Drawing: a 2-way passive crossover schematic, with a capacitor, inductor and L-pad resistors feeding the tweeter and an inductor and capacitor feeding the woofer.

Passive crossovers explained

Short answer: a passive crossover is a small circuit of capacitors, inductors and resistors between the amp and the drivers. It sends lows to the woofer and highs to the tweeter, but just as importantly it levels, shapes and corrects each driver so the pair sound like one speaker. A good one is designed from measurements of the drivers in the actual box. That's why a calculator crossover usually sounds a bit off.

What a crossover actually does

Splitting the signal is the obvious job. A tweeter can't handle bass, and most woofers get ugly at the top. But a well-designed crossover is doing four or five jobs at once:

  1. Dividing: lows to the woofer, highs to the tweeter (and mids to the midrange in a 3-way).
  2. Level matching: tweeters are often more sensitive than woofers, so the tweeter gets turned down with resistors (an L-pad).
  3. Baffle step compensation: as frequencies get lower, sound wraps around the cabinet instead of beaming forward, so a speaker naturally gets thinner in the bass below a frequency set by the baffle width. The crossover tilts the response to balance it.
  4. Taming problems: a cone breakup peak, a rising impedance, a resonance. Notch filters and impedance correction (a Zobel) handle these.
  5. Phase alignment: making sure the drivers add up properly through the crossover region, rather than cancelling.

The three parts, and what each one does

Part Passes Blocks Typical job
Capacitor Highs Lows In series with the tweeter (high-pass). In parallel across the woofer (part of a low-pass).
Inductor Lows Highs In series with the woofer (low-pass). In parallel across the tweeter (part of a high-pass).
Resistor Everything, less of it Nothing Turning a driver down, shaping, and impedance correction.

Slopes, or "orders"

Crossovers are described by how fast they roll a driver off:

  • 1st order: 6 dB per octave. One part per driver. Simple, but the drivers overlap a lot.
  • 2nd order: 12 dB per octave. Two parts per driver. Very common.
  • 3rd order: 18 dB per octave.
  • 4th order: 24 dB per octave. Steep, keeps drivers out of trouble, more parts.

Here's the bit that trips people up: what matters is the acoustic slope, not the electrical one. A driver already rolls off on its own at the edges of its range. A 2nd-order electrical filter on a woofer that's naturally falling away can produce a 3rd- or 4th-order acoustic slope. That's why you can't design a good crossover from the circuit alone.

Why calculator crossovers usually disappoint

Online crossover calculators assume the driver is a plain resistor with a flat response. Real drivers aren't:

  • Impedance isn't flat. A woofer's impedance rises with frequency (voice-coil inductance) and peaks at resonance. A tweeter's impedance peaks at its Fs. The filter values a calculator gives you assume a constant impedance, so the actual crossover point and slope end up somewhere else.
  • The response isn't flat. Drivers have peaks, dips and breakup. The box and baffle add their own shape.
  • There's no baffle step correction or level matching in a basic calculator.
  • The drivers aren't in the same place. The tweeter and woofer sit at different depths and heights, which affects how they add up.

A calculator gives you a crossover. Measurements give you the right crossover.

How a crossover gets designed properly

  1. Build the box (the real one, not a test box) and mount the drivers.
  2. Measure each driver in the box: frequency response and impedance, using one measuring method from start to finish. Measuring speakers on a budget covers the gear.
  3. Load the measurements into crossover software. VituixCAD and XSim are free and widely used.
  4. Design to an acoustic target: the slopes you want, a flat summed response, sensible phase alignment, and a smooth off-axis response.
  5. Build it, measure the finished speaker, listen, adjust. Expect to swap a resistor or two.

Choosing parts: where the money matters

The values and the design matter far more than exotic parts. That said, a few choices are worth getting right.

Capacitors - Film capacitors (polypropylene) are the standard for anything in series with a tweeter or midrange. They're stable, low-loss and last decades. - Bipolar (non-polarised) electrolytics are fine for large values where the signal isn't passing straight through, like a shunt cap across a woofer, and they're much smaller and cheaper at high values. Never use a polarised electrolytic in a crossover. - Tolerance: for a pair of speakers, how closely the left and right match matters more than hitting the exact value. 2% parts make matching easy.

Inductors - Air-core inductors have no core to saturate, so they stay clean at any level. The trade-off is resistance (DCR): a big air-core value needs a lot of wire. - DCR matters most on the woofer. It sits in series with the woofer and acts like an extra resistor, softening bass control and shifting the baffle step correction. Heavier gauge wire means lower DCR. Heavy gauges (14 to 16 AWG) or cored inductors are common for woofer series coils; 18 and 20 AWG suit smaller values and the midrange and tweeter sections. - Cored inductors (iron, ferrite) get high values with low DCR, at the risk of saturating at high levels. Plenty of good designs use them on woofers.

Resistors - Wire-wound and cement (ceramic) resistors are the usual picks. What matters is the value and the power rating: a resistor in a tweeter L-pad can see a fair bit of power at volume.

Our own crossover range covers the lot: 2% film capacitors, 5% film capacitors, axial electrolytics, 18 AWG and 20 AWG air-core inductors, and wire-wound and cement resistors.

Building one: practical tips

  • Mount inductors at right angles to each other and a few centimetres apart, so their magnetic fields don't couple.
  • Use a solid board (MDF, ply or perfboard). Fix parts with cable ties and a dab of hot glue or neutral-cure silicone, so nothing rattles.
  • Solder, don't twist. Twisted joints work loose and corrode.
  • Follow the polarity on the schematic. Some designs wire the tweeter in reverse on purpose.
  • Mount it where it won't vibrate loose, and away from the woofer's magnet if you're using cored inductors.
  • Label everything before it goes in the box. Future-you will thank you.

Passive or active?

The alternative is an active (DSP) crossover: one amp channel per driver, with the filtering done digitally before the amps. It's flexible and easy to tweak, but needs more amp channels and a DSP. Passive crossovers need nothing but the parts, work with any amp that suits the speaker's impedance load, and once they're designed, they just keep working. Both are valid. For hi-fi speakers, passive is still the simplest way to a finished, plug-in-and-play speaker.

Quick questions

Can I upgrade the caps in my crossover for better sound? Replacing old, drifted electrolytics with new parts of the same value is a sensible repair. Swapping good parts for boutique ones usually changes far less than people hope. A better design beats better parts every time.

Can I use a crossover from one speaker with different drivers? No. A crossover is designed around specific drivers in a specific box. See Why you can't just swap out drivers.

What crossover frequency should I use? It depends on the drivers: where the woofer is still well behaved, and comfortably above the tweeter's resonance (often around double its Fs or more for a gentle slope). Measurements decide it.

Do I need a Zobel? Sometimes. It flattens a driver's rising impedance so a low-pass filter behaves as designed. Crossover software will show whether it helps.

Related guides

Back to blog