Wire sizing question

[tkamiya]

Being originally from Japan, I'm more used to metric expressions.  In Japan, we either express wire size by diameter in mm if it is a solid wire, or cross section area by mm^2.

That said:

I've been in US now and all we use here is AWG or gauge.  I am used to these now, that 24 gauge wire is very suitable for hand wiring perf board, and 18 or 20 is good for power line connection (for example).

Fairly recently, I started buying inline fuse holder from eBay, and they mostly come from China.  Most of them say 18 gauge 20A and some of them say 30A.  When I actually receive them; though, they are no where near what I'd think 18 gauge should be.  Insulation is thick resembling the looks of #18 but wire inside is anywhere from #20 to #22 gauge.  I know inside chassis wire requirement and current load is different from electrician wiring requirement.  But there is no way I would try to pass 20A on this!

I know AWG (American Wire Gauge) is an US thing but is there a Chinese equivalent?  Am I ordering incorrectly or am I being duped?  What's going on?

[Monkeh]

It's just China selling you rubbish.

18AWG is closest to 0.75mm². Which isn't good for 20A in most environments to begin with.

[edpalmer42]

In general, current ratings are based on the maximum temperature of the wire insulation.  If you have teflon or silicone insulated wire, ratings up to 200C are not unusual.  One table I saw said that with either teflon or silicone, 18 AWG wire could handle 24 amps and 20 AWG could handle 17 amps.

But Monkeh is probably correct.   

[Monkeh]

Sure, silicone insulated wire is good for 20A (obviously volt drop is a concern), but then everything it connects to and near it needs to be rated for the same temperatures. This is why in most practical terms, even 90C rated PVC or XLPE cable is downrated to 70C figures.

[ArthurDent]

This might help. There is a lot of info in this on line calculator.
https://www.rapidtables.com/calc/wire/wire-gauge-chart.html

[Jwillis]

Sure, silicone insulated wire is good for 20A (obviously volt drop is a concern), but then everything it connects to and near it needs to be rated for the same temperatures. This is why in most practical terms, even 90C rated PVC or XLPE cable is downrated to 70C figures.

All I can say is buyer beware . Personally I have not experienced issues with Chinese products from Ebay. But many cases the seller may not even know what they are selling. If all they sell is wire then you'll probably be OK . If it's mixed in with other cheap designer fashion and crap , then you takes your chances. Do diligence in your searches and compare prices . Going for the cheapest possible is not always the best and cheapest.
   
XLPE, PEX, or Polyethylene (PE) is the same thing and is rated for -65°C to 80°C ambient. But can go up to 115°C or more before melting
PVC (Polyvinyl Chloride) is rated for -55°C to 105°C ambient . But can go as high as 260°C
The ambient rating is the working range between fracturing at the lowest temperature and holding rigidity before softening at the highest ambient temperature. Manufactures will derate the wire so it'd held well below the softening temperature of the insulation.

The NEC ( Nationaelectrical Code) is adopted as a standard for electrical manufacturing even though there is no actual law. The NEC is designed for civil electrical installation , specifically homes. House wire can be rated for either 60C , 75C or 90C and this rating is for an ambient temperature of 30°C. This is over building for an obvious reason. So people don't burn down their homes overloading the wiring. 

Automotive and electrical manufacturing does not need to follow the NEC . Although many have adopted the code as a base line . For example , Typical PVC automotive wire can be rated as high as 125°C

Now under the NEC 18 gauge wire is rated for 16A @ 90°C with a 20°C ambient. It does not mean that it won't conduct 20A and remain stable. It will only be a little warmer.  And the voltage drop will be based on 20.95mOmhs per metre. At 20A the voltage drop will be around 400mV / per metre. The actual voltage drop will be a bit more because of the heating wire. The resistance will increase with temperature depending on the temperature coefficient of the conductive material .  Copper is around +0.393 percent higher resistance  per degree above 20°C

For example typical hook up wire for bread boards is around 24 to 22 gauge. I use 22 gauge with a rating of 7A and it can handle 12A easily . It gets warm so I wouldn't recommend it for long periods of time or for final design. But it doesn't suddenly flash into flame.

When selecting wire for your project you need to know the length of wire you will be using between 2 points . The typical ambient temperature . The except-able voltage drop across the wire between 2 points. The maximum rated voltage of the insulation.
If higher currents are required go with a larger wire.

