Speaker Wire and Calculating Guage

[DW1961]

How do you calculate what size wire you need for speakers and amps?

I am familiar with this site http://www.roger-russell.com/wire/wire.htm#wiretable

However, when I ran my original speaker wire, I didn't know it was CCA wire (Amazon brad), so now I want to redo my calculations. Also confounding my calculation is , how do I know what the resistance is for CCA wire? I've read that the skinning affect actually reduces resistance is CCA wire (since there is some copper on the outside) so it's not just like calculating for copper or aluminum.

[bob91343]

Don't overengineer it.  Just use a heavy enough gauge to avoid adding too much resistance to the circuit.  After all, it's often an 8 Ohm setup so any resistance greater than a half Ohm or so begins to take away from the output power.  Skin effect is seldom important.

16 AWG is about 0.004 Ohms per foot.  So 100 feet of cable has 200 feet of wire in it, for a resistance of around 0.8 Ohm.  That is beginning to be significant but you are unlikely to hear any difference.

[themadhippy]

How do you calculate what size wire you need for speakers and amps?

Either use ohms law , work out the current  the cables handling,decide on  the maximum acceptable  volt drop,with these you can calculate the maximum acceptable cable resistance and use the resistivity of the desired cable material to find  the  cross sectional area of the cable that  meets that.  Or just chuck in the biggest lump of copper you can find/afford,no need to go exotic,plain old 2.5mm twin n earth*   is just as good as  virgin  free oxygen dusted   audiograde cable
*romex i believe is the equivalent in foreign parts

[DW1961]

How do you calculate what size wire you need for speakers and amps?

Either use ohms law , work out the current  the cables handling,decide on  the maximum acceptable  volt drop,with these you can calculate the maximum acceptable cable resistance and use the resistivity of the desired cable material to find  the  cross sectional area of the cable that  meets that.  Or just chuck in the biggest lump of copper you can find/afford,no need to go exotic,plain old 2.5mm twin n earth*   is just as good as  virgin  free oxygen dusted   audiograde cable
*romex i believe is the equivalent in foreign parts

HAHA, yeah Romex would be fun to work with! lol Romex in the US is solid core plastic PVC coated wire. Nasty stuff unless you are running house wires.

[james_s]

I've never bothered to calculate it. 16AWG lamp cord works fine for the sort of lengths you'll find in a typical house at the sort of power levels I'd ever be interested in having. If you're dealing with unusually long lengths, really high powers or something exotic then you might want something heavier. These days amplifier power is cheap and typically the RMS will be only a few watts, I know audiophiles will cringe when I say this but a bit of loss doesn't really matter. Just don't use ridiculously thin wire like the stuff that comes with a lot of cheap consumer gear and you'll be fine.

Also, CCA is garbage. I got screwed a while back when I ordered a box of CAT5e thinking "wow this stuff has gotten cheap!" and then realized after I went to use some that it's CCA which doesn't even meet the spec to be called CAT5 in the first place. I probably should have complained and sent it back to Amazon but I never got around to it.

[DW1961]

Don't overengineer it.  Just use a heavy enough gauge to avoid adding too much resistance to the circuit.  After all, it's often an 8 Ohm setup so any resistance greater than a half Ohm or so begins to take away from the output power.  Skin effect is seldom important.

16 AWG is about 0.004 Ohms per foot.  So 100 feet of cable has 200 feet of wire in it, for a resistance of around 0.8 Ohm.  That is beginning to be significant but you are unlikely to hear any difference.

Russel is saying, "It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again."

So .05 * 8 = 0.4ohms

 But if 16AWG is only.004ohms per foot, you could run 100 feet total wire and still meet his requirements.

In his chart he has it at 48 feet--BUT--I'm not sure if he means one way or 24*2.  Here is what he says: "In the early speaker manuals, starting with the XR5, I included a chart for estimating the maximum wire lengths for various sizes of copper wire . . . . It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again. Then in his chart, he has 16AWG @ 48'.

http://www.roger-russell.com/wire/wire.htm#wiretable

Using your resistance that would be right at 100 feet total wire calculated as length x2. (48*2=96.) Is that what he is saying ?

Your resistance per foot got me thinking about this "48 foot" chart length. He must mean 48' is one way?

If so, I misread his chart when I calculated for my run thinking I had 24 feet for the run and doubled it.

