Calculate a second order speaker crossover from a defined low-pass or high-pass topology, natural corner frequency, load model, and alignment Q. Report its capacitor and inductor separately, then check the complete electrical and acoustic response rather than treating filter order alone as a component formula.
Open the Second Order Crossover Calculator.
How to Calculate a Second Order Speaker Crossover
A second-order electrical section contains two reactive energy-storage elements and approaches a 12 dB-per-octave stopband slope. That description does not uniquely specify the alignment, corner attenuation, circuit arrangement, or acoustic result.
The calculator offers ideal second-order Butterworth with Q equal to approximately 0.7071 and second-order Linkwitz-Riley, LR2, with Q equal to 0.5. These alignments produce different component values and different attenuation at the natural corner.
Its low-pass topology uses a series inductor followed by a capacitor across the resistive load. Its high-pass topology uses a series capacitor followed by an inductor across the resistive load. Both assume a zero-resistance source and ideal components.
Real driver impedance, inductor winding resistance, other crossover parts, and acoustic behavior change the finished response. Use the calculation as a circuit reference, then model and measure the actual loudspeaker.
Draw the Selected Topology Explicitly
For low-pass operation, the source feeds series L, followed by a node where C and the driver load R connect in parallel to the return. Output voltage is taken across R.
For high-pass operation, the source feeds series C, followed by a node where L and R connect in parallel to the return. Output is again across R. The inductor and capacitor occupy different positions from the low-pass circuit.
The topology is part of the specification. Do not connect a shunt component directly across the source in a way that bypasses the intended series element. A list of two component values cannot replace a complete labeled schematic.
Also identify any added resistors, attenuation pads, or impedance-correction parts. These can change the effective source or load seen by the section. The two-component ideal result does not automatically survive their addition unchanged.
Choose Alignment Before Interpreting the Corner
The calculator’s frequency is the natural electrical corner of the specified second-order response. At that frequency, either branch’s amplitude magnitude is Q.
For Butterworth Q approximately 0.7071, the corner magnitude is approximately −3.01 dB. For LR2 Q equal to 0.5, it is approximately −6.02 dB. Calling both frequencies the minus-3-dB point would therefore be incorrect.
The ideal low-pass magnitude is 1/sqrt[(1 − x²)² + (x/Q)²], where x is evaluation frequency divided by natural corner frequency. The high-pass magnitude has x² in the numerator over the same denominator.
Linkwitz’s active-filter discussion explains the Q and crossover-alignment distinctions. Its active circuit implementations differ from the passive topologies here, but the intended transfer-function definitions provide a useful comparison.
Calculate the Two Ideal Component Values
Let ω0 = 2πf0. For either of the calculator’s defined topologies, the numerical component values are L = R/(ω0Q) in henries and C = Q/(ω0R) in farads. Their circuit positions depend on the selected low-pass or high-pass option.
Multiply L by 1,000 for millihenries and C by 1,000,000 for microfarads. The same numerical pair for an alignment does not mean the low-pass and high-pass circuits are wired identically.
You can independently check the values through f0 = 1/(2πsqrt(LC)) and Q = R sqrt(C/L) under the stated circuit assumptions. These checks are useful for identifying unit mistakes or an incorrect alignment coefficient.
Do not multiply isolated first-order parts by an arbitrary factor or sum L and C into one component value. The correct relationship follows the selected circuit’s complete transfer function.
Worked Example: 8 Ohms and 2,500 Hz
Assume an illustrative 8-ohm resistive model and a 2,500 Hz natural electrical corner. Begin with the Butterworth option, Q equal to 1/sqrt(2).
Calculate L = 8/[2π × 2500 × 0.7071068], approximately 0.720253 mH. Calculate C = 0.7071068/[2π × 2500 × 8], approximately 5.626977 microfarads.
Select the low-pass series-inductor output to obtain approximately 0.720253 mH and its shunt-capacitor output for approximately 5.626977 microfarads. For high-pass operation, the same capacitor becomes the series component and the same inductor becomes the shunt component.
