Skip to content
FiftyOhm
Esc
  • Air-core coilSingle-layer air-core solenoid inductance from Nagaoka's exact coefficient, compared with Wheeler's formula. Solve for the number of turns for a target inductance.Components & magnetics
  • Attenuator PadsResistor values for Pi, T, bridged-T and minimum-loss L-pad attenuators, with unequal impedances, power dissipation per resistor and E24/E96 rounding.Couplers, dividers & attenuators
  • CoaxCoax Z0, capacitance and inductance per length, velocity factor, loss, exact TE11 cutoff and breakdown voltage. Solve for D, d or εr from Z0.Transmission lines
  • CPW / GCPWCoplanar waveguide and grounded CPW impedance, effective permittivity, dispersion, delay and loss by conformal mapping. Solve for centre width or gap.Transmission lines
  • dBm ↔ W ↔ VConvert RF power between dBm, dBW, watts, Vrms, Vpeak, Vpp, dBµV and dBmV for any system impedance. Instant, accurate, with formulas.Fundamentals & conversions
  • Differential pairEdge-coupled microstrip and stripline differential impedance, Z0 even/odd, common-mode impedance, coupling and delay. Solve for spacing or width from Zdiff.Transmission lines
  • Filter OrderMinimum filter order to meet a passband/stopband attenuation mask for Butterworth, Chebyshev and elliptic responses, with the attenuation actually achieved.Filters
  • Free-space path lossFree-space path loss from the Friis equation, received power with antenna gains, and a far-field check. Solve for distance or frequency.Propagation & link budget
  • L / Pi / T MatchingDesign lumped L, Pi and T matching networks between complex source and load impedances. Both L solutions, loaded-Q Pi/T, component values and bandwidth.Matching & Smith chart
  • LC Filter DesignerDesign lumped LC ladder filters: Butterworth, Chebyshev or Bessel; low-pass, high-pass, band-pass, band-stop to order 10, with exact insertion loss and group delay.Filters
  • MicrostripMicrostrip Z0, effective permittivity, dispersion, delay and loss. Solve for trace width from target impedance. Hammerstad–Jensen + Kirschning–Jansen models.Transmission lines

Differential Pair Impedance Calculator

Edge-coupled microstrip and stripline differential impedance, Z0 even/odd, common-mode impedance, coupling and delay. Solve for spacing or width from Zdiff.

6 cited sources59 reference checksRuns in your browser

Inputs

Edge-coupled microstrip: two traces of width W spaced S apart on a substrate of height HWSHtεr

Dielectric thickness between the traces and the reference plane.

Results

Differential impedance(Z_diff)
104.97
Common-mode impedance(Z_cm)
31.985
Odd-mode impedance(Z₀ₒ)
52.486
Even-mode impedance(Z₀ₑ)
63.97
Coupling coefficient(k)
0.09861
Odd-mode εeff(εeff,o)
2.3982
Even-mode εeff(εeff,e)
3.042
Odd-mode (differential) delay(τₒ)
131.21
Even-mode (common) delay(τₑ)
147.77
  • Finite copper thickness: against a 2-D field solver the even/odd impedances are within ≈ 3 % for t/H ≤ 0.35 (zero thickness: ≈ 1 %).
  • Results are quasi-static (no dispersion, no solder mask). Solder mask typically lowers Z_diff by 2–6 Ω on outer-layer pairs.

