• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 33 subscribers
  • Views 571 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

SN65DP159: Native AUX on SN65DP159

Stephen Crocker
Prodigy 20 points
Part Number: SN65DP159

I am attempting to test a board that uses the SN65DP159 with an NVIDIA RT 1030 GPU.  It is able to receive valid video during POST, so I am confident that the card is outputting a compatible format.  However, when the driver loads (I am currently using Ubuntu Linux with the Nouveau drivers), I see from turning up the driver debug level that the AUX channel performs a command 9 (native read) which then fails.  I can dump the EDID using I2C over AUX using the i2cdump command, which causes commands 4 and 1 to show up in the kernel log (I2C write and read).

Should the SN65DP159 be responding to native requests over the AUX bus does the AUX channel function only for I2C over AUX?

  • 0
    TI__Genius 13310 points

    Hello, Stephen,

    The source can either use the DDC pins on the DP159 or the AUX pins to command the I2C registers of the device.

    How are your device's DDC pins configured?

    Thanks,


    Zach

  • 0 in reply to Zach Dryer
    Prodigy 20 points

    We have the option of wiring the AUX lines (pins 15 and 17 on the DisplayPort connector) either to the AUX pins (pins 45 and 44) via 0.1uf capacitors or to the DDC pins (pins 46 and 47).  I have tried boards with both configurations; in both cases, I can see a differential signal on the AUX lines.  When I use pins 44 and 45, I can use the Linux i2cdump tool to dump the EDID from the EEPROM behind the DP159.  When I use pins 46 and 47, I get no response.  I have tried this test both on a GT1030 PCIe card hosted on a Ubuntu desktop PC and on a Jetson Xavier development kit. 

    The GT1030 displays a BIOS screen during POST so I am confident that the card supports DP++.  The Xavier claims to support dual standards on the DisplayPort output but I have not yet seen any display from it.

    As I say, the Nouveau driver running on the GPU that I am using to test the board seems to be sending native AUX commands to the DP159 but not receiving a response.  I am not clear from the datasheet on how the DP159 should respond to such requests.

  • 0 in reply to Stephen Crocker
    TI__Genius 13310 points

    Hello, Stephen,

    AUX signals from the source need to go to the DP159 pins 44 and 45 and not to pins 46 and 47. In the datasheet, the 9.2 Functional Block Diagram shows the I2C over AUX to DDC bridge which takes in (from pins 44 and 45) AUX electrically but output is I2C electrically. 

    SDA_SNK and SCL_SNK pins are I2C electrical.

    Thanks,

    Zach

  • 0 in reply to Zach Dryer
    Prodigy 20 points

    Yes, I had worked this out.  My question regards how the DP159 is supposed to respond to native requests.

  • 0 in reply to Stephen Crocker
    TI__Genius 13310 points

    Hello, Stephen,

    Figure 31 shows the connection for both DDC and AUX GPU connections with the SNx5DP159RGZ. Either the DDC or AUX option should produce read/write commands to the DP159 registers. However, the DDC lines should be not be driven simultaneously as the AUX. 

    Thanks,

    Zach

  • 0 in reply to Zach Dryer
    Prodigy 20 points

    The Nouveau driver appears to be timing out when sending native AUX transactions but not when sending I2C over AUX transactions.  Is there something that I need to do to enable native AUX transactions?

  • 0 in reply to Stephen Crocker
    TI__Genius 13310 points

    Hello,

    DP159 doesn't support Native AUX. DP159 only supports I2C-Over-Aux.

    Thanks,

    Zach