• Resolved

Issue on TPS65835----------3D glass solution

 forum members:

I'm now working on a 3D glass solution and demanded that only one KEY is allowed to achive the POWER ON/OFF fuction and  

mode-switch.so we chosed the following method.



The problem is that I can't  turn of f the system (the glass is always working if it's in  charging staus.

so ...anyone can  help answer in what way the MSP430 can know the VIN has connect an AC adapter (in Battery charging staus) if I don't want to cost an I/O to detect it?

 then my SW can disable the BOOST circuit to solve this issue.

Thanks in advance!

  • You need to setup an interrupt with the MSP430 on the GPIO connected to the Schottky (P2.1 in the picture above) then drive P3.0 (SLEEP) low when ready to power down. 

  • In reply to Daniel Acevedo:

    Hi Daniel,

    Thanks for your reply.However,the approach you provided  works only when the VIN pin is in NC【not connect】state.if an AC adpator or USB connects to VIN pin,and in charging process,this mean fails. That's why I come here for help.

     Besides,I was puzzeld why place a  Schottky on the GPIO. Any clues will be appreciated! 


  • In reply to SeaFesse:

    Hi Benny,

    The diode is to prevent the MSP430 from being back-powered and turned on by this external connection to the SWITCH pin. The SWITCH pin has an internal pull-up if configured in the push-button mode and this diode connection will block this pull-up from powering the MSP430 on when the device should be off. A diode that has a lower forward voltage [VF < 1/3 * VLDO] is ideal to get the MSP430 GPIO to see the VIL input low logic level when the switch is pressed.

    To do what you want you just need to monitor VIN with the ADC or the comparator of the MSP430. It would require a resistor divider from VIN into one of the GPIOs that you choose. Then you can think of a table of operating conditions and how the key should react and the system should behave depending on key press, VIN, boost status, H-bridge switching, charging status, and so on.

    Let us know what other questions you have about this.


    Chris, TI PMU Systems Engineering

  • In reply to Christopher Johnson (TI-PWR):

    Hi Christopher,

    Thanks very much for your answer. It really took me too much time on this SWITCH pin and how to turn the device into sleep mode or  to wake it up...etc.

     I still have doubts on the effect of the Schottky Diode you said. If I take away the diode and configure the GPIO into high-impedance state.Besides, GPIOs

    of MSP430 are initlizated as  input state【high-impedance】when the device is powered on ....so  the diode  seems like a little redundant. 

    I 'm also confused when U say"The SWITCH pin has an internal pull-up if configured in the push-button mode ".  As my experiments say,TPS65835 will not be shut

    down if P3.0 is pulled up , regards of the state of SWITCH  pin. The SWITCH pin only works when P3.0 is driverd to low.【Usually,I  enable the pull-down resistor】

    so the pull up resistor connect to SWITCH pin  you mentioned above is  a part of PMIC, and the pull up resistor connect to P3.0 is part of MSP430,am I right?

    I really think a schematic of SWITCH pin and SLEEP【p3.0】 pin will answer all my  issues.

    ps: TPS65835  has not much  competitive power ,we face a great pressure from Taiwan .so any unnecessary components ,a diode、crystal osscillator ..ect.will be seriously treated .  ^_^!


  • In reply to SeaFesse:

    Hi Benny,

    It is not meant to be difficult to use. If you would like some more assistance or have other questions please let us know.

    The SWITCH pin has an internal pull-up resistor that is different from the pull-up / pull-down resistors of the MSP430. I have attached a more detailed drawing of this part of the device. In the drawing the relevant parts of the TPS65835 are shown (PMIC = power functions of the TPS65835).

    The diode is only necessary if you do not want push button press on and off timings set to 16 ms (tON = tOFF = 16 ms). As I said before, the diode is used to block current flow so the device will not power on the MSP430 when you do not want it to be on. Some customers want to do some custom functions like you are and this configuration leaves it flexible and easy to use the part for different 3D glasses platforms you may be working on.

    A crystal is not necessary to operate the device. It is something you can add and connect to make the timing very accurate. Depending on your application, the DCO of the MSP430 may be just ok. Any competitor's part with an MCU would likely need a crystal in order to get the same timing accuracy.


    Christopher Johnson, TI PMU Systems Engineering


  • In reply to Christopher Johnson (TI-PWR):

    Hi Christopher,

     The drawing explains a lot. Thanks.

    My  system【CC2500+TPS65835】  costs about 450uA on average in active state which benifit from the 32Khz crystal and SW algorithm.so,the crystal matters .

    Thank you once more. ^_^

  • In reply to SeaFesse:


    Hi Christopher,

    is there any more details diagram for TPS65835  ?

    if there is no the Diode, I  can turn off  the VLDO  only by press down the SWITCH over than the  Toff   time.

    but   the VLDO can not be turn off even pressed time  over than  Toff  ,unless config   to sleep with software. 

  • In reply to Arthur Han:


    To use the external SWITCH pin in push button or slider configuration you need to keep the SLEEP signal low. If you want to trigger an automatic shutdown using SLEEP please keep SLEEP low and then toggle it low-high-low. The PMU will shutdown on the high-low edge of the SLEEP signal.

    It sounds like you are setting SLEEP to high and trying to shut off the device by pressing the push button. In that configuration I would expect exactly what you are describing. When SLEEP is set to high, the outside button does not have any effect on the PMU.


    Christopher Johnson

    TI, PMU Systems Engineering