Almost Everything you wanted to know, or did not want to know about speakers

Thanks for clearing it up Freq Labs :D

I'll elaborate on speaker power ratings, since freq labs already brought it up.

"If "pink" is meant to be "pink noise" then it has no specific frequency, see here. I've never heard of any other pinks relating to audio."

Yes, exactly. And as far as we are talking about the thermal power ratings of speakers or loudspeaker elements, the actual frequency isn't a concern. Thermal power rating is what loudspeaker manufacturers usually quote, and in home audio it isn't of consequence. In PA applications, where loudspeakers are often designed to output the most sound pressure for the money, this power rating can mean something. That is, in those applications it is _sometimes_ the case, that voice coil overheating and melting is the limiting factor.

 

In home audio, where loudspeaker designs need not output high levels of sound to big concert halls, speakers are designed to have extended frequency response to the lower octaves. In PA application a 15" element might reproduce frequencies down to 40-50 Hz, a 6.5" midbass element (often augmented with reflex port) can be tuned to have usable output to the same 40 Hz or even lower. This is where the excursion or linear travel of the driver becomes the limiting factor, in thiele-small parameter sheets of drivers, this is often termed xMax. For every octave you go down in frequency, twice the displacement (the amount of air moved) is needed. And since the area of the cone of loudspeaker element stays the same, it is the amount the cone moves that needs to be doubled for every octave. So if at 160 hz an element needs to move +-2mm to produce 90dB(spl), at 80 Hz that is 4mm, at 40 Hz it's 8mm and at 20 Hz it's 16mm.

 

So what is limiting the cone to move +-16mm at will? The way dynamic loudspeakers work (that's the ordinary loudspeaker) is a voicecoil, or just coil that is suspended in a magnetic gap. Just in the same way an electric generator produces current from moving magnets against a coil of wire with inductance, the loudspeaker element produces mechanical movement from a changing current the amplifier provides. Magnet's aren't cheap though, and you need dense magnetic flux to have good efficiency. This done by directing the magnetic flux using ferromagnetic materials (iron is a good, cheap one) to a small magnetic gap. The voice coil sits suspended in this gap until it's exposed to a changing current, which it will try to resist by storing the energy into it's own magnetic field, resulting in motion as two magnetic fields repel and attract each other.

 

But, as soon as the voice coil partially leaves the magnetic gap, the force acting on it diminishes, resulting in distortion. There are two ways loudspeakers are constructed to counteract this. One is called underhung and the other overhung. These are basically like this,

 

---- magnetic gap ----

       underhung

---- magnetic gap ----

 

     ---magnetic gap---

o     v    e     r    h    u    n    g

     ---magnetic gap---

 

So as you can see, in both cases, the voice coil does not need to leave the magnetic gap even partially to allow some movement. But just as clearly, there is a only finite travel possible. Due to real world limitations (such as cost) we can't make the magnetic gap as large as we'd want, and for maybe less obvious reasons we neither can make the voicecoil arbitrarily long. The most important reasons for the latter is 1. lengthening the coil increases inductance  (duh) to the point where the voice coil becomes too good of a inductor, meaning huge amounts of voltage would be needed to drive the element at higher frequencies at all and 2. the voice coil is part of the moving mass, increasing the force needed to accelerate the damn thing.

 

As a general physics sidenote, these two are actually very much related, since it turns out inductance is the electrical equivalence of (mechanical) mass. Just as mass resists acceleration and tends to go keep moving once it's going, inductance does the same for current. As long as current in a circuit is in steady-state inductors are happily conducting, but as soon as it changes, they will resist it. In fact, you can model whole loudspeaker systems in a program like electronics workbench just by carefully adding the electronical equivalents of the loudspeaker voice coil mass, cone mass and such, along with self-damping properties with resistance and spring-forces with capacitance. This how the old analog computers were used to.. say calculate resonance frequency of a bridge. Just add the equivalent amounts of capacitance, inductance and resistance and see what output looks like with different input frequencies.

 

TL:DR,

Don't care about speaker power ratings, in home audio they are almost completely useless. Go listen to some speakers and pick the ones you like the most.