Wire gauge calculator

Drop, resistance and power for any gauge, computed from the AWG definition rather than looked up. Bundle derating and the temperature budget alongside, because a wire fails on two counts and one usually hides the other.

Fig 1 — Drop is 23.9 % of the 5 V rail; current is 2 A against a derated allowance of 3.50 A.

Gauge (AWG)

Current (A)

Run, one way (m)

Conductors bundled

Supply (V)

Ambient (°C)

Jacket rating (°C)

Chart current (A)

Conductor: 0.644 mm · 0.3255 mm²
Resistance: 59.7 mΩ/m at 60 °C · 597 mΩ for the loop
Voltage drop: 1.194 V · 23.9 % of 5 V
Power in the wire: 2.39 W
Bundle derating: × 0.7 for 6–15 conductors
Temperature headroom: 45 K between 60 °C ambient and the 105 °C jacket
Thermal verdict: 3.50 A allowed, 2 A drawn — inside
For 5 % drop: 15 AWG or wider

Drop is 23.9 % of the supply. On most rails this decides the gauge long before heating does.

This wire passes the thermal check and fails the voltage one. That is the ordinary case below 24 V, and it is why the drop is computed first here.

How this is calculated

Standard: AWG geometric definition; NASA/NESC NESC-RP-17-01264 for derating; Alpha Wire for the bundle factors

$d(n) = 0.005 \cdot 92^{\frac{36 - n}{39}} \text{ in}$: The whole of AWG. #36 is 0.005 in, #0000 is 0.460 in, and there are 39 geometric steps between, so the ratio between adjacent gauges is 92^(1/39) = 1.1229. Every dimension on this page comes from this expression — no gauge table is stored, so none can be stored wrongly.
$\frac{R}{L} = \frac{\rho_{Cu}}{A}\left(1 + \alpha_{Cu}(T - 20)\right)$: With ρ = 1.68 × 10⁻⁸ Ω·m and α = 0.00393 /K — the same constants the trace width and via calculators use, so a wire and a trace never disagree here about what copper does. A conductor 100 K above where it was characterised carries 39 % more resistance.
$V_{drop} = I \cdot \frac{R}{L} \cdot 2 L_{run}$: The factor of two is the mistake worth naming: current goes out and comes back, so a 5 m run is 10 m of copper. Half of all drop miscalculations are that missing 2.
$I_{allowed} = I_{chart} \cdot k_{bundle}$: The only thermal arithmetic here. The chart current comes from your wire, and k is 1.6 for a single conductor, 1.0 for 2–3, 0.8 for 4–5, 0.7 for 6–15 and 0.5 for 16–30 — a spread of 3.2 to 1 on identical copper.

Assumptions

This tool does not compute an ampacity. A wire has no current rating; its insulation does, and the published charts are empirical — the NASA NESC assessment found the standards behind them carry considerable unstated margin. Normalising a √(ΔT·d³) scaling law to one published point and asking it to predict another lands 26 % low, so a derived figure would be a confident wrong number sitting next to the exact ones. Bring the chart current from your wire and this derates it.
Resistance is computed at the ambient you give, not at the conductor temperature the current will produce. The real conductor runs hotter and therefore drops more, so the voltage figure is optimistic by however much the wire heats.
Solid conductor geometry. Stranded wire of the same AWG has slightly more copper in a slightly larger envelope, and its DC resistance is typically a couple of per cent higher than this because the strands spiral.
DC and low frequency only. Skin effect is ignored, which is correct below roughly 10 kHz for the gauges here and wrong above it.
Continuous duty. A short circuit is an adiabatic problem with an entirely different answer — the fusing current calculator solves that one.
The bundle table stops at 30 conductors. Beyond that the factor is held at 0.5 rather than extrapolated, and the tool says so.

What sets a wire's current and voltage drop

Two independent verdicts on the same wire, because it can pass one and fail the other, and the one it fails is usually the one nobody checked.

The voltage answer is exact. Gauge fixes the copper area, area and resistivity fix the resistance per metre, and the drop is that times the current times twice the run — twice, because the current has to come back. Nothing here is estimated.

The thermal answer is not, and this tool refuses to pretend otherwise. A wire has no current rating; its insulation does. Feed it the chart current from your wire's datasheet and it applies the bundle derating and reports the temperature headroom your ambient has left. Leave the chart current at zero and it returns no thermal verdict at all, which is honest rather than unhelpful — an invented ampacity would sit in the same panel, in the same typeface, as numbers that are exact.

AWG wire gauge chart

The common gauges, computed by the calculator above from the AWG definition: diameter, copper area, resistance per metre at 20 °C, and the voltage a one-amp current drops across one metre of two-wire run, out and back. There is no current column, and that is the point of the page: the current a wire may carry is set by its insulation and the conditions the ampacity chart was measured under, not by its copper.

