Use of Ferrite Cores for EMI Reduction???


I've been doing a lot of research on the net lately about the use of Ferrite Cores to reduce ElectroMagnetic Interference (Yes, I are an engineer). It started when I bought a digital camera and there was a Ferrite Core at each end of the USB Cable AND one at the end of the power supply going into the Camera. The directions said that it was for noise reduction through those 2 cables! So I thought, HEY, I have cables in my Audio System...

This really got me thinking about the use of Ferrite Cores to significantly reduce the influence of any stray EMI that could be degrading the sound quality. Does anyone have an idea of this would help, OR does the EM shielding that already comes with most cables do the trick?

Any hardcore Electrical Engineering input would be GREATLY appreciated. Am I on to something worth trying, or just full of bunk??
hager_charles
Hello Charles. Have you visited this website? It has a very good discussion of noise and ways to control it. Also a *very* good DIY noise control section:

http://www.magnan.com/
Check out the Audiopoints.com site and see what they are using with their cables and AC cords!
I was getting taxi cab and cell phone calls through my speakers. I solved the problem by using rf stoppers on the interconnects from my pre amp to amp. The heavy duty ones 2 per interconnect.
be careful what you do with those ferrite clamps. I clamped 1ea. on a pair of Audioquest Lapis interconnects; they killed the highs & PRaT was all but gone. Sold the whole pack forthwith.
Bob, out of curiosity, where was this cable located at within the confines of your system ? I have found that ferrite works best on digital based products. In specific, i'm talking about the power cords of transports, dac's, cd players, digital tuners, digital amps, etc... Like anything else though, try things in moderation. You CAN get "too much of a good thing" if you try ganging up multiple ferrite beads or clamps.

As many of you know, digital components are VERY noisy devices. Since they can pump their digital "hash" back into the power lines and other components, placing ferrites on the power cords at the component chassis helps to minimize this. Besides reducing the potential for the "dirt" to re-enter the AC system, this also reduces the chance of the device using the power cord as an "antenna" to re-radiate the rf based digital signal. This also works in reverse, as rfi is less apt to enter the device via the "power cord antenna" due to the impedance bump that the ferrite creates.

One thing that we have to make sure of when using ferrite based filters is that the cable remains centered in the ferrite choke. If the ferrite actually rests against or is closer to one section of the cable than another, it can produce an erratic filtering action. Not only does this reduce the efficiency of the ferrite as a filter, it can produce various loading conditions on the different polarities of the signal due to proximity effect. I would "assume" that this could produce a slightly "disjointed" or "un-natural" effect to music, possibly resulting in the lack of PRAT and tonal balance that Bob mentioned. This is not to say that this was what caused the problems he noted in his specific situation, only that it could be a possibility.

Since many ferrite chokes or clamps fit loosely over some cables, a ( cheap and dirty ) solution is to fill in the gap between the cable and ferrite with something that is "non-offensive". While the use of a paper towel cut into strips and then wrapped around the cable comes to mind, anything that is non-metalic or prone to static build-up should work fine. With the paper towel, etc... you can make as many turns necessary to take up the needed space. The clamp is then applied over the "spacer".

Not only can this keep the clamp more secured and keep it from sliding around on the cable, it should keep the gap between the ferrite core and the cable pretty evenly spaced. This is not a "test proven method" by any means but something that came to mind. If others have found better solution to this problem, PLEASE contribute your suggestions as you see fit. Sean
>