The sizing of the speakers with respect to the power output of the amplifier

My goal is to clarify what the best sizing of a loudspeaker versus the stated power of an amplifier might be, dispelling some myths and trying to provide some clarity that can help the enthusiast make an informed choice.

Let's start with the declared specs

A common Class AB amplifier has an efficiency around 60%-70%-the rest of the energy is dissipated in various ways. So how is it possible that a modern amplifier-and just to give an example-that claims 350 watts per channel, then draws a maximum of 850 watts from the grid? Let’s do the math. 350 watts for 2 channels is 700 watts, with 60% efficiency is 980 watts drawn, no doubt very dissimilar to the 850 watts claimed. So where is the scam, compounded also by the fact that an amplifier also has numerous circuits, between preamplification and services, that consume additional power? We are certainly not back to the 1980s, when Car Audio was all the rage, and when incredible powers were claimed for apparatuses, not only lilliputian in relation to the claimed powers, but especially powered at 12 volts without any surdevoltage circuitry (step-up of the supply voltage).

Actually, each power section of those 2 mentioned in the example are capable of delivering 350 watts, relative to the voltage supplied by the power supply circuit of the syntoampli, but they simply cannot do so in unison. The concept, pass me the expression, is “sound,” in the sense that it is impossible for there to be an identical power demand–for 350 watts–for all 2 channels. What becomes important at these junctures is to have well-sized filter capacitors, which will always act as a ready and available energy reserve under all conditions, so much so as to cope with continuous absorption demands, no doubt more pressing than can be with a normal stereophonic amplifier. Let us try to figure out-academically-what the real continuous power on all 2 channels of the amplifier might be. 850 watts of absorption, take away a 40% of varied dissipation and we are at 510 watts, which, divided by 2, does not reach 255W continuous. This, however, is not the figure to keep in mind when designing the acoustics to be connected to the amplifier. The output stages and power supply are sized-usually-to deliver the stated power, at the specified distortion. This is not “cheating” on the power output, but rather sizing the amplifier much more than amply, which will simultaneously be able to deliver about 255W per channel ! Then if there will be special power demands there are the filter capacitors to come to the rescue of the power supply, being reservoirs that discharge impulsively while still making the machine deliver the maximum declared watts for a few moments. Those who “cheat”-and this is done more in low-cost products or in the home theater world-do so blatantly-declaring power figures on 6 ohms, which is certainly higher than that on 8 ohms commonly used. 100 watts on 6 ohms is plus or minus 70 on 8 ohms at best. All this is to say, that one only has to look at the manufacturer’s stated power for each channel, add it up, and see the maximum power draw of the amplifier to understand the “real” sizing of that machine.

Some excellent examples

The Yamaha A-S3200 claims 100W per channel into 8 ohms, a power consumption of 350W and 700W maximum, with a power transformer of 625VA. So we are dealing with EXTREMELY conservative specifications, which make us realize how very conservative the Japanese engineers have kept themselves, guaranteeing that the power output is indeed 100W per channel, and both channels simultaneously. And this in listening can be felt, which is after all why one of the things that is taken care of most in absolute level products is precisely the power supply. First of all as sizing. Another shining example is Denon’s “old” PMA-SA1, from “only” 50+50 W into 8 ohms, but with a 230W power draw. If you look at the photos of the internals of both products, you will notice the equipment as far as transformers and filter capacitors are concerned, which immediately suggest that the adjective “oversized” is reductive.

Speaker sizing, too many myths and few certainties

Let us come to speaker sizing and dispel some myths. To do this we must briefly discuss distortion and its deleterious effects. When a 10-watt amplifier finds a musical peak at its input that forces it to deliver more power than its maximum it, being unable to do anything else, will deliver for the time of the peak, in continuous operation, only the 10 W of which it is capable. This time will last not only the moment of the musical peak, but as long as the rise and fall time of the wave of that peak, which is a time considerably longer than the peak itself, is called “shearing” (refer to Fig. 1 for clarity)

