Because of the Thanksgiving holiday in the U.S., TI E2E™ design support forum responses may be delayed from November 25 through December 2. Thank you for your patience.

  • State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 64 subscribers
  • Views 170 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.

TPSM82816: Maximum output capacitance considerations.

Shady Gross
Prodigy 50 points
Part Number: TPSM82816
Other Parts Discussed in Thread: TPSM8287A06, TPSM82866A, TPSM8287A12, TPSM8287A10

Hello,

I have three question in blue below.

Design requirements:

  • Vin = 5 V
  • Vout = 3.3 V
  • I_LOAD max < 6 A
  • C_LOAD = large and application dependent.
    • Application in question: About 370 uF of nominal capacitance from ceramic capacitors distributed on the motherboard (where the TPSM82816 3.3 V regulator is placed) + 800 uF of tantalum capacitors on a daughter board (PMC module) connected to the motherboard through multiple board-to-board connectors.
      • The 370 uF on the motherboard is fixed.
      • The 800 uF on the daughter board could be much less depending on the third-party daughter board model. 800 uF (from tantalum capacitors) is currently the maximum value encountered. 10 uF (from tantalum capacitors) is currently the minimum value encountered.

From the TPSM82816 data sheet:

  • The recommended maximum effective output capacitance is 470 uF.
  • Section 8.3.3: "The minimum output capacitance in Table 8-1 is for capacitors close to the output of the device. If the capacitance is distributed, a lower compensation setting can be required."
    • What is considered "close to the output of the device?"
    • Can additional capacitance beyond 470 uF be present if it is far enough, like on the daughter board in the example in the design requirements above?
  • R1, R2, and C_FF values in table 9-2 (88.7k. 19.6k, and 3 pF, respectively) are different from the online Webench values (453k, 100k, and 0.59 pF, respectively). Both sets meet the data sheet requirements. High efficiency at loads below 0.1 A is not required. The device mode is set to PSM.
    • Is one set of values better than the other?

If TPSM82816 is not suitable because of C_LOAD, please suggest an alternative that meets the given design requirements.

Thank you.

  • +1
    TI__Guru*** 147230 points

    Hi,

    Thanks for sharing many relevant details of your application, so that I can give a thorough answer.

    For your questions 1 and 2, it will probably work fine.  The 370uF on the same board is probably 'close'--assuming they aren't 1 meter away or something.  The 800uF on the other board is far away, especially since there is the impedance of the connectors and traces/planes in between.  As well, and very helpful for stability, is the tantalum caps instead of ceramic.  The ESR really helps stability by adding a zero in the loop.

    I recommend ordering an EVM and testing it with a mockup of what your final system would look like electrically.

    3. Either set will work fine.  If you decrease the resistors a little bit more, you can get a more reasonable value for the Cff (like 10pF).

    The TPSM8287A06 is another device to look at.  It has external compensation that can for sure be stable with that output cap.  It's also larger and more efficient.  You don't need to use the I2C interface.  Let me know if you have any questions on it.

    Thanks,

    Chris

  • 0 in reply to Chris Glaser
    Prodigy 50 points

    Hi Chris,

    Thanks for your input. I will test TPSM82816 with the EVM. I will try to run my test within the next couple of days before marking this issue as resolved in case more questions come up.

    Regarding TPSM8287A06:

    I was originally using TPSM82866A, which kept hiccuping and could not complete ramping all the way up to 3.3 V if the 800 uF daughter board was connected. I surmised that the extra capacitance was drawing too much current and the soft start time was not long enough, so I focused my search on buck modules that have an adjustable soft start time.

    I dismissed TPSM8287A06 because it shows fixed under "soft start" in the parametric table when browsing and filtering "Buck modules (integrated inductor)," but I see from the data sheet that soft start is adjustable over I2C. Perhaps that "fixed" soft start value in the table should be "software adjustable" or something like that for clarity.

    I still, however, have no means to change the TPSM8287A06 soft start time because 3.3 V powers the device that can talk to TPSM8287A06, so I am concerned that TPSM8287A06 will not be able to ramp up to 3.3 V with a 1 ms soft start ramp when the 800 uF daughter board is installed.

    Thanks,

    Shady

  • 0 in reply to Shady Gross
    TI__Guru*** 147230 points

    Ah, I forgot to consider the SS time for such large Cout.  It's been a while since I wrote this app note, which should match the issue you saw: https://www.ti.com/lit/an/slyt669/slyt669.pdf

    Yes, with TPSM8287Axx there is a 4msec SS time option if you have I2C to set it.  Or you could use the higher current, pin to pin, TPSM8287A10 or TPSM8287A12.  The higher current and higher current limit should remove any startup issues.

    Chris

  • 0 in reply to Chris Glaser
    Prodigy 50 points

    One more question about Cff for TPSM82816:

    Data sheet table 9-2 gives max Cff at min Cout, but the text later states "increasing the CFF above values given by Equation 10 can also improve the response with larger COUT."

    The EVM comes with R1 = 200k which corresponds to a 1.33pF Cff at min Cout using data sheet equation 10, but the EVM has a 10pF Cff.

    Is there a way to compute Cff for the specific application's Cout? What would happen if the chosen Cff value is too large?

    Thanks,

    Shady

  • 0 in reply to Shady Gross
    TI__Guru*** 147230 points

    Hi Shady,

    Here's an app note that discusses how to optimize the Cff: https://www.ti.com/lit/an/slva289b/slva289b.pdf 

    Chris