Does wire size affect signal quality...

[donnie7]

PP Forum,

Was wondering if wire size at line level, has anything to do with signal quality/gradation? The size of wire connecting from the pins of an opamp to its chip, is ten times smaller than a human hair. Could this be why opamps sound different than hardwired components? Does the size of the pipe/wire matter in line level audio transmission? Curious to know what you all think/know. Thank you all!

 

[Ethan]

The short answer would be YES.
The thicker the wire, the less resistance and the more current is able to travel through it. The longer the line the more the side-effects will become apparent. Capacitance becomes an issue with higher impedenace as well--i.e. high frequency loss.

>>The size of wire connecting from the pins of an opamp to its chip, is ten times smaller than a human hair. Could this be why opamps sound different than hardwired components?
Be more specific about what you mean by "hardwired components".
Nevertheless, my presumptuous short anwer to that question would be NO.

 

[donnie7]

By hardwired components, I mean older point to point wiring. Including the thicker PCB traces made in the 60/70/80's. Just wondering if those teeny wires connecting opamp pins/leads to the IC chip could stifle/constrain the sound? Thanks for your reply.

 

[SSLtech]

Thick conductors are needed for high current. You can prove this by wiring your speakers with magnet wire (human har thickness or less) then hearing the disastroous efect that louder signals suffer.

Impedance determines gurrent for a given signal level.

Chips are (ordinarily) very high impedance, indeed the power dissipated by most chips is rather small. For this reason, not much current actually flows through most pins. (ignoring IC power amps of course...)

So you ask a question that has no simple "Yes" or "No" answer, because it is overly simplistic.

Ohms law governs this (or Kirchoff's law as a variant). Ohms law is very simple, but requires that at least two variables be specified in order to provide an answer. If you tell me the voltage and the impedance, I'll tell you if you need a thick wire.

Most IC pro audio stuff (as a rule) is high input impedance, low effective source impedance. If you're asking if these need thick pins to do the job any better the answer is no.

At work I have a rack of amps for the main room I'm currently (no pun intended) working on. It is fed from four 30-Amp breakers at 120V. The load impedances which the amplifiers see are a couple of ohms. The wire needs to be thicker than a weightlifter's forearm because of the biblical current that these amps can be asked to produce.

So in this case, the answer is yes.

Your question specifies no condition, so no accurate answer can be given, however, I hope you can see that I think the answer that you seek is 'No'.

Keith

 

[Ethan]

Opamp designers are pretty clever people (much more so than I will ever be), and they know how "thick or thin" a wire connection needs to be. With that said...Some wires just don't "need" to be thick. In fact, some wires are better off being thinner to a certain extent. Mic level or even line level audio doesn't require super thick wire/traces. If you're thinking of the difference in the sound of opamps and say...vacuum tubes, there are many other factors at play. Russell O. Hamm wrote a paper back in the 70s investigating this 'phenomenon'. Google it you should be able to find it with little effort. It's still a debated topic but most of the arguments seem to revolve around distortion characteristics.

 

[Ethan]

OK Keef is talking now....Ethan will shut his mouf.

 

[soundguy]

since we are being general, I would think that the design of a circuit has much more influence on the sound of things than the diameter of the wire used in that circuit... There are designs that I have heard that use chips which sound fantastic. There are others that dont. There are discrete designs that sound fantastic, there are others that dont. Sure, components add up to play a huge part, but to make a blanket statement pointed at one thing like the wire inside an opamp probably isnt the best thing. Then again, perhaps it IS the case and designers have to work harder to make a chip sound good but regardless, it CAN be done, so in the end I think design trumps all.

dave

 

[PRR]

> Was wondering if wire size at line level, has anything to do with signal quality/gradation?

Use #18 wire for rock-n-roll, #16 wire for classical.

> The size of wire connecting from the pins of an opamp to its chip, is ten times smaller than a human hair. Could this be why opamps sound different than hardwired components?

The guts inside transistors are wired with the same hair-fine bond-wires.

For Audio systems from Hi-Fi to Studio, the concept you really want is "thickness to length ratio". If a wire is short, it can be thin and make no difference. If a wire is long, it may need to be fatter. But consider the complete circuit. In an interconnect, the input impedance of the next box is maybe 10K ohms. This is equivalent to 10,000 feet of very-thin #40 wire (0.003", 0.080mm). Clearly another few feet of #40 makes no significant difference to the electrons. However a loudspeaker voice coil has to be made of a short length of very sturdy wire, and tends to be 2Ω to 50Ω. 10 feet of #40 wire will double the total resistance, halve the voltage on the speaker, and reduce power to 1/4 which is -6dB and significantly softer than the speaker was designed to play.

> wiring your speakers with magnet wire (human har thickness or less) then hearing the disastroous efect that louder signals suffer.

Not "louder signals". ALL signals: the effect is perfectly linear.

If the speaker were merely a resistor, it would just play softer. But speakers are "lumpy resistors": 8Ω at 400Hz, 50Ω at 80Hz, 20Ω at 10KHz, with possible other lumps and dips at crossovers. So 8 feet of #40 wire would give 6dB less in the midrange, only 1dB less at 80Hz and 2dB less at 10KHz. If it was "flat response", it is now un-flat. At all power levels.

I've assumed common audio practice where heat in the wire is totally negligible. The only case where this is not commonly true is the wire in a speaker voice coil. Speakers do melt; with modern epoxy they hang together long enough for copper resistance to rise. Most "600W" speakers will double their resistance on a 600W test signal (so they actually take 300W, and play 3dB softer). If you wired such a speaker and amp with #40 wire, the #40 would get very hot and might melt before the speaker did. But that's stupid. Speakers "have" to use thin wire because of mass and magnet limits: there isn't much space in the magnet gap. But we have plenty of space in the room or conduit for wire far larger (and better cooled) than the stuff used in the speaker. #18 wire will not get hot even in severe speaker abuse. (You may need a larger size to reduce power loss and response shift in very long runs, but not to keep from melting.)