[DW1961]

I've never bothered to calculate it. 16AWG lamp cord works fine for the sort of lengths you'll find in a typical house at the sort of power levels I'd ever be interested in having. If you're dealing with unusually long lengths, really high powers or something exotic then you might want something heavier. These days amplifier power is cheap and typically the RMS will be only a few watts, I know audiophiles will cringe when I say this but a bit of loss doesn't really matter. Just don't use ridiculously thin wire like the stuff that comes with a lot of cheap consumer gear and you'll be fine.

Also, CCA is garbage. I got screwed a while back when I ordered a box of CAT5e thinking "wow this stuff has gotten cheap!" and then realized after I went to use some that it's CCA which doesn't even meet the spec to be called CAT5 in the first place. I probably should have complained and sent it back to Amazon but I never got around to it.

I just watched a Youtbue video of a guy explaining what happened to you. He said the same ting you found out that it isn't even CAT5 spec! This should never be happening. It's false advertising, it's not spec, it's illegal.  It's bullshit.

I did the same thing recently. I bought a 50' roll of Amazon 16 AWG wire that was CCA but it doesn't say CCA on it. I'll just calculate aluminum wire if I ever use it again, but with pure copper 16AWG wire from Germany I found on Amazon at 100' for 20 bucks, I don't need to worry about getting F----- again with CCA garbage.

I mean it's not garbage if you use it correctly, and aluminum per weight is more conductive than copper. They use it is aircraft, you know, so it's not garbage. As consumers we need to know what we are buying.
 
I have no use or interest at all for aluminum or CCA wire--none. It just confuses things.

[themadhippy]

Romex in the US is solid core plastic PVC coated wire

yep sounds like twin n earth,2.5mm is a little  bit bigger than 14awg.

If you want something "exotic" https://www.tnt-audio.com/clinica/ffrc_e.html there a nice purple colour and the platt makes them look fancy

[james_s]

I just watched a Youtbue video of a guy explaining what happened to you. He said the same ting you found out that it isn't even CAT5 spec! This should never be happening. It's false advertising, it's not spec, it's illegal.  It's bullshit.

I did the same thing recently. I bought a 50' roll of Amazon 16 AWG wire that was CCA but it doesn't say CCA on it. I'll just calculate aluminum wire if I ever use it again, but with pure copper 16AWG wire from Germany I found on Amazon at 100' for 20 bucks, I don't need to worry about getting F----- again with CCA garbage.

I mean it's not garbage if you use it correctly, and aluminum per weight is more conductive than copper. They use it is aircraft, you know, so it's not garbage. As consumers we need to know what we are buying.
 
I have no use or interest at all for aluminum or CCA wire--none. It just confuses things.

Aluminum wire has uses, but CCA CAT5 (and power cord) is garbage. The stuff I bought did say CCA on the box but it didn't even register to me what that meant since CAT5 is by definition copper wire. It works fine for data aside from being fragile but I wanted it specifically for a POE application which the higher resistance makes CCA completely unsuitable for.

[Bassman59]

How do you calculate what size wire you need for speakers and amps?

I am familiar with this site http://www.roger-russell.com/wire/wire.htm#wiretable

However, when I ran my original speaker wire, I didn't know it was CCA wire (Amazon brad), so now I want to redo my calculations. Also confounding my calculation is , how do I know what the resistance is for CCA wire? I've read that the skinning affect actually reduces resistance is CCA wire (since there is some copper on the outside) so it's not just like calculating for copper or aluminum.

Dunno what sort of speakers or amps you are using, but in the professional audio world, we use standard SO or SJO power cordage, from 16 AWG up to 12 AWG. Maybe 50 cents a foot for the common 14/2 cable.

[themadhippy]

but in the professional audio world, we use standard SO or SJO power cordage,

you might,but  van damme is more  common over here,and  it coils much easier than flex

[DW1961]

How do you calculate what size wire you need for speakers and amps?

I am familiar with this site

However, when I ran my original speaker wire, I didn't know it was CCA wire (Amazon brad), so now I want to redo my calculations. Also confounding my calculation is , how do I know what the resistance is for CCA wire? I've read that the skinning affect actually reduces resistance is CCA wire (since there is some copper on the outside) so it's not just like calculating for copper or aluminum.

Dunno what sort of speakers or amps you are using, but in the professional audio world, we use standard SO or SJO power cordage, from 16 AWG up to 12 AWG. Maybe 50 cents a foot for the common 14/2 cable.

So in Russel's web page, does he mean 48 feet one way for 16 AWG copper?

