22
\$\begingroup\$

Sorry if this is the wrong place but I'm interested in a scientific and electrical explanation and not some HiFi voodoo magic.

I guess you all have seen the discussions about expensive audio cables and "improved" sound quality.

My question is, what's the best cable for an audio signal? Loudspeaker cables are usually a pair of ~12 AVG cables terminated with banana connectors.

enter image description here

As far as I understand this must be affected by all sorts of noise. To minimize noise it's normal to use coax cable for "radio" signals (LF, MF, HF, ...). Wouldn't coax with proper connectors like APC-7 or equivalent be the best for audio too? Or are the frequencies too low for coax somehow?

Internet was a bit vague on this topic.

What I'm really wondering is why don't we use coaxial cables for loudspeakers in average home stereos?

Lets say a signal level of 50 dBm (100W).

\$\endgroup\$
15
\$\begingroup\$

There are two reasons why speaker cables are not shielded/screened:

  1. The signal is so powerful that any interference would not be noticed.

  2. Speakers are not very sensitive; it takes a lot of power to create sound on a speaker

This is why speakers are connected to an amplifier.

Input to amplifiers are very sensitive and so input should use shielded/screened cables.

P.S. If you have a 100W amp the good ones use about 48V. So you have at least 2A, with peaks higher. Peaks might hit 20A but for a very short time. To hear this your speaker cable needs to be thick.

\$\endgroup\$
  • \$\begingroup\$ I'll give you an up vote, but I'm not sure if its the right answer. How I read it, it doesn't matter as long as the cable is thick enough? But still, in a world of HiFi and magic shielding would still be better? \$\endgroup\$ – Duckers Feb 3 '11 at 14:10
  • \$\begingroup\$ @Duckers - This is an engineering site. Sure, you could keep some noise off with shielded cables, but that's only useful if the noise can cause an effect on the output. With high-power signals at audio frequencies, it's difficult to couple any noise onto the cable. You don't need to be concerned about interference higher than audio frequencies \$\endgroup\$ – Kevin Vermeer Jul 1 '11 at 20:08
11
\$\begingroup\$

Impedance is a large factor in this situation, as is sensitivity.

In the case of a speaker wire, your driving it with a very low output impedance amplifier (<< 1ohm) and driving a load with a fairly low impedance, a typical speaker may be from 3-50 ohms across the audio spectrum.

The voltage noise you see on a cable is very dependent on these impedances as the noise is really a coupled current, the voltage that is measured on the cable is a product of this current and its impedance path to ground.

In the case of a speaker wire this is a very low impedance so it takes a lot of coupled noise current to induce meaningful voltage changes in the signal. Shielding a speaker wire is really pointless unless its in a very high noise environment, like if you set a coil on top of a building mains transformer.

Another way to say this is that speakers are really current mode devices, the driving force is large currents and EMI coupled currents are really small in normal applications for speaker wire.

Line level signals are more sensitive for multiple reasons, one being that the signal they carry is generally later amplified, another being that the input impedance and source impedance for single ended line level audio are much higher. Input impedance is generally around 10k for single ended line level inputs, it takes far less current to create a large voltage noise on that input than it does with a speaker wire. This is why almost all line level audio cabling is shielded, be it RCA type running on coax or XLR running on STP.

Another way to say this is that line level transfer is almost always voltage mode, meaning that the receiving end is looking for voltage levels and drawing minimal current, as a result, small currents such as coupled noise become larger voltages and have a much bigger impact on the signal.

This same concept applies to many other issues. For instance with high impedance op amp inputs care has to be taken to minimize any coupled noise to that input signal for the same reasons that line level audio signals are more sensitive. Often guard rings or similar approaches are used.

\$\endgroup\$
6
\$\begingroup\$

For driving speakers at 100W, simply use cheap mains cable. It's a waste of money using anything else.

