The Lowdown on Subwoofers

As the director of the electro acoustics laboratory at the University of Wisconsin, I have daily contact with students who have a great interest in high-fidelity systems. They often ask me about subwoofers: What is a subwoofer, and how did this term originate? Why not use two subwoofers for a stereo system? Or why not use two loudspeakers that have been designed for flu-range bass response in the first place?

The subwoofer gained popularity because of the great number of small loud speaker systems designed with very small woofers, 5 to 7 inches or so. This tactic was the result of trying to make acoustic-suspension, bookshelf speakers smaller and smaller. Design alignments often give the best result with a relatively small driver piston size. In fact, it is quite possible to make a loudspeaker with a driver of, say, 5 or 6 inches in diameter that will have flat frequency response down to 30 Hz or so. However, it is not easy to get a very satisfying amount of power out of such a system at the lower end of the frequency range. This is because the laws of physics require relatively small pistons to move a large distance to generate the required volume velocity of air, which, in turn, gives the acoustic power output required.

Many of the small loudspeakers with small woofers now on the market are quite appealing for their size and price, as well as for the high quality of their sound performance. In recent years a good understanding of the physics and design of small loudspeaker drivers, and improvements in materials, has made this possible. These loudspeakers are able to emit adequate power in the upper two frequency decades, 200 Hz to 20 kHz, but often lack the necessary cone area to give satisfying power output, especially in the bottom half of the lower decade of frequencies, 20 to 100 Hz.

The power output delivered by a loud speaker goes down rapidly as frequency decreases and even more rapidly as the speaker’s radius is decreased. At low frequencies, the limited mechanical motion of the loudspeaker cone limits the acoustic output. The Table in the sidebar gives approximate levels of acoustic power out put for various driver sizes. Clearly, if we want to have as much as 0.1 acoustic watt of sound output at 20 Hz, we are hard pressed to do so with anything less than a couple of very-long-throw 10- or 12-inch drivers. At 50 Hz, the problem is much less severe. I have always solved this limitation by using a large cone area for the bass end of my loudspeaker designs. For most large living rooms (3,000 cubic feet or larger), this requires a couple of 12- or even 15-inch drivers and a floor-standing box of 3 to 4 cubic feet for each speaker in a system.

It may not be practical for many people to have such a large system, for reasons of but an interesting, solution to the problem was the introduction of the three-piece system. This generally consists of two quite small loudspeakers that cover the frequency range down to 100 Hz or so. Such loudspeakers, with a 5- to 6-inch piston, are fairly easy to design. But a system with the bottom two octaves missing is not very satisfying, so a third component is added, the subwoofer.

The rationale for using a subwoofer is that a loudspeaker can be designed to give appreciable power output and have optimal performance at very low frequencies because it does not have to reproduce the higher frequencies. In theory, the box can be placed in a more or less out-of-the-way location in the room since the low frequencies have very little directionality. It was generally thought that a single monaural channel was all that was needed, because the frequencies below 80 Hz are typically mixed to monaural anyway. (This latter factor was a result of the limitations of the mechanical reproduction process used in making vinyl records.)

The assumptions that justified using a single subwoofer have some basis in fact.

Three-piece systems have been, and are still, popular when space and room styling take precedence over, other acoustic considerations. However, I have some concerns and reservations about the technical justifications for the three-piece system. Certainly not all, or even most, loudspeakers need a subwoofer or two to make them Illustration complete in the extreme low bass. In fact, more and more floor-standing loudspeakers of the past several years have two 8-, 10- or even 12-inch drivers that can provide entirely adequate bass response for a relatively large listening room. It may be the wisest decision to get a truly full-range loudspeaker in the first place, since it will have been designed as a unit and have smoothly balanced response over its full range. However, in the case of smaller bookshelf-type loudspeakers, considerable advantage can be had by removing the lower two octaves of the sound and feeding them to a good subwoofer. The advantages are that the subwoofer now handles the hard-to-reproduce very low frequencies, and the smaller loudspeakers are freed of the power and displacement demands made by the low frequencies. This can result in lower distortion in the small loud speakers and considerably improved sound quality from them. It also allows planning for system growth; if you first invest in a good, small pair of loudspeakers, you can augment them later by increasing bass capability with a subwoofer.

A question of concern is whether pile or two subwoofers should be added. If one subwoofer is added, the result is a three- piece system with monaural bass below bout 80 to 100 Hz. If two subwoofers are added, the result is two full-range loud speakers with stereo bass. Does it matter if we have stereo or monaural bass? I would claim that it does, for at least two reasons. The first is that we have much better pro gram material today from the Compact Disc than we did with the vinyl record. We have greatly extended bass frequency response and have two entirely independent channels. That is, there is more realistic bass—and it is often completely stereo. We should try to reproduce this new and better sound, if possible. To do so requires two complete, full-range, stereo channels with two independent bass drivers.

The second reason is not quite so obvious: There is considerable interaction between the room and the bass loudspeakers. We all know that placing the bass loud speaker near a wall or in a corner of the room changes the amount and distribution of the bass frequencies in the room. At low frequencies, strong and widely spaced room modes are occurring. These modes are evidenced by the fact that as we move about the listening room, some locations have a lot of bass while others lack bass. The locations where more or less bass is heard move around within the room when the loudspeaker is moved from place to place. The problem with the three-piece system is that we have only one bass loud speaker. Thus, the room is only excited from one point, and we have a set of modes that are quite audible. When two bass loudspeakers are placed in the room and are excited by a stereo bass signal, each will have its own set of modes. It is not likely that the lack or excess of bass from one loudspeaker will fall at exactly the same point as for the other loudspeaker. Thus, a considerably better uniformity of bass response is obtained when the room is excited from two sources. (See “A Theoretical Look at Subwoofer Operation in Rooms” by John D. Bullock, AES Preprint No. 2861, 1989.)

For best-quality bass response, I would therefore suggest that you use two full- range loudspeakers, that you augment a three-piece system with a second bass loud speaker, and that when you enhance a smaller system with additional subwoofers, add two of them, not just one. In this fashion, you can take advantage of the full stereo sound at low frequencies available from the Compact Disc and other digital sources—or from analog cassettes, for that matter.

=============

Loudspeaker Driver Capabilities:

Approximately, loudspeaker drivers with the following advertised diameters will be able to generate the acoustic power levels listed below. The Table is based on an actual diameter of about 85% of the advertised diameter, a peak motion capability of about 5% of the radius (a generous amount for most drivers), and placement away from the corners of the room. Since all of these factors have a strong effect on the power capabilities of the loudspeaker, the approximate nature of the Table must be strongly emphasized.

The power levels listed look quite low, but fortunately an acoustic power output of 0.1 watt is very loud in a typical listening room. A reference level might be that a loudly played piano emits about 0.1 to 1 acoustic watt.

The most important aspect of the figures is to note that the loudspeaker runs out of power output capability very fast as the frequency is lowered, and it gains power capability very fast as the frequency i increased. Typically, one or two 8- to 10-inch drivers (or the equivalent) are needed to generate realistic acoustic power levels at 40 to 50 Hz in a typical listening room of, say, 2,500 cubic feet.

=============

Author: R. A. Greiner, a Fellow of the Audio Engineering Society is a professor of electrical and computer engineering at the University of Wisconsin in Madison, Wisc.

This article adapted from Audio magazine, August 1993.

============