The datasheet says 14 AWG. The tender says 2.5 mm². The electrician says 7/029. All three describe a wire by how much copper is in its cross-section, and none of them is the same size as the other two. AWG is a geometric series of diameters from an American drawing bench, mm² is the IEC area, and SWG and the 7/029 family are British survivals that still turn up in Pakistani wiring and on old drawings. This guide shows how each is defined, how to convert between them honestly, and the one thing a conversion never gives you: the current the wire may carry.
AWG: a geometric series from a drawing bench
The American Wire Gauge, standardised in ASTM B258, is fixed by two diameters: 0000 (written 4/0) is 0.4600 inch, 11.68 mm, and 36 AWG is 0.0050 inch, 0.127 mm. Between them are 39 equal ratio steps, so each gauge step multiplies the diameter by the 39th root of 92, which is 1.1229, and the area by 1.2610. From those two numbers the whole table follows:
Because the series is geometric, its arithmetic is in ratios. Three gauge steps double the area (1.261³ = 2.005): 13 AWG has twice the copper of 16 AWG, 10 AWG twice that of 13. Six steps double the diameter. Ten steps multiply the area, and so the resistance, by ten (1.261¹⁰ = 10.16): 20 AWG has ten times the resistance per metre of 10 AWG. The number runs the wrong way because it began as the number of times the wire had been drawn through a die: more passes, thinner wire, higher number. The series runs from 0000 (107 mm²) down to 40 AWG (0.0799 mm, 0.005 mm²); above 0000 American conductors are named in kcmil, thousands of circular mils, at 0.5067 mm² each, so 4/0 is 211.6 kcmil and the next size up is 250 kcmil, 126.7 mm². For building wire only the even gauges are stocked: 14, 12, 10, 8, 6, 4, 2 and the aughts. The odd gauges exist and are used in magnet wire and some instrumentation cable, and they matter for conversion because 11 AWG (4.17 mm²) is much closer to 4 mm² than 10 or 12.
mm²: the IEC way
IEC 60228 names a conductor by its nominal cross-section in square millimetres and by its class: class 1 solid, class 2 stranded for fixed installation, class 5 flexible and class 6 extra-flexible. The preferred sizes are 0.5, 0.75, 1, 1.5, 2.5, 4, 6, 10, 16, 25, 35, 50, 70, 95, 120, 150, 185, 240, 300 and on up to 2 500 mm². The ratio between neighbours is roughly 1.5 to 1.7 but not constant: this is a list of agreed sizes, not a formula.
What the standard actually fixes is not the diameter and not even the area but the maximum resistance per kilometre at 20 °C: 7.41 Ω/km for 2.5 mm² in class 1 or 2, 4.61 for 4 mm², 3.08 for 6 mm², 1.83 for 10 mm², 1.15 for 16 mm². The nominal area is a name. A 2.5 mm² class 2 conductor may legitimately contain a little less than 2.5 mm² of copper, because 7.41 Ω/km at 100 % IACS conductivity corresponds to 2.33 mm² of solid copper, and makers work close to the limit; class 5 flexible conductors are allowed a higher resistance again (7.98 Ω/km for 2.5 mm²) because of their fine strands. Two consequences. A conductor that measures below the nominal area is not automatically wrong; one whose resistance exceeds the table value is. And the honest way to compare an AWG conductor with a metric one is by area or resistance, never by name.
SWG and 7/029: what old drawings and older electricians mean
The Standard Wire Gauge was the British gauge, defined in BS 3737 as a list of diameters in inches with no single formula behind it. It survives in Pakistan on old drawings, on earthing and binding wire, and in the habit of naming wire by a number: 8 SWG is 4.064 mm (12.97 mm²), 10 SWG 3.251 mm (8.30 mm²), 12 SWG 2.642 mm (5.48 mm²), 14 SWG 2.032 mm (3.24 mm²), 16 SWG 1.626 mm (2.08 mm²), 18 SWG 1.219 mm (1.17 mm²), 20 SWG 0.914 mm (0.656 mm²), 22 SWG 0.711 mm (0.397 mm²). The numbers do not line up with AWG. 16 SWG happens to be the same copper as 14 AWG; 16 AWG is 1.31 mm², 37 % less than 16 SWG. A drawing that says "16 gauge" without saying which is a question, not an instruction.
