PCB Trace Resistance Calculator

PCB Trace Resistance Calculator

The resistance of a copper trace from its length, width, copper weight and temperature — with the voltage drop and power it costs at your current, the sheet resistance in milliohms per square that layout people actually think in, and the number of squares your trace is.

Trace resistance

Geometry + °C → Ω, mV, mW
Follow the copper, not the straight line: corners and detours count.
The finished width. A neck-down at a pad or a via is a separate, shorter trace in series — work it out on its own.
Outer layers are usually plated up beyond the base foil, so the finished copper can be thicker than the weight you ordered; ask your fabricator.
The copper’s own temperature, not the room’s. A trace carrying its rated current sits at ambient plus its temperature rise.
Used for the voltage drop and the power. Continuous or RMS.
The trace seen from above, not a circuit: current enters at one end and leaves at the other, and the copper's resistance is its length divided by its width — the number of squares — times the sheet resistance. Not to scale.
100.5mΩExample

a 50 mm trace 0.25 mm wide in 1 oz copper at 25 °C, carrying 0.5 A

Advertisement

Resistance, three ways round

R = ρT · L ÷ (w · t)   =   R□ · (L ÷ w)
R□ = ρT ÷ t   (0.4926 mΩ per square for 1 oz copper at 20 °C)
ρT = 0.017241 × (1 + 0.00393 (T − 20)) Ω·mm²/m    V = I·R    P = I²R
L, w, t
length, width and copper thickness: 1 oz/ft² is 1.378 mil, which is 35 µm
R square
sheet resistance — the resistance of any square patch of that copper, whatever its size. A trace is L/w squares in series
rho
resistivity of annealed copper, 0.017241 Ω·mm²/m at 20 °C (1.7241 × 10⁻⁸ Ω·m), the IACS reference value
alpha
0.00393 per kelvin at 20 °C. The 1910 Bureau of Standards determination gives 0.00394 for 100% IACS copper; 0.00393 is the tabulated engineering figure and the one this site uses throughout

Worked example

a 50 mm trace 0.25 mm wide in 1 oz copper at 25 °C, carrying 0.5 A
1 oz copper is 1.378 mil = 35.0 µm thick, so the cross-section is 0.25 × 0.0350 = 0.00875 mm²
At 25 °C, ρ = 0.017241 × (1 + 0.00393 × 5) = 0.017580 Ω·mm²/m
R = 0.017580 × 0.05 ÷ 0.00875 = 100.5 mΩ
The same answer per square: 1 oz copper is 502.3 µΩ per square at 25 °C, and this trace is 50 ÷ 0.25 = 200 squares
At 0.5 A that costs 50.23 mV of drop and 25.11 mW of heat
Heat it to 105 °C and the same trace measures 131.4 mΩ — a third more

Resistance per metre of trace, at 20 °C

Width0.5 oz1 oz2 ozSquares per mm of lengthNotes
0.15 mm6.568 Ω/m3.284 Ω/m1.642 Ω/m7A fine signal trace. 0.15 mm is a common minimum for volume board houses.
0.25 mm3.941 Ω/m1.97 Ω/m985.2 mΩ/m4The page default — a typical signal trace.
0.50 mm1.97 Ω/m985.2 mΩ/m492.6 mΩ/m2A small power trace.
1.00 mm985.2 mΩ/m492.6 mΩ/m246.3 mΩ/m11 mm at 1 oz is almost exactly 0.5 Ω per metre.
2.00 mm492.6 mΩ/m246.3 mΩ/m123.1 mΩ/m1A modest power rail.
5.00 mm197 mΩ/m98.52 mΩ/m49.26 mΩ/m0Approaching a pour. Below about half an ohm per metre it is usually easier to use a plane.
Per metre, so multiply by your own length in metres. The last column is the number of squares a millimetre of that trace is worth, which is the quickest mental check there is: squares × sheet resistance = ohms, and 1 oz copper is very nearly half a milliohm per square.

Ohms per square, and why it is the useful unit

Copper on a board is a sheet of fixed thickness, so its resistance depends only on the SHAPE of the trace, not on its size. A square patch of 1 oz copper measures 0.4926 mΩ between opposite edges whether the square is a millimetre across or a metre: make it twice as wide and you have halved the resistance and doubled the length in equal measure. That is what sheet resistance means, and it turns every trace-resistance question into counting: a trace 50 mm long and 0.25 mm wide is 200 squares, so it is 200 × 0.4926 mΩ ≈ 99 mΩ at 20 °C. Layout engineers do this in their heads, and it is worth learning because it also works for corners (a right-angled corner is worth a little less than one square, because the current crowds round the inside of the turn) and for pours.

