Starlight Tools

Audio Crossover Calculator — Frequency, Slopes & Component Values

Calculate exact capacitor and inductor values for an ideal passive 2-way speaker crossover, compare common electrical slopes, and find the nearest E12 or E24 parts.

Crossover settings

Impedance magnitude at crossover if known; otherwise nominal. Treated as a fixed resistance; phase and frequency variation are not modeled.

Impedance magnitude at crossover if known; otherwise nominal. Treated as a fixed resistance; phase and frequency variation are not modeled.

Your settings stay in this browser. No frequencies or driver details are uploaded or stored.

Passive 2-way network

Selected electrical design 2nd-order Linkwitz–Riley at 2.5 kHz
Electrical slope12 dB/octave
Branch level at crossover−6 dB
Starting polarityReverse one driver
Total reactive parts4 components

Woofer low-pass

    Tweeter high-pass

      Important: These are textbook electrical starting values for resistive 8 Ω loads. Real loudspeakers are reactive, so verify the design with impedance and frequency-response measurements before building or using it at power.

      How to use this passive crossover calculator

      1. Choose a safe crossover frequency. Check both driver datasheets and measured responses. Keep the tweeter comfortably above its recommended minimum crossover frequency and away from high-distortion regions.
      2. Select the electrical slope. Higher orders attenuate out-of-band signals faster but need more parts and become more sensitive to real impedance and component tolerances.
      3. Use impedance at crossover when possible. A nominal 8 Ω driver may be far from 8 Ω at the selected frequency. Use its measured impedance magnitude at crossover when available, or a manufacturer impedance curve for an estimate. Even with measured magnitude, the calculator approximates each driver as a fixed resistance; it does not incorporate impedance phase or variation with frequency.
      4. Read parts in signal order. “Series” parts go in the signal path from the amplifier toward the driver. “Shunt” parts connect from the indicated signal node to the branch return, the conductor joining the driver’s return terminal to the amplifier’s return terminal. In a second-order network, the shunt part connects at the node after the first series part, across the driver. In a third-order network, it connects at the intermediate node between the first and last series parts, not directly across the driver. The displayed sequence describes the ideal ladder order.
      5. Model, build, and measure safely. Refine the network in loudspeaker-design software using measured driver impedance and response data. Disconnect the amplifier before wiring, and use components rated for the expected voltage, current, and heat.

      The nearest E12 or E24 value is convenient for sourcing, but it is not automatically the best final part. Combining parallel capacitors or series inductors can approach an exact value; the components’ tolerance and the inductor’s DC resistance still matter.

      Which crossover slope and alignment should you choose?

      AlignmentElectrical behaviorTypical trade-offStarting polarity
      1st-order Butterworth6 dB/oct; each branch −3 dB at crossoverFewest parts and gentle phase rotation, but drivers must overlap broadly and tolerate more out-of-band energy.Same polarity
      2nd-order Butterworth12 dB/oct; each branch −3 dB at crossoverSimple and steeper, but the ideal in-phase voltage sum peaks after correcting the 180° branch relationship.Reverse one driver
      2nd-order Linkwitz–Riley12 dB/oct; each branch −6 dB at crossoverIdeal flat on-axis sum after polarity reversal, with only two reactive parts per branch.Reverse one driver
      3rd-order Butterworth18 dB/oct; each branch −3 dB at crossoverFaster attenuation and ideal flat voltage sum, but more components and greater sensitivity to load error.Same polarity

      “Starting polarity” describes the ideal electrical network only. Driver diaphragm offset, natural roll-off, polarity conventions, and acoustic phase can change the correct measured connection.

      Crossover formulas and assumptions

      Let f be crossover frequency in hertz, R the branch load in ohms, and ω = 2πf. Use the entered woofer impedance as R for every low-pass component and the entered tweeter impedance as R for every high-pass component, treating each as a fixed resistance. Inductance is calculated in henries and capacitance in farads. Multiply henries by 1,000 to obtain mH and farads by 1,000,000 to obtain µF. The general scaling is:

      L = kL × R / (2πf)
      C = kC / (2πfR)

      The numbers beside L1, C2 and the other component labels below are dimensionless normalized scaling coefficients, not component values in henries or farads. Use each inductor’s coefficient as kL and each capacitor’s coefficient as kC in the equations above. Components are listed in circuit order from the amplifier toward the driver: first-order branches have one series part; second-order branches have a series part followed by a shunt part at the driver node; third-order branches have a series part, a shunt part at the intermediate node, and a final series part leading to the driver. Every shunt part connects to the branch return.

