
AVS Forum  Home Theater Discussions And Reviews
(https://www.avsforum.com/forum/)
 Speakers
(https://www.avsforum.com/forum/89speakers/)
  Perspective  Speaker Wire  How much is enough?
(https://www.avsforum.com/forum/89speakers/2732169perspectivespeakerwirehowmuchenough.html)
Perspective  Speaker Wire  How much is enough?
What size speaker wire do you really need?
This question can be answered from several perspectives. One is simply Resistance/UnitLength. Practically applied you want the Resistive Losses to not exceed a certain percent of the Speaker Impedance. Charts for this can be found here  https://www.avsforum.com/forum/89spe...tml?highlight= Another method, though only more rarely considered, would be to calculate the impact of Inductance and Capacitance, and combine that with Resistive loses to get Total Losses. I touched on that subject at the link below (mathy). But at normal lengths, Inductance and Capacitance don't really have any impact in the audio range, which is why they are so rarely considered  https://www.avsforum.com/forum/89spe...rspective.html A recent discussion has lead me to consider speaker wire from a new perspective. How much Power can each wire functionally handle? That was the question I asked myself. Here is my analysis. Here is where I got the data for the Current carrying capacity of common wire sizes  http://www.powerstream.com/Wire_Size.htm Now to make this work, I had to do a bit of fudging of the numbers, or at least reinterpreting of them. On the chart we have "Maximum Amps for Chassis Wiring" which is the largest Amperage value you will find. I'm going to treat this as the Maximum Peak Current. Then we have the common working Current Rating for use in household circuits or in extension cords. I'm going to consider that the Peak Working Current. Then we have, for the Power Stream Chart, "Maximum Amps for Power Transmission", which will be the lowest rated Amperage value. I will consider this, for the purpose of this illustration, Sustained Current. Now, one last fudge to reframe this into a framework we can all understand relative to speakers, amplifiers, and wire. I'm going to push those rated Currents through an 8 Ohm Load to reframe things in Watts. While the result are not, and are not intended to be, definitive, they serve as an illustration to put various wire gauges into perspective. These are the results of my calculations  18ga wire would break down like this  Power = Current² x Resistance = 2.3² x 8 = 5.29 x 8 = 42.32 sustained watts of power Power = Current² x Resistance = 10² x 8 = 100 x 8 = 800 peak working watts Power = Current² x Resistance = 16² x 8 = 256 x 8 = 2048 max peak watts of power NOTE: It is difficult to determine the working current for 18ga wire because they don't make 18ga extension cords any more. So, I looked at some specs, and made a best guess. 16ga wire would break down like this  Power = Current² x Resistance = 3.7² x 8 = 13.69 x 8 = 109.52 sustained watts of power Power = Current² x Resistance = 12² x 8 = 144 x 8 = 1152 peak working watts Power = Current² x Resistance = 22² x 8 = 484 x 8 = 3872 max peak watts of power 14ga wire would look like this  Power = Current² x Resistance = 5.9² x 8 = 34.81 x 8 = 279 sustained watts of power Power = Current² x Resistance = 15² x 8 = 225 x 8 = 1800 peak working watts of power Power = Current² x Resistance = 32² x 8 = 1024 x 8 = 8192 max peak watts of power 12ga wire would break down like this  Power = Current² x Resistance = 9.3² x 8 = 86.49 x 8 = 692 sustained watts of power Power = Current² x Resistance = 20² x 8 = 400 x 8 = 3200 peak working watts of power Power = Current² x Resistance = 41² x 8 = 1600 x 8 = 12,800 max peak watts of power Again, especially on the peak working watts, I had to make some best guesses, though, this is more about perspective than absolute definitive information. As you can see, for a vast majority of needs, 14ga wire is well in excess of what is required. Even with 500 watt amps, you are very very very very unlikely to sustain 300w for any length of time. I provide this for perspective and you can make of it what you will. But I think for the casual user, it does put speaker wire into a workable perspective. FYI Steve/bluewizard 
What's the purpose of this? It doesn't say anything that hasn't been said much better before. Everything you ever needed to know about wire is in this link:
http://www.rogerrussell.com/wire/wire.htm and a very easy to use wire gauge calculator is in this link: http://www.bcae1.com/images/swfs/spe...rassistant.swf Both of these links have been posted here for years. 
Chassis means very short runs. Selfheating and electromigration is a bigger issue with longer runs thus the transmission (very long) runs chart. EM is less a problem with AC signals, natch. The chassis rating is closer for short'sh speaker runs than the longline figures (which are much more conservative). The UL current rating is sufficient for most of us. There are other charts around that better span the gap, or just start with basic principles. Or just use Roger Russell's chart as stated above...

