Talk:American wire gauge

From Wikipedia, the free encyclopedia

Contents 1 Conversion table data 2 Math notes 3 Various comments snipped from the article and saved here 4 Bizarre term 'ampacity' 5 Focus 6 Approximate stranded metric equivalents 7 Unused gauges of wire

[edit] Conversion table data

I think the conversion table should also include cross sections in mm², as this is the common way to classify wires outside of the USA.

You know what to do: Be bold! If you have or can generate the data, then please add it!

Atlant 13:16, 11 Apr 2005 (UTC)

[edit] Math notes

Note that the conversion from diameter D to AWG gauges could be written as

• AWG = 36 - 39 log₉₂(200D)

but that would probably confuse more people than it would help.

This is most useful in illustrating that for D = 0.005 inch

• AWG = 36 - 39 log₉₂(200·0.005) = 36 - 39 log₉₂(1) = 36 - 39·0 =36

and for D = 0.460 inch

• AWG = 36 - 39 log₉₂(200·0.46) = 36 - 39 log₉₂(92) = 36 - 39·1 = −3

where, for AWG gauges, −3 is written as 0000 or 4/0.

The formulas could also be simplified by defining a constant k equal to the 39th root of 92 (about 1.122932), but once again that would probably confuse more people than it would help. This is the constant ratio in the AWG classification; for example, the diameter of AWG20 is 1.122932 times the diameter of AWG21. Using this k, the formulas become

• D = 0.005 k^36-AWG
• AWG = 36 - log (200D)/log k, or
• AWG = 36 - log₁₀ (200D)/0.05035353

with the intermediate explanatory step

• log D = (36 - AWG) log k - log (200)
  

Gene Nygaard 17:32, 11 Feb 2005 (UTC)

This sounds almost as much fun as explaining why the aspirin pills used for aspirin therapy are 81 mg! "Where'd 81 come from?"

In any case, do we really need all these complex formulas? It sounds like you're leading to an explanation that would read something like the following:

Each size of wire in the American Wire Gauge is related to its neighboring wire sizes by the ratio 1.122932. (See footnote for the derivation of this value). For example, 0 (zero) gauge wire has a diameter of 0.3249 inches and the next smaller wire (1 gauge) has a diamter of 0.3249 / 1.122932 ~= 0.2893 inches.

We could then cast the formulas to simply say diameter = 0.3249 / (1.12932^gauge_number).

(Sorry, I don't know Tex or I'd try to make this example clearer, but you probably get my drift.)

And then in some tangled footnote we could bury the explanation of where we get the constant that approximates 1.122932.

Atlant 19:04, 11 Feb 2005 (UTC)

[edit] Various comments snipped from the article and saved here

The precise formula for American Wire Gauge of solid wire is

/* this is quite wrong...try it with no. 10 AWG, .1019 inches diameter, this formula then gives -10*log (pi *0.0081552)= 15.91 as the gauge number, plainly wrong. */

where the "Area" is the cross-sectional area of the wire in square inches (1 inch = 25.4 mm). Note that this uses the same "10log10" numbering system as is used with decibels. You may also use the following approzimations:

---

and log is a logarithm to any base.

I've cleaned up these comments. These formulas are wrong. I don't think we need many different formulas relating gauge number and diameter; anyone who's interested can do the math from what's given in the article, and festooning the equations with logarithms isn't, in my opinion, the best way to convey the idea. Forumulas are rare in practice, people interested in wire properties tend to use tables, in my experience.--Wtshymanski 02:17, 22 Jun 2005 (UTC)

[edit] Bizarre term 'ampacity'

I do not understand the word 'ampacity'. I investigated further and found that it meant current. I changed the word but it was changed back with the comment that it was the word used. Even if it is used in the technical documentation for specialists, is there a way in which we can make the article more accessible to readers? Bobblewik  (talk) 11:06, 22 Jun 2005 (UTC)

The word is definitely in current use (!) in the United States and is probably a portmanteau of "Amps capacity". The word "ampacity" gets 36,400 Google hits while the equivalent phrase "current carrying capacity" gets 32,400. I'm fairly certain "ampacity" is the term used in the National Electrical Code (US).

Now, as to how to make the article more accessible, I see that we have the word linked to "Ampere"; perhaps "Current (electricity)" might be another alternative. We could replace the word entirely, but it is a table heading so space is somewhat important. Or maybe we should just break down and write an Ampacity article. I guess I'm open to suggestions.

Atlant 16:09, 22 Jun 2005 (UTC)

The other headings have a description that does not use a unit word, followed by the unit. For example 'Area' and then 'mm²'. This heading has a description that embeds a particular unit, and then does not have the unit. If we added the unit, then it would certainly be a clue, particularly for those of us not familiar with the NEC terminology (perhaps lack of awareness is the reason they are searching Wikipedia).

