Winding coupled inductors

[evandude]

I am planning to build this:

**broken link removed**

as graciously provided by hantto.

I need to know what wire gauge I would have to use. I figure at 12v output, 8v input (worst case) and 80% efficiency the coil would have to withstand 15A average. I found a table of current capacities for the various AWG wire sizes. According to that, I'd need a 10-gauge wire for "power transmission" for that amount of current, but only an 18-gauge for "chassis wiring". Obviously 10 gauge is way too big for this, 18-gauge sounds more realistic, but I'd like to get a second opinion.

the wire he describes as actually using is 0.75mm, which corresponds to just under 20 gauge, which as far as I can tell, is only rated for about 1.4 amps. I am sure there are rules of thumb for how much you can overshoot these ratings, but I don't know them, and I don't want to blow things up. The best I can get at radio shack is 22 gauge which seems too narrow for this application, but I can probably get larger wire from the professors here at school.

As I said, I'd just like some info from those more experienced with these things. thanks![/url]

 

[stevez]

Think of wire as a conductor with some resistance. Thicker wire has less resistance than thinner wire - assuming materials of construction are the same and that we're at DC. Current flowing thru a resistance will result in some heat.

If you add heat to a material the temperature will increase - that's the easy part. How much of a temperature increase - until the heat flow out equals the heat flow in. In the case of a resistor (or wire) the flow of current, any current, will warm the resistor or wire until the heat lost equals the heat generated.

If we run a resistor/wire out in the open (bare) where the air can circulate well and flow current thru it the wire will warm to a point where the heat loss equals the heat generated by the current flow - I am deliberately repeating myself. If you take that same wire with the same current and add a layer of wire insulation the wire itself will get warmer because the electrical insulation will also act to reduce the heat flow out. Add enough insulation and the wire could get really hot with very little current flow. The limit will be when the wire melts or the insulation gives out. With this in mind the argument can be made that insulated wire takes less current to get to the same temperature. If you look at NEC tables you'll see that wires packed in conduits have a lower current carrying capacity than a single wire in open air. That's because the wire insulation and the conduit keeps the heat in better so less current is required to get to the same temperature.

We can go the other way and put our current carrying resistor or wire in an oil bath. The oil bath transfers heat faster than air so more heat can be added to the wire for a given wire temperature increase. I think I've illustrated the point enough.

Wrapping a wire around something like a core will result in trapping some heat. As with the other examples, the heat generated by the flow of current will equal the heat lost at some temperature. What temperature depends on a lot of things including the insulation, winding style, etc. While there are ways to calculate the final temperature of something like this it will take less time/effort to simply build one. Note that I've not mentioned the heating caused by the core itself and the wire at something other than DC - that adds to the problem.

How hot something can get before it gives out depends on a lot of things. Certainly when the copper melts is one limit. The core could also melt or otherwise fail too but another thing to consider is the temperature limit of the insulation. Letting something like a transformer get hot isn't a bad thing - heat loss or flow is proportional to the temperature difference between the object that is trying to cool and the surrounding temperature.
What you have to do is keep the temperature below a failure point.

If you really think that 15 amps is the current then take a look at a commercially made power transformer that is designed for about 15 amps. You can often see the winding - enough to get a sense of the wire size. That isn't the whole answer but it might confirm a place to start. Use wires that have a fairly high temperature insulation if you can.

Good luck.

 

[evandude]

thanks.

I'll be using typical magnet wire, with enamel (or similar) insulation so it shouldn't reduce heat transfer too badly, and the insulation itself can probably withstand more heat than I could ever throw at it...

But I guess you're right, I should probably just build the thing and see.