2

Recently a friend of mine spent several tens of euros to buy a set of gold connectors to connect to the audio amplifier, justifying this expense on the grounds that they were gold and therefore the signal propagated better in the cables (the final sound would therefore have better quality).

What I thought is:

The EM waves carrying the signal (simplified with the terms current and voltage) propagate into the twisted cables/wires insulation. That is, the wires are armatures (electromagnetic guides) that keep the EM field confined inside, all the way to its destination. This means that if you want to have a very good quality of the received signal, you must have less dispersion and therefore a very high quality of the insulator - so spend money on well-insulated cables.

  1. Is my reasoning correct?
  2. How important is the type of wire material (gold, copper, etc.) and what does it affect instead? The speed of the signal?
  3. What are the differences between a copper wire with gold connections at its ends or a gold "cable"?
winny
  • 13,064
  • 6
  • 46
  • 63
KaleM
  • 327
  • 7
  • `a set of gold connectors to connect to the audio amplifier` ... `type of wire material (gold, copper, etc.)` is the question about a copper wire having gold connections at its ends or a gold "cable"? – Rohat Kılıç Jun 20 '23 at 08:22
  • Good point! I actually wouldn't know, so I'm taking the opportunity to also ask for an explanation of this difference (updated post) – KaleM Jun 20 '23 at 08:40
  • On the related issue of cable from the amp to the speakers: https://electronics.stackexchange.com/questions/9671 – jonathanjo Jun 20 '23 at 08:45
  • 7
    There is a significant difference in your enjoyment of sound between buying romex/twin'n'earth for speaker cables and specialist linear-crystal oxygen-free grain-oriented pasture-fed audiophool gold plated cables, as the latter will deplete your budget that could otherwise be spent on better speakers and more audio material. Gold plated connectors, there's a good argument for those, just don't go overboard on the price. – Neil_UK Jun 20 '23 at 10:15
  • 4
    You left off "phase stabilized" in your description of good audio cables, – SteveSh Jun 20 '23 at 12:18

4 Answers4

9

For low level signal connections, it can be useful to have connectors plated with materials that either don't oxidize (gold) or have an oxide which is conductive (silver). For the same reason you can find gold or silver plating inside signal relays, switch contacts, RF connectors (SMA, BNC, etc).

Even for high level connections (speakers) this can be useful (Speakons are silver plated, and audiophiles love their gold plated bananas).

Speakons are of course the better option because you can't get the polarity wrong, and you can leave the speaker end of the cord disconnected on the floor without the two bananas shorting together and blowing up the amplifier (audiophiles consider protection circuits to be of bad taste).

The opposite end of the spectrum is encountered on old style loudspeakers and amps: a bare stranded copper wire stuck inside a hole and held with a spring contact. Unlike a proper wago, these contacts are terrible, and copper oxidizes readily, which means sometimes the contact is just bad. I don't mean "a slight veil in front of the speakers" bad, I mean scratching sounds in the speakers from intermittent contact kind of bad.

Now in your cable, about the material of the wires, there's been some debate, but so far I haven't seen anything that could be qualified as "scientific".

The insulator though, it really matters. If you wire a microphone on a stage and your cable sucks, then it will make a loud THUMP in the speakers every time someone steps on it. Some insulators generate charge when pressed or rubbed together (triboelectricity) and in addition to that, if there is DC bias like 48 V phantom power, then when the cable gets squeezed there is a change in its capacitance which also creates a signal. All these manifest as charge, in other words current, so they are most observable on low level signals from high impedance sources (i.e., some microphones). If the cable is driven at volt levels by an opamp with an output impedance of 50 ohms, good luck observing anything. If the cable is driven at millivolt levels by an instrument with an output impedance of 10 MΩ, then just rubbing a finger on it will probably generate an observable signal. This isn't just for audio, but for any kind of acquisition involving low levels and high impedance, like EEG/ECG. Special cables are manufactured for these applications: without these all the measurements would go YOLO every time the patient moves and the cable bends.

This only becomes audiophile bullshit if you take the second case (high impedance + mV) and apply it to the first case (low impedance + Volts).

Another thing about audio cables is that for cost reasons most home equipment uses unbalanced connections which are just terrible. There's always some current in the shield and it's also the ground which is the signal reference, so any voltage drop gets added to the signal, and "how it sounds" can thus depend on the resistance of the "shield" ground connection. It's much better to use balanced connections where the shield carries stray current but is not the voltage reference. Good quality balanced cables can be had at low prices.