So my first suggestion is to provide the unit, just like we do with 'mm²'. In this case, it would be the symbol for ampere 'A'. When I read documents in other languages, I can often decipher what they are talking about if they use the correct symbols (even in non-latin scripts such as japanese and arabic).

My second suggestion is to replace ampacity with some term like 'max current', 'current' or 'current capacity'. I am not fixed on the term, but I think the word 'current' should be in there.

My third suggestion relates to spacing. It is only now that I see that 'NEC' belongs to that heading. Copperwire, NEC and ampacity are not familiar terms to me, they are close to each other so I did not understand which was in which column. The same problem applies to units, the values in the 'Area' column overlap into the next one so reducing my confidence in what goes where throughout the table. The heading is a bit compact on my screen. It is on three lines and has two links in it. I am not a big fan of putting links on commonplace units, particularly when I am struggling to find the signal in the noise. I find that they add clutter, but I know other people have different opinions. Perhaps making the column wider would help so that no heading is more than two lines on respectable screen sizes.

Am I whining too much? I am just trying to put into text what I could point at in two minutes of speech. Bobblewik  (talk) 19:04, 22 Jun 2005 (UTC)

I have edited the table but feel free to put it back. It will need some further work anyway because it is half in one style and half another. We still need a term for the right hand column. It seems to me that the right hand column heading may contain too much information. The purpose of the table is, I presume, merely to translate between AWG values, mm values, resistance and current. The reference to 'NEC' and '60 degrees' and 'insulation' could all be taken out of the table because they are supporting information. What do you think? Bobblewik  (talk) 19:32, 22 Jun 2005 (UTC)

No, the extra conditions are absolutely required. Copper conductors can carry a lot of current before melting, but they'll start their insulation on fire a lot sooner than that. The standard insulation (so called "TW" wire) is rated up to 60C and those current ratings ("ampacities") are what will keep the conductor below 60C. So that column describes what the U.S. NEC stipulates for the maximum current allowed with 60C plastic insulation. We could write another column for 90C wire and a third for 105C (so-called "THHN") wire.

Atlant 23:53, 22 Jun 2005 (UTC)

You have reminded me of a conversation with an electrician about correct cabling and electrical devices in the home. He also mentioned the insulation melting as an issue. Your explanation is quite clear and helpful. Perhaps that is worthy of being in the article to explain why the current limits are set as they are. You also used the term 'current rating'. That seems quite a clear term to me. In order to make the table more self-explanatory, could we replace 'NEC ampacity' in the table heading with 'Current rating'? After all, confusion about that column heading is why I brought this topic up. Thanks. Bobblewik  (talk) 08:31, 23 Jun 2005 (UTC)

Please, use °C for degrees Celsius, always, and especially when, as in this article, you could well be referring to charge in coulombs whose symbol is C, or to the italic C often used for a variable representing some quantity related to this discussion and which isn't always italicized as it should be. You can write °C as °C in Wikipedia if you don't know how to get it on your keyboard or from the character list on the edit screen. And use a space between the unit and its symbol. Gene Nygaard 10:17, 23 Jun 2005 (UTC)

Bobblewik asks: Could we replace 'NEC ampacity' in the table heading with 'Current rating'?

I'd be in agreement with "Current rating with 60 °C insulation" (or whatever the proper Wikiform of "degrees celsius" is; I'm not sure I accept Gene Nygaard's request above as consistent with the accepted WIki style, although I can see a conflict between Celsius and Coulombs).

Now, can you fit all that in the table and still have it look pretty? :-)

(And remember, we really may want to add a 105 °C column as well! Heck, if we're going to make this great, we might as well make it GREAT!)

Atlant 11:50, 23 Jun 2005 (UTC)

I have done as you said. I am not sure what Gene means. I thought that we had put °C in the table. I a bit confused by all the various forms of little circles and hyphens or dashes. They all look the same to me.

I would also be interested in Gene's thoughts on the term 'ampacity'. I will leave it up to others to expand the body text a bit more. I think the bit in this talk page about insulation melting was helpful to me. Bobblewik  (talk) 11:58, 23 Jun 2005 (UTC)

[edit] Focus

The article is rambling quite a lot. Current carrying capacity, table of metric (forsooth!) standard wire sizes, discussion of I2R losses, etc. are all extraneous to the topic, which is adquately explained by a table. Has anyone proofed the current versions of these formulas, the last time someone put formulas in they were quite wrong. I don't know where the IEC sizes belong but it's surely not in an article called "American wire gauge". Perhaps in "wire" or elsewhere? --Wtshymanski 04:45, 18 July 2005 (UTC)

Considering that the AWG is a pretty odd measure in a world that is very much dominated by mm² wire specifications, I don't think it is helpful to remove the metric versions of the conversion formulas. After all, this is meant to be an encyclopedia for a global audience. People who look up AWG very often do so, because they have to convert an AWG spec into an IEC spec, or vice versa. I have carefully checked the formulas against several sources, including tables from the US NEC. I agree that there is a lot of rambling about current carrying capacity, and that this subject is far more complicated than what can be (and is) handled in this article. If you want to strengthen the focus, please start at that part of the article, rather than the metric variants of the conversion formulas. Markus Kuhn 09:55, 18 July 2005 (UTC)

