calculating max current through wire?

Ben Bradley wrote in news: snipped-for-privacy@4ax.com:

Well - ideally I'd like to use some .05" ribbon cable for this - and all the .05" ribbon cable I've seen is 28 or 30 AWG. My choice of .05" ribbon cable is due to me wanting as fine as possible of a ribbon cable to connect a couple boards. I'm thinking that what I could probabaly do, if necessary, is use multiple wires in the ribbon cable for this power line . Everything else except for the ground will be very low current - it's just this one wire and ground that will be carrying substantial amounts of current, so this makes more sense to me than using thicker ribbon cable.

It will be stretched out in open air.

-Michael

Why? Is this a mass-produced product where every fraction of a penny (coincidentally made of copper) must be shaved off the production cost? Or does the wire have to go through a very small area, such as a tube or conduit (this could affect max power it could dissipate - see below)?

It depends on what you want to do with the wire. I presume these table assume the wire is stretched out like a cable, thus the heat generated would be dissipated over a substantial area or volume. If you wrap the wire around a resistor so it has much less area and volume to dissipate heat, the power rating will be much lower.

-----

There is another table here which gives maximum current as well as resistance, that may help:

Just try it. put your finger on the wire to see if it is warm. Heat flow out the attachment points is a factor too. You may be making fuses. Your calculation based upon resistance is only correct for a range of current, not near the max at all.

From memory, all the 0.05" pitch ribbon cable I've seen is rated at a nominal 1 amp.

It's essentially nominal since working temperature is actually the determining criterion. i.e you have to derate with high ambient temp.

I see no problem with your case.

Graham

Flat Flex cable (FFC for the lookup) is rated somewhat lower, but makes a fine interconnect. I use it for LCD and touchpanel interconnects, amongst other things.

RS/Farnell carry a fair amount of it, and there's a small outfit in Germany that has all sorts of odd lengths/tinning options/number of ways - I'll dig out the website later.

Cheers

PeteS

I think you will be fine as long as you aren't running more than 1A CONTINUOUS. But try it for yourself. Ideally if you have a current limited voltage supply, short the terminals together with a piece of the 28awg cable you want to use, and set the current limit at 1.5A. Feel how warm the wire gets. If it gets very hot very quickly, you might not want to use it in your design. If it stays merely warm after like 10 minutes, you are probably fine. It is all about how hot you want the wire to get. If you search google, you might be able to find a temperature rise calculator based on wire size, current, and distance.

This is the rule of thumb that i have used to wind power transformers, and i have never had any IR problems: Start with the maximum current, in milliamperes for the wire in question. and look in a wire chart for the size with the closest value cross-sectional area in circular mils. Example: number 30 wire would be rated for use near 100mA maximum current (continuous). That same wire seems to be useable asa replacement for a one amp fuse. If that follows your tests, then the "fuse rating" would be roughly ten times my rule of thumb rating.

Why not look at using "flex"; that is, flat flexible "wire"?

Search for a copy of the AWG wire table. All will be explained. Its so simple that almost anyone can understand it. I did, when I was 10.

#define circular mil dimensional madness

ROT are a terrible way to design anything. As is current density (even more so when expressed in amps per circular mil). All that matters is temperature rise. Continuous current rating is invariably specified with an isolated wire in optimal thermal conditions; winding a whole bunch of terns in close proximity (if the feathers dont interfere) changes things significantly. See, for example, the various articles written in the [cant recall name] PCB design comic, or standards for PCB current density, where N adjacent identical tracks are considered the same as one track of N* width carrying N* current, due entirely to thermal coupling.

why not assume adiabatic heating to work out fusing current? the maths is trivial. Again, well covered in PCB design comics.

If the insulation is PVC, and you heat it up, it will outgass chlorine, which aint too good for crimped connections.

Cheers Terry

Not to mention units that are unique to the USA, so no-one anywhere else knows what you mean.

Graham

The reference was to American Wire Gauge so it doesn't matter if people outside the US don't understand circular mils, because they won't be using AWG wire anyway.

I have made a crude Excel spreadsheet that calculates current carrying capacity for wires as well as bus bar. For wire or bus bar in free air, the current carrying capacity is closely related to watts per square unit of surface area, so smaller wires can carry more current per cross sectional area than larger wires. For multiple conductors in close proximity (such as in a wound coil or multi-conductor cable), the capacity is more closely related to the cross sectional area, or watts per unit of volume. The temperature rating of the insulation also comes into play, so you need to determine the ambient temperature and the temperature rise due to power, factoring in the temperature coefficient of copper, and also the rate at which heat will be conducted or radiated from the wire. I made my wire chart based on the NEC ratings of 30 A for #10, 20 A for #12, and 15 A for #14, and found a constant 0.0037 A/sq mil of cross section area and 0.240 W/sq in of surface area. The surface area calculations correlate fairly closely with the published ratings for bus bar, so I am comfortable with those ratings for single larger wires.

Using this chart (which I just updated with approximate data for smaller wire sizes), the #28 AWG wire should be able to handle 1.3 amps, and the #30 AWG should handle 0.9 A. Based on cross sectional area, these ratings would be 0.5 and 0.3 A. I normally use this chart for much higher currents and very large wires and bus bars. My work with circuit breaker test sets involves continuous currents up to 6000 amperes and short pulse currents up to 100,000 amperes. At those levels, there are other considerations such as skin effect and both mutual and self inductance, as well as mechanical constraints to prevent conductors from jumping around due to magnetic effects.

The Excel spreadsheet is on my website at

Paul E. Schoen, President P S Technology, Inc.

"Robert Baer" wrote in message news:ot%Ee.8001$ snipped-for-privacy@newsread2.news.pas.earthlink.net...

can

and

peak

dropped

That's not wire - that's bar stock. ;-)

Cheers! Rich

I didnt say I didnt understand it, only that its spectacularly stupid. There is a reason MKS is referred to as "rational" units. Circular mils are a great example of irrational units.

Cheers Terry

Back in the days when you could order "10-inch" and "13-inch" pizzas, we used to calculate pizza surface area in circular inches.

Yeah, a real smart system. Whats the next wire size larger than 0000 AWG?

I stand by my original comments, that circular mils are a stupid, arbitrary and useless measure. WTF is wrong with conductor diameter? Dude, R = Rho(T)*length/area, baby physics. Circular mils serve only to complicate the matter unnecessarily.

Even the "cookbook" approach becomes more complex, more stupid numbers popping up everywhere, the almost complete lack of which is the main advantage of the MKS system (hence the term "rational").

Another great example of stupid numbers is the penchant for using the "transformer" equation. 4.44 is a pretty stupid number. E = wNBA is far more meaningful.

Cheers Terry

Hi Paul,

thats a damn good way of working it out. My only comment would relate to the thermal characteristics of the wire insulation, which is probably not constant over varying wire sizes. Radox, for example, uses a X-linked polymer with a nice low Rtheta, allowing even higher current densities (and it wont melt when you run the soldering iron across it, nor does it outgas Chlorine when hot).

Conversely it is highly unlikely that ribbon cable insulation is optimised for heat transfer (the insulation can be surprisingly thick), so I'd expect it to go horribly wrong with small wire sizes. But it certainly gives a reasonable upper limit, which can easily be followed up with suck-it-and-see experimentation.

Cheers Terry