The stranded sizes of pre-metric British cable are named by strand count and strand diameter in thousandths of an inch, and the Pakistani wire market still uses them: 3/029 is three strands of 0.029 inch (0.737 mm), 1.28 mm² in all; 7/029 is seven of them, 2.98 mm²; then 7/036 (4.60 mm²), 7/044 (6.87 mm²), 7/052 (9.59 mm²) and 7/064 (14.5 mm²). These are not the metric sizes they are sold as equivalents of. A 7/029 has 19 % more copper than a 2.5 mm²; a 3/029 has 15 % less than a 1.5 mm². The name is useful for exactly the reason it is precise: a micrometer on one strand tells you in a moment whether a "7/029" is what it says. Strands of 0.65 mm instead of 0.737 mm mean 22 % less copper.
Converting honestly: area, not ampacity
Two conversions exist for every size, and they answer different questions. The nearest metric size to an AWG conductor is what you would call it in conversation: 14 AWG is "about 2.5 mm²". The next size up is what you specify when the AWG figure is a requirement: 14 AWG (2.08 mm²) is met by 2.5 mm², 12 AWG (3.31) by 4 mm², 10 AWG (5.26) by 6 mm², 8 AWG (8.37) by 10 mm², 6 AWG (13.3) by 16 mm², 4 AWG (21.2) by 25 mm², 2 AWG (33.6) by 35 mm², 1/0 (53.5) by 70 mm² and 4/0 (107) by 120 mm². For those sizes the next metric size up is also the nearest, which is convenient; it stops being true at 18 AWG (0.823 mm²), where 0.75 mm² is nearer but smaller and 1.0 mm² is the size that meets it, and at 1/0, where 50 mm² is nearer and 70 mm² meets it.
Going the other way, a metric specification is met by the first gauge with at least that area: 2.5 mm² by 13 AWG (2.62 mm²), or by 12 AWG (3.31) if only even gauges are available; 4 mm² by 11 AWG (4.17) or 10 AWG (5.26); 6 mm² by 9 AWG (6.63) or 8 AWG (8.37). The even-gauge route is wasteful, 30 to 40 % more copper than needed, and a designer who has to build to a metric specification with American stock should ask for the odd gauges.
What no conversion carries across is the current rating. The NEC rates a 12 AWG copper conductor at 25 A with 75 °C insulation and then caps its overcurrent protection at 20 A under 240.4(D); IEC 60364-5-52 rates a 4 mm² PVC conductor in conduit on a wall at 32 A, in a thermally insulated wall at 24 A, and less again grouped. The two systems assume different insulation classes, different installation methods and different safety margins, and a rating read off one table for a conductor described in the other system is meaningless. Convert the size, then rate the cable by the rules that govern the installation: the cable sizing guide walks through the IEC method.
Resistance and why it is the number that matters
What a conductor does in a circuit is set by its resistance, and resistance is the resistivity of the metal divided by the area. Annealed copper at 100 % IACS has ρ = 0.017241 Ω·mm²/m at 20 °C, so a conductor of A mm² has 17.241 ÷ A ohms per kilometre: 6.90 Ω/km for 2.5 mm², 3.28 for 10 AWG, 1.72 for 10 mm². Resistance rises with temperature by 0.393 % per degree, so the same conductor at 70 °C, the working temperature of a loaded PVC cable, is 20 % higher. Voltage drop is the current times the resistance of the loop, out and back: 20 A through 10 m of 10 AWG drops 20 × 0.0655 = 1.3 V, which is where the voltage drop calculator starts.
Aluminium has ρ = 0.02826 Ω·mm²/m, 1.64 times that of copper, so an aluminium conductor needs 1.64 times the area for the same resistance: 8.62 mm² to match 10 AWG copper, in practice 10 mm². It weighs 2.70 g/cm³ against copper's 8.89, so the heavier aluminium conductor still weighs half as much as the copper one, which is why overhead lines and large feeders are aluminium and why its terminations need the right lugs and compound.
Mass is the other number that resistance leads to. One mm² of conductor over one kilometre is one litre of metal, so a copper conductor weighs 8.89 kg/km per mm²: 46.8 kg/km for 10 AWG, 22.2 kg/km for 2.5 mm². Cable is sold by the metre and made by the kilogram of copper, and that arithmetic is the whole economics of under-gauge cable. It is also the test: weigh a measured length of bare conductor, or measure its resistance with a four-wire meter, and compare with 8.89 g per metre per mm² or with the IEC 60228 maximum resistance. Either catches a short conductor that a printed "2.5 mm²" on the sheath will not.
