• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 64 subscribers
  • Views 416 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.

TPS54260 compensation

Grayson King
Prodigy 130 points
Other Parts Discussed in Thread: TPS54260

I'm trying to design a regulator using the TPS54260 and am finding some serious limitations in the datasheet and design tools.  Specifically, SwitcherPro appears to ONLY support designs clocked at 300kHz when using this chip, so I'm having to resort to running through good old fashioned hand calculations using the datasheet.  Fine, except that the datasheet has some errors/problems relating to compensation that I can't seem to reconcile.  Details....

I'm working with the TPS54260 datasheet marked "SLVSA86A –MARCH 2010–REVISED DECEMBER 2010", and am specifically looking at the "Compensation" section of the "Step-By-Step Design Procedure" on pages 33 & 34.  There seems to be a problem with equations 41 & 42.  Using the example design of the "Step-By-Step Design Procedure" we have the following:

Ioutmax = 2.5A  (per initial design criteria, p.28)
Vout = 3.3V  (per initial design criteria, p.28)
Cout = 40uF  (per "Compensation" section statement: "For Cout, use a derated value of 40uF", p.33)
Resr = 1.5mOhm  (per "Output Capacitor" section, bottom of p.30)

Using these numbers in equations 41 & 42 I come up with the following results:

fpmod = 3014Hz
fzmod = 2.653MHz

But the text in the "Compensation" section states "For the example design, fpmod is 1206Hz and fzmod is 530.5kHz".  Something's not right here.

I think I know where they came up with the figure of 1206Hz.  In equation 41, if you were to replace the numerator with "1" instead of "Ioutmax" you'd come up with 1206Hz using the above stated values for Vout & Cout.  So is the equation wrong, or is the value in the text body wrong??  And I really have absolutely no idea where they came up with the figure of 530.5kHz for fzmod.

Can somebody please help with this??  If the design tools don't support the frequency I need to operate at, and I can't trust the equations in the datasheet, I'm not sure how I'm supposed to come up with a solid compensation network.  Thanks in advance for any assistance!

  • 0
    TI__Guru**** 191445 points

    There may be errors in that section.  It may have been cut and pasted from another datasheet, I'll have to check.

    But more importantly, you can choose the switching frequency in switcherpro.  When you create your design, use the "advanced inputs" tab.  There is a field for Fsw there.  Let me know if you have trouble with it.

  • 0 in reply to JohnTucker
    Prodigy 130 points

    Hey, thanks for the quick response John!  I see that now in SwitcherPro.  Just confusing that that feature isn't visible when you design for a device that you've already selected.

    Another question with SwitcherPro:  After I go and manually change my inductor, output capacitor, etc to better suit my specific design needs, is there a way I can get SwitcherPro to re-select only certain components (like the compensation network for example) to better match the components I modified manually?

    And I'd still like to get an answer to the issues with the datasheet too.  I always like to have a couple sources to verify a design.

  • 0 in reply to Grayson King
    TI__Guru**** 191445 points

    I tried this.  If you change the output filter components, switcherpro re-calculates the compensation automatically.  I am not sure what algorithm switcher pro uses.  If you want you can manually adjust the compensation in the "what if analysis" section.  I have my own technique that I use with "what if" and the power stage gain and phase plots that you can view.  Remember power stage + compensated error amp = total loop.

    I'll have a look at the datasheet, but it may not be today.  I can tell you that the datasheet uses a very simple model for calculating compensation based on theoretical current mode control with no slope compensation.  When you add slope compensation into the transfer function it gets complicated fast.  I use manual calculations based on the power stage gain and phase, either measured or simulated in switcherpro or psipce.