• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 63 subscribers
  • Views 531 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.

LP38690: Stability and Output Capacitor ESR

John Garrett
Expert 4210 points
Part Number: LP38690

I have a customer who is using a LP38690 with Vin = 6.6V from a 400kHz switching power supply and Vout = 5.0V.  They are using a 4.7uF ceramic capacitor on the input and both a 4.7uF, 0603 located right at the output pin of the part and a 47uF, 1210 located at the load which is an inch away through a .025 inch wide trace. Further away there are also some 0.1 uF’s and a few 100pF scattered about on the load line. The output load current varies between 20 mA and 450mA .  I have ESR curves for the ceramic caps they are using . 

The customer would like to know if this LDO would be stable in this configuration. The ESR curve for the 4.7uF output capacitor shows an ESR of over 100 m-Ohms at 400kHz and the 47uF which is an inch away shows about 20 m-ohms of ESR   The data sheet would indicate this is not a problem as it mandates an ESR range of 5 - 500 m-Ohm.  Please confirm.  

If this circuit will not work as is, please recommend capacitor changes for the 4.7uF and 47uF capacitors that will enable stability across the load range. 

Thanks, John Garrett, AFA

  • 0
    TI__Guru 55940 points
    Hi John,

    The ESR range specified in an LDO datasheet is for the resonant frequency of the capacitors. This will be the lowest point on the ESR curves. Keep in mind that the trace resistance does factor into the total ESR that the LDO sees.

    As it sounds like you already have boards made and a good understanding of the transients that you are expecting, often the easiest way to determine if the capacitor selection is adequate for the application is to test the transient response. If you perform your worst case transients and watch the response of the output rail, is the magnitude of the transient response acceptable? Is there too much ringing on the output?

    Very Respectfully,
    Ryan
  • 0
    Guru 140200 points
    Hi John,

    the LP38690 does not show a good input voltage ripple rejection. This is typical for LDOs. So, feeding the input with a switcher could make trouble. I would recommend the use of a pi-filter between switcher and LP38690, like discussed here:

    e2e.ti.com/.../669266

    The LP38690 is uncritical if it comes to the ESR of output capacitor. But take care: When paralleling caps with very low ESR, heavy resonances can occur, resulting in sharp and very high impedance maxima. At the frequencies where thess impedance maxima occur, there's no longer any decoupling! This can be very critical in fast circuits and can result in instability.

    To avoid these resonances small resistances should be inserted in some of the connections between these low ESR caps, either in the supply line or in the connection to ground. I use 0R22/0805 resistors for this purpose. The idea is to make some of the caps "lossier". But only do it where the ESR of the Cap is not critical.

    In the good old days when electroyltics where used for decoupling, at least partially, the non vanishing ESR of these caps played this role. Another good idea is to parallel a low ESR cap (e.g. 0805) with an electrolytic. Then you have both, the low inductance of the low ESR cap and the ESR of the electrolytic.

    Of course, you can also add series resistors between the decoupling caps in the supply line and by this form RC-filters. Then you have additional filtering and the avoiding of resonances.

    Kai