• Resolved

TPS65311-Q1: TPS65311-Q1 Not working external FET

Part Number: TPS65311-Q1

     It is configured as below, and when power is applied at the first, the gate voltage is not lowered and the FET does not operate.

    That is VIN and GPFET are rising as the same voltage.

    In such a case, what causes it?

  • Hello,

    Could you please let me know more about this project and customer?

    Are you checking this on EVM or you have developed your own board? If you apply 12V input, is this working?

    Could you please share the schematic? have you Is VIO voltage levels are correct?

    "

    The external PMOS switch is disabled if:
    • The device detects VIN overvoltage
    • The device is in ERROR, LOCKED, POR, INIT, TESTSTART, TESTSTOP or LPM0 mode

    "

    Regards,

    Murthy

  • In reply to Krishnamurthy Hegde:

    Hello, Murthy

    Thanks for your fast reply.

    I've developing my own board for power module with TPS65311Q.

    When you apply 14V to Vin as recommended, the voltage of GPFET pin does not go down as shown below.

    That is, the GPFET maintains an OFF state continuously. [ blue line : GPFET, Yellow line : VIN , scale : 5V/step ]

    The voltage of the GPFET is not lowered so that the FET can be turned on near the power on reset voltage.

    It is known that the GPFET pin is driven by a constant current source of 20 uA. Therefore, the pin does not have any load other than the external FET.

    What parts should be investigated for normal operation?

    I would appreciate your help. And attach some of our drawings.

     

    External FET : DMG4413-LSS-13

    Regards,

    Jeho.

  • In reply to Jeho Shin90:

    Hi Jeho.

    Thanks for the details. Could you please let me know the customer details?

    You can email me the details to my email ID (kmurthy.hegde@ti.com), i can send the design checklist for your reference.

    I had a quick look at the schematics (some part of the schematics is missing) and you may have issues with monitoring. Figure 18. Operating Mode Transitions in the datasheet explains the state machine and GPFET turns on during VTCHECK state if device passes previous states.

    - In case the LED or high-side switch control is not used in the application, HSSENSE must
    be connected to VINPROT.

    -In case any of the supply rails for BUCK2, BUCK3, LDO or BOOST are not used in the application, the
    respective VMON2 and VMON3 or VSENSE4 and VSENSE5 pin of the unused supply must be connected to VMON1. Alternatively, the VSENSE4 pin can also be connected directly to ground in case the LDO is not used.

    -Also for proper device operation, Watchdog signal needs to be applied correctly.

    -VIO signal needs to be correct.

    So, i would request you to check your schematics again and see whether device is configured correctly.

    Regards,

    Murthy

  • In reply to Krishnamurthy Hegde:

    Hi  Murthy

    Thanks to your help, the external FET problem has been solved.

    There are, however, other problems continue to come out.

    1)  As shown in the figure below, the desired voltage stabilizes after about 30ms.

    Not only 5V, but also 3.3V and 1.2V stabilize after about 30ms.

    Commonly, during the initial 30ms, each output voltage is 1V higher than the desired output voltage, and then returns to the normal voltage.

    The EV board does not show this phenomenon. How do I resolve it in my board?

    Regards,

    Jeho.

      

  • In reply to Jeho Shin90:

    Hi Jeho,

    Could you please let me know more details about the customer and project? You can email me if you are not confident to share this information on E2E.

    Unless I have the detailed schematics, I cannot really tell why this happening. So far, no customer has complained about this behavior and i have not seen this behavior myself. Please try to bring your schematic close to EVM schematics and see whether this issue gets solved. Also it is very difficult to do such remote debug.

    so suggestions to try:
    * I noticed that you have only 4.7uF output capacitors and if this is really true, it may be too small. Please keep them at least around 20uF.
    * send some zoomed in pictures of start up behavior showing, VIN, VINPROT, VBUCK1, VBUCK2
    * You can try to apply some small load on the regulator (some 20-50ohm or so on BUCK1 ) and try
    * You can try to decouple this regulator from other loads by removing FB8 component on 5V.. just to see whether something from load side is causing this issue.

    Regards,

    Murthy
  • In reply to Krishnamurthy Hegde:

    HI, Murthy

    Thank you for your continued support.

    As I modified it according to your recommendation, the required voltages are coming out normally.

    However, the problem that the initial voltage of BUCK1, BUCK2, BUCK3 is slightly higher is not solved. I attach the investigation (waveform) related to the problem below.


    We added 94uF to the outputs of each BUCK1, BUCK2, and BUCK3.

    The waveform without FB8 beed is the same. EVB : Evaluation board.

    < VINPORT-GPFET >

    < VINPORT-BUCK1 >

    <VINPORT-BUCK2>

    < EVB-VIN-GPFET >

    < EVB VIN-BUCK1 >

    Regards,

    Jeho

     

  • In reply to Jeho Shin90:

    Hi Jeho,

    Could you please provide the customer name and project details? I asked few times, but i didn't get the answer yet.

    Regards,

    Murthy

  • In reply to Krishnamurthy Hegde:

    Hi Murthy

    I always appreciate your answers.

    I am a robot engineer, and we are developing our safety related controllers for certification.

    I am very worried that my Project is too delayed because of this PMIC.

    I know that designing the power section is not easy.

    However, recent buck type regulators can get excellent voltage by adding fewer parts to the external.

    Anyway, loads of the PMIC are only Safety CPU and Ethernet Controller In my system..

    You can see that the output of each BUCK is slightly higher until VINPORT rises completely.

    But the EV board is different.

    I hope to get God's hand from you as soon as possible.

    Regards,

     Jeho

  • In reply to Jeho Shin90:

    Hi Jeho,


    Thanks for the details. It is difficult to debug remotely. However i can ask you to check the below. But not sure whether this works for you.

    1. change BUCK2/3 output capacitors to  22uF. change Buck-1 output capacitor to 47uF and re check the compensation calculations for BUCK1.

    2. Send me the exact schematics (complete PMIC page) in PDF format with all the changes you have made so far. Schematic you have sent is not complete. you can send it to my email ID (kmurthy.hegde@ti.com) and we keep it confidential.

    3. fix the bottom sense resistor= 1.6k ohm and recalculate the top side sense resistors for BUCK2/3.

    4. what have you done with COMP2/3?

    5.short Vin and VINPROT (for testing purposes only). Also check with 12Vin

    6. check PH/SW node of BUCK1/2 (zoom in to see few clock cycles clears during overshoot) during start up and send some pictures to me showing VINPROT, VBUCK1, VPH1, VPH2.

    7. Add 500mA load (fixed resistor is ok) on each regulators and check the behavior.

    8. check sense pins to see whether there is some noise coupling happening. If you have current probe, check the BUCK1 inrush current during start up.

    9. if you have another board, modify the board according to the design checklist and try on another board. TPS65310A design checklist is also applicable to this device.

    www.ti.com/.../slva609.pdf

    Regards,

    Murthy

  • In reply to Krishnamurthy Hegde:

    Hi Jeho,

    Were you able to solve this overshoot issue during the start up?

    What was the root cause of this problem?

    Regards,

    Murthy