Controlled Impedance Stackup Calculator

How it is calculated

The tool treats the trace as a surface microstrip over a ground plane and applies the IPC-2141 approximation: Z0 = 87 / √(Er + 1.41) × ln(5.98·H / (0.8·W + T)). W is trace width, H is dielectric height under the trace, T is copper thickness (all in mils, i.e. thousandths of an inch) and Er is the relative permittivity of the laminate. A second formula gives the effective dielectric constant, Eeff = (Er+1)/2 + (Er−1)/2 · 1/√(1 + 12H/W), and the propagation delay is 84.72 · √Eeff in ps per inch. If thickness is left empty it is taken as zero, which raises the answer by a few ohms. Width and height must be above zero and Er between 1 and 100. When the trace is so wide that 0.8W + T reaches 5.98H the logarithm drops to zero or below and the tool shows no result. The formula is only an estimate for a fairly narrow trace, so confirm final widths with your fabricator's field solver.

Common mistakes

• Entering millimetres instead of mils. All lengths must use the same unit, but the formula was derived for mils and the delay is quoted per inch, so convert first (1 mil = 0.0254 mm).
• Using the headline Er from a datasheet at 1 MHz for a multi-GHz signal. Laminate permittivity drops with frequency, so the impedance shifts.
• Putting the height of the whole board in H. H is only the dielectric between the trace and its reference plane, not the total thickness.
• Applying the result to a stripline (trace buried between two planes) or to a differential pair. This is a single-ended surface microstrip estimate only.
• Leaving the copper thickness empty. It counts as zero, so the same trace shows 51.47 Ω instead of 45.60 Ω; enter the real thickness (for example 1.4 mil for 1 oz copper).