[Czilla9000]

Out of curiousity, how does ferrite suppress EMI/RF? Is it because it is magnetic? IS the more magnetic the substance the better it is as suppressing EMI/RF?


Thank you.

 

[Nisbeth]

I'm not completely sure about this, but I believe it's because of the inductance of the chokes. The chokes form a low-pass filter that filters away EMI and noise which is normally very high in frequency. The reason for the ferrite core is that the ferrite increases the inductance of the choke as compared to an air-core coil/choke. More inductance = lower cut-off frequency = less noise

/U.

 

[Demolition]

To complement Nisbeth's post, I've cobbled together a simplistic explanation of what a ferrite is actually doing when you place it on a cable:

When current flows through a cable, a magnetic field forms around it. Because a ferrite is magnetically permeable, when one is placed on a cablee, it increases the flux density of the magnetic field around that section of cable. A higher flux density equals a higher impedance (with a correspondingly higher resistive component) and, as Nisbeth mentioned, a higher impedance allows low frequency signals to pass through more easily. At the same time, higher frequencies (e.g. RFI/EMI aka "noise") are absorbed and dissipated harmlessly as heat and vibration.

D.


Edit: Corrected the transposition of 'inductance' for 'impedance'.

 

[Chipko]

Just to complete the above answers:

Common mode chokes lets even high frequency signals through. A pair of wires carrying a signal results in a net magnetic field of zero as the signal travels both ways. An outside disturbance only travels in one direction through the choke.

There's a great page describing all this but I can't find it.

 

[Budgie]

Excellent explinations by Demo and Nis.

Simplified- The ferrite creates a high resistence to the frequencies to which it is reactive, at the spot it is located on the cable, attenuating the level of those frequencies passing thru the cable. There are different formulations of the ferrite materials, that change the band of frequencies it is most reactive too.

http://www.netmotion.com/htm_files/ot_ferrites.htm

 

[Czilla9000]

The reason I asked is I am wondering if using a Neodymium magnet ring would be better. Since Neodymium-Iron-Boron is more stronger magnetically, I am wondering if it would be better than ferrite.

BTW, shoudn't inductance be good then (it is usually refered to as bad? Why not simply add an inductor?

 

[kheldar]

Inductance isn't bad, but it has its place in a circuit. Almost every good power supply will have inductance to reduce noise in the output.

I dont know if a permanant magnet would be good for a choke, the function of a choke is somewhat similar to a capacitor in that it holds a slight charge and releases it when the magnetic field is reversed (in an AC circuit) so when the current flows one way, it will have an extra amount of resistance, but the otherway will have reduced resistance due to the pre-existing magnetic charge.

 

[kheldar]

Inductance isn't bad, but it has its place in a circuit. Almost every good power supply will have inductance to reduce noise in the output.

I dont know if a permanant magnet would be good for a choke, the function of a choke is somewhat similar to a capacitor in that it holds a slight charge and releases it when the magnetic field is reversed (in an AC circuit) so when the current flows one way, it will have an extra amount of resistance, but the otherway will have reduced resistance due to the pre-existing magnetic charge.

 

[Czilla9000]

Quote:

Originally posted by Demolition To complement Nisbeth's post, I've cobbled together a simplistic explanation of what a ferrite is actually doing when you place it on a cable: When current flows through a cable, a magnetic field forms around it. Because a ferrite is magnetically permeable, when one is placed on a cablee, it increases the flux density of the magnetic field around that section of cable. A higher flux density equals a higher inductance and, as Nisbeth mentioned, a higher inductance allows low frequency signals to pass through more easily. At the same time, higher frequencies (e.g. RFI/EMI aka "noise") are absorbed and dissipated harmlessly as heat and vibration. D.

So is the more the metal is reactive to magnetism the better the EMI filteration?

 

[Demolition]

Quote:

Originally posted by Czilla9000 So is the more the metal is reactive to magnetism the better the EMI filteration?

Are you asking if a stronger magnet makes a better filter? If so, then the answer is no. As kheldar mentioned, a permanent magnet is not an ideal ferrite choke.

The keyword here is one that I mentioned earlier: permeability. A magnetically permeable material is one which absorbs and/or alters a magnetic field that it comes in contact with. It's not necessarily magnetic itself. As mentioned in my earlier post, a ferrite merely alters the cable's existing flux density (i.e. concentration or strength of a magnetic field) at the point that the ferrite is placed.

Changing the flux density affects a cable's impedance (combination of inductive reactance, resistance, and capacitive reactance). As high frequency noise travels to the spot in the cable where the ferrite is placed, the resistive component of the cable's impedance increases, which effectively filters the noise out, yet still allows the lower frequency signals to pass through.

(Note: kheldar's post twigged something in my brain, which made me look back at my previous post. I accidentally transposed 'inductance' for 'impedance'. I'll go back and change that after I finish this post.)

D.

 

[Czilla9000]

Demolition, I have abandoned the idea of a permanent magnet.

So in other words, the more a material responds to a magnetic field, the better it is as a choke?

 

[Demolition]

Quote:

Originally posted by Czilla9000 Demolition, I have abandoned the idea of a permanent magnet. So in other words, the more a material responds to a magnetic field, the better it is as a choke?

Yes, that's basically the idea. However, to be more precise, the permeability of a choke determines which frequencies it is better at filtering. So, for example, if you have a specialized application that requires exactly 145 ohms of impedance and an attenuation of 7.5 dB in your cables to target and filter certain frequencies, then you could conceivably find ferrites that will do that.

Anyway, before I launch into another Physics lecture, I think that for most purposes that we will run into, almost any ferrite choke will do the job as long as it fits on the cable.

A popular one is from Radio Shack. I've bought snap-on ferrites like this one from Digikey, as well. I use them on power cords and on cables around my computer (such as the ones leading to multimedia speakers). As mentioned above, the idea is to attentuate high frequencies which often manifest as noise. There's a lot of stray EMI/RFI around power supplies and computers, so ferrites help here.

I've also heard that they work well on headphone cables to prevent them from picking up stray RF. On the other hand, I've also heard that they dull the highs. I haven't used a ferrite on a headphone cable since I haven't had a problem that might require one. Maybe someone with experience in that regard could chime in on that?

D.

 

[Chipko]

Found the link to a good article on ferrites as noise suppressors

http://www.elmac.co.uk/ferrite.htm

 

[rkt31]

I bought a flat USB oyaide cable . is there any flat ferrite clip on choke available to use with the flat USB cable ? can I use 2 flat ferrite magnet pieces instead of ferrite chokes. I have few small flat ferrite magnets which are powerful.

 

[Speedskater]

But ferrite magnets are not ferrite chokes.  Ferrite chokes have a hole in the middle where the signal cable passes through.
 
Why would you think that your USB cable needs a filter? Are you having data errors?
 
Ferrite chokes only work on RFI.  EMI is at too low a frequency for the filter.
 
The go-to expert on all things EMI/RFI is Jim Brown, he has several papers on using ferrite chokes.
 
http://www.audiosystemsgroup.com/publish.htm