"In the early speaker manuals, starting with the XR5, I included a chart for estimating the maximum wire lengths for various sizes of copper wire needed for 4 and 8 ohm loads. I have expanded it on this page to include 2 and 6 ohm loads as well. It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again. "

Then teh table has:

16 AWG Max. 48'

I think he must mean 48 feet one way judging by Bob's comment above, '16 AWG is about 0.004 Ohms per foot.  So 100 feet of cable has 200 feet of wire in it, for a resistance of around 0.8 Ohm."

Given that, 50' would be .4 Ohm, and that is what Russel must mean--48' one way includes both wires?

If so, even though I have that crappy Amazon CCA, my 16AWG run is 24' one way, so I'm WAY under that .4 Ohm limit, right?

[Bassman59]

How do you calculate what size wire you need for speakers and amps?

I am familiar with this site

However, when I ran my original speaker wire, I didn't know it was CCA wire (Amazon brad), so now I want to redo my calculations. Also confounding my calculation is , how do I know what the resistance is for CCA wire? I've read that the skinning affect actually reduces resistance is CCA wire (since there is some copper on the outside) so it's not just like calculating for copper or aluminum.

Dunno what sort of speakers or amps you are using, but in the professional audio world, we use standard SO or SJO power cordage, from 16 AWG up to 12 AWG. Maybe 50 cents a foot for the common 14/2 cable.

So in Russel's web page, does he mean 48 feet one way for 16 AWG copper?

"In the early speaker manuals, starting with the XR5, I included a chart for estimating the maximum wire lengths for various sizes of copper wire needed for 4 and 8 ohm loads. I have expanded it on this page to include 2 and 6 ohm loads as well. It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again. "

Then teh table has:

16 AWG Max. 48'

I think he must mean 48 feet one way judging by Bob's comment above, '16 AWG is about 0.004 Ohms per foot.  So 100 feet of cable has 200 feet of wire in it, for a resistance of around 0.8 Ohm."

Given that, 50' would be .4 Ohm, and that is what Russel must mean--48' one way includes both wires?

If so, even though I have that crappy Amazon CCA, my 16AWG run is 24' one way, so I'm WAY under that .4 Ohm limit, right?

How about telling us your application. Are you building a mastering studio or are you setting up speakers in your living room?

How long are your speaker cable runs?

For most practical purposes, 16 AWG is more than sufficient for home-stereo listening. Skin effect is entirely irrelevant. Damping factor matters only in the limit, like if you were driving 18" subwoofers for an EDM show.

[DW1961]

How do you calculate what size wire you need for speakers and amps?

I am familiar with this site

However, when I ran my original speaker wire, I didn't know it was CCA wire (Amazon brad), so now I want to redo my calculations. Also confounding my calculation is , how do I know what the resistance is for CCA wire? I've read that the skinning affect actually reduces resistance is CCA wire (since there is some copper on the outside) so it's not just like calculating for copper or aluminum.

Dunno what sort of speakers or amps you are using, but in the professional audio world, we use standard SO or SJO power cordage, from 16 AWG up to 12 AWG. Maybe 50 cents a foot for the common 14/2 cable.

So in Russel's web page, does he mean 48 feet one way for 16 AWG copper?

"In the early speaker manuals, starting with the XR5, I included a chart for estimating the maximum wire lengths for various sizes of copper wire needed for 4 and 8 ohm loads. I have expanded it on this page to include 2 and 6 ohm loads as well. It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again. "

Then teh table has:

16 AWG Max. 48'

I think he must mean 48 feet one way judging by Bob's comment above, '16 AWG is about 0.004 Ohms per foot.  So 100 feet of cable has 200 feet of wire in it, for a resistance of around 0.8 Ohm."

Given that, 50' would be .4 Ohm, and that is what Russel must mean--48' one way includes both wires?

If so, even though I have that crappy Amazon CCA, my 16AWG run is 24' one way, so I'm WAY under that .4 Ohm limit, right?

How about telling us your application. Are you building a mastering studio or are you setting up speakers in your living room?

How long are your speaker cable runs?

For most practical purposes, 16 AWG is more than sufficient for home-stereo listening. Skin effect is entirely irrelevant. Damping factor matters only in the limit, like if you were driving 18" subwoofers for an EDM show.

