Wire Size Calculator

Wire size has two limits, and the cable has to clear both

Almost every wrong answer on wire sizing comes from treating it as one question. It's two, and they have nothing to do with each other.

Ampacity protects the cable. Current through resistance makes heat, that heat has to leave the conductor through the insulation, and if it can't leave fast enough the conductor sits above its rated temperature. Nothing blows up on day one. The insulation slowly turns brittle until it cracks. Ampacity depends on the metal, the insulation rating, how the cable is installed and how hot the surroundings are. It doesn't care at all how long the run is.

Voltage drop protects whatever sits on the far end. The wire's resistance eats part of your supply voltage before it arrives, and that loss grows in direct proportion to distance. Voltage drop doesn't care in the slightest how the cable is installed.

So a short high-current run, like a 40 A range circuit five meters from the panel, is a pure ampacity problem, and the voltage-drop answer looks almost comically thin. A long light run, like lighting at the back of a property 150 meters out, is a pure voltage-drop problem, and the ampacity answer would be a wire you couldn't physically terminate. Anything in between is governed by whichever is larger. That's why this calculator won't show you just one number.

An undersized cable doesn't fail. It ages.

This is the part that makes wire sizing feel abstract until it isn't. Run a conductor 20% over its ampacity and nothing happens. No trip, no smell, no flicker. The insulation just runs hotter than its rated temperature, and thermal aging adds up: every hour above rating uses up part of the material's life.

Years later the insulation has hardened. It cracks where the cable bends inside a box, at a staple, or where it passes a joist. Now you have conductor against conductor, or conductor against a metal enclosure, inside the wall, with a breaker that never saw a fault current big enough to react. The failure looks like a mystery. It was decided at install time, with a cable one size too small.

That's also why "it has worked fine for six years" proves nothing. Six years of running hot is how the damage happens. It doesn't defend the install.

The breaker protects the wire, not the appliance

An overcurrent device has exactly one job: open the circuit before the conductor behind it overheats. Your equipment has its own protection, so the breaker isn't there for it, and the breaker isn't sized to the load either. It's sized to the cable.

That leaves you a rule with three terms. The breaker rating has to be at least the load current, or it nuisance-trips, and at most the conductor's corrected ampacity, or it fails to protect. When no standard rating fits between the two, don't pick the closest one. Go up a wire size, which raises the ceiling until a rating fits.

NEC 240.4(D) is the rule people miss here, because it contradicts what the ampacity table seems to say. A 12 AWG copper conductor reads 30 A in the 90 °C column. You still can't protect it above 20 A. Same for 14 AWG at 15 A and 10 AWG at 30 A, however good the insulation is. The 90 °C column exists mostly so derating starts from a higher number. It does not let you hang a bigger breaker on a small wire. The calculator applies this cap automatically, which is why selecting 90 °C sometimes changes nothing at all for small conductors.

Voltage drop is why the motor won't start

A few volts sound like nothing. For resistive loads they nearly are: a heater at 5% low just runs about 10% weaker and nobody notices. Motors are another story.

Starting torque in an induction motor falls with the square of the applied voltage. Ten percent low at the terminals means roughly 19% less starting torque. The inrush current during that start is several times the running current, so the drop is worst at the exact moment the motor needs voltage most. You see it as a well pump or compressor that hums, draws heavily, gets hot and finally trips on thermal overload, while the panel reads perfectly normal voltage with the motor off. Electronics on long runs behave differently again: switch-mode supplies hold their output and just draw more current as voltage sags, which makes the drop worse.

The NEC's suggested limits are 3% on a branch circuit and 5% total including the feeder. Both live in informational notes, so neither is enforceable. They're good engineering practice that an inspector can't cite you for. Design to 3%, and know that going to 4% on a lightly used lighting run is a trade you can defend. Nobody can write it up as a violation.

The ampacity derating that quietly disappears from spreadsheets

Published ampacity assumes 30 °C ambient and no more than three current-carrying conductors bundled together. Real installations break both assumptions all the time, and each one multiplies the capacity down.

Look hard at that last line. A 10 AWG THHN conductor reads 40 A in the 90 °C column. Put it in a 50 °C attic (0.82) sharing a conduit with two other circuits (0.80) and it's a 26 A conductor. That is below the 30 A that 240.4(D) would have allowed anyway, so now the derating sets the limit, not the code cap. The cable didn't change. The conditions did. The calculator prints both factors and their product as visible numbers instead of folding them silently into the result, because a derate of 0.60 should make you uncomfortable.

AWG wire size is a logarithmic scale, not a lookup table

American Wire Gauge is defined geometrically: 36 AWG is 0.005 inch in diameter, 4/0 is 0.46 inch, and there are 39 equal ratio steps between them. Each gauge number is the previous one multiplied by a fixed factor. That definition gives you some handy rules:

• Going down three gauge numbers multiplies the cross-section by 2.005, near enough to double.
• Going down six gauge numbers doubles the diameter, and quadruples the area.
• Nothing in the scale lands on a round metric number, ever.

That's why the popular equivalence tables mislead. 12 AWG is 3.31 mm², not 3.5. 10 AWG is 5.26 mm², not 6; a 6 mm² cable carries 14% more copper. 4/0 is 107 mm², not 120. Above 4/0 the scale is dropped entirely and sizes are quoted in thousands of circular mils, where 250 kcmil is 127 mm². This tool reports every result in both systems, including the fractional AWG equivalent, so a size like 6 mm² shows up as roughly 9.4 AWG instead of being forced into a gauge it doesn't equal.

Copper or aluminum wire

Aluminum has about 61% of copper's conductivity: 0.0282 Ω·mm²/m against 0.0172. For the same voltage drop it needs 1.64 times the cross-sectional area, which usually means two full sizes up. It really is cheaper and much lighter for long feeders and service entrances, and it's the normal choice there.

You can't swap it in at the same gauge, though. Aluminum also creeps under the pressure of a terminal screw and forms a resistive oxide when a joint is disturbed. So terminations must be rated AL or CU-AL, torqued to spec, and generally treated with antioxidant compound. The old problems with aluminum branch wiring were connection problems, not conductor problems. Small metric sizes below 16 mm² are left out of the table here on purpose, since aluminum isn't used at that scale in building wiring.

Reading the wire size result

The headline number is the smallest size that clears both limits. The badge under it tells you which limit produced it, and that's the part to remember. If it says voltage drop, the answer changes when you move the load. If it says ampacity, it changes when you add circuits to the conduit or the attic gets hot. The two cards show each limit's verdict separately, so you can see how much headroom you have on the side that didn't govern.

The neighboring-sizes table is there so the decision isn't just yes or no. One size up often takes a 2.9% drop down to 1.8% for a modest amount of copper, and you want to know that before you pull the cable, not after. One size down shows you exactly which limit it fails and by how much.

Everything runs in your browser, and none of it replaces the datasheet for the cable you're actually buying, your local code, or a licensed electrician. Table values here are typical published figures for common insulations at 30 °C ambient. Conditions of use in a real installation are for the person who signs off on it.