd said:

Could you please be more specific rather than "refer to datasheet?"

I have given you a direct link to the page/section which describes how to configure the BQ device without external mosfets.

Benjamin Todd said:

If I decide to use the BQ24292i charger with one VBUS input only with TPS65988 instead of BQ25792, it is possible, as long as I use the two PFETs to ensure both chargers dont supply VBUS sink to the charger?

Can you provide a block diagram of what you are trying to do?  If you need a place to start consider the block diagrams found in the datasheet starting with Figure 9-7. TBT Notebook with PD Charging.  You can disregard the high-speed lanes and focus on the power.  Figure 9-11 will be more like the EVM referenced in the previous post.  If you want to 'OR' the power paths to the charger, then you would need an EC to manage which power path you want to sink from in the event that both ports are connected to sources.  Alternatively, you could update the unused port in real-time to only support source mode to guarantee only one power path to the charger is enabled at a time.

https://www.ti.com/lit/ds/symlink/tps65988.pdf#page=50

Regards,
Chris

Hi Chris,

Apologies for the lack of clarity, and thank you for your explanations.

I can provide a block diagram if needed, but my implementation will likely follow Figure 9-7 quite closely.

My main question is regarding the ORing of the external sink power paths. As you mentioned, to achieve this, I assume I need to implement the FETs shown on the EVM — specifically, the ones with gates connected to PPEXT1 and PPEXT2 on the TPS65988.

I’m likely not going to have an embedded controller in my system, so I was hoping to rely on the TPS65988 reading from the SPI flash for configuration and handling the toggling of the PFET gates. 

My understanding is that, in this setup, if one port is sourcing VBUS, the TPS65988 would automatically prevent the other port from sourcing VBUS or passing power to the charger in the instance that both sources are connected. Is that correct, and is there any risk of accidentally passing two VBUS sources to the charger with this configuration, if setup correctly with the config tool. 

I plan on using two external FETs for the sink paths, and the source paths will be the internal paths as per Figure 9-7 block diagram. 

Thanks again,
Ben

Hi Ben,

Could you still draw out a simple block diagram? I am a little confused on how you are using the BQ device. Is it only being used for sinking? How are you sourcing power and how will you be sinking power?

Benjamin Todd said:

if one port is sourcing VBUS, the TPS65988 would automatically prevent the other port from sourcing VBUS or passing power to the charger in the instance that both sources are connected.

What do you mean when you say sourcing? Typically "source" is used to say that the port will provide power to the type-C port to an external device. I.E. sourcing 5-V from the Type-C port to a peripheral. The term "sink" is used when power is being drawn from the port into the system.

In the block diagram in figure 9-7, both port1 and port2 are capable of sourcing 5-V and this is done somewhat independent of the other port. If one port is acting as a source, there is nothing automatically preventing the other port from also sourcing 5-V.

In general, the two ports on the TPS65988 act independently of one another. There are some fairly limited behaviors that can depend on the other port, but typically an MCU is required for more complex behaviors and dependencies on other ports. If you are expecting ports to not support sourcing when the other port is sourcing, it may not be something easily implemented with the PD controller alone.

Thanks and Regards,

Chris

Hi Chris, 

Correct, I will only use the BQ to sink VBUS, won't be using its internal boost at all. So my plan was to short the PPHV1/2 paths together to a 5V VBUS source, as I did with TPS65987D (only using one port). 

I plan on implementing the exact same setup here (i.e using the internal power paths for sourcing VBUS, which will be from an external 5V DC/DC boost). The external paths will be used for sinking VBUS (i.e charging the battery/powering the load). 

My question is, is there any risk of the two VBUS sink path switches being closed at the same time, therefore shorting the two VBUS lanes if two same chargers are connected? Or if the TPS65988 is configured correctly in the GUI, this situation cannot happen? (i.e once one path is closed, the other will not also close?). 

I am wondering if this is all handled by the firmware or if I need some external EC as you suggested. 
If you could confirm that would be great. 

Kind regards
Ben