The example in numbers
A packaging machine from the United States arrives with its motor circuits wired in 10 AWG copper, and the panel must be extended with local cable over a 10 m run at 20 °C.
- Diameter: 0.127 × 92(36 − 10) ÷ 39 = 0.127 × 20.38 = 2.588 mm; area π × 2.588² ÷ 4 = 5.26 mm².
- Metric: between 4 and 6 mm²; the nearest is 6 mm² and so is the next size up. Specify 6 mm², which has 14 % more copper; a 4 mm² substitute would have 24 % less.
- Resistance: 17.241 ÷ 5.26 = 3.28 Ω/km; 0.0328 Ω for the 10 m run and 0.0655 Ω for the loop. At 20 A the loop drops 1.3 V.
- Mass: 5.26 × 8.89 = 46.8 kg/km, 0.468 kg for the run. In aluminium the same resistance needs 8.62 mm², so 10 mm².
The wire gauge converter reproduces these figures from any of the three starting points and adds the resistance at the conductor's working temperature.
Where mistakes happen in Pakistan
Most of the equipment that arrives in Karachi and Lahore is wired in AWG: American and Japanese machinery outright, Chinese machinery in AWG when it is built to UL for export, and the manuals give AWG in their wiring tables. It is then connected to local cable in mm², through terminals made for one system or the other, by people who have been told the two are equivalent. The errors are always the same ones.
- The "equivalent" is the nearest, not the next up. A manual that calls for 12 AWG (3.31 mm²) gets 2.5 mm² because that is what 12 AWG is "about"; it should get 4 mm². The reverse happens on export jobs: a 4 mm² specification met with 12 AWG, 17 % short. The converter's next-size-up figure is the one to write on the drawing.
- Terminals sized for the other system. A terminal rated for 2.5 mm² will accept a 12 AWG conductor only under strain; a 4 mm² conductor may not enter a terminal made for 12 AWG at all, and the answer on site is to cut strands, which turns a correct conductor into an under-gauge one at the one point that carries all the heat. Order the ferrules and terminals for the conductor actually being fitted.
- "16 gauge" with no system named. 16 SWG is 2.08 mm², 16 AWG is 1.31 mm², and both are read as 1.5 mm² by somebody. Old drawings and older electricians mean SWG; imported manuals mean AWG; ask.
- Under-gauge cable sold by nominal size. Cable printed 2.5 mm² whose conductor measures 1.5 mm across contains 1.77 mm², and a 7/029 whose strands measure 0.65 mm contains 2.32. The test is a micrometer on one strand, or the weight of a metre of bare conductor, or its resistance, against the figures above; none takes longer than the argument that follows.
- A rating carried across with the size. An American panel schedule that pairs 12 AWG with 20 A breakers is a design for its conditions; the local extension in 4 mm² must be rated for its own installation method, grouping and ambient by the IEC tables, and an overload setting that was right for the machine's internal wiring is not automatically right for the feeder to it.
Ahmedonics builds control panels and machine wiring to IEC sizes and, where the equipment is specified in AWG, converts upward and documents the conversion on the drawing, so that the next person to open the panel is not left to guess. It is a small discipline that removes a whole class of site problems.
References
- ASTM B258-18, Standard Specification for Standard Nominal Diameters and Cross-Sectional Areas of AWG Sizes of Solid Round Wires Used as Electrical Conductors — definition of the gauge and tabulated diameters and areas
- BS 3737:1964, Specification for Standard Wire Gauge (SWG) — the imperial gauge diameters; withdrawn but still the reference
- IEC 60228:2004, Conductors of insulated cables — nominal sizes, classes and maximum resistance at 20 °C
- IEC 60028:1925, International standard of resistance for copper — the IACS reference resistivity
- IEC 60364-5-52:2009, Low-voltage electrical installations — Part 5-52: Wiring systems — current-carrying capacities by installation method
- NFPA 70, National Electrical Code 2023, Table 310.16 and 240.4(D) — the American ampacity table and the small-conductor overcurrent limits, for comparison