By far the best table for wire gauge conversion from metric and current capability is here  https://en.wikipedia.org/wiki/American_wire_gauge#cite_note-ampacity-8

[andy3055]

Try this. It has the current carrying capacity as well.

https://www.engineeringtoolbox.com/wire-gauges-d_419.html

[engrguy42]

I always get a chuckle when I see discussions like this. It's what I call "spec monkey heaven". 

It's where people tend to get blinders and don't step back and look at things in perspective and try to understand them, they just look at a table and say "OHMYGOD !!! It says 20A rated and I'm pushing 21A !!! It's going to explode and burn the house down !!!".

Anyway, as others have mentioned, it's about temperatures. You can run any wire at any current as long is it doesn't either melt itself or melt something it's touching. If it's copper it won't melt until it reaches some crazy temperature (1,000+ C or something?). The problem is stuff it might come in contact with, like insulation, or heat shrink, or you, or whatever. For example I think hot glue and heat shrink melts around 80C according to my table of melting points. BTW, I recommend that everyone opens up Excel and starts to make a table of melting points of stuff. Great way to get a good perspective on thermal capabilities of stuff you work with. Rubber is somewhere around 180C (though I think jwillis has better numbers for PVC and polyethylene insulation). And if you touch something it feels really hot around 40+C.

But if you force cool this stuff with a fan and/or add a heat sink or something it may be fine. Also, keep in mind we're talking stuff like 20A. How the heck is the average user gonna put together a circuit pushing 20A? Intentionally, that is...

Anyway, this is just one more area where I think people would learn a lot if they actually could test real wire, connectors, etc. and see how they respond, rather than the mindless "Oh, it's Chinese junk". Which, BTW, the OP did an excellent job of when the issue of cheap Chinese variacs came up.   

Put a thermocouple from your meter on the wire as you push current and see what happens (under controlled conditions of course, blah, blah, blah). You may find that either: "Hey, I can't even figure out a way to make a circuit that will push that much current, so there's no way I'll need that amount of current handling capability in the first place", and/or, "Hey, it just get's real hot but doesn't melt anything".

[engrguy42]

FWIW, I just did a quick test as an example:

28AWG breadboard jumper. The tables say the amp rating is somewhere around 1.5-3 amps, depending on who you believe and what temperature rise to expect.

I applied 4 amps DC, and had a thermocouple Kapton-taped to the outer insulation, and another connected to the pin on the end of the jumper. I couldn't get the temps on either to exceed 52C, which they flattened out to after a few minutes. Hot to the touch, but certainly not damaging to the insulation. Unless it was a lot hotter inside the wire insulation, which I doubt.

So I encourage folks to do their own simple tests. Quick and easy, and you can learn actual facts rather than rely on speculation. 

[T3sl4co1l]

I applied 4 amps DC, and had a thermocouple Kapton-taped to the outer insulation, and another connected to the pin on the end of the jumper. I couldn't get the temps on either to exceed 52C, which they flattened out to after a few minutes. Hot to the touch, but certainly not damaging to the insulation. Unless it was a lot hotter inside the wire insulation, which I doubt.

Oh, it'll be a lot hotter inside the insulation -- I have no doubt!

Even with the Kapton tape outside, I mean, it's not much insulation, it's still a heat dissipating surface.  Some of which is dropping past the thermocouple itself (and along itself, give or take what size wire it's made of), some of which is through the insulation.

Generally speaking, radial gradients between coaxial cylinders give logarithmic potentials.  That is to say, for this case, the temp rise goes as ln(R/r).  The inner temp rise can easily be several times the outer temp rise.  (Very unlikely to be, like, tens or hundreds of times higher -- log is a very slowly rising function! -- but a small number like 2 or 3, absolutely.)

To measure it, you'd have to pierce the insulation and shove a very tiny thermocouple down in there, beside the wire.  Not impossible, but not doable without a very fine thermocouple.

Might be better off measuring the wire's resistance and calibrating against an oven temperature.  Which, is its own matter, because who knows if you have pure copper, copper-clad aluminum, or copper-clad steel?   Hard to extrapolate under those conditions.

Tim

[engrguy42]

I applied 4 amps DC, and had a thermocouple Kapton-taped to the outer insulation, and another connected to the pin on the end of the jumper. I couldn't get the temps on either to exceed 52C, which they flattened out to after a few minutes. Hot to the touch, but certainly not damaging to the insulation. Unless it was a lot hotter inside the wire insulation, which I doubt.

Oh, it'll be a lot hotter inside the insulation -- I have no doubt!