The run is 24' one way.
It's just a home system with RM51 Klipsch speakers and a Chinese D amp rated at 100 watts RMS at 4 Ohms. The Klipsch speakers are SENSITIVITY 93 dB @ 2.83V/1M. So that really helps.
Klipsch Specs: https://f072605def1c9a5ef179-a0bc3fbf1884fc0965506ae2b946e1cd.ssl.cf2.rackcdn.com/product-specsheets/R-51M_Spec-Sheet_v01.pdf
ThIs is the little Amp:
https://www.amazon.com/Nobsound-Bluetooth-Amplifier-Audio-Receiver/dp/B07QNJPC31/ref=sxts_sxwds-bia-wc-nc-drs1_0?cv_ct_cx=Nobsound&dchild=1&keywords=Nobsound&pd_rd_i=B07QNJPC31&pd_rd_r=07857196-922e-4acf-9f6c-c01cca9d82da&pd_rd_w=FX7ab&pd_rd_wg=PQrhL&pf_rd_p=43f4b3f0-0b04-46ba-8a08-2e851d035e17&pf_rd_r=D4751P6417T95SNKSDM7&psc=1&qid=1595447770&sr=1-1-f3947b35-9c59-4d7a-9603-b751e6eed25b

I get what you're saying about needs for power and bass etc. 

I never exceed 50% power level on the D Amp, as I mainly listen to Jazz at lower levels (I set the Amp at 50% then adjust the volume from my computer software So if I turn up the software volume to 100% the Amp is doing 50% power. I should probably turn it up to 80% and use a lower software volume, but I don't know if that matters or not.) The Amp uses the Texas Instruments 3116D amp chip. Since the speaker are 8 Ohms I'm only getting 25 Watts per channel, unless I misunderstand power and Ohms. Still sounds good for my needs.

If this is correct below, then even with CCA wire, I should be golden:

"Russel is saying, "It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again."

So .05 * 8 = 0.4ohms

 But if 16AWG is only .004ohms per foot, you could run 100 feet total wire and still meet his requirements? And that would mean his statement is 48 feet two wires one way?"

[TimFox]

He's talking about a two-wire cable, where the total wire length is twice the cable length.
Back in sixth grade arithmetic, we had a trick question where we had to know that the required length of rail is twice the length of the track in a railroad.

[Bassman59]

The run is 24' one way.
It's just a home system with RM51 Klipsch speakers and a Chinese D amp rated at 100 watts RMS at 4 Ohms. The Klipsch speakers are SENSITIVITY 93 dB @ 2.83V/1M. So that really helps.

Sensitivity just tells you the output you can expect to get for a given power input. Higher sensitivity speakers are more efficient. But it's not interesting for sizing wires.

I never exceed 50% power level on the D Amp, as I mainly listen to Jazz at lower levels (I set the Amp at 50% then adjust the volume from my computer software So if I turn up the software volume to 100% the Amp is doing 50% power. I should probably turn it up to 80% and use a lower software volume, but I don't know if that matters or not.) The Amp uses the Texas Instruments 3116D amp chip. Since the speaker are 8 Ohms I'm only getting 25 Watts per channel, unless I misunderstand power and Ohms. Still sounds good for my needs.

A "reasonable" way to set levels is to run the software level at maximum and then set the amplifier to "somewhat louder than you'd ever listen to it." Then you can use the software controls to set the computer/whatever output level to whatever is comfortable. This way, if the computer went wonky for some reason, the level won't blow your ears or your speakers.

If this is correct below, then even with CCA wire, I should be golden:

"Russel is saying, "It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again."

So .05 * 8 = 0.4ohms

 But if 16AWG is only .004ohms per foot, you could run 100 feet total wire and still meet his requirements? And that would mean his statement is 48 feet two wires one way?"

You're still waaaaaaaay overthinking it.

The losses in the wire for the short distance you describe is not going to matter. I mean, you're not using 30 AWG wire-wrap wire for this, right?

[donlisms]

As I understand it, the issue with speaker cables, particular for subwoofers, is damping factor, rather than just power loss.  Damping factor is the ratio of the source resistance to the load resistance.  We want the highest value possible, e.g. around 40 would be a reasonable minimum if the goal is tight, proper transients in the bass.

(I may not be able to explain this very well, but this is the kind of thinking I live with...)

As the driver voice coil moves according to the applied voltage, because it's a moving coil in a magentic field, it becomes a generator, essentially producing voltage back toward the amplifer.  If this back EMF is shorted out by nearly zero ohms looking back toward the amp, it does not contribute to further motion of the driver; the voice coil is "stuck" at the position it was driven to by the input signal.  You can observe the effect by shorting out a generator or DC motor and trying to turn it, and then again with the connection open.  It will be obviously stiffer while shorted.

