• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 3 answers
  • Subscribers 39 subscribers
  • Views 585 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.

WEBENCH® Tools/TPS54335A: Forced constant conduction mode in webench

Olli Mannisto
Expert 1750 points

Part Number: TPS54335A

Tool/software: WEBENCH® Design Tools

It seems that Webench won't run operating values on DCM, at least not on this part. It switches to forced constant conduction mode, i.e. anti-green mode where the inductor is charged and discharged by the IC. The part does claim that it's got low-load efficiency mode and this appears counterintuitive. If the controller uses FCCM to maintain tight control over output at all situations, how can it be saving energy at the same time?

The datasheet is a bit strangely worded here:

7.3.2 Light-Load Operation
The device monitors the peak switch current of the high-side MOSFET. When the peak switch current is lower
than 0.5 A (typical), the device stops switching to boost the efficiency until the peak switch current again rises
higher than 0.5 A (typical)


This does appear to be the opposite of FCCM, i.e. pulse skipping mode. Is the webench providing incorrect results here and in reality the part will go into pulse skipping mode, rendering the simulations invalid when webench transitions from CCM to FCCM? This does happen in both Flash and HTML5 versions.

  • 0
    Expert 1750 points

    Ah yes, it pops up a disclaimer if you try to model a <0.5A max current design but doesn't react if you tweak operating parameters inside design.

  • 0 in reply to Olli Mannisto
    TI__Mastermind 28720 points

    Olli,

    Thanks for your question. The device does seem to be going into discontinuous current mode at currents lower than 500mA. I quickly checked the datasheet and found figures 31 and 32 at 100mA and no load ripple conditions showing the switching node.  WEBENCH tool does not model this discontinuous mode of operation for this device. This is a limitation in the model of the device. However, the actual device should enter discontinuous mode and boost efficiency for lower current conditions. Please let me know if you have further questions about the device operation.

    Regards,

    Amod

  • 0 in reply to Amod Vaze
    Expert 1750 points

    Amod Vaze said:
    Olli,

    Thanks for your question. The device does seem to be going into discontinuous current mode at currents lower than 500mA. I quickly checked the datasheet and found figures 31 and 32 at 100mA and no load ripple conditions showing the switching node.  WEBENCH tool does not model this discontinuous mode of operation for this device. This is a limitation in the model of the device. However, the actual device should enter discontinuous mode and boost efficiency for lower current conditions. Please let me know if you have further questions about the device operation.

    Regards,

    Amod

    There is actually one thing..

    The no-load scenario shows quite large gap between pulses, about 180µs. That translates to 5.5kHz switching frequency. I've got an application that has abrupt and severe load step from virtually no load (~20mA)  to 1.8A  in active mode. It's an IoT cellular module which behaves that way in 2G mode, even if you use it to transmit constantly the time division functionality means it looks like a PWM load with fsw of about 200Hz. 

    I'm concerned how long does it take for the switcher to wake up. If the switcher will wait up to 180µs before engaging, the 470µF tank capacitor will already be down by 0.7V at 1.8A step load, which would likely reset the cellular module. My design has ~15kHz cross-over which presumably will wake up the controller eventually but if the switching remains clamped until the next power save pulse..

  • 0 in reply to Olli Mannisto
    TI__Mastermind 28720 points

    Thanks for your question Olli. I have requested a product expert to reply to your question.


    Regards,

    Amod

  • 0 in reply to Olli Mannisto
    TI__Genius 10070 points

    Hi Amod,

    The switch frequency changed with load in skip mode, the larger the load current, the higher the frequency. It is not always 180us. And you can test the load transient waveform as Figure 5 of below user's guide to check your application.

    http://www.ti.com/lit/ug/slvub76/slvub76.pdf

    Shawn

  • 0 in reply to Shawn Nie
    TI__Mastermind 28720 points

    Thanks Shawn.

    Olli, hope that helps answer your question.

    Regards

    Amod

  • 0 in reply to Amod Vaze
    Expert 1750 points

    Yes, thanks. In other words I can make the switcher more responsive by a ballast load but it doesn't have to be 500mA..