PA Systems (Part 2)
Part two of Ken Dibble's explanation of P.A. complexities.
It is often said that a good loudspeaker should be like a musical instrument, but in fact, nothing could be further from the truth. The purpose of a musical instrument is to generate or create its own sound according to the whim of the musician. Not so with a loudspeaker, however, whose one and only purpose is to reproduce sound fed into it as faithfully as possible, adding or subtracting nothing, and certainly not generating sounds of its own! If it does, for example, by creating a boom at the bass end, or by becoming tinny in the treble region, it is a poor loudspeaker indeed and should have no part to play in professional public address. As stated last month, the arrival of the 4x12 column was a major step forward in improving the quality of P.A.s, especially when compared to the puny little columns of 6" and 8" loudspeakers that first appeared on the scene. To non-critical ears at that time, they really did sound good. Most 4x12 columns contained four 12", 25 watt units in either a straight or staggered vertical array, in an open back cabinet rated at 100 watts. A number of manufacturers preferred to use a closed-back or "infinite baffle" design, and both techniques have their respective merits and disadvantages.
The prime purpose in leaving the cabinet back open is to reduce the booming and often distorted bass sound that is usually associated with closed back cabinets, and in this, some success usually results, but not without some sacrifices in other areas. An open back cabinet has, for example, very low acoustic suspension properties, and therefore offers little or no acoustic impedance to the movement of the loudspeaker cone, thereby making necessary the use of a loudspeaker drive unit with a very stiff cone system in order that the units own mechanical impedance can limit excessive cone movement at low frequencies. It is unfortunately dictated by the various laws of physics that such a cone will, of necessity, have a rather high natural resonant frequency, and this will restrict (among other things), the downward extension of the loudspeaker's response capabilities. Other drawbacks include a reduced life expectancy from the actual loudspeaker drive units due to cone fatigue, and the desirability of de-rating the units power handling capability to avoid blowing out cones under overload conditions.
The sealed box, or infinite baffle cabinet as it is generally termed, will usually need to be considerably larger than its open back counterpart if it is to have a similar bass response. It relies on a volume of air trapped inside the cabinet to control the movement of the loudspeaker cones instead of on the cones themselves, and for this reason, a greater freedom of choice of loudspeaker drive units exists, enabling softer cone units with improved low frequency characteristics to be employed. As the loudspeaker cones are driven back into the cabinet, the trapped air is compressed in a similar way to a spring under compression, and so further cone movement is restricted. As the cones are driven forward, a vacuum is formed and again, cone movement is restricted in a similar way to stretching the spring. In this type of cabinet, loudspeakers will run their normal useful life and at their full rated power levels without undue fear of damage, and overall, this would seem to be the better of the two cabinet configurations to use. When a number of loudspeaker units are mounted one over another in a vertical column, certain interactions take place between the units so that the response of the column is not similar to the response of a single unit. If the column is properly designed as, for example, the linesource loudspeakers used in Westminster Abbey and St. Pauls Cathedral, this can be turned to great advantage, but if speaker units are placed in a haphazard manner in a convenient box without consideration of the rules, the overall performance of the column is likely to be considerably worse than that of a single unit working on its own. Unfortunately, the latter was generally the case with most band P.A. columns, and the reasons for this will be considered.
Sound can be expressed in two ways, either by reference to its frequency in Hz. (or cycles per second, as it was once known) or by reference to its wave length. Almost all loudspeaker design is based on wave length. Certain dimensions of a loudspeaker system are bound to coincide with the wave length of some musical notes, and this is the condition that gives rise to most of the problems. For example, the actual cone diameter of a 12" loudspeaker is about 10", and this is the approximate wave length of the E two octaves above middle C on the keyboard.
As a 12" loudspeaker reproduces music below this frequency, it will radiate the sound in a pretty uniform manner, depending upon the quality of the actual loudspeaker being used. However, if high frequency sound of wave lengths smaller than this is fed to the loudspeaker, a progressive narrowing of the radiation takes place until all the treble output of the loudspeaker is contained in a narrow, penetrating beam of very harsh sound that is only heard as an ear-shattering din by listeners standing on the central axis of the loudspeaker. The only treble heard by the majority of the audience is that which is reflected off walls or the ceiling, and this is very spasmodic and usually sounds as an echo.
