Engineers frequently attempt to add ferrite beads to power traces to reduce ripple. As we have stressed, ripple results because the demand for current requires many reflections. The bead approach is similar to that used in the low frequency filtering of analog signals, where RC or LC filters are used. Ferrite filters fail in most cases. First, the inductance is very small and second, raising the transmission line impedance at a point can only add another point of reflection.
A single-turn ferrite inductor might have an inductance of a few tenths of a nanohenry. At 1 GHz, the reactance is around 1 ohm. This is much lower than the characteristic impedance of most transmission lines. If a shunt capacitor is added, it is obviously in the right direction. The capacitor alone might be of some help. If multiple turns are used on the bead, the resulting parasitic capacitance will lower the natural frequency of the inductor to well below 1 GHz.
Ferrite materials have very little permeability above a few megahertz. The impedance they offer is in the form of losses. In fact, the inductance is usually not much different from the same air-core geometry. In high current applications, any core material will probably saturate. Air-core inductors do not saturate.
The presence of a ferrite bead changes the geometry of a transmission line usually in the direction to increase the characteristic impedance. This is in the direction to add reflections and increase the noise on the ...