• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 62 subscribers
  • Views 319 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.

Original question:

LM3478: Load transient

LM3478: Symmetry of Load Transient

user6055081
Prodigy 40 points

Part Number: LM3478

Tool/software:

I used LM3478 for load transient and found that the Vout (ac) waveform was asymmetric. I am curious about the reason and whether it can be improved?

Many competitors will intentionally increase the voltage scale of Vout (ac) when doing load transient, making the asymmetric waveform difficult to see.

  • +1
    TI__Mastermind 33510 points

    Hello,

    Thanks for using the e2e forum.
    The LM3478 is an asynchronous controller.
    This means if the output voltage is too high, the device cannot "take back" power from the output side, as the diode will block negative currents.
    The device will stop switch instead and waits until Vout drops by itself.

    You can see this in the load transient waveforms:
    IOUT rises --> VOUT drops --> device reacts with active regulation of the duty cycle
    IOUT decreases --> VOUT rises --> device react by stopping to switch. No active regulation until Vout drops by itself

    This behavior is normal and expected.
    You may improve the undershoot (IOUT rises), by adjusting the compensation to reduce the undershoot and the oscillation afterwards.

    Best regards,
    Niklas

  • +1 in reply to Niklas Schwarz
    Prodigy 40 points

    If I change to a synchronous rectifier, how do you think the path will flow?

  • +1 in reply to user6055081
    TI__Mastermind 33510 points

    Hello,

    If changed to a synchronous rectifier switching in FPWM mode (forced pulse width modulation), I would expect the overshoot and overshoot behavior is the same.
    Right now your load transient jumps from zero load to 2A. You will also see a different behavior in an asynchronous controller, if the lower value is more than zero. E.g. 200mA to 2A. Then the device will start switching again much earlier and the rising and falling transient behavior is more similar as well.

    Best regards,
    Niklas

  • +1 in reply to Niklas Schwarz
    Prodigy 40 points

    For synchronous rectifiers and non-synchronous rectifiers, the overshoot and undershoot of synchronous rectifiers will be symmetrical at 0A~2A, while the overshoot and undershoot of non-synchronous rectifiers will not be symmetrical. If it is 200mA~2A, both synchronous rectifiers and non-synchronous rectifiers may be symmetrical because the test conditions are not so strict? Is my understanding correct?

  • +1 in reply to user6055081
    TI__Mastermind 33510 points

    Hi,

    Your understanding is almost correct.

    user6055081 said:
    because the test conditions are not so strict?

    It is not because the test conditions are less strict, but because you never operate with 0 load.
    Without load, it takes much longer for Vout to go down by itself if an overshoot happens. Like said before, in asynchronous, current cannot flow back through the diode, so if there is no load to consume the power, Vout drops slowly just through leakage and the device does not switch and waits. If you have a load, the output voltage does not stay high after an overshoot as the load consumes current, so the device does not have to stop switching and keeps operation.

    Best regards,
    Niklas