winny
  • 13,064
  • 6
  • 46
  • 63
bobflux
  • 70,433
  • 3
  • 83
  • 203
  • Forgive my ignorance, what do you mean with "low level signal connections"? Some examples? What happens at the semi-microscopic level if the material (copper for example) oxidizes? – KaleM Jun 20 '23 at 16:33
  • "signal" means everything that doesn't carry high current, in audio that's everything except loudspeaker and headphone jacks. If the connection is bad you can get extra resistance, if it is very bad and oxidized you can get a point contact diode. – bobflux Jun 20 '23 at 16:41
  • 1
    Low level mean low voltage. You need enough voltage to break through the oxide that forms on copper. That's not a problem with a gold plated contact. I'm not sure exactly where the breakpoint is between needing gold and copper. But for mV or hundreds of mV stuff you want gold. For 10's of volts copper works OK. – SteveSh Jun 20 '23 at 18:36
  • I love my gold plated bananas. And not only that, I also use gold plated HDMI cables for my monitors, since I have to move them around a lot, they keep getting connected, disconnected and spend some time "hanging" on the air, I love the fact that they don't get a bit "rusty". They are a little bit more expensive but if you move them a lot, most connectors can benefit from a little bit of gold plating I think! – Benito-B Jun 21 '23 at 07:47
5

As @bobflux writes, gold plating is beneficial for low level signals since gold doesn't oxidize, with exceptions...

Gold on gold works well. Gold on tin is a disaster due to the intermetallics that form when gold is in contact with tin. The intermetallic formed on the gold forms an insulating layer that is hard to remove. Beginner audiophools will make this mistake with patch cables that use gold-plated contacts only to mate them to a connector on their equipment using tin-plated connectors.

10s of euros is pretty cheap for a good patch cable, just make sure the contacting surfaces are compatible.

qrk
  • 7,474
  • 1
  • 5
  • 20
  • I make a lot of custom cable assemblies for prototypes, and I use both gold-plated contacts and tin-plated contacts. So now you've got me wondering if I should avoid mating the two metals together, and also avoid crimping a gold-plated crimp barrel onto a tinned wire. – Cassie Swett Jun 21 '23 at 03:45
  • @CassieSwett Avoiding dissimilar metals if the product needs to last more than a year would be a wise choice. There is lots of information on the Web about gold-tin intermetallics. Selective plating is normally available for many connector styles. – qrk Jun 21 '23 at 13:46
4

Addressing just one point of your reasoning:

The transmission line (EM wave interaction with conductors and insulators) reasoning does apply to cables in general, but is wholly irrelevant at low (audio) frequencies, and the cable lengths used in AV systems; and any effects are only just perceptible at analog video frequencies (i.e., some horizontal sharpness or color accuracy may be missing with the poorest quality PVC (lossy dielectric) cables).

Shield quality is a somewhat different issue, which can affect audio and picture quality in terms of interference from external noise sources. Shield quality depends on the braiding density, foil coverage, and number of layers of each. Or for balanced feeds, it depends on the quality of that balance (both of cable construction, and the transmitter/receiver circuits thus connected), and the use of shielding (optional, but preferred in very noisy environments).

Materials will generally be low-loss, i.e. bare copper, or copper or silver plated steel or aluminum for conductors, and PVC, PE, PP and PTFE for dielectrics (again, with PVC not mattering for audio, and barely mattering for analog video).

At low frequency, loss is dominated by wire DC resistance. Video is generally matched-impedance so the overall line loss matters, while line-level audio is generally high impedance and so wire resistance has little effect.

(I should probably add a disclaimer that I mean NTSC/PAL conventional analog video; I am aware of professional analog formats with much higher bandwidths, which will therefore be more critical of cable quality. They may also need to be transmitted over longer distances, i.e. throughout a studio. I suspect the number of remaining applications is few and far between, so I doubt it is worth doing more than mentioning these for completeness sake. I would also expect that few professionals working with such a setup would be looking for related information in this thread.)

There is of course no effect on digital audio or video, so long as the cable is delivering a recoverable signal at all. Which needs to be at least some minimum quality say to handle HDMI, and such cables are comparable to Ethernet CAT-5 for example, in several characteristics.

Tim Williams
  • 22,874
  • 1
  • 20
  • 71
  • The effect is present on audio cables as well. You can likely hear a guitarist complaining about too long cable attenuating high frequencies and I once bought maybe 10 meter audio cable between computer and amplifier. The cables have capacitance and it will form an RC filter with the output impedance. Electric guitars have quite high output impedance so cable capacitance will make a difference. – Justme Jun 20 '23 at 18:31
  • @Justme Yes -- a consequence of the high impedance signals, and a lumped-element consequence; no transmission line theory needed. – Tim Williams Jun 20 '23 at 19:38
0

The EM waves carrying the signal (simplified with the terms current and voltage) propagate into the twisted cables/wires insulation. That is, the wires are armatures (electromagnetic guides) that keep the EM field confined inside, all the way to its destination.

This is basically along the right lines. As well as keeping out induced unwanted signals. That's the design theory behind twisted-pair network cable, or coaxial antenna cable, for example.

A side effect is that the dielectric constant of the insulator can affect signal speed. Not that it makes any significant difference for audio.

pjc50
  • 46,540
  • 4
  • 64
  • 126