Odd measure? I'm sure a significant part of the world's annual wire sales is specified in AWG. The table gives the square mm areas. People who work in the industry rarely use the formulas anyway, so two versions seems to be overkill. I'll try to find more relevant homes for the text in the article - we do have articles on electrical wiring that may be a better fit. --Wtshymanski 13:49, 18 July 2005 (UTC)

"Odd measure" was meant in the sense of "difficult to understand unless you know the definition or have a reference table". Markus Kuhn 14:26, 18 July 2005 (UTC)

The problem with the current-carrying capacity is that this is not simply a physical property of the cable. It is also an engineering judgement that involves ventilation and generous safety margins. For example, the current carrying capacities listed in the US NEC are a bit more conservative/generous than those used in Europe, possibly because in the US (where wood-frame buildings dominate) there is more worry about fires created by electrical accidents than in Europe (where brick-and-mortar buildings dominate). It is mostly because of such complexities that I suggest to move the current-capacity information into a separate article and start discussing it there properly. An interesting discussion of regional differences in current ratings is in one of the NEMA references at the end of the Electrical wiring article. I guess, Electrical wiring would be a good starting point from where such information should be easy to find. Markus Kuhn 14:26, 18 July 2005 (UTC)

Well, this is the article about the AMERICAN Wire Gauge, so it's entirely reasonable to state ampacities in terms of what the AMERICAN regulatory authority specifies. Really, the column is completely well specified: Copper wire, 60°C plastic insulation, and the NEC. Run more amps through that and your installation is not conformant to the regulations.

Atlant 15:47, 18 July 2005 (UTC)

I'll find a better home for the following :

Current-carrying capacity of wires

The ability of a wire to carry electric current (its ampacity) is limited by two factors:

• The amount of voltage lost to electrical resistance in the form of V = IR voltage drop and
• The amount of heating of the wire caused by the P = I2R power dissipated in the wire as a result of that voltage drop.

Both factors are affected by the cross-sectional area of the wire and that cross-sectional area is denoted in America by its AWG (gauge) number.

While copper (or aluminium) can carry substantial current before melting, practical plastic wire insulation will melt (allowing short circuits) or even catch fire at much lower temperatures, limiting the current-carrying capacity of any insulated wire to a much lower value than that point at which the wire itself will melt.

For relatively high-voltage circuits, the temperature of the insulation thus provides a limit to the current-carrying capacity of the wire and the wire gauge to be used must be selected to provide reliable, safe operation of the insulation. (And various kinds of insulation can withstand varying degrees of heat.) For low voltage circuits, it is instead the voltage drop as a result of IR losses that usually dominates the choice of wire gauge.

and

Outside Canada and the U.S., wire is typically specified in terms of its area in mm². The international standard manufacturing sizes for conductors in electrical cables are defined in IEC 60228.

--Wtshymanski 01:39, 21 July 2005 (UTC)

[edit] Approximate stranded metric equivalents

SGBailey recently added a column approximate stranded metric equivalents to the table. I find these somewhat odd and in need of explanation, for a number of reasons:

• Could you please add an explanation and reference for the n/d notation used (I assume it is number of stands / diameter of a single strand in mm, but the table should say so).
• Where is this particular notation used? Metric cables are normally labeled in mm² cross section, therefore there is no need to look up anything in the column you added.
• Where do the particular figures that you added come from? Is is a selection from a particular product standard? If so, please add a reference!
• Why are there < and > signs? This does not make sense to me.

Unless you can answer/fix the above, I would suggest to revert these additions as they look very arbitrary and not very useful in their present form. Thanks.

Markus Kuhn 13:37, 9 August 2005 (UTC)

I've added some commentary on the metric stranded equivalents. Should it be decided that this is not the place for this information, I'll make a separate page along the lines of "equivalent wire sizes". -- SGBailey 13:47, 10 August 2005 (UTC)

In your added explanation, you say that the d in n/d is the area. But the equivalences that you have chosen match the mm diameter column, not the mm² area column. This now looks definitely wrong. If the notation you use is really based on area, then there seems no point in providing an equivalence table. Everyone who knows how to multiply can see that a 5x1 mm² strand is (for low frequency purposes) equivalent to a 5 mm² wire, etc. I don't think, we need or want to have a full multiplication table here. You also haven't answered my other questions. Markus Kuhn 14:31, 10 August 2005 (UTC)

[edit] Unused gauges of wire

Why is 11,13,15, etc listed as wire gauges? I'm pretty sure you can't buy those sizes, and even if you could, they would be so rare, why even bother to list them?

     Magnetic wire (enamel coated wire for inductors and transformers) is available in both even and odd gauges.
     Hookup wire is normally available only in the even gauges as you note.