speaker wire

Most designs of the ferrites just clamp over the wires. Do make sure that the wires for one channel aren't connected to the other channel as some amps use differential drive to get the power up for the output. Even just if you have a "correct" hot and ground, the current is high enough that one channel will feed to the other if you tie the grounds together anywhere in the system. The wire size is a good one for higher power amps as you will be driving a 8 ohm load generally and any resistance between the speaker and the amp will drop the quality of the sound by allowing the speaker to resonate more as the impedance of the wire is increasing the source impedance that it sees.

-- Why isn't there an Ozone Hole at the NORTH Pole?

The rf signal does not show up as differential mode on the speaker wires. It shows up as *common* mode. Both speaker wires of the pair have the same voltage impressed on them.

Think of the speaker wire as a *single* conductor that sits in the rf field that a local tranmitter sets up. The wire is a driven element in the field so it has so-and-so many volts per meter of rf voltage impressed on it by virtue of being immersed in that field.

So there will be some large rf voltage between the end of the speaker wire(s) and ground in the chassis of the amplifier or whatever teh speakers are connected to.

This is fine at this point because the output transistors don't care about a few millivolts of rf being there, and because it's common mode, the speakers never even *see* the rf.

The trouble happens because the internal layout of consumer grade electronics is pretty poor. Once the rf gets inside the chasssis, it can easily re-radiate and be picked up by more sensitive stages. If there is any non-linear device in the input stages of, say, a low-level preamp, the rf can be demodulated down to audio frequencies at a point where it *does* matter.

How do you fix a problem like this?

1) make the driven element (speaker wires) a worse antenna. This is done by putting chokes at some point along their length (wrap the speaker wires around a ferrite torroid a few times to make a common mode inductor) so as to change the length of the antenna, away from the problem frequency. 2) bypass all the speaker leads, inside the chassis. This is done with a low inductance capacitor (short leads, typically 0.01 mfd or so) so that the speaker terminals present a very low impedance to ground for rf. The rf currents in the driven element never get inside the chassis then. 3) improve the layout and isolation inside the amplifier or reciever by using better shielding on sensitive front end stages. This is one area that most manufacturers avoid because it adds complication and expense.

Note that even incoming power leads can bring common mode rf into a chassis. They should in principle be bypassed right at the point of entry as well.

Jim

Even if you are, you won't know the difference. Time domain reflectometry comparison between Monster Cable and #14 zip cord showed no difference in signal quality in a band from 0 to 100MHz.

I don't know about _your_ ears, but mine don't go quite that high... The point about solid wire looking like power wiring is valid, though. As low voltage cable, it'd have to be in different junction boxes and so on. Something to consider - if you run that blue flexible conduit ('Smurf tubing' I've heard it called), then you can always pull new cables in as technology and your needs change. That's the one thing I didn't do when I built my house, that I regret not doing.

Dave Hinz

A typical answer from a monster cable advocate

Don't you know that expensive cables improve "believability" of sound?

i

Good point. However, a fine point marking pen will write on the jacket. I would suggest writing "SPEAKER" every 6' or so along the wire.

Ted

Heh...I've saved a ton of money over the decades by not being able to hear that alleged last 5% of sound quality improvement that makes something 10 times more expensive.

As soon as you can put "believability" into engineering terms I can measure, I'll be happy to measure it. I used to work with a guy who was an freak amongst audiophiles - granite turntable, magnapan speakers, tube amps, cables the size of your arm, and all that. Yeah, sure, it sounded good, but...the Bose system for a few hundred bucks sounds pretty damn good too, y'know?

With the controls of the system set for normal listening, stop the CD, tape or whatever. Turn up the volume control about 5 or 10 db and listen carefully. If you don't hear any stray signals, as they say, "Don't fix it if it ain't broke." If you do hear a bit of a local brodcast station or ???, try the ferrite beads.

Ted

Have you got faster hardware that can use better such as fiber optic?

Ted

That's why I suggest the flexible conduit. Yank out the cat5, put in the gigabit ethernet fiber. Next time it'll be something else.

It's not the voltage but the signal that is the problem. The signal will 'ride' on whatever is available in the cables. The field does not need to be 'hugely strong' either, even passing cars with radio transmitters (CB, Ham, taxis, emergency services etc.)can affect audio systems with long speaker wires.

