• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 4 answers
  • Subscribers 63 subscribers
  • Views 3519 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS40056: Startup waveform of TPS40056

Masami Michino
Genius 5070 points
Part Number: TPS40056
Other Parts Discussed in Thread: TPS40055

Dear, All

Customers are conducting evaluations using TPS40056.
Their circuits are made variable output voltage by externally inputting voltage to EA_REF.


When observing the waveform at startup, they observed the following problem when VOUT is 1.5V or less.
- HDRV pulse width is about 50ns
- There is no dead time when LDRV rise.
Waveforms of interest are not observed in normal waveforms. It occurs only at startup.


Although the charge current to the output capacitor was suspected, the same waveform is observed even if the capacitance of the output capacitor is made extremely small.
Please tell me the cause and measures.

Thanks, Masami M.

  • 0
    TI__Expert 3870 points
    Hi Masami,

    Please share with us the input voltage and load current conditions.

    Best Regards,
    -JP
  • 0 in reply to Juan Paulo Fung
    Genius 5070 points
    Hi, Juan Paulo-san,

    The input voltage of the power supply is 24V.
    A load is not connected and it can occur even with only a capacitor.
    The operation at the time of occurrence is as follows.
    First, release SS pin from 0 V input.
    At the same time, EA_REF is gradually increased in units of 33 mV.
    The customer slows down the rise of EA_REF from the rise of the SS voltage.
    A phenomenon occurred under this condition.

    Thanks, Masami M.
  • 0 in reply to Masami Michino
    Genius 5070 points


    Hi, Juan Paulo-san,

    I checked the operation with TPS40055 because there is no EVM on TPS40056.
    As a result, I confirmed that the same phenomenon will occur in TPS40055EVM - 001.
    The waveforms are attached below.


    Green - - TP9 (Vout)
    Blue - - TP 7 (SW)
    Light blue - TP4 (HDRV)
    Purple - - TP 5 (LDRV)

    The power supply uses the regulated power supply. I am measuring with no load, EVM is unmodified.
    Please investigate.

    Thanks, Masami M.

  • 0 in reply to Masami Michino
    TI__Expert 4365 points

    Thank you for providing the additional details on the application.

    Given the timing resistor RT value of 150kOhm the switching frequency is set to 300kHz.

    For 1.5Vout  with 24Vin the required duty factor of the switch is 1/24 (lossless case) =4.2%

    300 kHz period is 3.3 usec * 0.042 = 140 nsec On-time during normal operation.

    However, during soft start, even shorter on-times are required which are probably shorter than the controller section can produce.

    One avenue to get closer to proper operation is to lower the operating frequency by raising the timing resistor value.

    I'd try dropping to 100kHz as an experiment realizing that doing so will impact the proper inductor value and many of the other component values in the application. But a lower switching frequency may be necessary for the lowest Vout desired.

    With controllers, it is difficult to state guaranteed minimum on-time due to dependence on MOSFET selection.

  • 0 in reply to Alan Martin
    Genius 5070 points

    Hi, Alan-san,

    Thank you for your reply.
    Measures to shorten the on-time are to slow down the switching frequency, but I am concerned that similar waveforms will occur during soft-start.
    What customers are worried about is not to shorten on-time.
    The dead time can not be seen between the time when the High side FET turns off and the time when the Low side FET turns on.
    For this reason, they are worried about the risk of a through current between the High side FET and the Low side FET.
    Please give me advice.

    Thanks, Masami M.

  • 0 in reply to Masami Michino
    Genius 5070 points

    Hi, Alan-san,

    Thank you for your support.
    Please tell me the circumstances of your investigation on the case that there is a cycle with no dead time at soft start.

    Thanks, Masami M.

  • 0 in reply to Masami Michino
    TI__Expert 6860 points

    Michino-San,

    The duty cycle that needs to be supported in your application is very small which is closer to the Min on time specs of the device .Now during turn even lower on time is required and that may be the reason of what you are seeing in your waveform .
    Can you please change the Switching frequency to 100KHz and verify the Drives .

    ---Ambreesh

  • 0 in reply to Ambreesh Tripathi
    Genius 5070 points


    Hi, Ambreesh-san,

    Thank you for your reply.
    I use TPS40055EVM to change the switching frequency to 100KHz and measure the observation.
    Now, please give me some time as I can not take the time to measure.

    Thanks, Masami M.

  • 0 in reply to Masami Michino
    Genius 5070 points

    Hi, Ambreesh-san,

    Thank you for your support.
    When changing the switching frequency of TPS40055EVM to 100kHz under the condition that 12V was supplied to Vin, the phenomenon disappeared.
    However, the customer's Vin input voltage is 24V.
    Since the absolute maximum ratings of the input capacitor of TPS40055EVM is 16V, we observed it by supplying 15V to Vin.
    In that case, it was observed that a phenomenon occurred.

    Please give me advice.

    Thanks, Masami M.