Designer’s Guide Community

‹ Previous topic | Next topic ›

Pages: 1

Cutoff frequency for circuit in Figure 9 of Power Supply Noise Reduction Guide (Read 1469 times)

nkinar

New Member

Offline

Posts: 2

Cutoff frequency for circuit in Figure 9 of Power Supply Noise Reduction Guide
Nov 05^th, 2010, 9:09am

Hello,

I just had a read of the extremely interesting Power Supply Noise Reduction guide available on this website, and I would like to use the circuit in Figure 9 to provide damping at the output of a SEPIC switching converter. The switching frequency is approximately 300 kHz, and my system has a number of ADCs sampling at a maximum rate of 1 kHz.

I am wondering if there is a design equation available to calculate the cutoff frequency of Figure 9. What should I choose as the cutoff frequency of this filter? I am assuming that the filter is constructed from both parallel and series RCL, and is a low-pass filter.

I would assume also that L should be rated for the maximum current being passed through the circuit, Cdamp and Cbyp should be rated for the maximum voltage, and Rdamp should be rated for maximum power. Is there a design equation available for calculating the required power of Rdamp?

IP Logged

nkinar

New Member

Offline

Posts: 2

Re: Cutoff frequency for circuit in Figure 9 of Power Supply Noise Reduction Guide
Reply #1 - Nov 7^th, 2010, 5:09pm

I think that there are actually two equations for the cutoff frequency. �For the parallel RCL circuit, the lowpass cutoff frequency (in Hz) will be

fc_parallel = 1 / ( 2 * pi * R *Cbyp )

Alternately, for the series RCL circuit, the lowpass cutoff frequency (in Hz) will be

fc_series = R / (2 * pi * L)

The designer selects both cutoff frequencies so that attenuation is provided. �I believe that it is important to select the fc_parallel cutoff frequency prior to the fc_series cutoff frequency. �These are two filters that are being cascaded.

Then, the selection of all other values in the circuit is an optimization problem based on the two equations above and the constraints for Rdamp given in the paper. �

The required power for Rdamp is calculated simply using the ubiquitous P = I * V equation for power.

As mentioned in the paper, if Cdamp is very large (i.e. 200uF), then the ESL of the capacitor will serve in lieu of using a resistor R; no resistor is required.

IP Logged

Pages: 1

‹ Previous topic | Next topic ›