Elite x2 1012 G1 does not successfully negotiate USB PD contract with sources w/ more than 1 PPS PDO

[zhongfu]

(tl;dr: possible firmware bug in TPS65982 type-C/PD controller; does HP have any plans to update firmware?)

I've been experimenting with PD chargers and my Elite x2 for a while now, and amongst other discoveries (e.g. the Elite x2 only draws power if it manages to negotiate a contract for 12V and a minimum of 2A), I've found out that the Elite x2 won't even successfully negotiate for any PD contract if the source has more than one PPS PDO -- not even 5V 3A.

Here are some examples of PD sources/chargers I've tried:

1. 18W charger with 5V 3A, 9V 2A, 12V 1.5A fixed PDOs: doesn't charge, but negotiates 5V 3A contract
2. 45W charger with 5/9/12/15V 3A and 20V 2.25A fixed PDOs: charges
3. 60W charger with 5/9/12/15/20V 3A fixed PDOs and 3.0-21V 3A PPS PDO: charges
4. 100W charger with 5/9/12/15V 3A and 20V 5A fixed PDOs, and 3.3V-21V 3A PPS PDO: charges
5. 60W charger with 5/9/12/15/20V 3A fixed PDOs, 3.0-5.9V 3A and 3.0-11V 3A PPS PDOs: charges (*)
6. 100W charger with 5/9/12/15 3A and 20V 5A fixed PDOs, 3.3-11V 4A and 3.3-20V 5A PPS PDOs: doesn't charge, doesn't negotiate PD contract

When I plug the tablet into charger #6, the source sends a (valid) Source_Capabilities PD message advertising all 7 PDOs, but the tablet doesn't respond with a Request message [footnote #1]. At the same time, Vbus seems to pulse on and off between 0V and 5V. (hard reset? idunno)

As for charger #5, which does have more than 1 PPS PDO: it seems that it doesn't advertise any PPS PDOs when plugged into my tablet, which would fix this problem. It's still nothing more than a workaround though, and it would require every charger to implement this workaround (even though they would otherwise be compliant with the PD spec!)

This issue also seems to be present in one other PD device that I've tested -- a Dell XPS 15 9550, which also has a Alpine Ridge Thunderbolt 3 controller as well as a TI TPS65982 Type-C/PD controller. None of the other PD devices I've tested -- four different powerbanks, three different phones, two different laptops, and even a HP t430 thin client, which uses a Cypress PD controller -- have issues drawing power from the charger, which leads me to think it's an issue with the TPS65982, and possibly old firmware. I'll be doing more testing if I manage to get my hands on other similar devices, or a proper PD analyzer (I should have a CY4500 on the way now)

Here's some forum threads I found off TI's support forums, that are related to this issue:

• http://e2e.ti.com/support/interface/f/138/t/828192?TPS65982-behavior-when-attached-USB-C-charger-w-P...
  • The first charger (the one with issues) advertises two PPS PDOs, while the second charger (which works fine) only advertises one
• http://e2e.ti.com/support/interface/f/138/t/857196?TPS65987D-Cannot-negotiate-with-Samsung-travel-ad...
  • This is for a different chip -- the TPS65987D -- but the issue seems to be similar. According to TI, older firmware versions only support chargers that advertise at most one PPS PDOs, while newer firmware versions are fine with chargers that advertise any number of PPS PDOs

So, with this in mind, I've already tried updated the Thunderbolt firmware to the latest version available on HP's support site (NVM fw version 25, PD controller fw version 1.07.03 iirc). Why the Thunderbolt firmware? Well, it seems like the PD and Thunderbolt controllers share the same SPI flash for firmware, so I guess that's why. It still doesn't work, though, which is why I'd like to find out:

• if HP is aware of this issue;
• if it is possible to rectify this issue (perhaps with new firmware provided by TI?); and
• if HP has any plans to rectify this issue.

This issue is a little vexing, and that's on top of other minor annoyances like not being able to power the laptop with any power source that doesn't advertise 12V at at least 2A (annoying for multiple reasons, but I shan't expound on this here). And while it would probably be really easy for anyone to just say "why don't you just use the provided power adapter", I'd appreciate if you didn't use that convenient excuse, because that is obviously not what I want to achieve here -- I just want to use a PD-compliant charger that supplies 12V 3A with my (supposedly PD-compliant) tablet.

Anyway, thanks for taking your time to read through this wall of text. I hope this does help someone who's tearing their hair out about why their PD charger won't work -- hopefully HP does something about this, if they're able to.

#1: the PD analyzer I currently have isn't very fully featured; I'm not able to see every single PD message sent (e.g. GoodCRC, Soft_Reset messages aren't shown). I have a proper one on the way though.