Formulas

Zdiff=2Z0o,Zcm=12Z0e,k=Z0e−Z0oZ0e+Z0oZ_\mathrm{diff} = 2Z_{0o},\qquad Z_\mathrm{cm} = \tfrac12 Z_{0e},\qquad k = \frac{Z_{0e}-Z_{0o}}{Z_{0e}+Z_{0o}}Stripline (t=0): Z0e,o=η04εrK(ke,o′)K(ke,o), ke=tanh⁡πW2btanh⁡π(W+S)2b, ko=tanh⁡πW2bcoth⁡π(W+S)2b\text{Stripline }(t=0):\ Z_{0e,o} = \frac{\eta_0}{4\sqrt{\varepsilon_r}}\frac{K(k_{e,o}')}{K(k_{e,o})},\ k_e = \tanh\frac{\pi W}{2b}\tanh\frac{\pi(W+S)}{2b},\ k_o = \tanh\frac{\pi W}{2b}\coth\frac{\pi(W+S)}{2b}Stripline (t>0): 1Z0e=1Z0(t)−Cf′(t)Cf′(0)[1Z0(0)−1Z0e(0)],  1Z0o=1Z0(t)+Cf′(t)Cf′(0)[1Z0o(0)−1Z0(0)]\text{Stripline }(t>0):\ \frac1{Z_{0e}} = \frac1{Z_0(t)} - \frac{C_f'(t)}{C_f'(0)}\left[\frac1{Z_0(0)}-\frac1{Z_{0e}(0)}\right],\ \ \frac1{Z_{0o}} = \frac1{Z_0(t)} + \frac{C_f'(t)}{C_f'(0)}\left[\frac1{Z_{0o}(0)}-\frac1{Z_0(0)}\right]Cf′(t)ε=1π[21−t/bln⁡ ⁣(11−t/b+1)−(11−t/b−1)ln⁡ ⁣(1(1−t/b)2−1)]\frac{C_f'(t)}{\varepsilon} = \frac1\pi\left[\frac{2}{1-t/b}\ln\!\left(\frac{1}{1-t/b}+1\right)-\left(\frac{1}{1-t/b}-1\right)\ln\!\left(\frac{1}{(1-t/b)^2}-1\right)\right]Microstrip:  εeff,e=εr+12+εr−12(1+10v)−ae(v)be(εr),εeff,o=[εr+12+ao−εeff]e−cogdo+εeff\text{Microstrip: }\ \varepsilon_{\mathrm{eff},e} = \frac{\varepsilon_r+1}{2}+\frac{\varepsilon_r-1}{2}\left(1+\frac{10}{v}\right)^{-a_e(v)b_e(\varepsilon_r)},\quad \varepsilon_{\mathrm{eff},o} = \left[\frac{\varepsilon_r+1}{2}+a_o-\varepsilon_\mathrm{eff}\right]e^{-c_o g^{d_o}}+\varepsilon_\mathrm{eff}Z0e=Z0εeff/εeff,e1−Z0η0εeff Q4,Z0o=Z0εeff/εeff,o1−Z0η0εeff Q10Z_{0e} = \frac{Z_0\sqrt{\varepsilon_\mathrm{eff}/\varepsilon_{\mathrm{eff},e}}}{1-\frac{Z_0}{\eta_0}\sqrt{\varepsilon_\mathrm{eff}}\,Q_4},\qquad Z_{0o} = \frac{Z_0\sqrt{\varepsilon_\mathrm{eff}/\varepsilon_{\mathrm{eff},o}}}{1-\frac{Z_0}{\eta_0}\sqrt{\varepsilon_\mathrm{eff}}\,Q_{10}}

What this calculator does

A differential pair is two identical traces of width WW separated by a gap SS, driven with equal and opposite signals. Because the traces are close together, each one's field couples to the other. A coupled pair therefore has two characteristic impedances, one for each of its two natural modes:

  • Odd mode (equal and opposite currents, i.e. a differential signal): the field concentrates in the gap, so the capacitance to ground rises and Z0oZ_{0o} falls below the single-line value.
  • Even mode (equal currents, i.e. common-mode noise): the traces shield each other, so Z0eZ_{0e} rises.

The quantities engineers specify follow directly from these:

Zdiff=2Z0o,Zcm=12Z0e,k=Z0e−Z0oZ0e+Z0o.Z_\text{diff} = 2Z_{0o},\qquad Z_\text{cm} = \tfrac12 Z_{0e},\qquad k = \frac{Z_{0e}-Z_{0o}}{Z_{0e}+Z_{0o}} .

USB, PCIe, Ethernet and HDMI all specify ZdiffZ_\text{diff} (typically 85–100 Ω). Common-mode impedance matters for termination and EMI, and the coupling coefficient kk tells you how tightly coupled the pair is.

The models

Edge-coupled microstrip. The calculator uses the Kirschning–Jansen (1984) equations for the static even- and odd-mode εeff\varepsilon_\text{eff} and Z0Z_0. They are corrections to the Hammerstad–Jensen single-line model, which the calculator reuses from the microstrip calculator. Because part of the field is in air, the modes travel at different speeds. The odd mode has more of its field in the air above the gap, so εeff,o<εeff,e\varepsilon_{\text{eff},o} < \varepsilon_{\text{eff},e} and the odd mode is faster. That difference is the main source of far-end crosstalk in microstrip.

