Speaker Wire - What is correct or accurate?

[HDTV Forum]

Richard, going back to your earlier statement

As an example there is not an industry standard for input and output impedances of line level signals nor the cables and connectors for those impedances.

Yes, I guess there probably isn't. Connectors are probably driven by cost and what's easiest to hook wire up to. Some are better than others, but good conductivity and reliability are the most important. Doesn't really matter how you do it. For output impedance, lower is better. For max. power transfer it should match the load. But what's the load? Depends on the speaker design and quality. Just a hard thing to standardize, and if you did I bet you'd still have differences in apparent sound quality because of other reasons.

I'll have to email you for info on your 8 ohm cable. Interesting that some equipment blew up. Too much load capacitance (high peak currents) for the output transistors to drive perhaps?

[HDTV Forum]

I've been thinking about speaker cables and wires today and I've got some ideas for all you to think about. I have come to the conclusion that speaker wires actually can make a difference in the sound, specifically shielded cables or using coax. The engineer's immediate assumption is that it shouldn't make any difference, but some people, including Richard, claim it does. The engineers may be wrong, but also, those who claim it changes the sound could be drawing their conclusions on the wrong reasons also.

Here's what I think is going on, based on some very similar past experiences with amplifiers. Audio power amps are essentially differential amplifiers. They put out a balanced signal and are intended to drive a balanced load. Usually they use P or N-channel power MOSFETs in the final output stage. When the audio signal causes one side to go positive, it sources current through the speaker to the other side of the driver, which has gone negative and is sinking the current. Neither output is referenced directly to ground, and if you did, you'd probably blow the output driver or it just wouldn't work right. Follow me so far?

Now, those FET output transistors have what's called a gain-bandwidth product typically several MHz or 10s of MHz. In other words, you could probably make a low freq. RF amp out of them if you wanted to. In audio circuits, they could drive much higher frequencies if there weren't audio filter circuits to limit the bandwidth. Now days, everything is made faster, and you can't really find a transistor that will only work for audio frequencies.

So, what happens when you put a nice piece of coax cable on each speaker output? Say you cut a couple 20' pieces of RG-6. You want to have a sheild, so you pull off the shields into a pigtail and ground it to the back of your amp someplace. At the speaker end, you connect the center conductors to the terminals.

What you have done is connected a 20' open-ended transmission line to each amp output. So what you say? Well, RG-6 has about 16 pF/ft capacitance and 91 nH/ft inductance. The amplifier still has it's normal differential load of 8 ohms (or 4 or whatever) due to the speaker. However, each output now also has a single-ended load to ground which looks like a series L-C circuit. It will look like you have a 324 pF cap in series with a 1.8 uH inductor on each output.

The bad thing about series L-C circuits is they have resonances. At the first resonant frequency, the impedance goes to nearly zero. For a 20' length this will happen at about 6.5 MHz. Open-ended transmission lines have pretty high Q, which means the losses are low and the resonance is quite sharp.

If you put a reactive circuit like this on a transistor output, various things can happen. The worst thing is you turn your output stage into something resembling a feedback amplifier. If the transistor still has some gain, and since the phase goes from capacitive to inductive reactance at resonance, you can get instability or oscillation in your output. Lead and circuit board trace inductance around the transistors also plays a big role. If it is oscillating at RF you probably won't hear it, you may not hear any audio, and you might blow out the output stage.

Even if it does not become unstable, you are introducing a reactive circuit which may have some effect in the frequency response of the amplifier at lower frequencies. This is where I think people hear a difference. Perhaps it's either rolling-off the high end response, or maybe even peaking it. Whatever, I could see someone thinking they have changed the sound and it has to be better, because you have just shielded your cables!

Sorry for this being so long. I'm sure I'll get some argument with this. But I have seen differential amplifiers do screwy things when they had too much transmission line hung on their outputs.

emc guy

[HDTV Forum]

The engineer's immediate assumption is that it shouldn't make any difference, but some people, including Richard, claim it does. The engineers may be wrong, but also, those who claim it changes the sound could be drawing their conclusions on the wrong reasons also.

That is a great point and a trap that every designer and subjective listener must be aware of and one of the reasons I have the PS Audio AC synthesizer. If you take a neutral/accurate CD player and drop it in your system and are unhappy with the sonic signature then what you are unhappy with is your system and not the player. There is

[HDTV Forum]

Richard, I've got a better understanding of what you were trying to do with the coax. What I described earlier would not necessarily apply to a single-ended amp design.

I was under the impression that coax was being used in some fashion as a replacement for speaker wire. I saw another thread about someone doing this. I was trying to point out that this can cause an RF short circuit to appear at the amplifier output at certain frequencies, and this could be very bad for the equipment. I was assuming that most people are using amplifiers that have floating balanced outputs meant to have the load across the +/- outputs. Most home and auto equipment I've seen is like this. I didn't think bipolars were used much anymore. Perhaps there's some nostalgia for bipolars just like for tubes.

If you are trying to do impedance matching of speakers, you've opened a can of worms. You've got a passive compensation network with the cable, and R-C at the speaker terminals. Were you looking at the transient voltage response at the speaker terminals? If so it might be that what you have done isn't necessarily a good thing. It is the current in the voice coil which should be optimized for transient and frequency response. Just interested in how you know that what you did was good, besides going by what it sounded like.

To add to one of your other comments -

the engineers knowledge can be a double edged sword. It can help you reason out an event but if you are not careful it can also prejudice one against an event with out actually testing the hypothesis of that event in a real world setting.

I think the prejudice comes from hearing the reasons given as 'fact' for when something, like expensive speaker wire or power conditioners, are touted as an amazing miracle for improved audio (or video) quality. What I'm learning here is:

1. There may be perceptible changes in audio/video quality but there are often other scientific explanations for those changes. And understanding what actually happened may offer an alternative which is far less expensive or easier to do. This unravels the black magic and snake oil claims behind a lot of this stuff.

2. Engineers should be open-minded about claims of improved or changed performance, but that does not mean accepting the given reasons behind those claims. Especially if those claims are from the equipment manufacturer. They are in the business to make money.

3. It really takes someone with good engineering background and an active interest is audio/video/home theater to figure it all out. That's what's needed to sort out the truth from the BS. So how many people out there do we have like that? Probably not too many ...

emc guy

[HDTV Forum]

I didn't think bipolars were used much anymore. Perhaps there's some nostalgia for bipolars just like for tubes.

Not at all. Bi-polars are still common as dirt. Typically only hi-end gear has used FET

[HDTV Forum]

I didn't mean to cause such a stir, but I found it very interesting.

JZ

[HDTV Forum]

Don't give it a second thought. Your post was well received. The testing of different zip cord was well done. Unfortunately zip cord is not real speaker wire.

The conclusion will end up being that nothing has changed and each individual will have to select the product that provides optimum results for the stuff in their system because there are no standards in audio - just synergy.

Richard

[HD Library]

SPEAKER CABLES: Science or Snake Oil
viewtopic.php?p=19668