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: inserts HDMI display under AC S5 state, there will be leakage 386mV to +3.3VSUS;

Part Number: TPS65982

Hi ,
This USBC port inserts the previously verified dock (including HDMI/USBC/USB/ CRT D-sub interface), then inserts HDMI display to  dock under AC S5 state, there will be leakage 386mV to +3.3VSUS; But I install USBC adapter directly under AC S5 state, only leakage 14mV to +3.3VSUS. Based on below data, TPS65982DC seem do not tristate bus in AC S5 state then external bus voltage leakage 386mV to +3.3VSUS of system. What is your comment ? Thanks!
Dock measure voltage
EL1.1 = 0.497V
EL1.4= 0.497V
EL2.1 = 0.18V
EL2.4= 0.19V
EL3.1 = 0.468V
EL3. 4= 0.483V
EL4.1 = 0.023V
EL4. 4= 0.022V
EL5.1 = 0.008V
EL5.4= 0.002V
EL6.1 = 0.016V
EL6.4= 0.01V
U7. 4(7) = 0V
U7.5(6)= 0.018V

CC_TPS65982DC.pdfUSB TYPE-C_1_PT.pdf
USBC adapter measure voltage
EL1.1 = 0 V
EL1.4= 0V
EL2.1 = 0V
EL2.4= 0V
EL3.1 = 0V
EL3.4= 0V
EL4.1 = 0V
EL4.4= 0V
EL5.1 = 0V
EL5.4= 0V
EL6.1 = 0V
EL6.4= 0V
U7. 4(7) = 0V
U7.5(6)= 0V

  • Hi Steve,

    Not sure what the issue is here. Is this causing other issues in your system?
    Are you refering to VIN_3V3 when you say +3.3VSUS?

    Thank you,
    Eric
  • there will be leakage 386mV to +3.3VSUS: install USBC adapter to dock, then dock to MB for AC supply
    +3.3VSUS does not connected to TPS65982DC but connected to another IC such as CPU in system, then TPS65982DC bus connected to another IC bus, TPS65982DC seem do not tristate bus in AC S5 state , so leakage 386mV to +3.3VSUS under AC S5 state

     When I install USBC adapter to dock, then dock to another NB for AC supply to NB with different PD IC soultion , there is no leakage voltage to +3.3VSUS under AC S5 state. I think this issue should have soultion. 

  • Hi Eric,
    Our platform is NB. I found the pin B2(GPIO_0) of TPS65982DC is 3.3V when I install USBC adapter to dock and dock to MB for AC supply, then GPIO_0 voltage go through D14 to signal USB_OC0# ( up to 0.386 mV) and USB_OC0# pull high to +3.3V_SUS in system design, so there will be leakage 386mV to +3.3VSUS.
    Could you help command GPIO_0 to 0V under AC S5 state such as by I2C. Thanks!
    early experience, GPIO_4 (DP_HPD) had leakage voltage during sleep mode, TI provide proposal to exit DP mode (by sending AMEx command ) to fix this issue . We maybe can follow this debug logic .
  • Hi Eric,
    Sorry! The root cause is not the pin B2(GPIO_0) . I keep debuging. Thanks!
  • Hi Eric,
    Follow MUX suggest, TI Antony generate a test version FW to disable MUX DP mode and USB mode during AC S5 state for us try firstly, then try to disable the 2 pins L2&K2(DEBUG) of TPS65982 that connect to CPU and pull high to +3.3V_SUS. Thanks!
  • Hi Steve,

    Is further support needed for this request? It seems you have found the source of the leakage in your system.

    Thank you,
    Eric
  •  Sorry! My text confuse you. I do need further support for this request now. There is still leakage 386mV to +3.3VSUS after we try many FW proposal include disabled all GPIO of TPS65982.

    But one time, I make mistake to use off-line FW( flashimage file) replace on-line FW( lowregion file)  to program TPS65982 on Windows for test FW update. The result should fail, but good symptom as below steps:

    1. +3.3_SUS leakage= 386mV this moment

    2. After I drop dock and install USBC adapter to MB directly, +3.3_SUS leakage= 14mV, but

    3. I re-install USBC adapter to dock, then dock to MB again, +3.3_SUS leakage= 14mV, that mean fix this issue

    4. FW version show A.CE, Antony said never release  FW  A version to us.

    5.     VBUS is 5V, not 20V now.

    The only side efffect is when AC only , system just power on to show battery low icon then power off, can not boot into Win10. I need install battery module and boot into Win10, then reprogram on-line FW many times to fix this side efffect .

    It is one clue to fix this leakage issue by PD FW ? Thanks!

     TPS65982

  • Hi Steve,

    I have not been able to reproduce this phenomenon on a TPS65982EVM. Do you have an EVM available for testing?

    Thank you,
    Eric
  • Hi Eric,

    Thank you ! Maybe you do not have the dock, so cannot been able to reproduce this phenomenon on a TPS65982EVM. 

    1. Could you help check who release version A.CE FW  ?

    2. How PD can define dock + USBC adapter is same as USBC adapter only? Because AC S5 mode does not need HDMI/USBC/USB/ CRT D-sub function, like under my wrong operation , one workaround FW force VBUS back to 20V from 5V, then our MB can boot normally and PD also back normal FW .


     BR
     Steve

  • Hi Steve,

    I'm not sure I understand your question. What is A.CE FW? You should be using the TPS65982 FW from TI.com

    Thank you,
    Eric
  • Hi Eric,

    +3.3V_SUS does not have leakage after we remove EL6 and EL5 (break these pins K6&L6 and K7&L7 traces to TypeC connector). Do you know what register we can set by I2C to tristate these pins K6&L6 and K7&L7 under S5 ? I think C_USB_TP/N and C_USB_BP/N do not need work under S5.

    By the way. production line need off-line program FW to  TPS65982  ROM now that waste much resource(money). TI has method such as SW tool to on-line auto. program ROM ?

    Thank you,

    Steve

  • Hi Steve,

    You can disable the USB_RP and USB_EP fields during this scenario. This should disable the USB2 mux.

    Thank you,
    Eric
  • USB2 mux do not need work under S5, that make sense ?
  • Thanks! But I heard one side effect after disable the USB_RP and USB_EP fields :the CC connection at the type-c port will be reconnected that mean 20V adapter connected, it would drop to 0V while you write 0x28 register and back to 20V very soon as a re-connection result.
  • Hi Steve,

    Yes, when register 0x28 contents are changed, a disconnect is forced. I would suggest changing the USB EP and RP settings before entering sleep mode.

    Thank you,
    Eric
  • Hi Eric,
    Thanks a lot ! changing the USB EP and RP settings before entering sleep mode maybe has one risk: When AC only ( no battery), NB maybe shutdown before entering sleep mode cause no AC 20V.
    it would drop to 0V: I think put this side effect into dock+USBC adapter scenario only but USBC adapter scenario. PD can separate “dock+USBC adapter” and “USBC adapter” two scenario ? or need EC tell PD by I2C bus ?
    Steve
  • Hi Steve,

    This would need to be controlled by the EC. However, in the case where no battery connected (AC only). There is no need to go into Sleep mode. And also no need to change the USB EP and RP fields. Since the PD adapter would always be connected in that case. Whenever the battery is present and sleep mode is enabled, then the USB EP and RP fields can be updated.

    Thanks,
    Eric