How this page differs from the trace width calculator. They answer opposite questions and use different physics. The PCB trace width calculator is a THERMAL calculation: you give it a current and a temperature rise you will accept, and it returns the width that keeps the copper that cool, from the IPC-2221 curve fit. This page is an ELECTRICAL calculation with no IPC model in it at all: you give it the geometry you already have and it returns the resistance, the drop and the loss — at whatever temperature you name, which the width page cannot do because there the temperature is its own answer. Use the width page when you are choosing a trace and heating is the constraint; use this one when the trace exists and the question is what the voltage at the far end will be. The two agree exactly where they overlap: the same resistivity, the same 0.00393 per kelvin, the same 1.378 mil per ounce.

A worked power trace. Take a 5 V, 2 A rail run 120 mm across a board on a 1 mm trace in 1 oz copper. That is 120 squares, so 59 mΩ at 20 °C — but the trace is carrying 2 A, so it is not at 20 °C: it is at perhaps 45 °C, where it measures 65 mΩ. The drop is 130 mV and the loss 0.26 W, and the return path through the ground plane adds its own. On 5 V that 130 mV is inconvenient; on a 1.2 V core rail the same trace would be indefensible, and on a 3.3 V rail feeding an ADC reference it would be a measurement error. This is the calculation that decides where the regulator goes and where the sense point is taken, and it is not the same calculation as “will the trace get hot” — a trace can run stone cold and still ruin a rail.

What the model leaves out. Etching undercuts, so a real trace is a trapezium and is a few per cent more resistive than this rectangle; the effect grows with copper weight. Plating on outer layers adds copper of slightly higher resistivity than the annealed reference. Above a few megahertz the skin effect confines the current to the outside of the conductor and the resistance rises with the square root of frequency — at 100 MHz the skin depth in copper is about 6.6 µm, so a 35 µm trace is using only part of its copper. And none of this counts the vias in the path, which for a high-current net are often a bigger share of the total than the trace: the PCB via current calculator does those. For wire off the board, the AWG wire size calculator uses the same resistivity, and the Ohm’s law calculator turns any of these resistances into whatever else you need.

Advertisement

Frequently asked questions

What is the resistance of a PCB trace?

R = ρ × L ÷ (w × t), with ρ = 0.017241 Ω·mm²/m for copper at 20 °C and t = 35 µm for 1 oz. A 50 mm trace 0.25 mm wide in 1 oz copper is about 99 mΩ at 20 °C and 100 mΩ at 25 °C. The same answer by squares: 200 squares × 0.4926 mΩ.

What is the sheet resistance of 1 oz copper?

0.4926 mΩ per square at 20 °C — that is 0.017241 Ω·mm²/m divided by 35 µm. Half an ounce is twice that, 0.985 mΩ per square, and 2 oz is half, 0.246 mΩ. Multiply by the number of squares (length ÷ width) and you have the trace.

How much does temperature change a trace’s resistance?

0.393% per kelvin for copper, referred to 20 °C. A trace at 85 °C has 26% more resistance than the same trace at 20 °C, and at 105 °C a third more. That matters because a trace carrying current heats itself, so the resistance you measure in circuit is always above the room-temperature figure.

Should I use this or the trace width calculator?

The width calculator when you are choosing a trace and the limit is heating: it gives you the width for a current and a temperature rise. This page when the trace already exists and the question is electrical — how many millivolts it drops, how much power it wastes, what it measures at temperature.

Does a right-angled corner add resistance?

Less than counting it as a whole square would suggest: the current crowds round the inside of the turn and the outer copper carries little, so a 90° corner is worth somewhat less than one square. On a signal trace it is irrelevant, and on a 100-square power trace it is a fraction of a per cent either way.

Does this apply at high frequencies?

No. Above a few megahertz the skin effect pushes the current to the surface and the resistance rises as the square root of frequency — the skin depth in copper is about 6.6 µm at 100 MHz against 35 µm of trace. This page is the DC and low-frequency resistance, which is the one that sets rail drops and I²R loss.

Related calculators

References

  1. IEC 60028, International standard of resistance for copper (the International Annealed Copper Standard). Fixes 100% IACS annealed copper at 1.7241 × 10⁻⁸ Ω·m at 20 °C, which is the 0.017241 Ω·mm²/m used here and in this site’s wire pages. The standard itself is a paid document; the value was taken from the IACS definition and is the same one already stored in the AWG record.
  2. Dellinger JH. The temperature coefficient of resistance of copper. Bulletin of the Bureau of Standards, vol. 7 no. 1, 1910. The original determination: the 20 °C temperature coefficient of a sample is its percent conductivity times 0.00394. This page uses 0.00393, the tabulated engineering figure for commercial annealed copper and the value already used by this site’s trace width, voltage drop and cable pages — a 0.25% difference in the coefficient, which is 0.02% in the resistance at 100 °C.
  3. IPC-2221B. Generic Standard on Printed Board Design, IPC, 2012. Defines copper weight in ounces per square foot: 1 oz/ft² is 1.378 mil, which is 35 µm. This page uses that conversion so that its answers agree exactly with this site’s trace width record, which is built on the same standard’s conductor charts.