For copper thickness, Jansen's (1978) split of the width increment (an extra sidewall term Δt=t/(g εr)\Delta t = t/(g\,\varepsilon_r) for the odd mode) is combined with Hammerstad–Jensen's single-line approach: one width for Z0Z_0 and an air width for the reduction of εeff\varepsilon_\text{eff}, applied mode by mode. This combination is our own construction rather than a single published model, so the accuracy table below reports how it compares with a field solver.

Edge-coupled stripline. For zero thickness, Cohn's (1955) conformal mapping is exact:

Z0e,o=η04εrK(ke,o′)K(ke,o),ke=tanh⁡πW2btanh⁡π(W+S)2b,ko=tanh⁡πW2bcoth⁡π(W+S)2b.Z_{0e,o} = \frac{\eta_0}{4\sqrt{\varepsilon_r}}\frac{K(k'_{e,o})}{K(k_{e,o})},\quad k_e = \tanh\frac{\pi W}{2b}\tanh\frac{\pi(W+S)}{2b},\quad k_o = \tanh\frac{\pi W}{2b}\coth\frac{\pi(W+S)}{2b}.

Finite thickness uses Cohn's fringing-capacitance correction, built on the single thick strip from the stripline calculator. The odd mode switches to a separate formula for S<5tS < 5t, which adds the parallel-plate field between the strip sidewalls. In a homogeneous stripline both modes have εeff=εr\varepsilon_\text{eff} = \varepsilon_r, so there is no far-end crosstalk from mode-velocity mismatch.

How to use solve-for

Most designs start from a target ZdiffZ_\text{diff} with the stack-up fixed. Click ⇄ Solve next to Trace spacing and enter 100 to find the gap that gives 100 Ω at the current width. You can also solve for the width at a given spacing. Solving for SS is usually the more practical choice, since fabricators prefer a minimum trace width. You can also target ZcmZ_\text{cm}, Z0oZ_{0o} or kk, or back out εr\varepsilon_r from a measured ZdiffZ_\text{diff}.

If no solution exists, the target is out of reach at that width. ZdiffZ_\text{diff} can never exceed twice the single-line impedance (the limit S→∞S \to \infty), so the width has to come down first.

Worked example

An outer-layer pair on a 0.1 mm (4 mil) prepreg with εr=4.0\varepsilon_r = 4.0, W=0.13W = 0.13 mm, S=0.15S = 0.15 mm and 1 oz copper (t=35 μt = 35\,\mum), as microstrip (the defaults):

  • Zdiff=104.97Z_\text{diff} = 104.97 Ω, Zcm=31.985Z_\text{cm} = 31.985 Ω, Z0o=52.486Z_{0o} = 52.486 Ω and Z0e=63.970Z_{0e} = 63.970 Ω, with k=0.09861k = 0.09861.
  • εeff,o=2.3982\varepsilon_{\text{eff},o} = 2.3982 and εeff,e=3.0420\varepsilon_{\text{eff},e} = 3.0420, so the differential (odd) delay is 131.21 ps/in and the common-mode delay is 147.77 ps/in.
  • For exactly 100 Ω, solve for the gap: S=0.1063S = 0.1063 mm. Alternatively, keep SS and widen the traces to W=0.1447W = 0.1447 mm.

The same traces as a centred stripline with b=0.45b = 0.45 mm give Zdiff=90.055Z_\text{diff} = 90.055 Ω and Zcm=31.457Z_\text{cm} = 31.457 Ω, with both modes at 169.45 ps/in. Here 100 Ω needs S=0.2287S = 0.2287 mm or W=0.0929W = 0.0929 mm. The stripline pair couples more tightly (k=0.1657k = 0.1657). Each plane is about 0.2 mm away, twice the microstrip's 0.1 mm, so the neighbouring trace takes a larger share of the field.

Accuracy and validity

We compared the models with an independent 2-D finite-volume field solver (tooling/validation/laplace2d.py), which reproduces Cohn's exact coupled stripline to better than 0.05 % and Hammerstad–Jensen microstrip to about 0.15 %.

