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??
Charles, the use of ferrite is pretty widespread in various electronic fields. There are many in's and out's as to where it can and should be used. The one main drawback to use of ferrite's has to do with saturation. Knowing the magnetic properties of the specific ferrite being used can come in handy. As such, you might want to experiment with use of ferrites on various cabling. Some manufacturers already include beads or clamps on some of their products.

I would not go crazy with quantities and pay attention to where you place them at. I would also mention that their use on power cords for high current items ( like power amps ) are more questionable than anywhere else, so keep that in mind. Sean
Commercial products you might want to consider are Audioquest's RF Stoppers and RF Stoppers, Jr. Radio Shack also sells ferrite cores appropriate for audio systems. I was getting RF noise through my turntable, and using three (one on each phono interconnect plus one on the power cord) of AQ's RF Stoppers worked well. Admittedly, the core on the power cord theoretically shouldn't make any difference since there is no conducting path between the motor and the tonearm; I used a core anyhow to minimize any rebroadcast interaction possibility (the AQ's came in a pack of four).
If you want to experiment, I believe that Audioquest sells cores in two sizes for use on power and interconnect cables. I think they are called RF Stoppers or something similar. They are relatively cheap and come in simple snap on cases. I have them on my power cables to everything but in truth I never A/B'd to see if there was a difference. I think Stereophile gave them a thumbs up in their accesories category.
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:
You can buy a "billion" ferrite clamps in various sizes from either Digi-Key or Mouser ( can't remember which ) for the price of one small package of the Audioquest stuff.

Do a search through the old Greg Weaver articles on Soundstage and he even gives you part numbers. Some of these might not be valid anymore, but the dealer can probably cross reference them to current stock. They do have a minimum order, but it would still be less than buying a package of the AQ parts. If you can't find the part numbers, i'll try to look them up tonight or tomorrow. Sean
Check out the 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
Bob, I already got the Highwire Wrap. It's very good, especially for the price. The improvements are just what Bluenose posted.
Those RF Stoppers were used on a much different rig than my present setup. I was having a lot of RFI problems from trucks on a nearby highway; I'm surprised that my amp or speakers didn't blow up when some of those clowns went by with their (non)linear amps broadcasting simultaneously on all 40 channels. Their $hi! was supersaturating an older preamp, then of course the hash was amplified by my PA. I put those RF stoppers on the analog interconnects driving the PA; I clamped them on the cables within a few inches of the amp. The RFI was indeed substantially reduced to the point of being tolerable, but they just ruined the sound, & these were just the 'junior model' RF Stoppers. They do have spacers on the inner diameters of the clamps in order to help avoid the aforementioned assymetrical positioning problem. I used some electrical tape for that purpose but I do like the wrapped-paper idea because then the clamp can be moved up or down the cable for positional optimizating.
I've seen ferrite cores installed on power supply cables & AC cords (the Audioquest AC12) which maybe makes a little more sense, but when used on analog interconnects they were awful.
Respected designer Charles Hansen (of Ayre Acoustics) actually told me that when Ayre was experimenting with ferrite AC filtering, they found that the initially realized improvements were degrading over the long term due to a "magnetic buildup" within the cores, which had to be degaussed periodically. He also mentioned that ferrite was causing some sort of 'grunge effect' on the sound, another reason that they didn't like it.
Francisco thx for the Powerwrap report; I need to order one to play with but I'm hoping to find an alternate source vs. the near-north retailer that sells them in our area. I still have a bad taste in my mouth leftover from my latest dealing there; it's not a BFD but next time I'm going elsewhere if possible. Does the package show the manufacturer's name & address? Please advise if you have this info? thx
I bought some rat shack ferrite cores for my CDP 'power cord'tuner power cord and pre power cord.cant tell any difference but I sure sleep better at night!
In my case the rf stoppers work best close to the pre amp.
I just added the paper towel wrap. Sounds the same to me.
Ferrite beads should only be used on wires not carrying the audio signal, where you are filtering out unwanted noise.(If you use them at all) The reason that made these peoples systems sound bad is the same reason shielded cables sound bad. The audio signal is AC which generates a varying magnetic field around the conducting wires. If you attempt to collapse or restrict this field, you will influence the signal that is producing it. If you MUST shield your interconnects, then try using unshielded interconnect (30ga. solid) wire inside a MUCH oversize flex tube wrapped with appropriate shield material and a drain. Preferably, each wire + and - on each run will have separate shielded tubes as their fields also interact, especially when buched right next to each other as in a standard interconnect. Standard interconnect topology is designed for convenience, not sound, even though high end mfrs. would have you think otherwise. Use tube at least 1 inch diameter with teflon discs to center the wire in the tube. Be sure to connect the shield to one RCA plug end only , as a drain.Or drain directly to a chassis mount lug. The dynamics will stun you.
Twl, the cables that you mention would be quite high in inductance. Then again, the treble roll off that takes place might somewhat balance out the somewhat leaner and brighter sound passed on by using such a fine gauged conductor.