Now choose LR2 with Q equal to 0.5. The values become approximately 1.018592 mH and 3.978874 microfarads. This alignment has larger inductance and smaller capacitance than the Butterworth example, while retaining the same natural frequency.
| Alignment, 8-ohm model at 2,500 Hz | Inductance | Capacitance | Magnitude at natural corner |
|---|---|---|---|
| Butterworth, Q = 0.7071 | 0.720253 mH | 5.626977 microfarads | 0.7071, or −3.01 dB |
| Linkwitz-Riley, Q = 0.5 | 1.018592 mH | 3.978874 microfarads | 0.5000, or −6.02 dB |
These are ideal circuit examples, not recommended crossover frequencies or protection settings for unspecified drivers.
Verify the Transfer Function with the Actual Circuit
For the low-pass section, the shunt impedance is the parallel combination of R and the capacitor. Dividing that impedance by its sum with the series inductor yields the load-voltage transfer function.
For the high-pass section, the shunt impedance is the parallel combination of R and the inductor. Dividing it by its sum with the series capacitor yields the corresponding high-pass function.
Both ideal denominators reduce to s² + s/(RC) + 1/(LC) after normalization. The low-pass numerator is 1/(LC), while the high-pass numerator is s². This independently confirms the shared numerical LC relationship for the stated topologies.
At the natural corner, the low-pass output has phase −90 degrees and the high-pass +90 degrees in this ideal model. Their relative phase is 180 degrees. The magnitudes alone therefore do not determine whether their sum is constructive.
When practical component values are substituted, recalculate natural frequency and Q. Changing one value affects both; retaining the original printed corner label does not make the modified circuit equivalent.
Understand Pair Summation and Polarity
For an ideal LR2 pair with matched level and timing, one branch is inverted for the intended flat summed magnitude. At the corner, two 0.5-amplitude contributions then add constructively to one.
A Butterworth second-order pair has 0.7071 magnitude per branch at the corner. Inverting one for constructive summation produces approximately 1.414 amplitude there, or a 3.01 dB peak, under the ideal matched assumptions.
These are transfer-function comparisons, not universal wiring instructions for every loudspeaker. Real driver phase, natural roll-off, acoustic-center spacing, and response correction can change the required design decisions.
Rane’s Linkwitz-Riley primer explains the relationship between alignment, relative phase, and summation. Check the completed acoustic response rather than choosing polarity solely from a generic statement about second order.
Include Component Loss and Frequency-Dependent Impedance
The zero-resistance source and constant R assumptions simplify the equations. Amplifier and cable resistance, coil winding resistance, capacitor losses, and additional networks change the real section.
A loudspeaker’s impedance is also complex and frequency-dependent. A DC resistance measurement cannot describe its whole transition-band load. Use measured impedance in a suitable circuit simulation to predict the installed filter’s voltage response.
Inductor resistance can change both branch level and damping. Capacitor tolerance changes the actual value, while voltage suitability and current-related behavior remain separate part requirements. Check each physical component rather than selecting solely from its nominal capacitance or inductance.
The electrical section does not determine safe driver power. Review the driver’s permitted bandwidth, excursion, thermal conditions, and required filtering. An ideal 12 dB-per-octave slope does not make any chosen frequency suitable for an arbitrary tweeter.
Check the Finished Acoustic Crossover
Measure each filtered driver on the intended baffle, then measure their combined response with consistent microphone placement and timing reference. Inspect the useful listening angles as well as the main-axis transition.
Compare acoustic magnitude and phase with the intended target. The driver’s natural response can make the completed acoustic order differ from the electrical order. A correct nominal LC pair is only one part of establishing the desired handover.
Keep the topology, alignment Q, frequency definition, load assumption, calculated values, installed component properties, polarity choices, and measured results together. This makes the second-order calculation reproducible and prevents a later component change from inheriting an unsupported alignment label.