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.

TPS25750: BOOT status question[Urgent]

Part Number: TPS25750

Hi team,

My customer use TPS25750 on their project. In their production line test pattern, they will read the BOOT_STATUS  register to check the if PatchConfigSource is 5h, and I2cEepromPresent is 1h confirm the TPS25750 work with EEPROM well. The normal good board would show value 5h for PatchConfigSource  and 1h for I2cEepromPresent; however in their recent second source verification build of other circuit on the same board, they find the register show PatchConfigSource 7h, I2cEepromPresent 0h.  Both two boards could output 9V voltage which is exactly what they configured the part. Full read back register as below: 

OK PCBA      10100010111100100000000000011100 (A2F2001C) byte4-1

NG PCBA      11100010111100100000000000000100 (E2F20004) byte4-1

 Get following questions:

  1. PatchConfigSource 7h show reserved mean, what does this exactly mean?
  2. Does it possible to get 9V output while TPS25750 report no EEPROM? What may cause this issue?
  3. Any suggestion for further test?

Could you help share comments on this?

Best regards

Will

  • Will,

    1.  The error that you are showing that there was an error loading the patch

    2.  The EEPROM boot flow is limited to a 5V contract that does not close the PPHV switch to allow power until after the EEPROM has been loaded.  There are 2 other modes of operation, but they are not likely to be useful.  The AlwaysEnableSink mode will allow the PPHV switch to be closed under a 5V contract and the part will be in patch mode to allow an I2C load of the patch.  The EnableHighVoltage mode is similar to the AlwaysEnableSink, but it will allow the highest voltage offered by the DFP.  This can be up to 20V.  Both of these additional modes preclude the use of the EEPROM.

    3.  HAs the customer verified that the EEPROM was written correctly with the FUllFlash Binary?

    How are they configuring the EEPROM for their testing?

  • Hi Chuck,

    Thanks for your feedback.

    Do you get any details of the error? Unable to load the patch or what's any other error?

    Does the part could report what cause the error?

    1.  The error that you are showing that there was an error loading the patch
  • Will,

    The error code that you are getting is I2cEepromPresent is false.  This means that the TPS25750 does not detect that the EEPROM is attached to the I2C bus.

    This will prevent boot.

    This is usually triggered when the EEPROM has not been programmed or has been programmed incorrectly.

  • Hi chuck,

    I am the EE of this project. Glad to have your support.

    In our product, TPS25750 is configured as power sink, AlwaysEnableSink mode with max power 9V/27W. FUllFlash Binary file is used to write EEPROM. 

    This error occurs when the power is switched from the battery to the adapter. A delay exists between battery disconnection and adapter access. The detailed waveforms of OK product and NG product are shown in the figure below.

    Comparing with two waveforms, falling time and rising time of 3V3 are different; conversion time of VBUS from 5V to 9V of NG product is shorter; 3V3 follows 9V to start in OK product while about 300ms delays exits in NG product.

    After swapping 3V3 power ciucits between OK product and NG product, error disappears in NG product and occurs in OK product. So, error follows 3V3 power circuit.

    Then, three other verifications were performed. First, change the boot time of 3V3 on the NG product so that 3V3 follows 9V like the OK product. The error still occurred. Second, change the TPS25750 to safemode. The error still occurred. Third, swap EEPROM between OK product and NG product. Error still occurred on NG product while OK product still passed the test.

    Could you please share comments on this? or any suggestion for further test? Thanks.

  • Henry,

    From a very high level, both of these two sets of waveforms are showing a successful boot and successful PD negotiation.  You iwll not see 9V if the part does not boot correctly.

    It would be very helpful if you could do another capture that shows the VBUS and the I2Cm_SCL and I2Cm_SDA bus. 

    This will allow me to see what is happening on the boot.

    Based on your schematic, you should also change the ADCIN1 and ADCIN2 resistors so that only PDR12 and PDR13 are populated and the other 3 are not.  This will ground ADCIN1 and ADCIN2 .  This will put the TPS25750 into SAFEMODE with an I2C address of 0x21 which is how the part is intended to be used with the EEPROM.

    I suspect that if you do this, then the boot issue will be resolved, but I will wait to hear back from you to close this thread.

    Regards,

    Chuck

  • Hi chuck,

    The following tests have been performed. Hope this information will help to analyze the isuue.

    1. Safemode dose not work and error still occurs.

    2. Short Pmos PDQ1 and error still occurs.

    3. Remove 0ohm PDR1 to leave TPS25750 always in dead battery condition, error disappears.

    4. Waveforms of power rails belong to TPS25750 were captured as below.

    Looking forward to your reply. Thanks.

  • Henry,

    Because you have grounded PP5V, you will never be able to be a DFP (source), it is a good system solution to always leave VIN_3V3 disconnected and keep the TPS25750 in dead battery mode.

    This is the intended functionality in this case.

    Regards,

    Chuck

  • Hi chuck,

    It seems that there are some solution to solve the issue.

    Based on the above information, could you please help to find the root cause of this issue? Thanks.

  • Henry,

    The system does not have a consistent Voltage setting.

    If you are a sink only, then there is no need to connect VIN_3V3 and we suggest that you float PP5V.

    Please look at this FAQ:  https://e2e.ti.com/support/interface-group/interface/f/interface-forum/988215/faq-tps25750-how-do-i-create-a-sink-only-usb-c-pd-port-to-replace-a-legacy-barrel-jack-connector

    This shows that VIN_3V3 is connected to a 10uF capacitor, but nothin else in a sink only application.  If you are doing anything to change VIN_3V3, then the system may not function correctly.

    Regards,

    Chuck