• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 63 subscribers
  • Views 380 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.

Simple Switcher versus Compensation Calculator Tool

Bill Drumheller
Prodigy 135 points
Other Parts Discussed in Thread: TPS40304

Hello,

While doing a Point-of-Load design using the TPS40304, there seemed to be a discrepancy between the calculated phase margin and cut-off frequency of the control loop using the Compensation Calculator Tool (Rev. D Excel spreadsheet) and the Simple Switcher software.  The design process was to first use the Power Stage Designer, then perform the compensation design using the Excel file and then finally input the entire circuit into the Simple Switcher tool.  According to the Excel file, the bandwidth of the control loop was around 80kHz (which seems reasonable given the switching freq of 600kHz).  When putting the same values into Simple Switcher, the bandwidth was only about 40kHz.  The phase margin was also different but not as significant.  Any ideas why there would be such a large difference between the two tools?

I have not tried various input/output parameters (designs) nor have I actually performed any control loop testing in the lab, but will be doing that this week.  Any help in trying to understand why this would be very helpful as I would like some confidence in the compensation design in future designs.

Thanks!

Bill 

  • 0
    TI__Expert 4350 points

    Hi Bill,

    Thank you for your interest in TI parts. This is a common question. As with any switcher loop design, there are many possible and viable answers for compensation. Even with everything else in the design fixed by a particular BOM, there are many combinations of the 5 compensation components (for a Type 3) that will work and be stable. Simple examples would be:

    1) Fc = 80kHz, PM = 60deg

    2) Fc = 80kHz, PM = 61deg

    etc. Both of these would be stable and nearly indistinguishable, but they imply a different compensation set. The granularity of available standard component values may not provide such fine differences, but in theory there are almost limitless solutions.

    The XL loop tool uses one approach to stabilizing the loop. It chooses Fc based on a fraction of Fsw, then calculates the 5 values aimed at providing a predetermined target for PM. The approach used by Simple Switcher is different, so it will yield different values. Even if you force the rest of the BOM to be the same (everything except the compensation), and specify ripple and load transient requirements to be the same, there are still multiple possible solutions. Of course if the power stage is different (mostly the output choke and output capacitor bank) or the requirements are different, you would expect the compensation to be different.

    Both tools will provide viable solutions on paper. For an optimized loop comp, it is recommended that you start with one such paper design, and then actually measure the plant with real hardware and apply any compensation adjustments based on actual measured versus desired performance. There are parameters in the physical design such as device parasitics and even layout that can affect the loop and can't easily be predicted by these tools.

    Regards,

    MC.