I apologize in advance for the silly question. Does wire gauge matter if using banana plugs?

Quote:
Some banana plugs may not be physically able to accept 12 gauge wire. most will accept 14 gauge wire. Spade lugs which are soldered to the wire are the connection that will give the lowest contact resistance, and resist oxidation over time. 
For additional reference, I also calculated the above for 10ga (5.26mm²) and for the nearest AWG gauge to Euro 4mm² (11ga = 4.17mm²).
I can make the calculation for 10ga Wire (5.26mm²) using the Power Stream chart  1800 Sustained Watts 5000 Peak Working Watts (best guess) 24,200 Max Peak Watts The nearest to Euro 4mm² wire, would be AWG 11ga or 4.17mm²  1152 Sustained Watts 4232 Peak Working Watts (best guess) 17,672 Max Peak Watts I would say that is very substantial overkill. For a vast majority of people AWG 14ga (2.08mm²) or Euro 2.5mm² seems well more than adequate. Steve/bluewizard 
Quote:
https://www.avsforum.com/forum/89spe...l#post32129745 Knowledge is never wasted. Steve/bluewizard 
Quote:
OK, put banana plugs on these tiny wires and connect them between your Amp and your Speakers, then ask yourself the question  Does wire size matter if you are using Banana Plugs? Of course wire size matters. It matters in terms of Resistive Losses, and it matters in terms of current carrying ability. A typical 100w/ch amp has about ±45 volts on the output, if we push that onto an 8 ohm speakers, the current can be in the range of 5.625 Amps. So, you need that current carrying capacity minimum. Generally for safety you want to overrate the wire by a factor of two, which means you probably want wire capable of about 12 Amps. In terms of Resistive Losses, I provide a link to charts showing appropriate wire length based on Gauge, Speaker Impedance, and desire level of loss. 5% loss is generally considered acceptable, though on the Front Speakers, it is not hard to achieve well under 5% loss. Steve/bluewizard 
Quote:

Just to put it all out there: 100 W average power into 8 ohms is 28.28 Vrms = 40 Vpk = 80 Vpp = 3.54 Arms = 5 Apk = 10 App

I think the focus is not on capacity but resistant. All these wire have the capacity to deliver the power but the higher gauge will cause the amp to have to deliver more power than the lower gauge.

Quote:
100w/ch is derived using the RMS method, that means the 28.3 volts is RMS voltage, converted to peak we multiply by 1.414 or divide by 0.707  2.83v x 1.414 = 40.0162v That means 2.83v RMS equals roughly 40v Peak, I through in a couple extra volts for losses and headroom on the power supply. I think if you check the power supply voltage on most 100w/ch amps, it is going to be in the neighborhood of 45 volts. Steve/bluewizard 
Still making it up as you go along. Try using this for the correct answer:
http://www.sengpielaudio.com/calculatorohm.htm We don't use peak voltage, and the power supply voltage is moot. As for my 'alleged' knowledge of audio, acoustical engineering is my day job. What's yours? :rolleyes: 
Quote:
Remember this  Quote: Originally Posted by Bill Fitzmaurice https://www.avsforum.com/forum/images...s/viewpost.gif Quote:
https://www.avsforum.com/forum/89spe...l#post49961025 To which I responded with this  https://www.avsforum.com/forum/89spe...l#post49960353 Which is another speaker wire post were I had done some basic Current/Power calculations. And I thought, since you were so interested in Current Carrying Ability of Wire, I would find the Currents and post it reformed into WATTS, which leads us to where we are now. So, everyone, if you found this an interesting perspective on Speaker Wire, be sure to give BILL a Thumbs Up, because we owe it all to him. :) Steve/bluewizard 
Quote:
Just admit, at least to yourself, that you were wrong. You'll feel so much better, and your blood pressure will drop 20 points. Steve/bluewizard 
Quote:

Quote:
Quote:
Quote:
Retired  Bachelor of Science  Electronics Engineering Technology Just short of a Electrical Engineering Degree, but my degree is more related to Applied Technology than theoretical Electrical Engineering. :) Steve/bluewizard 
Quote:
Ip = Ep / R Ohm's Law  read it and weep. The goal was to determine the maximum likely Current to go through a Speaker Wire assuming a pretty standard 100w/ch amp. It 5.625 Amps PEAK. Even if we scale it back to the Calculated Peak Voltage of 40v, the current is still 5 Amps. Shake your fist and shout  Get off my lawn  all you want, it is 5.625 Amps PEAK. Steve/bluewizard 
Quote:
This posting (mine that is) is an interesting side note. It does help put speaker wire into perspective, but much like Inductance and Capacitance considerations, which rarely come into play when using Speaker wire, Current and Power capability are rarely taken into consideration. Likely this is the first time it has been mentioned in this context. So, most commonly people will simply take Speaker Wire Losses into consideration, but there are many approaches to this. BowersWilkins uses a fixed number; wire losses should not exceed 0.1 ohms. Roger Russell, in his often cited chart, uses a fixed 5%  http://www.rogerrussell.com/wire/wire.htm Myself, to expand the perspective, I have calculated various losses (1%, 2%, 3%, and 5%) for various Gauges and for various Speaker Impedance, and have cited that at the top of this thread. https://www.avsforum.com/forum/89spe...l#post32129745 By all parameters, very common AWG 14ga wire will serve the need of a vast majority of people in a vast majority of circumstances, and with a very considerable reserve of capacity. 14ga wire at 5% loss is 40ft to 4 ohms and 79ft to 8 ohms. More than enough. But for most, 14ga is very very reasonably priced, and is more than affordable. So, that's what most people buy. Quote:
Small high gauge wire will not cause the amp to deliver more power (within a context) but rather the losses will be higher. More signal will be lost in the wire, and less available to the speaker. BUT ... even at 5% signal loss, what you actually hear (or don't hear, as you will see) is very very small. Likely in the range of 0.5db. That is impreceptible by most accounts. Even 10% signal loss is still an inconsequential 1db. Part of the seeming imbalance is because there is a Square (or log?) function that makes voltage and perceived sound nonlinear. The common illustration of this is that ±3db is a doubling or halving of the applied power, yet ±3dB is a very small change in the sound you hear. Just barely noticeable. But, on general principle, I don't want to lose any more of my signal than I absolutely have to. Fortunately at common Stereo Cable length, you are well under 5%. Using my Charts again, for 1% loss to a 4 ohm speaker, you can still use 8 feet of 14ga wire. To a more common 8 ohms speaker, even with a 1% loss target, you can still have 16ft of wire. The dB, or perceived sound loss, of 1% signal loss is about 0.09dB, which is unimaginable microscopic. (Assuming I am using the right formula) But in principle you are right in what you say. Resistance per Length is the practical determining factor. But, it doesn't hurt to have more information and more perspectives on the available choices. Here I have simply presented one more perspective. Thanks to BILL's encouragement. Steve/bluewizard 
Quote:
1 % loss is 0.044 dB for power, 0.087 dB for voltage. 
Quote:
I've been using dB = 20 log(Vo/Vi) Steve/bluewizard 
I just used log of 0.99...
dB = 20*log (Vo/Vi) = 10*log(Po/Pi) Let Vi = 1 V then 1% less is 0.99 and thus the ratio is 0.99; similarly for power let Pi = 1 W and Po = 0.99 W. The factor of two in dB comes from the fact that power is related to voltage squared. 
Quote:
And.... typical 100W amps don't usually run at 100W 100% of the time, I would say they rarely run at 100W. The duration of time that a system will run at max capacity contributes to the calculation of what size wire to use. Will it run at ~20W most of the time and rise to 70W for just a few seconds? This needs to be taken into account. 
Quote:
Power dB = 10 x log(.9801/1) = .087dB, the same result as performing the voltage change dB calculation. This one reason why it is so nice to use dB. Distortion can also be stated in dB so that it's easy to compare values. 
this conversation is more complicated than applied mathematics. anyone want to do the math for feedback thru the wire at a fluctuating 148k ohm given a fluctuating 11000 watts with the left speaker at 5 feet and right at 7 ft.
even so called smart people in this field arent that smart and laymen are usually clueless so just spit out anything that sounds logical. 
Quote:
Quote:
Offtopic: There are places it (metal resistance) matters. The metal resistance and its thermal behavior was critically important in a highresolution DAC design I did a long time ago... It was amazing and quite vexing how a few squares of very wide traces on a chip could corrupt the linearity of a 16bit DAC. Audio DACs these days mostly use a different architecture so that is less an issue, but they don't run over a GHz either. It gets even uglier when you head on up in frequency and skin effect, eddy losses, surface roughness, and all that jazz starts to play a key role in the loss from source to sink. FWIWFM  Don (the dumb one) 
Quote:

I am not sure where you are going... Speaker/amplifier simulations have been used in design for a long time and they include all sorts of things, including driver and crossover parameters, usually simple wire models, and of course the amplifier circuits. It's been awhile since I did audio circuits, but I had models of all the components (transistors and passive elements) in the amplifiers I designed, the crossover network used realistic models, and of course speaker (driver) models. You simulate, measure, and iterate to get your model to match the measurements. But, it is normally all done during the design phase, not something presented to the final customers.
It's even worse than by the millisecond, however; the period of a 20 kHz signal is actually 1/20000 = 0.00005 s = 50 microseconds. The original post based wire choice on current capacity limits under various conditions. I doubt anyone on this thread really thinks the wire matters all that much unless it is much too small for the job. I suppose we should be thankful nobody has brought up how the dielectric (insulation) impacts the wires. It does, but far below the level of audibility, at least for most of us. 
Quote:
Your second point is a good one that gets complicated since you need to know the average power and timevarying statistics to calculate expected selfheating and failure rates. That gets back to @torii 's comment about going too deep down the rabbit hole (paraphrasing). 
Wow ... this took a strange turn. Still ... knowledge is power.
Since we are talking about a hypothetical 1% loss, we might as well normalize to 100. So 1% loss is  dB = 20 log(Vo/Vi) = 20 log(99/100) = 20 log(0.99) = 0.087296 dB But let's face even a full 1dB is insignificant, which is about 10%. dB = 20 log(Vo/Vi) = 20 log(90/100) = 20 log(0.9) = 0.91515 dB 5% loss is an inconsequential 0.44553 dB Assuming I'm doing this right, a very considerable 30% loss would be about 3.098 dB. A slight change in sound for a pretty substantial loss in signal. All nice to know, but I think, on general principle, we want to keep our signal loss as low as is reasonably possible. So, while most will simply consider Resitance/Distance relative to nominal speaker impedance, it is nice to know the other considerations. Knowledge is never wasted. Steve/bluewizard 
All times are GMT 7. The time now is 11:39 AM. 

Powered by vBulletin® Copyright ©2000  2019, Jelsoft Enterprises Ltd.
vBulletin Security provided by
vBSecurity (Pro) 
vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
vBulletin Optimisation provided by
vB Optimise (Pro) 
vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
User Alert System provided by
Advanced User Tagging (Pro) 
vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.