      • 1st-order Butterworth: low-pass L1 = 1; high-pass C1 = 1.
      • 2nd-order Butterworth: low-pass L1 = √2 and C2 = 1/√2; high-pass C1 = 1/√2 and L2 = √2.
      • 2nd-order Linkwitz–Riley: low-pass L1 = 2 and C2 = 1/2; high-pass C1 = 1/2 and L2 = 2. Illustrative electrical calculation: for f = 2.5 kHz = 2,500 Hz and two ideal 8 Ω loads, use R = 8 Ω in each branch. Each inductor is L = 2 × 8 / (2π × 2,500) H ≈ 0.00101859 H = 1.01859 mH. Each capacitor is C = (1/2) / (2π × 2,500 × 8) F ≈ 0.00000397887 F = 3.97887 µF. The woofer branch has series L1 = 1.01859 mH followed by shunt C2 = 3.97887 µF from the woofer node to the branch return. The tweeter branch has series C1 = 3.97887 µF followed by shunt L2 = 1.01859 mH from the tweeter node to the branch return. These rounded values illustrate the ideal equations; this is not a validated loudspeaker design.
      • 3rd-order Butterworth: low-pass L1 = 3/2, C2 = 4/3, L3 = 1/2; high-pass C1 = 2/3, L2 = 3/4, C3 = 2.

      These are constant-resistance, zero-source-impedance ladder equations. They calculate the electrical filter response, not the finished acoustic crossover. A real driver’s impedance curve, voice-coil inductance, sensitivity, enclosure, baffle diffraction, physical offset, and natural response all modify the result. For that reason, this page is a starting-value calculator rather than a substitute for loudspeaker simulation and measurement.

      Electrical and design limits

      Passive crossover parts can carry hazardous amplifier voltages and significant current. Turn off and disconnect equipment before changing wiring. Confirm amplifier load compatibility, driver power limits, capacitor voltage and type, inductor current capacity and DC resistance, and resistor heat ratings where additional compensation is used. If you cannot measure the loudspeaker or assess component ratings, consult an experienced loudspeaker designer or technician.

      Audio crossover calculator FAQ

      Which impedance should I enter?

      Use the driver’s impedance magnitude near the crossover frequency when measured data is available. Nominal impedance is acceptable only for a first estimate because the actual value changes with frequency. Even a measured magnitude is treated here as a fixed resistance: the calculator does not incorporate the driver’s impedance phase or frequency-dependent behavior. Use measured impedance and frequency-response data in loudspeaker-design software to refine the network.

      What is the difference between Butterworth and Linkwitz–Riley?

      A Butterworth branch is 3 dB down at its cutoff. The 2nd-order Linkwitz–Riley option is 6 dB down at crossover, allowing its two ideal electrical branches to sum to unity after one branch is polarity-inverted.

      Why does a 12 dB/octave network suggest reversing one driver?

      The ideal second-order low-pass and high-pass branches are 180° apart at crossover. Reversing either driver corrects that electrical relationship. Measurements determine whether the real system follows the ideal case.

      Are nearest E12 and E24 values exact equivalents?

      No. They are the closest values in a preferred-number series. The CSV preserves both the exact result and the preferred alternative so you can compare them.

      Does this include voice-coil inductance, Zobel compensation, or an L-pad?

      No. The model treats each driver as a fixed resistance. It does not calculate impedance compensation, sensitivity padding, notch filters, baffle-step compensation, or response equalization.

      Can I use the result for a car or home loudspeaker?

      The ideal equations are not installation-specific, but real impedance, temperature, available space, amplifier loading, and component ratings matter. Model and measure the installed system.

      Are my settings uploaded?

      No. Calculations, copying, and CSV generation happen locally in your browser.

      Technical references

      • Rane — Linkwitz–Riley Crossovers: A Primer, especially “1st-Order Amplitude Response,” “Butterworth Alignment,” and “Linkwitz-Riley Alignment,” for electrical slopes, branch phase relationships, summing behavior, and LR2 polarity reversal. “Time or Phase Correction” explains the driver time-alignment assumption. These discussions do not establish the exact passive ladder coefficients listed above.
      • Linkwitz Lab — Crossovers, specifically “B) Allpass crossovers,” items 4 and 6, for LR2/LR4 transfer relationships and driver polarity. These transfer-function discussions do not establish the exact passive ladder coefficients listed above.
      • Analog Devices — Basic Linear Design, Chapter 8: Analog Filters, “Passive LC Section,” printed pages 8.65–8.66, for passive filter sections, and the design example on printed pages 8.122–8.123 for frequency and impedance scaling. These passages support the general implementation and scaling method; they do not by themselves verify every coefficient listed above for the stated zero-source-impedance, fixed-resistance-load networks.

      Explore more tools