Wire gauge and voltage drop

Both limits at once — the current the wire can carry without cooking, and the voltage that never reaches the far end.

AWG / mm²Voltage dropBundle deratingHarness design
Conductor
Material
Return path

Quoting one-way drop on a two-wire run is the single most common error in harness calculations.

Load
Environment
Only wires carrying current at the same time count.

Voltage drop

V

Drop as % of supply
Cross-section
Loop resistance
Power lost in the wire
Derated ampacity
Smallest gauge that works
Drop against gauge — where your budget line crosses is the smallest wire you can use
Ampacity against bundle size — the loom, not the wire, sets the limit

The physics

Gauge is logarithmic

d(mm) = 0.127 · 92^((36 − AWG) / 39)

AWG runs backwards — bigger number, thinner wire — and each three gauges roughly halves the cross-sectional area. That is worth internalising: going from 18 AWG to 15 AWG halves your resistance, and 18 to 12 quarters it.

Voltage drop

R = ρ · L / A          V_drop = I · R · n_paths

With a two-wire feed the current goes out and comes back, so the drop is counted twice. This is the single most common error in harness calculations — quoting one-way drop on a two-wire run and then wondering why the load is a volt low.

Copper is a thermometer

ρ(T) = ρ₂₀ · (1 + 0.00393·(T − 20))

Copper resistance rises about 0.4% per degree. A wire that starts at 25 °C and settles at 85 °C has 24% more resistance than the cold figure — so it drops more voltage, dissipates more heat and gets hotter still. In marginal designs that feedback is what turns a warm wire into a failed one.

Bundle derating

Single-wire ampacity assumes free air. Inside a loom every wire is heating its neighbors and none of them can radiate. The approximation used here follows the general shape of the aerospace derating curves — roughly proportional to 1/√n — and reaches about 0.5 by 20 wires. For anything safety-related, use the derating table from the applicable standard, not this.

Where this stops being accurate

  • Ampacity figures are for chassis wiring in free air at 25 °C with a 30 °C rise. Power-transmission tables give much lower numbers for the same gauge, for different reasons.
  • Bundle derating here is an approximation. Aerospace and automotive standards give specific curves — use them.
  • DC only. Skin effect and proximity effect matter above a few kilohertz and change the effective resistance considerably.
  • Connector and crimp resistance is not included, and a bad crimp will exceed the whole wire's resistance.
  • Insulation temperature rating is a ceiling for the conductor, not a design target. Continuous operation near it shortens life sharply.

Design guidance, not a manufacturing instruction. Verify every result against your fabricator's stack-up and the applicable standard before release. Closed-form models are approximations; the fab's field solver and process window are the authority.

Questions

What people ask about this

Why does a wire that carries the current still fail?

Because ampacity and voltage drop are different limits, and one of them usually bites first. A 24 AWG wire will happily carry 2 A without overheating, but over 5 m it drops nearly a volt — which is fatal on a 3.3 V rail and irrelevant on a 48 V one. Check both, then take the larger gauge.

What does bundle derating actually do?

A wire in free air sheds heat in every direction. The same wire in the middle of a 30-way loom is surrounded by other wires doing the same thing, so its heat has nowhere to go. Derating factors of 0.5 or lower are normal for large bundles, meaning a wire rated 8 A alone may only be good for 4 A in the loom. This is a leading cause of harness failures in vehicles and aircraft, and it is invisible on a bench test with a single wire.

Should I use AWG or mm²?

Whatever your standard and your supply chain use. AWG dominates North America and consumer electronics; mm² dominates Europe, automotive and industrial. The conversion is shown both ways below because harness drawings routinely mix them, and a metric wire that is "about the same" as an AWG size is often 20% smaller in copper.

Does stranding change the current rating?

Barely for DC — total copper area is what matters, and a stranded wire has slightly less copper than a solid one of the same nominal size. Stranding matters for flex life, termination method and skin effect at high frequency. For a harness that moves, stranded is not optional.

Designing a harness that has to survive vibration and heat?

Gauge is the easy part. Strain relief, connector selection, sealing and service loops are what decide whether it lasts. Send us the envelope and the loads.