• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 32 subscribers
  • Views 1626 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.

TPS65986: GPIO Events not working properly

JoFu10
Intellectual 2385 points

Part Number: TPS65986

Hi,

My customer is using GPIO events to wake up their SoC, but some of the events we tested do not function properly.

For example, 

Using VBUS detect, the GPIO pin goes high for around 60ms every time we either reset or plug in the barrel jack into the TPS65986EVM (no type C cable inserted).

\

We have set the GPIO to default 0x0 and no pullup/pulldowns.

On another test, we also tried several different GPIOs to wake the SoC when in dead battery mode (no barrel jack attached) and a few did not function as specified in the GPIO events app note. We want to be able to insert a type C cable into the EVM (no barrel jack) and have the device power up from VBUS and pass a GPIO enable to the system to wake it.

We tried SINK_HVEXT, PP_HV ACTIVE, and SUPPLY_HV EVENT, but none of the GPIOs asserted high when a type C cable was plugged in. All GPIOs were powered by LDO3V3, and we installed the 0 ohm resistor on the board. We probed the VBUS pin with the type C plugged in and VBUS did have 20V as well.

We eventually settled on PWR_SINK_SOURCE as our GPIO event, as that is the only one that worked. Do you have any explanation why the other GPIOs do not work properly? We seem to fit all the behavior by the GPIO event app note.

Thanks,

Josh

  • 0
    TI__Mastermind 47060 points

    Hi Josh,

    Have you tried testing this on another 986EVM. VBUS detect would only assert high if there is a voltage present on VBUS. It's surprising to hear that when barrel jack is plugged in and nothing else is attached, VBUS gets pulled high as there should not be any voltage contract made at this point. Also I am not clear what the green probe in the waveform is. Is the yellow probe GPIO for VBUS detect and green probe for VBUS voltage on EVM board?

    What are you connecting the 986EVM to in order to perform dead battery? If the 986EVM does not have a barrel jack connection, and you insert a PD adapter to operate in dead battery mode, the 986EVM would operate in dead battery and VBUS will be pulled to the highest advertised voltage dictated by the PD adapter. This can be simulated by having two TPS65986EVM,one 986EVM confuged as Type-C PD charger and the other as the sink device.

    Regards,

    Jimmy Hua   

  • 0 in reply to Jimmy Hua
    TI__Intellectual 2385 points
    Hi Jimmy,

    Yes, I've tested this on both my EVM, their dev board, and their EVM and we all see the same pulse. That is surprising to us as well, as the power path FETs should be open to not allow the barrel jack power to pass through to VBUS.

    Disregard the green probe, that is not connected to anything. VBUS_DET is the yellow probe. VBUS is assumed 0V as no type C cable is plugged in.



    When we plugged in several type C chargers (chromebook charger, Anker hub, some others) into the EVM with no barrel jack to simulate a dead battery condition. The Anker hub and Targus charger trigger VBUS_DET fine, but the Chromebook 60W charger and j5create hub do not (they trigger VBUS_DET for varying ms from 40ms to 400ms, but VBUS_DET eventually goes low), so we are not sure what exactly is going on behind the scenes to cause different chargers to work differently in dead battery mode.
    But for all chargers, we plug in VBUS, and VBUS does say 20V, but VBUS_DET, PP_HV ACTIVE, nor SUPPLY_HV EVENT do not assert.

    I could get on a call to explain if that is easier.

    Thanks,
    Josh
  • 0 in reply to JoFu10
    TI__Mastermind 47060 points

    Hi Josh,
     
    If possible can you provide PD traces for different chargers so I can take a look at the messages being between DFP and UFP for the working case and the non working case? The team and I will look into this further situation and get back to you.

    Regards,
    Jimmy Hua

  • 0 in reply to Jimmy Hua
    TI__Intellectual 2385 points
    Hi Jimmy,

    I'll work on getting those for you.

    Were you able to reproduce the VBUS_DET pulse upon reset/barrel jack insertion into the EVM? I can send you the .bin file if you need.

    Thanks,
    Josh
  • 0 in reply to JoFu10
    TI__Mastermind 47060 points
    Hi Josh,

    Yes can you provide me with the bin file and project file so that I can see what your FW settings are and try to reproduce the issue.

    Regards,
    Jimmy Hua
  • 0 in reply to Jimmy Hua
    TI__Mastermind 47060 points

    Hi Josh,

    I am not able to reproduce the issue that you are seeing. On my end, I mapped the GPIO#0 event to occur on a VBUS Detect Event and was able to see it fault and stay HIGH after a source(PD adapter) was connected to it.

    Make sure that you are using the latest Application Customization tool version( 3.08) since this was recently released within the past week and can be found on the following link www.ti.com/.../tps6598x-config.

    If the issue still occurs, I'd recommend getting a new FTDI board to reflash and test the setup again.

    Regards,
    Jimmy Hua

  • 0 in reply to Jimmy Hua
    TI__Mastermind 47060 points
    Hi Josh,

    Since we have taken this thread offline, I would like to close this thread now.

    Regards,
    Jimmy Hua