In the case of loudspeakers using twin-cone or dual-concentric units, similar problems exist. With these, a mechanical crossing-over takes place due to the difference in mass, size and natural resonance of the two cones. Generally, the rate of crossover is much slower than with an electrical crossover, and takes place after the beaming of treble from the large cone has started, although it does reduce the extent of beaming by removing the higher treble frequencies from the bass cone and transferring them to the small centre cone. However, further up the treble range, we reach a point where the wave length is now similar to the diameter of the smaller cone, and beaming again starts, this time at around top C on the piano keyboard. A speaker of this type will certainly provide better quality sound than a similar single-cone unit, but will otherwise pose similar problems.
When several loudspeakers are mounted in a vertical column, the problem becomes worse. The critical factor here is the spacing distance between the loudspeakers fitted, for when the wave length of the sound reproduced is smaller than this spacing distance, further spurious beams of sound are produced in the vertical place, and these beams move around the column, according to the wave length of the sounds being reproduced at any one instant. Usually, these spurious beams succeed in reaching the microphones on stage and feedback results, even though the overall sound level in the hall may not be particularly high. A good quality radial horn loudspeaker, however, is generally capable of spreading treble evenly over a wide area, and is not usually prone to generating these spurious beams referred to. Such a unit will have a frequency response in the order of another octave upwards of the response of a 12" cone type loudspeaker. The obvious solution is to roll off the signal fed to the column of 12" loudspeakers at a point before the wave length is equal to the cone diameter, and to feed all the remaining middle and treble to a suitable horn unit by means of a cross-over network. This will result in a smoother, tighter sound over a wider range of frequencies that will be relatively free of unwanted beaming, thereby providing better sound quality for the audience and reducing the likelihood of feedback into the bargain. The same would apply to a single 15" or 18" bass unit, only the crossover point would have to be lower due to the larger effective cone diameter of such units. As evidence of this, take a look at the floor monitor speakers currently used for foldback in some larger P.A. rigs, and either a 12" or 15" unit will be found with a top quality horn alongside. The reason for such an expensive horn "just for foldback" as one might think, is not that the ultimate quality is needed for foldback, but that the better horns will more evenly spread the treble, and eliminate the beams that would otherwise cause certain feedback.
The favourite design concept for a P.A. loudspeaker would then seem to be an infinite baffle cabinet of fairly substantial proportions housing one or more loudspeaker drive units of 12" diameter or larger, depending on individual needs. To this is added a really good radial horn assembly, fed via a crossover network that not only prevents any bass from reaching the horn and destroying it but — and of equal importance it must also remove the treble from the bass loudspeaker at a frequency where wave length is greater than either the effective cone diameter, or the spacing distance between speaker units, whichever is the larger dimension. Such a loudspeaker system, if properly built using good quality components, would be capable of providing good quality sound over a wide frequency band and at fairly high power levels, and is still a relatively simple piece of equipment, but surprisingly few manufacturers seem to market such an item. In fact, every conceivable shape and variety of loudspeaker seems to be readily available, except the obvious simple solution to good quality P.A. I wonder why this should be?
There are, of course, other types of cabinets besides infinite baffle and open back, some of which are probably better from an acoustic point of view than either of the two types discussed. One such type is known as a vented or distributed port enclosure, and is somewhere half way between the two mentioned, as the vent is only large enough to partially relieve the compression of air in an otherwise closed cabinet. The Shure Vocalmaster column is probably an example of this type of system at its best. The rear vents not only regulate the loudspeaker cones within the cabinet, but also have a significant effect on the directional characteristic of the loudspeaker. One major drawback with this type of design is that the response of the loudspeaker varies depending on the physical position of the loudspeaker relative to the nearest wall, and can often cause otherwise inexplicable feedback.
When really high power levels are required for a large auditorium or outside stadium, the answer is not to stack columns, infinite baffle or reflex cabinets to the ceiling and drive them with as many 100 watts amps as can be found. All that will be achieved in most cases is acute distortion and deterioration of quality due to phase differences between the multiple amplifiers and loudspeaker units. In order to achieve high quality sound at very high power levels, a different approach to the problem is necessary, and explaining this will provide more than enough material for next months article.
Feature by Ken Dibble
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!