Gauge	Diameter	Area	Resistance	Drop per amp per metre of loop
10 AWG	2.588 mm	5.261 mm²	3.2 mΩ/m	6.4 mV
12 AWG	2.053 mm	3.309 mm²	5.1 mΩ/m	10.2 mV
14 AWG	1.628 mm	2.081 mm²	8.1 mΩ/m	16.1 mV
16 AWG	1.291 mm	1.309 mm²	12.8 mΩ/m	25.7 mV
18 AWG	1.024 mm	0.823 mm²	20.4 mΩ/m	40.8 mV
20 AWG	0.812 mm	0.518 mm²	32.5 mΩ/m	64.9 mV
22 AWG	0.644 mm	0.326 mm²	51.6 mΩ/m	103.2 mV
24 AWG	0.511 mm	0.205 mm²	82.1 mΩ/m	164.1 mV
26 AWG	0.405 mm	0.129 mm²	130.5 mΩ/m	261.0 mV
28 AWG	0.321 mm	0.081 mm²	207.5 mΩ/m	414.9 mV
30 AWG	0.255 mm	0.051 mm²	329.9 mΩ/m	659.8 mV

Every three gauges halves the area and doubles the resistance, and every ten gauges is a factor of ten, which is what the 92 in the definition encodes. The last column is the number that usually binds first on a low-voltage run: 24 AWG drops a volt every six metres at one amp, and a 5 V rail cannot spare that.

Worked example: 22 AWG carrying 2 A over 5 m

A 22 AWG sensor cable, 2 A, 5 m each way, in a loom of eight conductors, inside a 60 °C enclosure on a 5 V rail. The wire's datasheet allows 5 A at a 35 K rise.

Geometry     d = 0.005 × 92^((36−22)/39) = 0.0253 in = 0.644 mm
             A = π(0.644/2)² = 0.326 mm²

Resistance   R/L = 1.68e−8 / 0.326e−6 = 0.0516 Ω/m at 20 °C
             loop = 0.0516 × 2 × 5 = 0.516 Ω

Drop         V = 2 A × 0.516 Ω = 1.03 V  →  20.6 % of a 5 V rail

Thermal      5 A chart × 0.7 for 6–15 conductors = 3.5 A allowed
             2 A drawn, so inside the rating
             105 °C jacket − 60 °C ambient = 45 K of headroom

The thermal check passes comfortably. The voltage check loses a fifth of the rail. This is the ordinary result below 24 V, and it is why the tool computes the drop first.

Where the wire rating stops being valid

Resistance is taken at ambient, not at temperature. A conductor carrying current runs hotter than its surroundings and therefore drops more than this says. The error is bounded by copper's coefficient: 100 K of rise is 39 % more resistance. If the wire is working hard, add the rise to the ambient and run it again.
Solid conductor geometry. Stranded wire of the same AWG carries slightly more copper in a slightly larger envelope, and its DC resistance is typically a couple of per cent higher because the strands spiral — more length per metre of cable.
DC and low frequency. Skin effect is ignored. That is right below roughly 10 kHz for these gauges and wrong above it.
Continuous duty only. A short circuit is the opposite regime: over in milliseconds, no heat leaving the copper, and the question is whether the conductor survives rather than how warm the jacket gets. The fusing current calculator solves that one, and the two answers differ by more than an order of magnitude.
Wire, not board copper. A round conductor in air sheds heat differently from a ribbon bonded to laminate. Matching 24 AWG in 1 oz copper takes a trace 5.9 mm wide, so a wire chart transferred to a PCB is wrong by a large factor. Thetrace width calculator applies IPC's curves, which were measured on laminate.

Common wire gauge mistakes

Forgetting the return conductor. A 5 m run is 10 m of copper. This is the single most common error in wire sizing and it is a factor of two.
Sizing for the supply's rating instead of the load's draw, then finding the gauge was chosen for a current that never flows.
Taking an ampacity figure from a chart that assumed a single conductor in free air, and then lacing the wire into a loom. That is 1.6 down to 0.5 — the same wire losing more than three times its allowance to its neighbours.
Using room temperature as the ambient for a wire inside a sealed enclosure. The ambient spends the temperature budget before any current flows, and what is left is what the rise is allowed to be.
Reading the jacket rating and ignoring the plating. Tin is 150 °C, silver 200 °C, nickel 260 °C; a silver-plated conductor inside a 150 °C jacket is a 150 °C wire.
Sizing a wire for a fault with an ampacity chart. Different regime, different equation, answers that are not close.