For me, when I first started making speakers, crossover design was the most intimidating part of the whole process. I thought there was definitely some magic going on behind the scenes that I just didn’t have access to.
Now that's actuality kind of true, just not in the way I was thinking about it.
There are a few ways to design a crossover. The first, and the old school way is to build your speakers, is to sit in from of them, listen to some music, add the components you think the speakers need, listen again, add or subtract more components, listen, components, listen, components, and so on until you’re happy with the sound.
This is obviously a long process and not the most efficient use of time in 2018.
Luckily we now have software that can simulate different crossover components, values, and configurations, so we can do everything I just talked about, but without having to actually build the speakers first and sit there with a bucket full of components.
So there is a bit of magic going on, it's just all software magic.
Now, that being said, I can’t underestimate the importance of having an assortment of capacitors, inductors, and restores on hand in various values. We’ll talk more about this in later steps, but once you’ve designed your crossover in software and have your speakers and crossover built, you’ll still want to go back and make tweaks to make sure you get the sound you want.
I believe the only real test for great sounding speakers, is by listening to them live in a real room and making crossover tweaks with actual crossover components.
Alright, with that being said, let's talk about what a crossover is.
Intro To Crossovers
Your music source, your phone, or turntable, whatever… puts out an audio signal that contains the full spectrum of sound from real lows at 20Hz to real highs at 20,000Hz. A point source speaker system takes all that music information and plays it through one speaker driver. So the same driver plays the really lows and really highs at the same time.
Well, physics makes it so one speaker has a really hard time doing that. And in the real world, not many speakers do it well.
So, to make for better sound engineers design drivers that are individually suited to reproduce sound within a certain section of that low to high-frequency range. We call the drivers that handle the highs tweeters, and drivers that handle the lows woofers.
So how do we make sure the tweeters only get the highs and the woofers only get the lows? With a crossover! Yay!
A crossover uses a combination of capacitors and inductors to split the full audio spectrum and send the correct frequencies to the correct drivers in your system.
Alright, let's talk about the common components that make up a crossover.
All of these components impede the flow of electrons, this is called impedance. How they do this is what makes them different.
Resistors lower the volume equally across the frequency spectrum.
So if your tweeter is too loud, you would put a resistor on it to lower its volume.
Capacitors or “caps” have the ability to lower the volume of the frequency spectrum in relation to a given frequency point.
In the example of a high-pass filter, any frequency above a given point passes through the capacitor as if it wasn’t there, but the volume of the frequencies below that given point begin to decrease as the frequency gets lower, as you get further away from that point, you get to a frequency level where the volume is so low you can no longer hear it.
Capacitors increase impedance as a frequency decreases.
Inductors or cokes or coils are basically the inverse of a capacitor. They also have the ability to lower the volume of the frequency spectrum in relation to a given point.
Inductors used in a low-pass filter lets any frequency below a given point pass through it as if it wasn’t there and begins to lower the volume of the frequencies above that given point.
Inductors increase impedance as frequency increases.
Okay, now let's talk about orders.
The way we use these components in conjunction with each other represents a crossovers order.
The higher the crossover order, the steeper the cutoff slope of unwanted frequencies.
So let's take the inductor we just talked about. To use it as a low-pass filter we would place it in series with the driver, which means we would connect one end of the inductor to the positive lead of our audio amp and the other end fo the inductor to the positive lead of our driver. The negative lead of the driver would be connected straight to the negative lead of the audio amp.
This would be a first-order low-pass crossover. Meaning we only used one filter (the inductor) to lower the volume of the high frequencies after a given point on the frequency spectrum.
But the volume of those high frequencies doesn’t just suddenly go to zero after your crossover point. Its a gradual taper, or a slope until at some point the volume becomes zero.
So if you set your low-pass crossover point at 500Hz, you’re still going to hear some sound from 600Hz, but you’ll hear even less sound from 700Hz and even less form 800.
A first-order crossover has a 6db per octave slope. So for every octave past the crossover point, the volume of that octave will be 6db less from the octave before it.
A second-order crossover has a 12db per octave slope and a third-order has an 18 dB per octave slope.
The higher your crossover order, the faster you’re cutting off unwanted frequencies going to that driver. But, the higher the crossover order, the more components needed for your crossover, the more complex your build becomes, and the more comments you have in your signal path.
All, not great things.
There are LOTS of things you’re able to consider when selecting a crossover order, these are only a few examples. I definitely don’t have the depth of knowledge (or the time) to go over the advantages and disadvantages of each. And just like everything else in speaker design and building, you can get as deep and technical as you want.
I usually start out with either a first or second-order crossover, if that doesn’t get me the result I need, I’ll up it to a third-order. I rarely go above a third-order
Crossover Component Layout
We already went over the layout of a first-order crossover. It’s simply a capacitor in series with the driver for a high-pass filter and an inductor for a low-pass.
A second-order uses the same base as a first-order, but just adds a second component (the opposite component from the first) is parallel with the driver after the first component.
So a second-order high-pass filter looks like this:
A third-order uses the same base as the second-order but adds a third component to the filter (the same as the first component) in series with the driver after the second component.
So a third-order high-pass filter looks like this: