GOODSOUND!GoodSound! "How To" Archives

Published August 1, 2001


Watts Up with Amplifier Power?

How many watts do I need? Is the power rating the only amplifier spec I should concern myself with? Is bigger always better?

Amplifier power behaves logarithmically. Each time an amplifier drives a loudspeaker to an increase in sound pressure level of 100% (or adds 3dB of gain to the signal), it must double its power output. This means that the loudness ratio between 2W and 4W is exactly the same (3dB) as that between 100W and 200W, even though the difference in wattage is far more extreme.

Put differently, at full output, a 200W amplifier can only play 3dB louder than a 100W unit. However, it's just as rare to drive an amplifier to full output as it is to push a car to full redline. Besides, if you did drive your amplifier to full output, you'd know it -– it would audibly clip and you'd hear distortion. So the 3dB advantage of the larger amp would probably end up just sitting there. So why buy more amplifier power than you might need on average? Because music frequently has wide extremes in dynamics (which is another way of saying that it ranges greatly from the quietest passage to the loudest), and while an amplifier might cruise along happily outputting 10W, some passages might require ten times that amount of power for an instant.

Put in the most basic terms, four different factors determine how much amplifier power you need.

  1. Speaker sensitivity. This was covered in detail in our last installment. Briefly, with the same power, a more sensitive speaker will play louder than one with less sensitivity.

  2. Room size. Everything else being equal, a larger room absorbs more power than a small room. That's because the speaker sound we hear is the sum total of direct and reflected sounds, and sound pressure diminishes with distance. Hence, to overcome the increased volume of a large room, an amp needs to put out more power.

  3. Listener distance. This can be independent of room size. If you sit close to the speakers, even in a large room, you will be using less power than if you sit farther away.

  4. Your average playback level. Are you a background music type or party animal? Average listening levels (neither background nor party but, well, normal) require less power than you may think -- the average listening level requires somewhere around 10 watts in a standard 14'W x 20'D x 9'H room.

In addition, most popular music recordings compress their dynamic range electronically to keep the differences between the quietest and loudest passages to a minimum. The average power requirements barely fluctuate. Classical music can feature extremes that range from pianissimo to triple forte. Without altering the average power requirement, this type of material might demand spontaneous peak power bursts that could be 100W or more.

This brings us to the way power ratings must be read. Continuous or RMS power simply tells us how much power a given amp can deliver day-in/day-out on a steady basis. (RMS stands for root means square, an equation that specifies average power.) An amplifier's peak power might be three times its continuous rating, or even greater, but that rating is only available for milliseconds. There aren't many manufacturers left who'll try to fool you by passing off peak power ratings as average, but there are still quite a few who will specify an amplifier as having an output rating at a specific frequency rather than over an range. When you see a spec like 100W @ 1kHz, you should be suspicious -- this is not the same thing as an amplifier that delivers 100W from 20Hz-20kHz. The first is power output at only one frequency. The second is power output at the full bandwidth of what’s considered important for human hearing. Also look for the words both channels driven, since any amplifier can deliver more than its true peak output when only one channel is driving a loudspeaker (and for home-theater enthusiasts, look for something like all channels driven since there may be five or six channels of power).

Make sure you're not comparing apples with oranges when studying power specs. This means you must consider the impedance rating to which the power rating is attached. Since most speakers (especially in the affordable arena) are rated at 8 ohms, amplifiers and receivers most frequently list their power output into 8 ohms. However, speaker impedance generally varies with frequency, so you'll see some speakers specified as having a "nominal" 8-ohm impedance. And sometimes the specification will simply be an average taken over a wide frequency range, so be careful and ask your stereo salesman if a given speaker's impedance is a real figure, a nominal reading or an averaged reading. If the speaker you like has an impedance rating that drops below 4 ohms, you'll want to look at an amp with a robust power supply.

Amps with robust power supplies will sometimes feature a 4-ohm rating. Ideally, this rating should be double that of the 8-ohm figure, which indicates that such designs are stable and compatible with low-impedance loads. Certain amps will also specify their power ratings into 2 ohms and even 1 ohm. The true brutes among these will double power into each lower impedance, all the way into 1 ohm, giving rise to astonishing figures of 1000W or higher. But, since affordable speakers tend to be either 8 ohms or 4 ohms, you won't need to worry too much about power output into lower impedance loads.

For completeness' sake, we now need to cover amplifier current. A reasonably close car analogy is the horsepower/torque equation. A high torque rating is useful when negotiating steep inclines and towing trailers. Depending on how you use your car or truck, high torque could be more important than ultimate horsepower specs. Similarly, decent current delivery in amplifiers is necessary to control speakers with large and/or multiple woofers. Such speakers make greater current demands and (within the context of our analogy) could be thought of as steeper hills or heavier trailers.

The subject of current delivery gave rise to the popular notion that not all watts are created equal. In a limited sense, this is true, as long as it is not taken literally to mean that Denon watts are different from Kenwood watts. It simply means that an amp's current capability and power rating need to be looked at in conjunction with one another.

Our earlier explanations indicated how much (or, more accurately, how little) average power we tend to use. You can now appreciate that a 50W/high-current amp may outperform a 200W/low-current design when mated to demanding high-end audio-type speakers.

While massive amplifier current ratings are one way to evaluate a power amplifier's ability to drive all loudspeakers, they are not essential if the loudspeakers you have chosen are reasonably sensitive, your room is of moderate size and your listening habits are relatively sane.

The truth is that a simple listening session will tell you whether a particular amp/speaker combination is copacetic or not. Try to listen to an amplifier at home, so your room and other equipment can be part of the audition. Play some of your favorite records -- the ones you are familiar with, not special "test" discs (and not an "approved" type of music -- if you normally listen to rock, audition with rock, not classical). First listen at your regular loudness level and see if the amp/speaker combination can play at the level you prefer. Listen for the things you like about your musical selections. Is the bass punchy and tight and deep? Are the vocals clear? Is there a big difference between the loudest and softest passages (keeping in mind that not all recordings have such a distinction)? And if it all sounds good at your regular playback loudness, turn it up to about the loudest you'll ever listen at and check for the same properties. You'll be surprised at how much your ears can tell you in your own room.

Besides, that's where you'll be listening to it once you've bought your ideal amplifier, so listen carefully and put it through its paces before you buy it.

GOODSOUND!All Contents Copyright 2001
Schneider Publishing Inc., All Rights Reserved
Any reproduction of content on
this site without permission is strictly forbidden.