The effectiveness of the ferrite is a matter of the characteristics of the
ferrite and the "wave impedance" (the ratio of the total voltage [forward
and reverse waves] to total current [again, forward and reverse waves]) at
the point along the wire where the ferrite is placed. For example, if you
place over a horizontal ground plane a vertical wire that is 1/4 wavelength
long at a given frequency, the impedance looking into the wire at the
ground plane end is about 36 ohms. If you place a ferrite at the ground
plane end of the wire that in effect inserts a 50 ohm resistor, you will
reduce the emissions from the wire by about 7.6 dB.
Now take the same setup, but make the wire 1/2 wavelength long. The
impedance looking into the wire will be on the order of 1000 ohms. That
same 50 ohm ferrite will have almost no effect.
Note that the wave impedance along a wire will change as a function of
frequency, and as function of the position of the wire relative to the
driving source and other conductors.
Don Borowski
Schweitzer Engineering Labs
Pullman, WA
[email protected] wrote on 03/27/2004 04:35:44 AM:
>
> Dear EMC experts,
> For a beginner like me this is really a treasure. I get
> answers to all my questions. I am thankful to everyone who replied me.
> Today I ran radiated emission testing for an USB device in
> fully anechoic 3 meter chamber. In horizontal polarization the unit
> failed at 72 MHz. Then by adding a ferrite core on the USB cable
> the unit found to pass with good margin. But the observation is like
this:
> When the ferrite core is near to the EUT, test fails and
> the result is almost similar to the test without a ferrite core.
> Then I tried by moving the ferrite core gradually towards the other
> end of the cable. I noticed that as the ferrite core is slowly moved
> away from the EUT, the emission level also gradually reduced and
> when the ferrite core was kept at the farthest point of the cable
> (cable length is about 1 meter) the emission was minimum. In my
> previous scans (in other devices) I don't remember a ferrite core
> helping in horizontal polarization.
> So what is the logic behind this? I request you all to clarify
me.
>
> Thanking you all sincerely
>
> K.Balasubramanian
> Sr.Engineer, Hardware.