We have therefore subjected the speaker and the loudspeaker to a continuous signal for certainly a much longer time than was actually required by the peak. It is very important to understand this concept. A 50-watt amplifier would have handled that peak without any uncertainty, exposing the speaker only for a moment to that power, while a less powerful amplifier-having to deal with a driving condition beyond its limits-gives what it can, as far as it can, delivering less power to the speaker, but for a longer time. At that time the listener will experience a strong feeling of dynamic compression, not due to the mechanical limit of the speaker, but to the electrical limit of the amplifier. This “wasted” energy can cause various kinds of damage to the speakers. The copper wire of the voice coil, the motor of the speakers, has a variable cross-section according to the frequencies handled by the transducer. That of the woofer is the largest, sometimes rectangular in cross-section; those of the midrange when present in models with more than 2/ways-and of the tweeter are correspondingly smaller in cross-section, in Fig. 2 the voice coil of a tweeter.

In a hypothetical energy breakdown, if 100 watts go to the woofer, 40 watts go to the midrange and 15 or less go to the tweeter. Damage of electrical origin caused to loudspeakers is generally of two types: a first involves the interruption of the voice coil, resulting in the silencing of the component (also called the “fuse effect”). The second is referred to slangly as “fritting” Fig.3, with the insulating paint around the copper “cooked” by the high temperature reached by the metal conductor at the moment of the electrical overload, which – by enlarging the coil – creates a clearly audible rubbing, even though the speaker – it really has to be said – is still “croaking” but sometimes literally seizes up as if it were a piston in its jacket but without oil. The former phenomenon is more common in tweeters (characterized by a small cross-section of electrical wire), the latter in midranges and woofers.

Usually, a “dumb” loudspeaker not only has the voice coil winding interrupted, but has also been subjected to “frying,” considering that high power for an excessively long time creates first the melting of the insulation, and then the interruption of the winding itself. These phenomena are absolutely dependent on the time duration (and, of course, on the impulse power given by voltage x current), to which the voice coil is subjected, with associated electrical and mechanical stress. An undistorted spike, very powerful and of very short duration, a “burst” (a single pulse) does not usually create damage, while a spike in the “clipping” phase, thus distorted, albeit of less power, but more prolonged in time, is highly likely to create damage because it exponentially raises the temperature of the coil, which in the absence of excursion of the speaker cannot even ventilate. The speed of the undistorted pulse, on the other hand, does not give the voice coil time to overheat, producing the “fuse effect” (Fig 4);

upon the immediate cessation of the pulse itself, the movement of the coil and the resulting aeration allow the generated heat to dissipate. A pulse prolonged in time, on the other hand, much more easily “fries” the moving coil, creating an “avalanche” effect. The coil gets nailed, the force ratio between the generated magnetic field and the permanent magnetic field of the magnet changes, and the “fuse effect” arrives to burn out the speaker irreparably. It is best to rely, for a more realistic evaluation of the component being analyzed, on the manufacturer’s stated RMS power rating. If supplied with a value of 6 ohms, the power will have to be “subtracted” by 30 percent to obtain the reliable value on a load of 8 ohms. Once this figure is obtained, we can reasonably rely on an “equivalence” of values: to a speaker with an impedance of 8 ohms capable of accommodating 100 watts RMS, we can associate an amplifier declared for 100 watts RMS on the same resistance value. All this reasoning also implies the following assumption: it is much more likely that a small power amplifier will burn out the voice coil of a large speaker than the other way around. The causes of this should now be clear. A low-powered amplifier distorts much more easily than a high-powered one. It should also be kept in mind that a speaker declared to accept 100 watts RMS is certainly capable of withstanding 200 watts, albeit for short periods. The important thing is that there is no distortion phenomena going on. The same loudspeaker, exposed to a distorted signal generated by a 20-watt amplifier, could be damaged instead. Therefore, never be afraid of pairing powerful amplifiers with smaller, nominally more delicate speakers. It is, in fact, more difficult to exploit to the point of distortion a high-powered amplifier, rather than one of lesser power output, with which it may happen – with some ease – to exceed “12” hours with the volume knob, with the risk of incurring – more easily – distortion and potential damage to the speakers.