4
\$\begingroup\$

Old senior design project and in schematic I see he uses ferrite bead and capacitor after charge pump voltage output befor connecting to opamp input

Circuit very simple:

  1. charge pump uses MAX232 to get +- 12V output from 0-5V input
  2. +- 12V output rails from MAX232 having ferrite bead and capacitor in line
  3. output of above connected to Vdd/Vss of dual rail opamp

Any way of knowing how much ripple reduced from output of charge pump because of using ferrite bead and capacitor in line?

Any measurement to select value of ferrite bead and capacitor?

I also interested as I see ferrite bead and capacitor in line on Arduino board also but they give no value in schematic!

\$\endgroup\$
1
  • 1
    \$\begingroup\$ I had to go look it up (and found it here on TI's support site. The switching frequency you need to filter is about 50 kHz. \$\endgroup\$ Aug 16, 2011 at 14:07

1 Answer 1

7
\$\begingroup\$

A ferrite bead on a conductor creates a very small inductor. The inductance is a function of the size of the bead, the ferrite material used and the number of turns through the bead.

It is extremely unlikely that the ferrite bead will be able to filter switching frequency - they tend to be used to suppress noise well up into the megahertz. I doubt that it will have any effect on ripple (usually tens or hundreds of kilohertz) - it is quieting high-frequency spikes (megahertz and higher) which may be disturbing digital circuits or generating EMI.

Sizing an LC filter for this sort of noise reduction involves measuring the frequency and amplitude of the noise you're trying to suppress. The type of ferrite material as well as the type and value of the capacitor is critical for HF suppression. Once you know the noise you need to quiet, and the characteristics of the LC filter, you can determine how much attenuation you will have .

It can be difficult to design these HF LC filters up-front without knowing what sort of noise you're dealing with. Often, you'll put footprints in the PCB layout for the LC and empirically determine what values you actually need once you get the hardware running.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Another related point is that the original designer may have been relying on the opamp's PSRR spec to reject noise near the switcher's fundamental frequency. Because PSRR degrades rapidly at higher frequencies, it'd be a good idea to check the spec when you decide how much more filtering (if any) to add. \$\endgroup\$
    – John Miles
    Aug 16, 2011 at 21:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.