Even with the Kapton tape outside, I mean, it's not much insulation, it's still a heat dissipating surface.  Some of which is dropping past the thermocouple itself (and along itself, give or take what size wire it's made of), some of which is through the insulation.

Generally speaking, radial gradients between coaxial cylinders give logarithmic potentials.  That is to say, for this case, the temp rise goes as ln(R/r).  The inner temp rise can easily be several times the outer temp rise.  (Very unlikely to be, like, tens or hundreds of times higher -- log is a very slowly rising function! -- but a small number like 2 or 3, absolutely.)

To measure it, you'd have to pierce the insulation and shove a very tiny thermocouple down in there, beside the wire.  Not impossible, but not doable without a very fine thermocouple.

Might be better off measuring the wire's resistance and calibrating against an oven temperature.  Which, is its own matter, because who knows if you have pure copper, copper-clad aluminum, or copper-clad steel?   Hard to extrapolate under those conditions.

Tim

You're certainly free to go down the techno rabbit hole, and you're free to speculate and "have no doubt" about whatever you want, but the facts remain. You're also free to do your own tests if you need more info, but I measured both the conductor itself and the outside of the TINY insulation thickness. It seems reasonable, IMO, to assume that after a few minutes the internal temperature would make itself to either or both of those thermocouples. But the real test whether the insulation melts or smokes. Which it didn't.

BTW, I just added a 4" diameter fan, spaced 6 inches from the jumper, and with the same 4 amps flowing the temps on both thermocouples don't exceed 32C.

The point here is that it's far more beneficial, IMO, to do your own tests rather than rely on speculation.

[engrguy42]

BTW, for those who might want to erase all doubt, you might try stripping the wire, then get some heat shrink and shrink it over the thermocouple resting next to the wire. That not only gives you some stand-in insulation to keep the heat in, but also gives you a known data point: heat shrink should melt at around 80C (as I recall), so if it starts to melt you have a visual picture what the internal temps are.

Of course it might be worst case, since the heat shrink thickness is probably a lot  more than the jumper thickness (at least in this case)

[engrguy42]

Okay, well I erased all doubt...

Stripped the 28AWG, wrapped it around my thermocouple, heat shrinked some slightly larger diameter heat shrink over both, then ran 4 amps thru it. It's been 10 minutes and it has flattened out at 54C. Which is 2 degrees hotter than the thermocouple measuring the outside of the actual insulation.

So the "I have no doubt" is, well, doubtful. 

[engrguy42]

I'm curious...do people here actually need to push like 15-30 amps in practice? Is it for tubes and stuff? I know computer GPU's can draw a lot of amps (something like 200W at 12V? = 16A), but I'm wondering if it's just the rare exception where big current like this is required.

[Jwillis]

I'm curious...do people here actually need to push like 15-30 amps in practice? Is it for tubes and stuff? I know computer GPU's can draw a lot of amps (something like 200W at 12V? = 16A), but I'm wondering if it's just the rare exception where big current like this is required.

Not really a rare instance when working on solar systems.  Recently I needed to build an electronic load to test a 50 A solar charge controller for a friend. Solar systems can be several hundred Amps. Heated beds for green housing that need to run off grid . Electroplating can use quite high currents. Experiments in EDM drilling requires high currents.  Just a few examples where high currents are required  or present.

[tkamiya]

Thanks everybody. 

I used to be very active in electronics in 1970s to early 80s.  I re-discovered this "new" hobby few years ago and still try catch-up of what I missed.  I was in Japan then and US now.

I remember nearly identical item back then and seeing 18 gauge being what it looked like, I was surprised at what I got back from Chinese sellers.  The covering is either PVC or Vinyl.  I can't tell, but it sure isn't high temperature stuff.  I think it's vinyl.  I'm using it for just few amps, so it's fine, but when I buy #18, I want #18.  When I deal with mains side wiring inside projects, I want sturdy wire, not this flimsy stuff. 

I can test this, but spec is supposed to be the spec.  If I have to test everything myself, it kind of negates advertising #18 wire.  Now, I am starting to see some sellers put "REAL 18 gauge wires" in their ads to distance themselves.

By the way, in Japan, we used to call AWG 18 type wires, "30 strands", and it was for 7 amps max.  When we counted number of wires, they were usually 30.  AWG 16 type wires were "50 strands" and was up to 10 amps.

When I buy stuff on eBay, I usually check what ELSE the seller is selling.  In some cases, shops sell this kind of stuff, and make up, and nick-nacks, and adult toys...  I am not familiar with how business is run in China but it seems very odd to me.

