Cable Size Calculator

A cable has to pass two tests, and this calculator runs both and gives you whichever demands the larger conductor. It has to carry the current without cooking its own insulation, and it has to reach the accessory with enough volts left to run whatever is plugged into it. Those are unrelated failure modes and either can be the one that governs.

Test one is the current rating. Here is the first real difference from American practice. The NEC publishes one ampacity per conductor and then multiplies it by correction factors for the conditions. BS 7671 publishes a separate rating column for each installation method — so how the cable is installed decides which number you start from, not what you multiply. A 2.5 mm² copper cable is rated 27 A clipped direct, 24 A in conduit and 30 A in free air: three different starting points for one cable. Choosing the method is therefore part of the sizing, not an adjustment applied afterwards, and the four presets here stand in for BS 7671's reference methods B, C, D and E/F.

Test two is voltage drop, and it uses mV/A/m. This is the bigger conceptual difference. The NEC-based pages on this site reconstruct the cable's resistance from first principles every time: a resistivity constant, the length, and the conductor area in circular mils. The metric standards skip all of that by publishing the finished answer for each size — the millivolts that cable drops per amp carried per metre run. The calculation collapses to one multiplication:

Vdrop (mV) = (mV/A/m) × Ib × L, divided by 1,000 for volts, then taken as a percentage of the supply voltage.

There is no conductor area in that formula and no resistivity constant, because the table already accounted for both. There is also no √3. Three-phase drop is its own published column rather than a multiplier applied to the single-phase figure, which is why nothing in this calculator's code multiplies by 1.732 — the phase relationship was handled by whoever compiled the table. The trade is transparency for reliability: you cannot see the physics working, but there is correspondingly less to get wrong.

The drop target is different too. BS 7671 gives 3% for lighting and 5% for other uses, measured from the origin of the installation. The 4% this page defaults to is not in the regulations — it is the general working figure a lot of UK design practice lands on, and it is the default because the calculator does not ask what the circuit feeds. The NEC's commonly cited 3% is a different number from a different document measuring from a different point, so it is not a figure to carry across.

A 25 A design current on a 230 V single-phase supply, copper cable clipped direct,20 metres from the board, against a 4% drop target.

The current-rating path. Reading down the clipped-direct column: 1.5 mm² is rated 20 A, short of 25. 2.5 mm² is rated27 A, which clears it with2 A to spare. The rating-constrained answer is2.5 mm².

The voltage-drop path. The budget is 4% of 230 V =9.2 V, or 9200 mV. Spread over 25 A × 20 m = 500 amp-metres, that allows a cable of no worse than18.4 mV/A/m. 2.5 mm² is18 mV/A/m, which just fits. The drop-constrained answer is 2.5 mm² as well.

The answer. Both checks land on2.5 mm², so neither one "wins" — this is the balanced point for the circuit. The actual drop is18 × 25 × 20 =9,000 mV =9 V, which is 3.91%of 230 V. The load sees 221 V.

And now the part worth remembering. That result clears the 4% target by 0.09 of a percentage point. At 25 A this cable runs out at about 20.4 m — so a route that turns out to be 21 metres rather than20 needs 4 mm², one full size up, and the binding constraint flips to voltage drop. A single metre of unmeasured route between the board and the accessory is the entire margin. This is exactly why measuring along the cable rather than across the drawing matters more on metric domestic work than the numbers suggest: at 230 V the volt budget is half what a 400 V or 480 V circuit gets, and circuits sit much closer to their limits.

This is the table the calculator reads. It is worth having open rather than running a calculation when you are checking a design someone else produced, or when you want to see how much a different installation method would buy you before committing to a route.

Read the four rating columns as four different cables. They are the same conductor in four thermal environments, and the spread between the worst and the best is around 40% at every size — a bigger effect than one step up the size ladder. The mV/A/m columns, by contrast, do not move with installation method at all: voltage drop is a property of the conductor, not of what surrounds it. That asymmetry is the single most useful thing on this page. If a design fails on the rating check, changing how it is installed may fix it; if it fails on drop, only a bigger cable, a shorter route or a higher voltage will.

Aluminium is offered from 16 mm² up and is not listed below that, so the copper table is shown here. Its ratings run about 22% lower and its mV/A/m about 64% higher than the copper figures at the same size — switch the material in the calculator to see it applied.

These numbers are provisional. They are illustrative placeholders that behave correctly relative to one another, not transcriptions from a published standard — BS 7671 Appendix 4 and AS/NZS 3008 are both commercially licensed, and the two do not agree with each other on reference conditions or installation-method definitions. Cross-check anything you intend to install against the current published table for your own jurisdiction.