Determining spaker wire size? McInnis?

I'm closing in on wiring up some speakers and want to determine appropriate wire size. My dilemma is that I'm going to be wiring up some tower speakers and want to keep wire size at a minimum to minimize holes in the tower. My application is a pair of 8OHM drivers to be wired in parallele. I have available to them 600 watts.

The research I have done has yielded keeping the wire-resistance below 5% that of the driver's total resistance - measured with a volt meter directly on the driver itself.

I wrote a spread sheet to calculate this out for me using standard wire resistance numbers.

Upload Speaker Resistance Chart.xls (17.9 k)

What is puzzling to me is that this formula doesn't take into account the wattage delivered over the wire. For example: If I were to wire these speakers in series creating ~16 ohm load, then I can use 18 AWG wire at 20 feet. So I can deliver 20 watts or 2000 watts over 18 AWG? I do understand that by increasing the driver impedance the wire current conversely decreases and so maybe this formula is sufficient.

Rod, I'd like to hear your thoughts on this subject.

Thanks

(Message edited by yosquire on May 03, 2006)

Most car audio amps will kick into protection mode after a few seconds seeing a 16ohm load.

14-16 ga should be more than enough for your application.

Run them in parallel with at least 14ga wire. You can get nice compact 13-14ga wire from Parts Express or AudioPile, it's a nice well insulated compact bundle with all 4 conductors within a single round jacket. Works quite well routed through round holes in towers and boats. At the run lengths in a boat the resistance of the wire (I2R heat losses) will be insignificant.

Craig:

First off, I really doubt that you would find an amp (suitable for installing in a boat) that would deliver your desired 600 watts into a 16 ohm load. Most amps rate their output at either 2 or 4 ohms and their output drops dramatically if you connect up a higher impedance.

You really want to put them in parallel, creating a 4 ohm load. They should also be indentical speakers, or at least speakers intended to be in parallel with each other.

Whenever you run wire there are two concerns that you should keep in mind. The first is the ability of the wire to withstand the current, and the second is the voltage drop at the load.

The ability for the wire to handle the current is independant of the length of the wire. It doesn't matter if the wire is one foot long or 100 feet long, there will be some current point where the heating in the wire causes the insultation to melt and bad things can happen.

For any significant length of wire you will usually encounter unwanted effects from the voltage drop well before you exceed the wire's ability to handle the current. This comes in the form of a voltage loss. Amplifiers will cut out at high volume, lights will be dim, etc.

For your speakers, the resistance of the wires will simply add to the impedance that the amplifier sees. A standard 18 awg wire should have a resistance of .00639 ohms per foot. So, if it took 10 feet to get up the tower and 10 feet back you would have 20 feet, or 0.138 ohms. This would raise your 4 ohm impedance to 4.138

Now the question is: will your amps deliever 600 watts into 4.138 ohms? If they will, then there is no problem, just turn up the gain on the amp a tiny bit more to compensate.

If you can manage a heavier wire then I would recommend running it. In this situation, bigger is better.

The example calling for 16 ohm speaker load was a factitious. I used that example to illustrate the failure of the ‘< 5% Total load formula.’ That formula would call for 18awg @ 25ft @ 16 ohms.

For my application I will run the drivers in parallel to achieve a 4 ohm total load.

Part of this is for a personal exercise in increasing understanding and a peer review of the research I've done --and wasn't terribly confident in. I'm really interested in the factors used to calculate wire size. I'd prefer to engineer this properly rather than take the approach of "that aut'a be nuff" and consume excess space in wire. There is a calculate-able point at which you begin seeing diminishing returns. I'm looking for the factors that plug into that calculation. If these drivers were inside the boat, I'd probably go excessive and run 10ga wire to them -- because I can.

Rod, as always your posts are very knowledgeable and insightful - thanks for the write up. The concept of adding the wire resistance to the speaker impedance was new and very useful for me. What I'm still a touch unclear on, is what performance degradation will be seen by a voltage drop between the amp and speakers? I would expect this to be a loss of amplitude of speaker travel, which would equate to a decrease in SPL?

Mike, thanks for the suggestion on the 13ga pre-bundled cable. Looks like a nice compact product, I think I'll end up using that.

Craig,

From an engineering perspective it's extremely difficult to calculate your load because it is highly variable. With high frequency signals it gets even more complicated because the electrons tend to travel on the skin of the wire strands. Ohms laws do not adequately address the issue. Steady state loads demand much larger wire sizes than variable loads. The power wires pretty easy calculate, but the speaker wire sizing is quite subjective.

I have tremendous respect for Rod's knowledge but I am not sure if I agree with the weakest point logic applied to speaker wires. The amplifier should have adequate tolerance to handle +/-5 to 10% for changes in impedence because a speaker's impedence is not constant across the frequency spectrum. All that said, I completely agree with Rod where you should minimize the wire's impedence allowing better power transmission from amp to speaker. Will it make a difference between 10, 12, 14, 16 ga.? In reality I doubt it, sometimes this is more of an emotional decision rather than a logical decision. I have done 4 boats with that 13ga finely stranded wire, with up to 500 watts per channel and I am still using it. My opinion/experience suggests the voltage to the amp will have a much greater impact on deliverd SPL than your speaker wire.

Mike,

I see what you're saying. In a nutshell, there's some basic common sense rules so you don't size your wire too small, from there you go bigger to the tune of your emotions. Ultimately it does sorta boil down to "that aut'a be nuff".

I've kind of picked up from your posts that it's important to use finely stranded wire. If that is correct, is this for fidelity reasons? (to increase the capacity at which the wire can support high frequencies?)

Btw: Do you have a Double E? If not, you sure sound like you do. Thanks for the informative response!

Craig,

Yes, I confess I have a BSEE with specialization in power engineering, minor in acoustics, plus 10 years of prior field experience as an electrician in commercial and industrial settings. Although my current employment is not specifically using my BSEE like Rod, he has much more knowledge on the 12v and charging side of the amps. My battery and charging system's design is owed to Rod's recommendations, it works very well. Since I also installed high-end car stereo's for a living waay back, I might understand a few more things on the speaker side of the amp and the acoustics.

The reason for the finely stranded wire is two fold. First, for flexibility and ease of installation; second, to reduce the "skin effect" where the electrons carrying the higher frequencies tend to ride on the surface of the wire strands. It also helps to reduce eddy currents. Back down from the 10,000 foot level again, it's doubtful that you will be able to hear the difference. But given a choice, go with the finely stranded oxygen free cabling since it's readily available at nearly the same cost. BTW, my in-boat speakers still use the stock 14/16ga speaker wires, the sub and tower use 10ga and 13ga (actually 2.5 sq.mm) respectively.

Have we killed this topic yet? LOL.

LOL - Yes, I think this topic is about dead.

I appreciate both your guy's write-ups. I know those posts take more than just a few minutes. Thanks!