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: no power after PC restart

Part Number: TPS65982

Hello,

we have made a board that sinks 5V 3A and uses DP alternative mode with 4 lanes.

When we restart the PC, board shuts down as expected, but when the PC is starting back up again , the board doesn't start.

PC doesn't recognize the DP connection and the device stays powered down.

Can this be fixed with some settings in FW?

Below I added PD analyzer file

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/trace_5F00_on_5F00_restart.7z

Best regards,

Eriks

  • Also added TPS65982 .pjt file

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/TPS65982_5F00_firmware.7z

  • Hi Eriks,

    The pjt file is not attached here, when you click the link it is going to the product folder. Can you re-attach?

    The PD log shows no response from sink side.

    Can you check if VBUS is present and LDO-3V3 comes up when PC is up? 

    If you have a flash on board, do you see SPI transactions when power up.

    Maybe attach your schematic as well.

    Regards,

    Peter

  • Okey , we have narrowed down the problem.

    It looks like , that , when you connect the device to a laptop, when restarting , it power cycles the device .

    But when the device is connected to a PC it doesn't power cycle the device , that's why it's not initializing correctly .

    Here is the project file : https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/5850.TPS65982_5F00_firmware.7z

    Here are the traces for a laptop and PC restarting : https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/restart_5F00_trace.7z

    The question is, how can I get the device to power cycle when PC is restarting ? And how to load the settings if its dead battery mode ?

    Regards,

    Eriks

  • Hi Eriks,

    I see in PC PD trace, it was working follows by a non-responsive sink. If you recycle power I would expect a disconnect to occur before starting the next PD messages.

    Do you see VBUS and CC lines went down and comes back up when PC power is recycled?

    Dead battery mode is when no VIN_3V3, but power is provided from external end partner through VBUS.

    Do you have power in the device or are you expecting to receive power from outside?

    What's your BUSPOWERZ and SPI_MISO connections like?

    Regards,

    Peter

  • Hello,

    I checked the VBUS and CC lines , they do go down , VBUS goes down for about 400ms and CC one line goes down for about 400ms also and the other stays low all the time , probably because It's the polarity.

    Yes, we use the dead battery mode , with no VIN_3V3 voltage, so the device is powered from VBUS. BUSPOWERZ is connected to LDO_1V8D and SPI_MISO pin is pulled high to LDO_3V3.

    The fact is , that the board works , but when the PC is restarted , it doesn't initiate for some reason. But then connected to a laptop and It's restarted it works as expected.

    Regards,

    Eriks

  • Hi Eriks,

    Ya we need to find out why it works on laptop but not PC, so we are checking the basic signals to ensure it's working per the flow.

    So you do see VBUS comes back up after 400ms and CC starts toggling, do you see PPHV and LDO3V3 voltages?

    It's also good to check the flash activities to ensure it's loading the configuration once VBUS is provided.

    Can you capture VBUS, CC and SPI lines?

    Regards,

    Peter

  • Hello Peter,

    I captured VBUS, PP_HV, CC and SPI lines when PC is restarting, as you requested, as it looks like , TPS65982 does not reload the configs from flash memory on reset.

    Any ideas why?

    Captured rails : https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/TPS65982_5F00_rails.7z

  • Hi Eriks,

    Did you see LDO_3V3? Why SPI_MISO is low?

    The Apps. Config will start after it enables the PPHV sink path.

    How's PC differs from laptop when booting up? Did you see SPI transactions in laptop?

  • Hello,

    I checked LDO_3V3 and It is fine. I don't understand why MISO is low also at the end of reading flash memory.

    PPHV sink path enables as it should and supplies the rest of the board with power.

    I took the SPI transactions for both PC and laptop , and they are the same so I'm adding one for you to check.

    SPI.zip

    Regards,

    Eriks

  • Hi Eriks,

    Can you read the Mode at address (0x03) when it happens?

    I checked with my EVM and connect to another EVM that is programmed with 3 PDOs (3A5V, 3A9V and 3A15V) using the pjt you sent and am able to see the contract. I have attached here the scope captured along with the traces.

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/TID_2D00_EVM.7z

  • Hello ,

    I checked the 0x03 address and it returns "APP" , so that means its booting correctly and reading the flash memory without a problem as I understand.

    Your test with 2 EVM are faulty , because the EVM acts different from a PC and Laptop so you can't test the restarting scenario, also are the one EVM using my firmware version in dead battery mode?

    As I see in your captured scope , with my sent TPS65982 firmware version your SPI_MISO goes up at the end. Could this mean there is a hardware problem that doesn't let your SPI_MISO go up?

    All the pullup resistors for the flash memory for TPS65982 are 100k except for the SPI_MISO pin that is 10k , are they too weak ?

    Regards,

    Eriks

  • Hi Eriks,

    Ya the APP mode shows it's able to boot up properly, so the MISO level is not an issue here.

    I'm using the EVMs with your pjt file in dead battery mode just to see if I can replicate the issue since the desktop has 3 PDOs and laptop has only 1.

    I have also tried to set the end partner port to be Source only instead of DRP as laptop most likely is DRP and desktop maybe just Source.

    In all the tests, I can see them working as expected.

    Have you tested this case maybe on an EVM or have you tested your system with another desktop? Is it only on this particular desktop it behaves this way?

    Regards,

    Peter

  • Hello,

    It has been tested on multiple desktop PCs , it only works on laptop.

    I checked when the TPS65982 flash memory is reloaded , it looks like , when PC is restarting it power cycles the device and it reloads the settings , but when the laptop restarts , it doesn't reload the settings from flash memory, probably that's why the device works after laptop restart , the settings hasn't changed for it.

    Any ideas how to fix it , or what to check ?

    Regards,

    Eriks

  • Hi Eriks,

    When powering up from dead battery, it will load the configurations from the flash. This applies to either laptop or PC, how do you know that the settings were not loaded. Are you saying the settings are different between laptop and PC from PD configuration point of view?

    Regards,

    Peter

  • Hello,

    that's not exactly what I mean , the fact is that when I power up the device from laptop, it loads the settings from flash memory .

    I can see this with logic analyzer on SPI lines, but when i restart the laptop , nothing happens on the SPI lines.

    But in desktop PC case, I see that the TPS65982 loads settings from flash when i power up the device and when the PC restarts.

    It looks line the laptop only works after restart because It doesn't reload the configs from flash memory.

    Could it be something to do with that?

    Regards,

    Eriks

  • Hello ,

    After some tests I found that the values of pull-up resistors were too high for the flash memory.

    We swapped our 100k and 10k pull-ups for 2.67k (only ones we had on hand) and looking at the SPI lines the MISO pin stayed high after loading configs from flash memory. Then when the desktop PC was rebooted , TPS65982 started as it should.

    As I see it , TPS65982 didn't like that the MISO pin was low at restart.

    Anyways thanks for the help.

    Regards,

    Eriks