Post Go back to editing

LTC3388 - DC bias derating for ceramic capacitors

The "typical application" reference circuit in the LTC3388 data sheet includes several capacitors:

  • 2.2uF 25V (Vin)
  • 1uF 6V (CAP)
  • 4.7uF 6V (Vin2)
  • 47uF 6V (Vout) -- this one will vary by application and is just included here for completeness

Per the "applications information" section, the Vin capacitor should be a ceramic capacitor, the Vout capacitor can be a ceramic capacitor, and the CAP and Vin2 capacitors are unspecified but presumably should be ceramic capacitors as well. Ceramic capacitors need to be derated for DC bias. This effect is not mentioned anywhere in the data sheet, and it's unclear how to account for this in relation to the listed capacitor values.

One clue is the evaluation kit data sheet, which provides specific part numbers for the capacitors. Based on that, the listed capacitor values appear to be nominal values -- the evaluation kit uses 2.2uF, 1uF, and 4.7uF-rated capacitors for Vin, CAP, and Vin2 respectively. That would suggest the listed values already include a derating factor. However, I'm not sure if it's a good idea to rely on that, as the derating curve can vary significantly between specific capacitor models (even from the same manufacturer), and the smaller the capacitor, the more pronounced this effect is. Can you provide more specific insight on capacitor selection? What I'm looking for is something along the lines of "for CAP, choose a capacitor that provides about 0.5uF at 4.8V, regardless of nominal value" (and any other relevant suggestions).

Most of the capacitors used in the evaluation kit are obsolete now, so I can't just use those exact parts.


  • Hi Eddie,

    It is a good question and often not addressed in datasheets. Many customers are not aware that this is even necessary with ceramic capacitors. As a result, we accommodate for derating in the datasheet values.

    A good clue of this, as you have observed, is that we use the same values on the evaluation boards.

    I recommend you use the values from the eval board. It is safe to rely on that because the eval boards are tested, working circuits. I recommend you ignore derating unless you are calculating a cap value for your specific application. For the VOUT cap, you might account for this - use a cap that provides the calculated capacitance at your programmed output voltage. For the rest of the caps, just use any ceramic cap with the recommended values in the case sizes used on the demo board.

    If you are hoping to use smaller case sizes, we may want to discuss this further. The idea there would be: "given the derated actual capacitance of the cap used on the demo board, how much rated labeled capacitance do I need in a cap with a package size of x?". There is some variation between models of certain packages, but there is not that much, so a generic estimate should do fine.



  • Hi Zack,

    Thank you for confirming that. I'm fine with the package sizes used on the eval board, so I'll stick with those.


Reply Children
No Data