Believe me it happens and I have had to deal with it several times. The signal may be applied to the 'output' in terms of the speaker wires but it is a 'closed loop' system and that signal can and *does* get back into the amp where it 'can' get into the amplification circuit and be amplified and sent out to the speakers as part of the 'mix'. A lot has to do with the design of the amp and how good it's rejection circuits are on the 'front end'. Most 'cheap' modern amps have very poor rejection circuits and are prone to picking up stray signals.

A few years ago there was a big scare up here in Canada when Industry Canada proposed some new rules with regard to domestic amplification equipment. The European Community had recently introduced much higher standards for equipment in terms of the ability to reject unwanted signals while Canada's proposals virtually eliminated the need for filtering by 'effectively' saying that the equipment in question had to 'accept' any extraneous signals. This raised huge problems for the Ham Radio community as it meant that Canada could be 'flooded out' with cheap Asian amps that could no longer be sold in Europe and which would mean that Hams would suddenly be 'causing' interference problems with neighbours who bought said amps despite the fact that they had changed nothing in their transmitting equipment!

I am not sure if the proposals ever went through but I do know of several cases of Hams suddenly getting into cheap 'home theatre' systems and having to help cover the cost of fitting ferrite loops to the systems in an attempt to cure the 'problem'. In most cases the Hams either paid for the loops themselves or helped to pay for them despite the fact that it was not really 'their' problem simply because the alternative was to face being 'taken' off the air for 'supposed or implied' interference problems!

Ah but Ted. You should know more than anyone that TDR measurements won't show the real effect taht true audiophools can hear. Their ears are so much more sensitive than those instruments.

How else could they justify buying those gold-plated power connectors?

Jim

Even at RF, that only works if the individual strands are insulated (Litz wire). Besides, skin depth in copper at 20KHz is ~0.5 mm so forget it.

Ted

The ferrite used in permanent magnets has a different composition than the linear ferrites used in attenuator cores and is already in saturation. In addition to adding a little inductance, the material in an attenuator core is "lossy" at RF frequencies, turning the RF energy into heat.

Kevin Gallimore

Right! Solid gold is what you need! Accept no substitutes! ;)

Kevin Gallimore

Yup....no point in fixing what ain't there! However, long speaker cable runs *are* prone to picking up signals and sometimes those signals may be short lived (a passing car transmitting etc.) As the ferrite 'filters' are fitted close to the respective equipment and not in the wall I would suggest you run the cables and try it. Then if you are getting problems you know how to fix it.

Thanks, Ted. Now I understand what I'd be looking for.

One question I've had for years is why I have some (very little, but noticeable) 60 hz in one channel and not the other. Been that way since I bought the Mc2300 way back in '75. Any clues?

Harold

The only exception I might make here is that your connections would certainly be better, considering gold doesn't corrode the way copper and other elements do. The only problem I've seen is that the typical gold connector is so thin that it almost wipes off with your fingers. If you could buy connectors with a thou of gold flashed on, I can't help but think you'd benefit by not getting noisy connections.

My choice in using stranded as opposed to solid is for other reasons. I've pulled enough wire through conduit to know solid is much more difficult. I also like the fact that it is easier to work with when wiring, and isn't as prone to breaking from bending. At my age, anything to make it easier works for me!

Thanks to everyone for the great information.

Harold

Gold is indeed only a poor third place in conductivity, but it comes in first place for connector plating, because of its freedom from oxidation. Gold contacts (gold on gold) are also self-lubricating which is another reason it is chosen for switch contacts.

It's not just a noise issue - connectors and switches for small signal applications need to 'dry switch.' By this I mean the contact resistance has to be very low, even for small signals. Most larger power switches will show a fairly large (fractions of an ohm) contact resistance until an arc forms and blasts the contact area clean.

The contacts in those switches are often made of an alloy of silver and cadmium. When I worked at GTE, there was a man there who did nothing but test different alloys to try to make contacts that worked as well as the existing ones, but *without* the cadmium!

Connector and switch manufacturers understand how much gold has to be put on to give a reliable number of operations for the device. The thicker the plating, the longer it will last in service.

Jim