Premium apps Part of Ohmforge

Ohmforge Microstrip & Stripline Impedance Calculator

PCB trace impedance Z0, εeff and delay (Hammerstad–Jensen, Wheeler), or the width for 50 Ω.

mm

Analysis only.

More options
mm

Microstrip: from the trace to its reference plane. Stripline: total distance between the two planes.

µm

35 µm = 1 oz/ft², 18 µm = ½ oz; 0 = ideal thin strip.

FR-4 about 4.2–4.6 (falls with frequency), Rogers RO4350B 3.66, PTFE 2.2. Use the laminate maker's value at your frequency.

Ω

Synthesis only.

MHz

For the guided wavelength; 0 = skip.

Try an example: FR-4: 0.3 mm trace on 0.17 mm prepreg · Width for 50 Ω on PTFE (εr 2.2, h 1.27 mm) — Pozar example 3.7 · 50 Ω stripline, planes 3.2 mm apart (εr 2.2) — Pozar example 3.5

More about this tool

How it works

Characteristic impedance of a PCB trace: surface microstrip with the Hammerstad–Jensen equations (thickness correction included) or a centred stripline with Wheeler's formula, plus effective permittivity, propagation delay and guided wavelength. Synthesis mode solves the trace width for a target impedance such as 50 Ω with Brent's root finder; every result is cross-checked against an independent formula and the IPC-2141 rule of thumb.

  • Microstrip: Hammerstad–Jensen (1980) closed-form equations — Z0 of the air line Z01(u) with u = W/h, effective permittivity εe(u, εr) with their a(u) and b(εr) fits (accuracy better than 0.2 % for 0.01 ≤ u ≤ 100), and the strip-thickness correction Δu (Wheeler's) applied separately to the air line and the dielectric. Z0 = Z01(ur)/√εe(ur).
  • Stripline (strip centred between two planes, spacing b): Wheeler's 1978 formula with the thickness correction (about 0.5 % accuracy for W/(b − t) < 10); for t = 0 it is compared with Cohn's exact conformal-mapping result η0/(4√εr)·K(k)/K(k′), k = sech(πW/2b), evaluated with the arithmetic–geometric mean.
  • Synthesis solves Z0(W) = target with Brent's method (Proofline numerical engine) over W/h from 0.01 to 100 (stripline: up to 20) on a logarithmic width axis.
  • Delay = √εeff / c, guided wavelength = c / (f·√εeff). The IPC-2141 rule-of-thumb formulas are shown for comparison only; they are less accurate, especially outside their range.

Quasi-static, lossless model: no dispersion (microstrip Z0 and εeff rise at high frequency), no solder mask, no roughness, no etch trapezoid and no nearby copper pour. Ask your PCB maker for their stack-up and impedance-controlled width; manufacturing tolerances are typically ±10 %.

Sources

Part of Ohmforge

  • Ohmforge Virtual electronics workbench

    Draw a circuit, switch it on, and measure it with real-feeling bench instruments.

    Open Ohmforge

Related tools

Calculate microstrip and stripline impedance, or the width for 50 Ω

Find the characteristic impedance of a PCB trace or the trace width for a target impedance, with effective permittivity, delay and independent cross-checks.

What it is for

An RF trace, a USB pair or an antenna feed only behaves when its characteristic impedance matches the system, usually 50 Ω. That impedance depends on the trace width, the dielectric thickness to the reference plane, the copper thickness and the board material. This tool calculates it for two common geometries: Microstrip (a trace on an outer layer over a plane) and Stripline (a trace centred between two planes).

Choose Impedance of a given width to check a layout, or Width for a target impedance to design one. Microstrip uses the Hammerstad–Jensen equations, stripline uses Wheeler's formula, and the width is solved numerically with Brent's root finder.

It is a quasi-static calculation. It does not model solder mask, copper roughness, etch shape, dispersion at high frequencies or a nearby copper pour, and it does not replace the stack-up data from your PCB maker.

Worked example

You need a 50 Ω microstrip on a four-layer board with 0.2 mm prepreg (εr 4.2) to the ground plane and 35 µm copper, for a 2.4 GHz antenna feed. Choose Microstrip, Width for a target impedance, enter Dielectric height 0.2 mm, Copper thickness 35 µm, εr 4.2, Target impedance 50 Ω and Frequency 2400 MHz.

The tool answers with a 0.3657 mm trace (W/h = 1.828), an effective permittivity of 3.067 and a delay of 5.842 ps/mm. The guided wavelength at 2.4 GHz is 71.32 mm, so a quarter-wave section is about 17.8 mm long.

The cross-check table shows how the copper affects the result: the same formula at zero thickness gives 52.463 Ω, and Wheeler's independent 1977 formula gives 52.349 Ω for that thin strip, 0.217 % away. The IPC-2141 rule of thumb reads 47.57 Ω, almost 5 % off, which is why it is shown for comparison only.

How to read the result

Z0 is the characteristic impedance; in synthesis mode it equals your target. Trace width W is the answer in synthesis mode. εeff is the permittivity the wave actually sees. A microstrip has field lines partly in air, so it is lower than εr. Delay tells you how long a trace takes to cross, which matters for length matching.

The Cross-checks table compares the result with independent formulas. The difference to the second formula should be a fraction of a percent; a large difference means you are outside the range where the formulas were fitted.

Tips and common mistakes

  • Use the dielectric constant your laminate maker gives for your frequency. FR-4 varies from about 4.2 to 4.6.
  • For stripline, enter the total distance between both planes as the height, not the distance to one plane.
  • Thin copper matters more on narrow traces: 18 µm copper on a 0.15 mm stripline changes Z0 by about 10 %.
  • Solder mask over a microstrip lowers Z0 by a few ohms; ask your board house for their controlled-impedance width.
  • Plan for ±10 % manufacturing tolerance unless you order controlled impedance.

FAQ

Why does my PCB maker suggest a different width?

They include solder mask, etch factor and their measured material data. Use this tool to understand the design and their numbers for production.

What does the frequency field change?

Only the guided wavelength. The impedance itself is calculated quasi-statically.

Is the result valid for differential pairs?

No. The tool calculates single-ended lines. Coupled pairs need a differential-impedance model.

Which formula is more accurate, Hammerstad–Jensen or IPC-2141?

Hammerstad–Jensen, by a wide margin. Its authors report errors below about 0.2 % over a wide range. The IPC-2141 rule is a quick approximation for a limited range.

Guide last updated:

For scripts and AI agents

  • JSON API: POST https://ohmforgemicrostripimpedance.usefulweb.app/api/v1/calculate (same calculation as this page)
  • Interface description: OpenAPI 3.1 · AI guide: llms.txt
  • MCP server (Streamable HTTP): https://ohmforgemicrostripimpedance.usefulweb.app/mcp · tool name ohmforge_microstrip_impedance
Example request
curl -s -X POST https://ohmforgemicrostripimpedance.usefulweb.app/api/v1/calculate \
  -H 'content-type: application/json' \
  -d '{"structure":"microstrip","task":"analyze","widthMm":0.3,"heightMm":0.17,"thicknessUm":35,"er":4.3,"frequencyMHz":1000}'
Feedback · ideas & bug reports welcome