GeometryRangeError vs. field solver
Microstrip, t=0t = 00.5≤W/H≤20.5 \le W/H \le 2, 0.2≤S/H≤10.2 \le S/H \le 1, 3≤εr≤9.83 \le \varepsilon_r \le 9.8Z0e,oZ_{0e,o} within 0.4 %, εeff\varepsilon_\text{eff} within 0.4 %
Microstrip, t/H≤0.35t/H \le 0.35W/H≥1W/H \ge 1, εr≤4.4\varepsilon_r \le 4.4Z0e,oZ_{0e,o} within ≈ 3 %, εeff\varepsilon_\text{eff} within ≈ 2.6 %
Microstrip, t/H=0.2t/H = 0.2, narrow, high εr\varepsilon_rW/H=0.5W/H = 0.5, εr=9.8\varepsilon_r = 9.8Z0oZ_{0o} +7 %, εeff,o\varepsilon_{\text{eff},o} −13 % (warned)
Stripline, t=0t = 0anyexact (Cohn)
Stripline, t>0t > 0, W/(b−t)≥0.35W/(b-t) \ge 0.35t/b≤0.06t/b \le 0.06Z0eZ_{0e} −0.8 % … Z0oZ_{0o} +1.3 %
Stripline, t>0t > 0, W/(b−t)≈0.3W/(b-t) \approx 0.3Z0oZ_{0o} up to +3 %
Stripline, t>0t > 0, W/(b−t)≈0.2W/(b-t) \approx 0.2Z0oZ_{0o} up to +5 % (warned)

Kirschning–Jansen's published range is 0.1≤W/H≤100.1 \le W/H \le 10, 0.1≤S/H≤100.1 \le S/H \le 10 and 1≤εr≤181 \le \varepsilon_r \le 18, with a stated accuracy of about 1 % for zero thickness. As S→∞S \to \infty the microstrip model reduces to the isolated line, which we checked against scikit-rf's independent Hammerstad–Jensen implementation to 0.02 %. Results are quasi-static: they do not include dispersion, which slightly raises both εeff\varepsilon_\text{eff} values above a few GHz.

Practical notes

  • Solder mask lowers ZdiffZ_\text{diff}. On outer layers a 15–25 µm mask fills the gap with εr≈3.5\varepsilon_r \approx 3.5 and typically lowers ZdiffZ_\text{diff} by 2–6 Ω. Fabricators' impedance tables include this, and this calculator does not.
  • Tight versus loose coupling. Tightly coupled pairs (S≤WS \le W) keep ZdiffZ_\text{diff} stable through neck-downs and reject common-mode noise better, but they are more sensitive to etching of the gap. Loosely coupled pairs (S≥2WS \ge 2W) behave like two independent 50 Ω lines.
  • Odd and even delays differ on microstrip. Here the difference is about 16 ps/in. Over long buses this produces far-end crosstalk and mode conversion. Stripline does not have this problem.
  • Intra-pair skew comes from length mismatch and glass-weave effects, not from the impedance. Match the lengths of the two traces to a few mils at multi-Gb/s rates.
  • Thick copper pulls εeff,o\varepsilon_{\text{eff},o} down, because the field between the sidewalls is in air. On outer layers with plated copper (t/H>0.3t/H > 0.3), confirm the final geometry with a field solver.

References

  1. M. Kirschning and R. H. Jansen, “Accurate Wide-Range Design Equations for the Frequency-Dependent Characteristic of Parallel Coupled Microstrip Lines,” IEEE Trans. Microwave Theory Tech., vol. MTT-32, no. 1, pp. 83–90, Jan. 1984.
  2. R. H. Jansen, “High-Speed Computation of Single and Coupled Microstrip Parameters Including Dispersion, High-Order Modes, Loss and Finite Strip Thickness,” IEEE Trans. Microwave Theory Tech., vol. MTT-26, no. 2, pp. 75–82, Feb. 1978.
  3. E. Hammerstad and Ø. Jensen, “Accurate Models for Microstrip Computer-Aided Design,” IEEE MTT-S Int. Microwave Symp. Digest, 1980, pp. 407–409.
  4. S. B. Cohn, “Shielded Coupled-Strip Transmission Line,” IRE Trans. Microwave Theory Tech., vol. MTT-3, no. 5, pp. 29–38, Oct. 1955.
  5. S. B. Cohn, “Characteristic Impedance of the Shielded-Strip Transmission Line,” IRE Trans. Microwave Theory Tech., vol. MTT-2, no. 2, pp. 52–57, July 1954.
  6. D. M. Pozar, Microwave Engineering, 4th ed., Wiley, 2012, §7.6 (coupled-line theory).

This calculator's results are checked against 59 independent reference values in automated tests on every change. How we check accuracy.