SizingKitRun check
Voltage drop calculator
Give it a conductor you have already chosen, the current it carries and how far it goes, and it returns the volts lost, the percentage of nominal and the voltage that actually reaches the load — with the multiplier, the constant and the circular mils that produced the figure printed beside it. It then shows how far every other size reaches at the same load. Free, no account, and it will not pick a conductor for you: that decision weighs two other limits this page cannot see.
- 100% free
- No signup
- Single and 3-phase
- Parallel sets
- Reach table for 21 sizes
The run as it is
A conductor you have already picked, a load and a distance. What comes back is what arrives at the far end of it.
10,383 circular mils · 5.26 mm²
Switching metal reloads K below, because the constant is a property of the metal.
The current that actually flows, not the breaker rating. Drop is proportional to it, so a circuit loaded to half its breaker drops half as much.
The percentage is taken against this figure, so 120 V and 240 V give the same volts lost and different percentages.
Measure the cable, not the floor plan. Two 90° bends up a wall and back down add more than most people allow for.
Multiplier 2 — A single-phase circuit's current traverses two conductor lengths, out and back, so the one-way run length is doubled.
Two identical sets halve the drop, because they halve the resistance. 1 for an ordinary run.
Default 12.9. Rises with conductor temperature: about 11.1 at 20 °C, 12.9 at 75 °C, 13.7 at 90 °C. A lightly loaded conductor runs cooler than 75 °C and drops less than K predicts.
Dropped down the run
13.05 V
As a percentage
5.44%
227 V arrives at the load under 30 A, down from 240 V at the panel. That is 5.44% against the 3% branch-circuit figure and the 5% total figure, which are recommendations in informational notes rather than requirements you can be failed on.
The arithmetic
2 × 12.9 × 30 A × 175 ft ÷ 10,383 cmil = 13.05 V
The multiplier is not the √3 of the power identity. Here it counts conductor length: on a single-phase circuit the current goes out and comes back, which is two run lengths of copper, and on a balanced three-phase circuit the return current splits between the other two phases so the effective figure is √3 rather than 2.
If the load insists on its power
A heater draws less current as the voltage sags and the drop partly corrects itself. A motor under load, a VFD or a switching supply does the opposite: it pulls harder to hold its power, which drops more voltage, which makes it pull harder again. Holding 7,200 VA constant, this run settles at 31.8 A and 5.77%. Two models of the extremes; a real load sits between them.
How far 30 A reaches on each size
| Size | Reaches 3% at | Reaches 5% at |
|---|---|---|
| 14 AWG | 38 ft · 12 m | 64 ft · 19 m |
| 12 AWG | 61 ft · 19 m | 101 ft · 31 m |
| 10 AWG | 97 ft · 29 m | 161 ft · 49 m |
| 8 AWG | 154 ft · 47 m | 256 ft · 78 m |
| 6 AWG | 244 ft · 74 m | 407 ft · 124 m |
| 4 AWG | 388 ft · 118 m | 647 ft · 197 m |
| 3 AWG | 490 ft · 149 m | 816 ft · 249 m |
| 2 AWG | 617 ft · 188 m | 1,029 ft · 314 m |
| 1 AWG | 779 ft · 237 m | 1,298 ft · 395 m |
| 1/0 AWG | 982 ft · 299 m | 1,636 ft · 499 m |
| 2/0 AWG | 1,238 ft · 377 m | 2,063 ft · 629 m |
| 3/0 AWG | 1,561 ft · 476 m | 2,602 ft · 793 m |
Reach is proportional to area, so each step up the gauge buys about 26% more distance and three steps buys double. It is not a size recommendation: a conductor also has to carry the current, and on the small sizes the 240.4(D) cap decides before either. The sizing page weighs all three.
3%: NEC 210.19(A), Informational Note (branch circuits): conductors sized to prevent a voltage drop exceeding 3% at the farthest outlet provide reasonable efficiency of operation. Informational notes are not requirements — NEC 90.5(C). The subdivision this note sits under has moved between editions; the article is 210.19.
5%: NEC 215.2(A), Informational Note (feeders): the maximum total voltage drop on both feeders and branch circuits to the farthest outlet should not exceed 5%. Informational notes are not requirements — NEC 90.5(C). Older editions print this as a Fine Print Note (FPN) and number the subdivision differently.
