# Ferrite bead in Ground path of Input Power Supply

I have been following the schematics of Colibri Toradex Carrer Board. I have used most of the components from their design. However, they have used ferrite beads in Positive terminal and Ground terminal of Input Power supply.

I am unable to understand how ferrite bead in ground connection helps in noise immunity. Please help me as I am planning to remove it as it consumes some space in my denser PCB.

• It all depends on what connects to the IN port and what connects to the OUT port and how the things that connect might be grounded (earthed) or not. Please describe the power source coming into the IN port and detail what is connected to the OUT port. Jan 19, 2019 at 16:11
• electric fields tend to couple to the largest mechanical structures with the stronger flux ----- this will be GROUND metal. Having a Bead in that GROUND is part of EMI FCC-mandate certification. Jan 19, 2019 at 16:24
• @Andyaka Input will be power from Wall adapter (12V 3A). At the Output, there are Buck Converters, PMIC, Power switches. Jan 19, 2019 at 17:13
• @abhiarora what you have just stated... Is that what your are "stealing" the design for or, is that what the Colibri thing is designed for? It makes a huge difference. Assume that I know nothing about the Colibri thing. Jan 19, 2019 at 17:26

@Andyaka Input will be power from Wall adapter (12V 3A). At the Output, there are Buck Converters, PMIC, Power switches.

Consider this modified picture: -

I think the likely scenario intended when applying a ferrite bead into the ground line is the reduction of common-mode switching noise.

The common-mode noise coming from many SMPS wall wart power supplies is notorious and this is mainly because they don't have a ground connection in their plug i.e. they operate with only live and neutral line connections. Some suppliers attempt to mitigate this problem by having a Y rated capactor internally from isolated DC output to primary rectified bus voltage but still, some noise can arise and, it will show itself on the output lines even though they are meant to be DC.

That noise is common to both positive and negative output lines hence the name common-mode. It comes about because the isolation transformer is being driven at circa 100 kHz (with fast rise and fall times) and primary couples to secondary due to internal transformer capacitance.

If the natural load capacitance to ground is low, you need extra help by inserting inductance$$\^1\$$ in the through-going power lines shown in your circuit (and mine). You need to apply that inductance on both wires (because both wires transport that noise) hence you see a ferrite bead on power and ground.

So, if you are designing a product that needs to meet regulatory emissions then you either have capacitors to ground that can shunt the common-mode noise or, if the product doesn't naturally have a local ground connection, you have to rely on the loads parasitic capacitance to ground and supplement the filtering effect it might have by series blocking elements such as ferrite beads or inductors.

$$\^1\$$ Ferrite beads not only produce inductive reactance to help natural parasitic capacitance reduce radiated EMI from the load but also are quite resistive in the range 10 MHz to over 100 MHz and this is chiefly where a lot of SMPS common-mode noise can bite you.

All transient current induced voltage noise is based on a voltage divider of impedance values. The caps have ESR, and the ferrite has the same load current in ground.

So what is the ratio or difference in using caps with an ESR*C of 1us vs 200 us if the ferrite is 1 or 100 ohms at 10 MHz?

Do you know yet how to calculate this? If not then that is what you need to learn.

Impedance, Z(f) of inductors , capacitors and resistors and voltage divider ratios.