Like the other thread, the product was probably made to the price.  I'd rather pay real money and get what I bought.  For many things, I'm starting to buy from Digikey.  They do sell cheap stuff but I've never had misleading specifications.

Again, thanks everyone.

[engrguy42]

tkamiya, yeah you're right...specs are specs. But I think the problem is that most people don't really understand what the specs mean and what they're actually saying. So they blindly follow some table, not knowing the actual data behind it.

My point is that a good engineer knows the ACTUAL data, so he can compare that with the tables to see stuff like how much margin is assumed in the tables. And how much of a worst case the tables' data represents. Maybe that worst case doesn't apply to you. Or maybe by testing you can realize, "hey, wait a minute, I can just add a small fan to this circuit and my temperature worries go away".

I see a lot of tech discussions where people love making blanket assumptions and dismissals of stuff they don't really understand. And they'll spend hours and days doing stuff like reconditioning old equipment or playing with some Arduino circuit or making PCB's by hand, but won't spend 15 minutes testing something so they can learn how it actually works.

Knowledge is good. 

[wizard69]

Being originally from Japan, I'm more used to metric expressions.  In Japan, we either express wire size by diameter in mm if it is a solid wire, or cross section area by mm^2.

That said:

I've been in US now and all we use here is AWG or gauge.  I am used to these now, that 24 gauge wire is very suitable for hand wiring perf board, and 18 or 20 is good for power line connection (for example).

For any AC wiring before any sort of fuse you would want a minimal of 14 ga wire.   I know that there are uses out there where 18 ga or 16 ga cords are plugged into an outlet but from my perspective that is asking for trouble as common outlets in the USA are either 15 or 20 amp circuits.

Fairly recently, I started buying inline fuse holder from eBay, and they mostly come from China.  Most of them say 18 gauge 20A and some of them say 30A.  When I actually receive them; though, they are no where near what I'd think 18 gauge should be.  Insulation is thick resembling the looks of #18 but wire inside is anywhere from #20 to #22 gauge.  I know inside chassis wire requirement and current load is different from electrician wiring requirement.  But there is no way I would try to pass 20A on this!

Sadly I suspect somebody in China is cutting corners to save a few pennies or is an out right scam artist.   To put it mildly buying this type of stuff on E-Bay is a massive gamble.   That isn't to say there are not legitimate manufactures in China but lets face it they don't use E-Bay.   At the very least look for such products handled buy reputable distributors.

Now we need to note here the the NEC does permit smaller wires sizes in fixtures than the circuits 15 amp or 20 amp protection level implies.   This is why a lot of LED light fixtures have light lead wires to he AC supply.    This is great for a defined usage, however a fuse holder or an AC entry socket is by definition an undefined device as far as what current it will be used at.    It needs to be able to handle the rated current for the device it is protecting.

I know AWG (American Wire Gauge) is an US thing but is there a Chinese equivalent?  Am I ordering incorrectly or am I being duped?  What's going on?

Well you could be ordering incorrectly but I'd lean towards bad actors in China.   Honestly I would never order critical parts through E-bay as even domestically there are too many scam artists.   There are of course many good vendors in both the USA and in China but it isn't always easy to tell them apart from the dealers in trash.    So for critical or safety related stuff just order from an established distributor of name brand components.   It will save you a lot of grief and let you sleep easy at night.

A few points here.   

• Building electrical systems are governed by the National Electrical Code (NEC) in the USA. 
• UL the Underwriters Laboratory does compliance testing for all sorts of electrical products and for whatever standards may apply.   What standards that might apply for any box you build is something that you would need to research.
• Make sure you have a fuse or circuit breaker protecting the device set at a suitable current level.   Circuit breakers are much harder for some bozzo to change out with an over rated device.
• I've actually have seen a few ISO AC entry sockets with fuse holders fail.   It could be a Chinese knock off that wasn't check for actual capacity
• Plain old fuse holders frankly don't always handle their ratings well in 24/7 operation.   This especially for 1/4" and 5mm glass fuse holders.   This even from legitimate manufactures.  I base this on experience with fuse holder running under their ratings (but close) 24/7.   Sometimes a more robust fuse and holder is justified from the standpoint of reliability. To put it mildly there are a lot of really cheap fuse holders out there that can not support a full load.