The K method models the conductor as pure resistance. It ignores inductive reactance, which on large conductors is a large share of the impedance, and it ignores power factor, conductor spacing, and whether the raceway is steel or PVC. It is close enough on small branch circuits — 14 AWG through about 1/0 at unity power factor — and progressively wrong above that, understating drop on long feeders of 250 kcmil and larger and on any load below about 0.9 power factor. Where the answer matters, use the AC resistance and reactance of NEC Chapter 9, Table 9, which tabulates both for steel and PVC conduit, or the impedance figures from the cable manufacturer.
How to check the drop on a run you have already pulled
Loaded with 10 AWG copper carrying 30 A over 175 ft at 240 V — the shape of run where the answer surprises people.
Describe the conductor as installed
Size, metal, and how many sets are in parallel. Two identical sets halve the drop because they halve the resistance, but the code will not parallel anything smaller than 1/0, so the field refuses that combination rather than returning a number nobody can install.
Enter the current that flows, not the breaker
Drop is proportional to current, so a 50 A circuit carrying 22 A drops less than half what its breaker rating suggests. Measure it if you can. Then give the one-way cable length, not the round trip and not the distance across the floor plan — two bends up a wall and back down add more than most estimates allow.
Read the reach table under the answer
It gives the one-way distance at which every size spends the 3% figure and the 5% figure at the load you entered. That is the trade in its plainest form: reach scales with area, so one gauge step buys about 26% more distance and three steps buys double.
Technical specifications
| Formula | multiplier × K × amps × one-way feet ÷ circular mils, divided by the number of parallel sets. The multiplier is 2 for a single-phase circuit because the current traverses two conductor lengths, and √3 for a balanced three-phase one |
|---|---|
| K defaults | 12.9 Ω·cmil/ft for copper and 21.2 for aluminum, both at 75 °C, both editable. A conductor at 20 °C is nearer 11.1 and 18.2, so a lightly loaded run drops less than the default predicts |
| Where the method stops working | It models the conductor as pure resistance — no inductive reactance, no power factor, no distinction between steel and PVC raceway. Close on branch circuits from 14 AWG to about 1/0 at unity power factor, and progressively optimistic above that |
| Sizes offered | 21, from 14 AWG to 1000 kcmil, each shown with its circular mils and its square millimeters so a metric cable schedule lines up |
| Parallel sets | 1 through 10. Sizes below 1/0 are rejected, and every conductor of a paralleled set has to match in length, material, size, insulation and termination method |
| The 3% and 5% figures | Informational notes at NEC 210.19(A) for branch circuits and 215.2(A) for feeders. NEC 90.5(C) states that explanatory material in an informational note is not an enforceable requirement of the code |
| Reach table | One-way feet and meters at which each of the 21 sizes reaches 3% and 5% at the current entered, with the size in the form highlighted |
| Data handling | The run stays in the browser: no circuit description, panel voltage or load figure is transmitted or logged |
Frequently asked questions
Is 3% voltage drop actually a code requirement?
No, and this is the most consequential misunderstanding in the subject. The 3% branch-circuit figure and the 5% total figure appear in informational notes — at 210.19(A) and 215.2(A) respectively — and NEC 90.5(C) says explanatory material in an informational note is not an enforceable requirement of the code. An inspector cannot fail a branch circuit for 4% drop on the NEC alone. What can bind you is a project specification, a local amendment, or an energy code, and some of those tighten it to 2%; equipment listings sometimes impose their own limit as well. The figures are good engineering and they are not law.
Do I use the one-way length or the round-trip length?
One way, because the multiplier already accounts for the return path. This is the most common arithmetic error on the subject and it is a factor of two in either direction: enter the round trip and the answer doubles, or drop the multiplier from a hand calculation and it halves. The 2 in the single-phase formula is exactly the out-and-back conductor, which is why a formula that already has it in cannot also take a doubled length.
Why is the three-phase multiplier √3 rather than 2?
Because the return current is shared. In a balanced three-phase circuit each phase conductor carries current whose return path is the other two phases rather than a dedicated neutral, and working the line-to-line drop through gives √3 ≈ 1.732 rather than 2. Watch which √3 you are holding: the one in the power identity, where three-phase power is √3 × V × I × pf, is a different quantity from this one, and mixing them produces an answer that is out by 15% and still looks entirely reasonable.
