• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 62 subscribers
  • Views 187 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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:

TPS543C20A: Inquiry of TPS543C20A loop gain characteristics

TPS543C20A: Determining Loop Gain for Multi-phase Configuration

Chuck Sampson
Intellectual 270 points
Part Number: TPS543C20A
Other Parts Discussed in Thread: TPS543C20

Hello,

I am using two TPS543C20A converters in a two-phase configuration. I am using the TPS543C20_Calculator. The calculator is not set up for multi-phase converter analysis. As a workaround, I used the standard, equivalent single phase model of the two phase converter. 

1. Single phase equivalent inductor is 1/2 the two phase inductor

2. Single phase converter switching frequency is double the two phase switching frequency.

3. Single phase equivalent converter input voltage is 1/2 the two phase converter input voltage.

My question is the following: Are these modifications sufficient for an accurate loop gain analysis? Is there something I missed.

As to the loop gain results, only the LC corner frequency and outer voltage loop gain crossover frequency is calculated and displayed. What about the current loop crossover frequency? How can I check that the inner current loop has a higher crossover frequency than the voltage loop?

To lower switching frequency I reduce RDAMP. Does this action reduce the modulator gain and then reduce the outer voltage loop frequency response? 

V/r,

Chuck Sampson

  • 0
    TI__Expert 5206 points

    Hi Chuck,

    The best way to model multi-phase design in the excel spreadsheet is to divide the IOUT of the multi-phase design by 2. Then fill out the spreadsheet as you would normally. The inductor value and COUT values will apply to each device individually. If the inductor value is 0.8uH, then you use 0.8uH inductor on each device in the multi-phase design. 

    There is a calculated bode plot shown to the right of the table. You can use this to better understand the loop frequency response.

    Where are you placing the mentioned RDAMP resistor?

    Best regards,

    Britton

  • 0 in reply to Britton Jones
    Intellectual 270 points

    Not sure I understand what you mean when you say stacked. The dc-dc are configured as a 2 phase, multi-phase design. Each phase is 180 deg. phase shifted. Therefore the effecting switching frequency is doubled and the equivalent inductance is 1/2 the inductance of each phase. This affects the LC corner frequency a lot. i agree that the current is reduced by one-half. 

    Thanks.

  • 0 in reply to Chuck Sampson
    TI__Expert 5206 points

    I meant a multi phase design. 

    Best regards,

    Britton

  • 0 in reply to Britton Jones
    TI__Expert 5206 points

    Hi Chuck, 

    I am closing this thread. As mentioned previously, you can use the excel tool for a multi-phase design by dividing IOUT by number of phases and then designing as you normally would for a single phase design. The resulting recommendations will be used for each phase in the multi-phase design.

    Feel free to reply to this message if you still need help with this request.

    Thanks,

    Britton

  • 0 in reply to Britton Jones
    Intellectual 270 points

    I sent an inquiry to my FAE, Steve Widener. I respectfully disagree with your answer. By dividing out the Iout current for each phase and running the calculator you will not get the right result for the control loop. A multiphase converter has an effective switching frequency that is double a single phase converter. The equivalent inductance is one half of the single phase. Therefore the output filter corner frequency increases. Please refer to the enclosed TI application note. I have used the analysis approach described in this application note before and measured correct loop gain plots. My only question was can this approach work with the TPS543C20A calculator? optimalDesign of Interleaved sync buck slyt139.pdf

  • 0 in reply to Chuck Sampson
    TI__Expert 5206 points

    Chuck,

    The analysis method of multiplying the inductance by 1/2 and the switching frequency by 2 is effective for voltage mode control devices. However, for current mode control devices, like TPS543C20A, the forward loop gain is not dependent on the switching frequency or inductance. Therefore, modeling a 2-phase design in the excel calculator using full load, full output capacitance, 2x the switching frequency, and 1/2 the inductance will not provide the best estimation for loop response. For these devices, the most accurate loop analysis is achieved by dividing the output current, transient load, and output capacitance by 2.

    Best regards,

    Britton

  • 0 in reply to Britton Jones
    Intellectual 270 points

    That goes against everything I have read about and designed and tested concerning multiphase dc-dc converters. It also conflicts with the TI app note I enclosed. We will just have to agree to disagree. 

    Respectfully

    Chuck