[wizard69]

In general, current ratings are based on the maximum temperature of the wire insulation.  If you have teflon or silicone insulated wire, ratings up to 200C are not unusual.  One table I saw said that with either teflon or silicone, 18 AWG wire could handle 24 amps and 20 AWG could handle 17 amps.

But Monkeh is probably correct.   

Actually the NEC has defined that 15 and 20 amp circuits must use specific wire ga sizes regardless of the insulation used.   I don't have recent copy of the NEC so I can't say  what the current wording is.   Even then you would need to determine if the regulation applies to whatever is being built here.   For an AC line coming into just about anything though I'd rather see minimal 14 ga to the first fuse in the system.   Obviously what is on the other side of the fuse is an engineering problem.

[tkamiya]

Yes, I know all that.  I am actually an licensed electrician, but my license was issued by Japan.  (it's no good here)  I do know few facts in NEC and local code though. 

I think I did mis-speak.  I meant to say #14 and #12 but instead, I said something else.  I blame my drink (that contained a wee bit of Vodka)   

Regulations really don't apply in my case; though, as there is nothing that regulates what DIYers use for their home made equipment.  You can use paper clip if you want.  Unwise, silly, but not illegal.  It may invalidate home owner's insurance but still not illegal.

Yes.  Buying Chinese merchandise from eBay is a challenge.  I had mostly good luck but few bad ones.  Fuse holders were 50/50 in luck gauge.  All I wanted to know was if China had different numbering system for wires.  Thanks!!

[engrguy42]

Yes, I know all that.  I am actually an licensed electrician, but my license was issued by Japan.  (it's no good here)  I do know few facts in NEC and local code though. 

I think I did mis-speak.  I meant to say #14 and #12 but instead, I said something else.  I blame my drink (that contained a wee bit of Vodka)   

Regulations really don't apply in my case; though, as there is nothing that regulates what DIYers use for their home made equipment.  You can use paper clip if you want.  Unwise, silly, but not illegal.  It may invalidate home owner's insurance but still not illegal.

Yes.  Buying Chinese merchandise from eBay is a challenge.  I had mostly good luck but few bad ones.  Fuse holders were 50/50 in luck gauge.  All I wanted to know was if China had different numbering system for wires.  Thanks!!

Speaking of blaming your drink...

The other day I was drinking a bit of tequila. Just a bit. And working on my bench. I had a couple sets of leads on the bench, and wiring up a circuit board to my power supply to test something, and measuring on my scope.

Turn on the power supply, and look at the scope. Nothing. WTF?? Check everything. Still nothing on the scope. WTF?? Polarities okay? Yup. Circuit board wiring okay? Yup. Maybe 15 minutes goes by. Checking and double checking.

Then I happen to look down at the leads from the circuit to the power supply. Banana plugs just sitting there not connected to anything. And another set of leads is plugged into the power supply. 

Yeah, that's just embarassing...

[tkamiya]

Haha....  I've done that too many times myself. 

As long as nothing releases the magic smoke, it's all good.

[Raj]

Best is to find a local retailer
If you can't, and have to buy from China,... Buy directly from source (alibaba) call the seller, even go as far as to arrange a video call (only problem is, usually, you won't get anything less than a kilogram in weight)

[tkamiya]

Hi, Raj

I'm kind of lucky in the sense that there IS a local surplus parts sources.  But with due caution with covid-19, I cannot travel to it these days.  My girlfriend and I are in "high risk" group, so we must be extra careful.  So I am buying everything via mail or some kind of delivery.  In US, "remote shopping" is kind of a norm as well.  How is it in India?  Do you have shops you can visit and buy electronics parts? 

[Raj]

Hi, Raj

I'm kind of lucky in the sense that there IS a local surplus parts sources.  But with due caution with covid-19, I cannot travel to it these days.  My girlfriend and I are in "high risk" group, so we must be extra careful.  So I am buying everything via mail or some kind of delivery.  In US, "remote shopping" is kind of a norm as well.  How is it in India?  Do you have shops you can visit and buy electronics parts?

Shops never closed down here despite government orders. During full lockdown, people could call the local retailer to package their stuff, and take it at their shops, since the shops are usually also the homes of these sellers. If someone asks where you're going, you could just say, you were gonna buy medicines/ food.
Now, the lockdown is over. The police goes around the city, ensuring that shops are enforcing social distancing.
it's easier to do that in India, since, in most shops, you stand outside and the shopkeeper brings the stuff to you (usually at a counter), like a McDonalds drive through, but with full view of what's inside.


What's funny, all government postal services have been suspended since March.