The Choke in the photo is for USB's cable, near the plug, it said. just "clip" it near the plug, not wires.

How it works?

Hallo Stanley,

what you have published is a clap-on ferrite. I would suggest you to use a real 6hole-choke, because clap-on ferrites will show much lower inductivity, unless you feed the cable over and over through the ferrite. A 6hole ferrite choke like '74275043' from www.we-online.de gives an inductivity of more than 20µH.

How does the ferrite works?
If a current is flowing through a wire, a magnetic field develops, which circumferes the wire. The magnetic flux Q of the resulting magnetic field is proportional to the current I, according to:

Q = L x I

As Q is rather small with naked wire, inductivity L is also rather small, about 1µH per meter.
But if you put a material arround the wire, where the magnetic field can pass through a medium of permitivity µr with µr >> µair, then the magnetic flux is heavily increased and according Q = L x I also L is increased.

This is the mechanism of introducing relevant inductivity by using a clap-on ferrite or by any other ferrite, where ferrite material circumferes a piece of wire.

Ferrite material has another advantage: If you choose a soft ferrite material, then at frequencies above about 10MHz ferrite looks like a series resistor, dissipating high frequency interference directly into heat. This is especially helpful, because an inductivity only could cause heavy ringing in combination with stray capacitane of cable. But with this developing series resistance ringing is heavily damped.

Why the hell can a clap-on ferrite decrease radiation of a cable, even if no shunt device is used, which could shunt interferenece to ground, like a RC low pass filter will do it?
Your cable looks like an antenna, showing a relevant stray capacitance to sourrounding ground. Keep in mind, that this sourrounding ground must not necessarily be close to the cable! Even if sorurringding ground is far far away, there's a relevant stray capacitance of about 180pF per meter present, which is called intrinsic space-charge capacitance. And this really relevant stray capacitance is charged by the driver circuit of cable or by common mode noise, when different ground points on PCB show different potentials!!!
A clap-on ferrite now helps to decrease the current flowing into this stray capacitance and by this decreasing the radiation caused by the cable.

An example: Asssume an inductivity of 3µH as result of 2µH from soft ferrite and 1µH from 1m cable. Then, with intrinsic space-charge capacitance of about 200pF, this yields a resosance frequency of about:

fres = 1 / 2 / pi / SQRT(L x C) = 6.5MHz

Ringing is prevented, if ohmic resistance in series with inductivity (think of the dissipative loss of a soft ferrite!) is bigger than

R >= SQRT(2 x L / C)

In our little example R >= 170Ohm would be needed to prevent ringing. A soft ferrite will not show such high ohmic resistance at 6.5MHz, but it's enough to keep the ringing heaviliy limited. And even if ringing would occur, 6.5MHz cannot relevantly be radiated by a 1m long cable, because wave length of electro-magnetic wave would be 46m.
And, for all frequencies greater 6.5MHz radiation is heavily decreased, especially at those frequencies, where the cable length would create a perfect antenna!

That's the design rule of ferrites with cables:
Choose a ferrite to show so high inductivity, that in combination with intrinsic space-charge capacitance resonance frequency is dropped to frequencies, where the cable cannot radiate much. Even if ferrite cannot provide enough ohmic resistance (dissipative losses) to totally prevent ringing, the system will highly profit from the ferrite, means radiation is heaviliy decreased.

Kai