TPS65987D: DisplayPort Alt Mode Discovery sequence

Hi Anderson,

I'll looking into this question and will respond you back in a day or two.

Thanks,
Rahul

Anderson,

The device does initiate a VCS if it has to talk to the cable, and it can as well be configured to automatically send 'Get Sink Cap' messages, but unfortunately not in the sequence that you listed. The sequence defined in Section-5.1.2 of the DP specification is not mandated by the PD specification, and therefore is not implemented in 987D’s firmware. May I understand the reason for this request? Do you see any failures w.r.t your DP use-cases?

-/Praneet

Dear Praneet
When TPS65987D setting prefer to be Source and UFP_U/D, I find a compatibility issue with MacBook.
1. Shut down MacBook.
2. Connect TPS65987D(prefer to be Source and UFP_U/D)
3. MacBook reject DR_SWAP
4. In the end TPS65987D is Source, DFP_U and UFP_D.
I think if TPS65987D follows the sequence defined in Section-5.1.2, this compatibility issue will clear.

Anderson,

Since the Macbook rejected the DRS-To-UFP request of the SRC (I'm assuming its the monitor here), 987D on the monitor side initiates the DP AM negotiation, and this is the expected behavior.

As I mentioned earlier, the sequence defined in Section-5.1.2 of the DP specification is not mandated by PD specification, and therefore is not implemented in 987D’s firmware. A suggested workaround for your immediate need is to disable ‘Automatic ID Request’ in ‘Port Control 0x29’ register – This will prevent the monitor from initiating the AM discovery sequence as DFP. The Macbook shall eventually initiate a DRS-To-DFP and initiate the DP AM sequence. Please test this configuration and share your observations.

-/Praneet

Dear Praneet
Thank you for your suggestion.
If disable ‘Automatic ID Request’ in ‘Port Control 0x29’ register, the MacBook issue will be solved but the TPS65987D behavior cannot meet DP Altmode CTS(10.3.14).
So, I don't want to change this setting much.
Do you have plan to implement the mandated items about DP altmode in 987D's firmware?

Anderson,

I'm not sure how the sequence of Section-5.1.2 will help in this case - The Macbook could still 'Accept' VCS and 'Reject' the subsequent DRS request from the monitor, and the ports will end up in the same current state (Monitor/SRC/DFPU/UFPD and Host/SNK/UFPU/DFPD). As I mentioned in my previous posts, if the host sends 'Reject' in response to a device's (whose preferred DP role is UFP_D) DRS request, the latter should continue w/ its AM discovery and entry sequence. Host should ideally respond w/ a 'Wait' in such cases, so the device doesn't negotiate and enter DP AM as UFP_D - Please take up this issue w/ the Macbook vendor.

Let me know if you need additional information.

-/Praneet

Dear Praneet
I am sorry, my question was not good.
I understand you mentioned.

You saied "the sequence defined in Section-5.1.2 of the DP specification is not mandated by PD specification".
But the sequence defined in Section-5.1.2 of the DP specification is mandated by DP specification.
If 987D don't meet DP specification, the device using 987D may be failed the compliance test.
I want to know whether you have plan to implement the mandated items about DP specification in 987D's firmware.