The rule that isn’t a rule
Ask an electrician for the voltage drop rule and you’ll hear “3 percent” — keep the drop on a branch circuit within 3% of nominal voltage, and within 5% for the feeder and branch combined. What surprises people is where those numbers live in the National Electrical Code: not in enforceable text, but in the informational notes to 210.19(A) and 215.2(A). The NEC is a safety code, and a circuit that wastes voltage isn’t inherently unsafe — so 3%/5% is a recommendation for “reasonable efficiency of operation,” not a violation you can be red-tagged for on its own.
It still runs the industry, because the consequences of ignoring it are real even when they aren’t code violations: dim or flickering lighting at the end of long runs, motors that run hot and die early, nuisance tripping of sensitive equipment, and energy burned as heat in the walls. Most project specifications simply adopt the 3%/5% figures as hard requirements, and the international counterpart, IEC 60364-5-52, commonly applies a 4% limit. In practice you design to these numbers; the fine print about enforceability matters mostly in arguments about existing installations.
Why wire drops voltage at all
Copper is a good conductor, not a perfect one. Every foot of wire has resistance, tabulated in NEC Chapter 9, Tables 8 and 9 — a 12 AWG copper conductor runs about 6.39 ohms per kilometre at the 75 °C design temperature, a 10 AWG about 4.02, an 8 AWG about 2.56. Push current through that resistance and Ohm’s arithmetic takes its cut before the load sees anything. For a single-phase circuit the drop is Vd = 2 × I × L × R / 1000, with I the load current in amps, L the one-way run in metres, and R in ohms per kilometre.
The 2 is the detail most back-of-envelope estimates forget: current flows out along the hot conductor and back along the neutral, so it traverses the run twice, and both passes drop voltage. Balanced three-phase circuits get a discount — the return currents cancel in the shared conductors and the line-to-line drop works out to √3 (about 1.732) times the per-conductor value instead of 2: Vd = √3 × I × L × R / 1000. Same wire, same current, same distance, and three-phase delivers power with about 13% less drop, one of several quiet reasons industrial loads run three-phase.
A 100-foot circuit, worked honestly
This guide’s running example is the classic detached-garage problem: a 120 V, 20 A single-phase branch circuit on 12 AWG copper — the standard pairing for a 20 A breaker — with a one-way run of 100 ft (30.48 m). The drop: Vd = 2 × 20 × 30.48 × 6.39 / 1000 = 7.79 V. As a percentage, 7.79/120 = 6.49% — more than double the 3% guideline — and the receptacle at the far end sees only 112.2 V while the circuit is fully loaded. Nothing about this violates the wire’s ampacity; the conductor is perfectly safe. It’s just delivering poor voltage.
That is the essential insight of voltage-drop design: ampacity and voltage drop are separate checks that happen to involve the same wire. Ampacity asks whether the conductor overheats at this current — a safety question, answered by NEC 310.16 and its derating factors. Voltage drop asks whether the load gets usable voltage at this distance — a performance question the breaker-and-wire pairing tables know nothing about. Short runs pass both checks with the ampacity-minimum wire; somewhere around the length of a long house, distance starts to govern, and the wire that satisfies the breaker stops satisfying the load.
Fixing it: the upsize ladder
The cure for voltage drop is copper cross-section, and the example shows how the ladder climbs. Upsizing the 100-ft run to 10 AWG (4.02 Ω/km) gives Vd = 2 × 20 × 30.48 × 4.02 / 1000 = 4.90 V — 4.08%, better but still over the 3% line. It takes 8 AWG (2.56 Ω/km) to clear it: 3.12 V, or 2.60%. Two full size steps above the ampacity minimum, purely for distance. The drop scales linearly in both current and length, so a lightly loaded circuit tolerates a long run, and halving the load halves the drop; for big feeders, running parallel conductors per phase divides the effective resistance the same way.
The solvable question hiding in the formula is “how far can I go?” — set Vd to 3% of the system voltage and solve for L. For the 12 AWG, 20 A, 120 V example that boundary lands at just 14.1 m — about 46 ft one way — which is why voltage drop is invisible inside most rooms and unavoidable at the garage, the well pump, and the gate. The voltage drop calculator runs this whole analysis from a built-in copper and aluminium resistance table in both AWG/kcmil and metric IEC sizes, grades the result against the NEC 3%/5% and IEC 4% limits, and plots drop against run length so the crossover point is visible before you buy wire.
Reactance, power factor, and the three-phase version
The full formula carries one more term: Vd = 2 × I × L × (R·cosφ + X·sinφ) / 1000, where X is the conductor’s reactance (about 0.13–0.19 Ω/km for conductors in conduit) and φ the load’s power-factor angle. At unity power factor sinφ = 0 and the reactive term vanishes, which is why the resistive shortcut works for heating and lighting loads. For lagging loads — motors, transformers — and for the larger conductor sizes where resistance is small enough that reactance competes with it, the R·cosφ + X·sinφ form is the honest one.
A three-phase worked example to close the loop: a 400 V feeder on 50 mm² copper (0.463 Ω/km) carrying 100 A over 60 m at unity power factor drops Vd = √3 × 100 × 60 × 0.463 / 1000 = 4.81 V, which is 1.20% — comfortably inside every limit, delivering 395.2 V at the load. Note what made it easy: three-phase’s √3 instead of 2, a fat conductor, and a modest run. The same current at 120 V single-phase over the same distance would be a very different story, because the percentage is taken of a much smaller base.
The guideline in one pass
Recap the garage circuit: 120 V, 20 A, 100 ft one way on 12 AWG copper drops 7.79 V — 6.49%, versus the NEC informational-note guideline of 3% branch and 5% feeder-plus-branch — leaving 112.2 V at the load. 10 AWG improves it to 4.08%; 8 AWG passes at 2.60%; and the 3% boundary for the original wire sits at about 46 ft. The rule itself is a recommendation, but the physics behind it is not negotiable: two lengths of resistance per single-phase run, √3 for three-phase, drop growing linearly with amps and feet, and the fix always spelled in copper. Check ampacity for safety, check drop for performance, and let whichever asks for the bigger wire win.