I think that the guy that basically came up with ideas like this and pushed the envelope as we know it was David Magnan. He has gone to great lengths to minimize time smear, skin effect, etc.. in cables like nobody else that i know of. I've got a couple sets of his cables that make use of 36 gauge conductors that are air insultated and somewhat spaced apart. His newer designs don't even use wire at all, as they use some type of conductive "paint" to minimize skin effect. Never heard or used them, but some say that they are the ultimate. Sean
Sean, please elaborate on your comment of high inductance characteristics for the cable I mentioned. I believe that as inductance increases with proximity, the 1/2 inch minimum separation of conductor to shield should be adequate to reduce inductance to a minimum. Larger separation would be better, but we must retain some level of ability to move the cable around. With the pos. and neg. each run separately in their own separate tubes, they would not interact with each other significantly. I am aware that inductance affects can be present from AC 120v power cables as far away as 1 foot, but in the smaller power levels present in interconnects I don't think the problem is anywhere near that magnitude. Also, I mentioned that I would prefer no shield, but was trying to address his RF splash problems.
Bringing conductors closer together moves them more towards being a capacitive load. Spacing them further apart makes them less capacitive and moves them more towards the inductive side of reactance.

When one can find equal amounts of capacitance and inductance at the same point, the reactance is effectively nulled. While this can be achieved, it typically varies with frequency. As frequency rises, wavelengths are shortened and the gap between conductors appears to widen. That is why zip cord is more inductive than a twisted pair given the same amount of insulation around each conductor. Both start out relatively even at very low frequencies but the gap between inductive reactance and capacitive reactance widens as frequency climbs.

The twisted pair has more intimate contact area between the two conductors as they spiral around each other. This increases capacitance / lowers inductance. The zip cord has only one continual point of contact along their lengths and is therefore more isolated from each other. As a result, inductance is increased and capacitance lowered.

As a side note, anyone that has tinkered with speaker crossovers knows that putting increased inductance in series with a speaker will roll off high frequency response. Hence, the lack of treble clarity and extension when using zip cord with thick insulation.

As one can tell, this is a very simplified example. If you want further proof, take some Goertz speaker cable and separate the conductors. Not only will it no longer be a high capacitance design, the total impedance will be drastically altered. Sean
Sean, thanx, but what I was really wanting to know was at what distance, of separation of the wires, eliminates inductance and capacitance effects entirely? And if the shield is not used to carry signal, but only drains, does any impedance effect on the shield affect the sound? I expect that power levels will influence the distance, but what would you consider a "safe" distance apart, regardless of convenience? Also, would having separate + and - wires with individual shields/drains affect the spacing issue?
Sean, in the above when i said impedance I meant to say inductance.
Twl, just running a "drain" aka "floating or telescoping ground" will alter impedance. I would therefore "assume" that putting ferrite over that would also alter it further. That is, at least within that small region where the ferrite was applied. After all, that what ferrite does. It creates an impedance "bump" within the nearfield.

As to specific ratios in terms of spacing vs capacitance / inductance, frequency and power levels, you've got the wrong guy. Better find someone that is both an EE and knows their physics : ) I'm WAY out of my league on stuff like that. Sean