 

[SSLtech]

[quote author="PRR"]Use #18 wire for rock-n-roll, #16 wire for classical.[/quote]
Just trying to reconcile this suggestion with this observation:[quote author="PRR"]
> wiring your speakers with magnet wire (human har thickness or less) then hearing the disastroous efect that louder signals suffer.

Not "louder signals". ALL signals: the effect is perfectly linear.[/quote]

:wink:

Keith

 

[andre tchmil]

speaking about cables .....

How long can a speaker cable be, running from my 100w marshall amp to my cabinet.

And another question.
I have some goood quality signal cable running from one room to another , about 10 meter long.
Can I wrap all wires (+/-/ground) and use this as one side for a speaker cable ?

 

[NewYorkDave]

How long is a piece of string?

:wink:

It depends on the gauge of the wire. Nobody has mentioned it yet, but there's a handy reference tool called a "wire chart." Here's an example:
http://www.cm-corp.com/dgStrandedChart.htm

I chose a stranded wire chart because stranded wire is what we usually use for speakers.

A good rule of thumb to follow is that the wire resistance should be much, much less than the DC resistance of the speaker voice coil. (My personal practice is to try to keep it within 1/100 of the speaker's DCR--you don't want any significant power to be dissipated in the wire!). An 8-ohm speaker cabinet has a DC resistance of about 6 ohms.

The DC resistance figures shown in the wire chart are for Ohms per 1000 feet. Suppose you're using 14-gauge stranded speaker cable. The DCR for 1000 feet is about 3 ohms.
3 / 1000 = 3 milliOhms per foot.
It would be wise to keep cable resistance below 60 milliOhms. In that case, you could use up to 20 feet of cable.

 

[BYacey]

Is there a certain type wire that works better for Polka Music? Perhaps a nice houndstooth insulation? Hmm...

 

[Kev]

[quote author="NewYorkDave"]How long is a piece of string?[/quote]

indeed
:roll:

long enough ... unless it is too short at one end.

 

[crazydoc]

On a slightly different tack, though still to do with wire size and signal quality, I found this posted on the web site of an audio designer/distributor:

Litz Cables.....What the heck is that?

--------------------------------------------------------------------------------

Well, a Litz Cable is a new patent pending cable by *.... We are only offering them in a 3pin and 7pin for now. We think these cables are better than any Blue cable, Monster, Mogami, or other expensive cable. Of course that is our opinion, but you would expect us to say that.

They will start showing up in stores soon, and really are great. What makes them so great you ask???? Well to begin...They incorporate:

? Hybrid Litz/Solid Core Construction for Maximum Resolution and Preservation of Sonic Detail(each audio pin has a solid copper wire paired with a stranded wire).
? Proprietary Patent Applied for Construction
? 99.99% Purity Oxygen Free Copper Conductors
? Quad Balanced Construction for Maximum Noise Rejection, and Wide Bandwidth
? Double Shielded for Maximum Freedom from Noise and Interference
? Carbon Impregnated Dielectric for Low Impedance Microphonics
? Corrosion Resistant Gold Plated Connector Contacts for Best Possible Connections
? Precision Formulated PVC Jacket for Maximum Durability and Long Life
? Polyethylene Dielectric Material for Low Loss and Superior Sound Quality

Litz construction means that each conductor strand is individually insulated. In most cables this is not the case, and what happens is the signal can randomly jump from conductor to conductor in an uncontrolled manner. Distortion results in phase and time domain, and although this may not be easily measurable, it can certainly be heard by the most sensitive instrument, the human ear.

So why not just use all Litz conductors of the same gauge, and leave it at that? Simply put, finer conductors tend to favor the high frequencies, and heavier conductors are better for the midrange and bass.The * Cables use a specifically designed combination of heavier conductors and fine gauge conductors, which gives the most realistic balance throughout the frequency extremes. This is combined in a Quad Balanced configuration that gives maximum rejection of noise, and gives you a wide bandwidth natural sound that results in a big improvement in overall sound quality compared to conventional microphone cables.

I understand there are issues at radio frequencies, but am unaware of any significant effect on audio signals.
Does anyone have experience/knowledge in this area to enlighten me?

 

[NewYorkDave]

Short answer: it's ********.

I actually laughed out loud when I read that piece of copy.

 

[BYacey]

Finer wire favoring high frequencies?? As you go up in frequency in the RF region the larger the diameter the better, due to skin effect, but eddy current losses increase. In fact in the old days they use to wind transmitter tank coils with copper tubing. With Litz wire it's not just the individual conductors, but the fact that they are woven in an entwined braid minimizes eddy current losses in the wire. This is useful to make high Q coils for working in the RF spectrum. Audio frequencies have little to no benefit from using Litz wire; no matter what some audiofools will tell you.

 

[cayocosta]

Speaker wire gauge calculator:

http://www.webervst.com/gauge.htm

 

[Ptownkid]

So what gauge hook-up wire do you all use?

 

[crazydoc]

[quote author="NewYorkDave"]Short answer: it's ********.[/quote]
That was my knee-jerk reaction too. But since I'd never heard of it being tried, I was wondering if anyone had.

History is full of the formerly impossible now possible.

 

[BYacey]

According to that calculator It says I should be using 50' of #5 AWG for a 4 ohm load at 500W power. I haven't seen much 5 guage wire used with touring production companies.