My feeder is 250 kcmil. Why the warning about the method?
Because the K method models a conductor as a resistor and a large conductor is not one. Above roughly 1/0, inductive reactance becomes a significant share of the impedance, and reactance does not fall as you add copper the way resistance does — so the K method progressively understates drop on large feeders, and understates it further on any load below about 0.9 power factor. The honest tool for that is NEC Chapter 9 Table 9, which tabulates AC resistance and reactance separately for steel and for PVC raceway, or the impedance figures from the cable manufacturer. This site does not reproduce Table 9, and rather than interpolate something plausible it tells you where the boundary is.
Does running two sets in parallel really halve the drop?
Yes, for the same reason two identical resistors in parallel have half the resistance: the current splits and each set carries half of it over the same length. NEC permits paralleling only in 1/0 and larger — the article has been renumbered across editions, appearing as 310.10(G) in 2020 and 2023, 310.10(H) in 2017 and 310.4 before that — and it requires every conductor of the set to be the same length, material, size, insulation and termination. Getting one of those wrong makes the sets share current unevenly, which is a heating problem rather than a voltage problem.
Should I calculate drop at the breaker rating or the measured load?
At the load, because voltage drop is a consequence of current that actually flows rather than current that is permitted to flow. A 60 A feeder serving a 27 A load drops what 27 A drops. Design work is a different question: for a circuit that will grow, or a receptacle circuit whose future load is unknown, calculating at the breaker rating is a deliberate margin. Just be clear which one you are doing, because the two answers differ by whatever the difference in current is.
Why does a motor draw more current when its supply voltage sags?
Because a motor under mechanical load takes the power the load demands, and power is voltage times current: lower the voltage and the current has to rise to compensate. That makes drop self-reinforcing on a motor circuit — more current means more drop, which means still more current — until it settles at a new operating point, which is what the compounding figure under the main result estimates. A resistive load does the opposite and is self-correcting. Long motor runs are where this matters most, because the reduced voltage also cuts starting torque roughly with the square of the voltage.
About voltage drop, and what the 3% figure really is
Voltage drop is Ohm's law applied to something people forget is a resistor. A conductor has resistance proportional to its length and inversely proportional to its area, so the volts lost on a run are set by four things — how much current, how far, how much metal, which metal — and by nothing else that this method can see. The constant K folds the metal's resistivity and a stranding allowance into one number: 12.9 for copper and 21.2 for aluminum at 75 °C. It is neither a code figure nor a physical constant, which is why it sits in an editable field here with its temperature dependence stated. A conductor running cool has less resistance than the default assumes, and one running at its 90 °C rating has about 6% more.
The reason this page refuses to hand back a conductor size is not modesty, it is that drop is the weakest of the three constraints on a conductor and the only one that is not enforceable. Ampacity is a requirement, the 240.4(D) cap is a requirement, and the 3% figure is a recommendation in a note. A tool that sized from drop alone would routinely return a conductor that the code forbids — 14 AWG on a 20 A circuit has plenty of area for a short run and is illegal anyway — so the sizing decision lives on the wire size calculator, which weighs all three and names the one that bound. If the current itself is what you need first, the kW to amps calculator and the three phase power calculator get you there, and the latter is also where the other √3 lives.
What voltage drop actually costs is worth being specific about, because the arguments for spending money on copper are usually made badly. Resistive heat is real but small: the wasted power is the drop times the current, so a 3% drop on a 30 A branch circuit is about 216 W turned into warm conduit while the load runs. The equipment effects are larger. Incandescent output falls roughly with the cube of the voltage, motor starting torque falls with its square, electronic supplies compensate by drawing more current and make the drop worse, and long control runs can leave a contactor coil unable to pull in while its holding current is perfectly happy. Against that, the fix — one or two gauge sizes over a long run — also costs raceway, and whether the conductors still fit is Chapter 9 arithmetic on the conduit fill calculator, with the geometry behind both on the wire gauge chart. None of these figures is a design approval; the numbers are here so a qualified person can check them against the code book and their own conditions.
Where the run details stay
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type — loads, lengths, nameplate ratings, the rates your utility charges you — is uploaded, logged or kept, which is also why the calculators carry on working in a mechanical room with no signal.
The reach table is recomputed on every keystroke from the figures in the form, which is why it responds instantly and why there is nothing to clear afterwards: close the tab and the run description is gone.