• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 6 answers
  • Subscribers 32 subscribers
  • Views 8609 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TPS65982: Display Port AUX Channel

guy lemire
Intellectual 950 points
Part Number: TPS65982

Hello, I am using the TPS65982 in a Display port application. For the aux channel, I wanted to know if any external components were required to bias the differential pair. I have include the DP AUX channel for the

source. Specifically are the 100kohm pullup/pulldown included in the TPS 65982. If they are how are they invoked.

Regards,

Guy

  • 0
    TI__Mastermind 36751 points
    Hi Guy,

    The TPS65982 should have the aux channel resistors internal. However, I would recommend having a "do not populate" resistor option for both pull up and down on both aux lines.

    Thank you,
    Eric
  • 0 in reply to Eric Beljaars
    Intellectual 950 points

    Hi Eric,

    can you clarify how these pullups get enabled in the TI part. I am a bit worried that you recommend having an external pullup/pulldown on the AUX_P AND AUX_N lines. This indicates that there might be a problem with the TPS controller in how these resistors are configured?

    I assume that once we configure the TPS65982 in DP Alt mode the pullup/pulldowns on AUX_CH_P and AUX_CH_N are automatically set.

    AUX_CH_P should be a 100k pulldown to GND and AUX_CH_N should be a 100k pullup to 3V3. This resistor is RPU100 in the datasheet I believe.

    Q. Is there any register that I can check to configure these pullup/down resistors are getting set correctly, or to over-ride these values, so that I can check the pullup/down status?

    When we measure the voltage on the AUX_P/AUX_N (SBU1/2) we are seeing an incorrect voltage. It appears that the pullup/pulldown are not set correctly in the TPS controller.

    I attached the project file I am using.

    Regards,

    Guy

    DP_0919_2017g.pjt

  • 0 in reply to guy lemire
    TI__Mastermind 36751 points
    Hi Guy,

    I only suggested the do not populate resistors as we have them on the TPS65982EVM. However, we have never needed to place any of the resistors so you would be safe without them. The AUX_CH_P and AUX_CH_N pullup/downs are set automatically when you enter DisplayPort Mode on the TPS65982.
    We do not have a register indicating the status of the pullup/down resistor on these channels.
    Are you having issues entering DisplayPort mode?

    Thank you,
    Eric
  • 0 in reply to Eric Beljaars
    Intellectual 950 points

    Hi Eric,

    we do not have issue going into display port mode, but it appears the DC bias point on the AUX_P and AUX_N are incorrect.

    Can you clarify what the TPS65982 is doing between the AUX_P/AUX_N IOs and the SBU1/SBU2 IOs. My understanding is that it should simply apply a 100k pulldown resistor for AUX_P and a 100k pull-up resistor to 3.3V for AUX_N.

    When dealing with a cable swap is there also a mux in this circuit? Is there any other control or setting needed to correctly configure this pullup/pulldown?

    Regards,

    Guy

  • 0 in reply to guy lemire
    TI__Mastermind 36751 points
    Hi Guy,

    The TPS65982 also contains a mux on the Aux/SBU lines to handle the cable flip. The TPS65982 will set the pull-up/down internally. Please note that the bias level and resistance value will be difficult to measure as the connected device will have the opposite pull-up/down on the AUX/SBU lines.

    Thank you,
    Eric
  • 0 in reply to Eric Beljaars
    Intellectual 950 points
    Hi Eric,
    understood abut the SBU lines and all of the associated pullups/downs on this line. However it appears to me that the AUX_P line is being pulled up with a 100K pullup resistor. We are measuring 2.4V on the AUX_P pin, and the only way to get this voltage is to have the 100k internal resistor pulled up to 3.3V. S o what setting in the TPS will cause this to be pulled up to 3V3?

    Regards,
    Guy
  • 0 in reply to guy lemire
    TI__Mastermind 36751 points
    Hi Guy,

    How are you measuring this? Please keep in mind that the pull-up/downs will not be applied until the DisplayPort Mode is entered. If you measure the lines beforehand, you would probably not see it being pulled down. When you are in DisplayPort Mode, the Aux line that is pulled down by the TPS65982 will be pulled up by the device you are connected to. In this case I can see why you would be measuring 2.4V.

    Thank you,
    Eric
  • 0 in reply to Eric Beljaars
    Intellectual 950 points

    Hi Eric,

    we are measuring after the device is configured. If you look in the DP Alt  specification, AUX_P should be less than 400mV and AUX_N should be > 2.4 V when biased. We are measuring AUX_P to be 2.4V and AUX_N to be 400mV. So it appears the pullup/pulldowns are reversed in the TPS. This is why I am asking what configuration registers will effect this pullup/down.

    Regards,

    Guy

  • 0 in reply to guy lemire
    TI__Mastermind 36751 points
    Guy,

    The Aux pull-up/pull-downs will only be enabled once the device enters DisplayPort Mode through PD communication. Measuring this voltage before connecting a DisplayPort device to the type-c port will not yield the correct results. There is no register to adjust the pull-up/pull-down values, this is done automatically by the TPS65982.

    Thank you,
    Eric
  • 0 in reply to Eric Beljaars
    Intellectual 950 points
    Hi Eric,
    it is measured after PD communication has occurred and the main lanes are transmitting data.

    regards,
    Guy
  • 0 in reply to guy lemire
    TI__Mastermind 36751 points

    Hi Guy,

    Can you highlight to me where in the DP Spec the Aux lines need to be the voltage you had mentioned? when reading the DP Spec, I find that the  AUX voltage can vary between -0.35V to 3.95V. Am I missing something here?

    Thank you,

    Eric