If the back EMF is not shorted out because the output resistance it sees is too high, it becomes part of the input signal again, and the driver is permitted to oscillate at the appropriate resonant frequency.  This becomes ncone motion that was not originally requested.  The result is that the kick drum, for example, instead of sounding like "Puh!" will now sound like "Fffluhh!".  The volume is not the issue; power is not the issue.  It is that the transient response is affected.  The punchiness and snappiness of the low end sound is degraded.

If we assume the damping factor should be 40 or better, just to pick a number, then we'd want the output impedence of the amplifier and wires combined to be less than 8 / 40, or 0.2 ohms, total of amplifier and wire (and crossover... and... ugh).

The issue of damping factor was (is?) quite prominent with tube amplifers, where the output impedence is generally not as high as can be easily obtained with solid state, due to the Big Old Transformer and so on.

Me (bass player / sound guy), I hate fluffy bass.  There are of course many other factors that can contribute to fluffy bass... we hates 'em all.   

[TimFox]

You should include the DC resistance of the voice coil in your circuit to determine the damping factor:  it is in series with the cable resistance and amplifier output resistance.  The DC resistance of the voice coil is, of course, less than the speaker impedance, which should be dominated by the electromechanical result of coupling the voice coil through the cone to the compressible air.

[DW1961]

He's talking about a two-wire cable, where the total wire length is twice the cable length.
Back in sixth grade arithmetic, we had a trick question where we had to know that the required length of rail is twice the length of the track in a railroad.

So it's how I thought it is then, 96 total feet of wire for his 48' run, which is just under a .4 ohm loss?

[DW1961]

You should include the DC resistance of the voice coil in your circuit to determine the damping factor:  it is in series with the cable resistance and amplifier output resistance.  The DC resistance of the voice coil is, of course, less than the speaker impedance, which should be dominated by the electromechanical result of coupling the voice coil through the cone to the compressible air.

Tim,

Yeah, I knew about the voice coil also, but I didn't account for that. It is, after all, wire too. But, no, I didn't.

[DW1961]

The run is 24' one way.
It's just a home system with RM51 Klipsch speakers and a Chinese D amp rated at 100 watts RMS at 4 Ohms. The Klipsch speakers are SENSITIVITY 93 dB @ 2.83V/1M. So that really helps.

Sensitivity just tells you the output you can expect to get for a given power input. Higher sensitivity speakers are more efficient. But it's not interesting for sizing wires.

I never exceed 50% power level on the D Amp, as I mainly listen to Jazz at lower levels (I set the Amp at 50% then adjust the volume from my computer software So if I turn up the software volume to 100% the Amp is doing 50% power. I should probably turn it up to 80% and use a lower software volume, but I don't know if that matters or not.) The Amp uses the Texas Instruments 3116D amp chip. Since the speaker are 8 Ohms I'm only getting 25 Watts per channel, unless I misunderstand power and Ohms. Still sounds good for my needs.

A "reasonable" way to set levels is to run the software level at maximum and then set the amplifier to "somewhat louder than you'd ever listen to it." Then you can use the software controls to set the computer/whatever output level to whatever is comfortable. This way, if the computer went wonky for some reason, the level won't blow your ears or your speakers.

If this is correct below, then even with CCA wire, I should be golden:

"Russel is saying, "It was based on the resistance of the speaker wire not exceeding 5% of the rated impedance of the system. The wire length is for TWO-CONDUCTOR wire. This includes one wire out to the speaker and one wire back again."

So .05 * 8 = 0.4ohms

 But if 16AWG is only .004ohms per foot, you could run 100 feet total wire and still meet his requirements? And that would mean his statement is 48 feet two wires one way?"

You're still waaaaaaaay overthinking it.

The losses in the wire for the short distance you describe is not going to matter. I mean, you're not using 30 AWG wire-wrap wire for this, right?

"A "reasonable" way to set levels is to run the software level at maximum and then set the amplifier to "somewhat louder than you'd ever listen to it." Then you can use the software controls to set the computer/whatever output level to whatever is comfortable. This way, if the computer went wonky for some reason, the level won't blow your ears or your speakers."

Yeah, definitely, I think I will do that.

"You're still waaaaaaaay overthinking it.

The losses in the wire for the short distance you describe is not going to matter. I mean, you're not using 30 AWG wire-wrap wire for this, right?"

Yeah, but I want to know how to calculate it myself. That's the point of starting the thread. And no, I'm suing 16AWG Amazon CCA wire.