\$\endgroup\$
  • 4
    \$\begingroup\$ It won't make any difference to the sound quality. Double-blind testing comparing expensive speaker cables to ordinary cables has showed that listeners couldn't distinguish between them. \$\endgroup\$ – Leon Heller Feb 2 '11 at 23:35
  • 1
    \$\begingroup\$ i wouldn't call it a waste of money, last time i looked 100ft of 12 AWG speaker wire from monoprice is cheaper than 100ft of 12/2 NM. Not to mention that 12/2 NM isn't all that likely to slide under your baseboards or bend around 90 degree corners so its not visible. \$\endgroup\$ – Mark Feb 3 '11 at 0:42
  • 1
    \$\begingroup\$ Isn't mains cable usually solid (by code?); that could be fairly impractical. \$\endgroup\$ – Nick T Feb 3 '11 at 2:11
  • 2
    \$\begingroup\$ Stranded mains cable, of course. Not the stuff that is used for household wiring. \$\endgroup\$ – Leon Heller Feb 3 '11 at 7:27
  • 1
    \$\begingroup\$ What do you mean by "best"? An audio amplifier is a system for presenting information to a human, and it's been proved scientifically that there's no human-distinguishable difference between lamp cord and any other type of large-diameter cable. So the only differences would be things like convenience (stranded vs solid), longevity (corrosion resistance), etc. \$\endgroup\$ – endolith Feb 10 '11 at 17:18
5
\$\begingroup\$

There are 3 main cases that you need to look at for the type of signal being carried:

Digital

In the case of digital, the quality of the cable doesn't have a huge effect on the audio itself. For the most part it either works or it doesn't work and the quality doesn't change at all. What you will find is that in very high noise environments or long lengths you might start hearing digital noise.

Low Level Analog

This is cables such as head phone cables, microphone cables, lines running to your amp, etc. These are the most susceptibility to noise, especially in long runs. For Pro audio XLR is used which is a differential line. This eliminates noise that is introduced in the cable. For consumer audio there are cables that are better shielded that can help with noise reduction, but the biggest key here is to try to keep your cables as short as possible and limit the number of connectors used.

High Level Analog

This is the situation where you have a cable that is carrying the audio to the speaker after the amplifier. This is an interesting area that could deserve a question in and of itself. You need to have a cable that is big enough to handle the current required, but a larger cable can cause problems with the signal if the level is too low. In this situation noise generally isn't an issue because the signal is much higher then the noise floor. However, there does need to be care taken in not making the line overly long and not passing it directly by sources of high noise energy.

The connectors here also make a difference, you mentioned Banana, but there is also terminals that you connect the bare wire to directly as well as speakon connectors that are used in higher end audio. The key is making sure the connector can handle the power being supplied.

Additional Information

The cable market is littered with overpriced cables that have many claims. I have seen one particular case of an XLR cable that claimed to make a signal "warmer" sounding. I had a reseller try to convince me to buy 50 of these cables to replace the ones I already owned. I asked for a "trial" with one of the cables so I could hear the difference. I ended up hooking them up to a high end audio recorder and found that there was essentially 0 difference between the different cables. The claim was that there wasn't a difference because my recorder was distorting the cable. So my advice, don't buy into that bull.

\$\endgroup\$
  • 1
    \$\begingroup\$ headphone cables are high-level analog. \$\endgroup\$ – markrages Feb 2 '11 at 22:15
  • \$\begingroup\$ "For the most part it either works or it doesn't work and the quality doesn't change at all" One exception is when the clock for the DAC is extracted directly from the digital signal. The digital information may get there with 100% purity in the presence of noise, but the clock does not. \$\endgroup\$ – endolith Feb 10 '11 at 17:20
  • \$\begingroup\$ @endolith: In those applications, you're supposed to have a narrow loop filter on the clock recovery PLL so the recovered clock doesn't jitter as much in the presence of noise. \$\endgroup\$ – Mike DeSimone Jul 17 '11 at 13:02
4
\$\begingroup\$

Speakers are relatively high power devices with very low impedance. This combination won't pick up any noticeable noise.
Use just any low resistance cable and you'll be fine. Low resistance is needed to achieve a good damping factor. Anybody paying 100 euros and up per meter of oxygen-free copper cable and the like just knows he's being swindled.

\$\endgroup\$
  • 1
    \$\begingroup\$ No, anybody paying 100 euros per meter probably doesn't know he's being swindled. \$\endgroup\$ – Rocketmagnet May 8 '12 at 20:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.