Speaker Power Ratings

jwatte's picture

One question that consumers often have relating to speakers, Hi-Fi and home theater systems, is "why are different systems rated for different amounts of watts, and why do watts matter anyway?"

It all starts with physics. A speaker is an electromagnet (wound wire around a permanent magnet, for example) that is powered by the current flowing through the wire. Depending on how much wire, how wound, and of what gauge, the electrical resistance of the speaker will be different. This resistance (or, really, the inductance, but let's simplify) is measured in Ohms. So, depending on how the speaker is constructed, it will draw more or less current at a given voltage; this can easily be measured with fairly simple instruments, and some weighted average over the speaker's frequency range is it's "Ohm" impedance rating.

The amount of current that a speaker draws at a particular voltage level is inversely proportional to the resistance (inductance), determined by Ohm's law. Ohm's law says "voltage = current * resistance," or, re-formulated, "current = voltage / resistance." Thus, you can expect 8 Ohm speakers to draw half the current of 4 Ohm speakers.

Now, the actual power is the product of voltage and current: "power = voltage * current." If you put the forumulation of resistance-based current into the formula, you get "power = voltage * voltage / resistance."

Power is usually measured in watts, and is the same kind of watts that power a light bulb, or an electric heater. If you've ever touched a 100 W light bulb after it's been on for a while, you should know that it gets pretty hot. The same thing happens to speaker elements. Different speaker elements are designed with different amounts of cooling, and different levels of heat tolerance. Thus, different elements will be able to take different amounts of power before they reach their design temperature limit, after which one of many things will happen:
1) nothing -- you're lucky
2) some component will blow, typically in the passive cross-over
3) glue in the winding will melt, or paper will char, and unwind the coils
4) the actual wire will overheat and break (in combination with the physical stress of moving)
5) something else that you're not going to like :-)

So far, so good. However, sound is not power -- sound is movement of air. The whole point of the speaker is to transfer electrical current into air movement. Because air has friction, that's where the power is needed. However, certain speaker elements are better than others at moving air, given a specific amount of power. This is called "sensitivity." A speaker with a higher sensitivity will move more air (and sound louder) than one with a lower sensitivity, at the same power level.

But we're not done! There is a limit to how far the speaker element can move in response to the current (and the magnetic force generated by the coil). At the end of the excursion, the membrane of the speaker will be stretched, and the movement of the membrane (and thus the air it pushes) will be non-linear to power. For a very sensitive element, this may happen before you get to the point where the element heats up too much. For those kinds of elements, as you put in more power past the stretch limit, you don't get more air movement; thus they don't get much louder, but instead get higher distortion levels. Some speakers are explicitly designed to allow this (think stacks of Marshall guitar amps/speakers at a rock concert).

A speaker manufacturer who wants a large "watt" number will rate his speaker at the limit where the speaker will either overheat, or where prolonged use will actually physicall break the membrane. Actually using the speaker at that power level will sound bad because of all the distortion that's introduced. Other speaker manufacturers will rate the element to the highest current draw (and thus wattage) that will generate sound without endangering the element, and without introducing more than X percent of distortion (typically 0.1% THD, which is equivalent to -80 dB off full signal).

It gets more complicated once you realize that the different speaker elements interact both in current draw and pushing air, and the impedance is really frequency dependent, and the coils in the speaker elements may interact with the components in the cross-over filter, and various physical characteristics of the speaker enclosure will resonate or dampen the elements at various frequencies. Which is why speakers aren't as simple to design as light bulbs -- the range of variables and compromises is a lot wider.