How big an amplifier should I get for my loudspeaker? While power ratings are normally used to answer this question, the real answer is found in three separate and distinct issues.
The first issue concerns loudspeaker power ratings, one of the first things people examine when looking at loudspeaker specifications. This rating means that the loudspeaker has passed a test where the test signal is increased until damage or failure results from exceeding the loudspeaker's thermal or mechanical limits. A power rating is determined based on the test results. However, this rating does not necessarily correspond to the best amplifier size or a “safe” amplifier size.
The power test signal, usually pink noise, is not representative of real audio signals or how a loudspeaker will react to those signals. Generally, power test signals are more stressful for a loudspeaker to reproduce than real music or speech signals. However, some real signals, such as rock, disco, and even some classical music, can have content that is more stressful than the test signal.
Commonly stated power ratings include: average, rms, continuous, AES, and EIA. These are the most important ratings and will be collectively referred to herein as the “average” power rating. Other commonly stated power ratings include: music, program, and peak. These ratings are typically two or more times the former. At best, they are an indication that a loudspeaker can handle peak inputs that are higher than the average rating.
Even though power tests and rating criteria differ among loudspeaker manufacturers, the results are usually surprisingly similar. More often than not, small differences in power ratings between two similar loudspeakers are a result of differences in testing methods, test equipment, or rating criteria, rather than actual differences in loudspeaker capabilities. As a rule-of-thumb, power ratings that are within a factor of about 1.5 of each other can be considered of similar capability. This factor equates to a barely noticeable difference in output level of 1.7 dB. For example, a loudspeaker rated for 600 W and a similar one rated for 900 W (1.5 times 600 W) are likely to provide similar power handling capabilities on real audio signals.
The second issue concerns selecting an appropriately sized amplifier for a loudspeaker. The usual assumption is that an amplifier equal to or less than the loudspeaker's rating is “safe” to use. The number of blown loudspeaker components that litter repair shops attest to the validity of this assumption. The appropriate amplifier size should be determined from the type of audio signals that will be reproduced and the sound levels required. However, few pay attention to such criteria and fewer still understand how to determine and correctly use them.
It is also assumed that a loudspeaker's power rating indicates how loud it will get. It does not. The maximum output is actually a function of amplifier power and the sensitivity of the loudspeaker. For example, 117 dB SPL is the calculated maximum output for a 100 W loudspeaker with a sensitivity of 97 dB (1 watt @ 1 meter) and for a 400 W loudspeaker with a sensitivity of 91 dB (1 watt @ 1 meter). However, with real audio signals, the maximum output one can achieve will invariably be significantly lower than this calculated result.
Loudspeaker components, particularly compression drivers, can usually handle momentary power peaks well in excess of those they are subjected to in power rating tests. Unlike test signals, music and speech signals typically have short-term peaks that may range from 5 to 10 times higher than their average level. This means they require 5 to 10 times more power than the average. This in turn means that, with real audio signals, the average power output from an amplifier may be no higher than about 1/10 of the peak power it is producing. And this means a 100 W amplifier, which can reproduce 200 W peaks, will likely be delivering less than 20 W average power to a loudspeaker. To fully exploit both the peak and average power capabilities of a loudspeaker and to avoid amplifier clipping on the peaks, an amplifier larger than the loudspeaker's power rating may be needed.
Contrarily, for audio signals with low dynamics, such as heavy metal or highly compressed music, an amplifier with a rating at or below the loudspeaker's power rating might be needed to avoid exceeding the loudspeaker's average power rating. On the other hand, a loudspeaker rated at 200 W might be selected to reproduce background music at low levels. Perhaps only a 25 W amplifier would be needed to reach the desired acoustic levels.
A hidden factor in all of this is that amplifiers are invariably tested with a sine wave signal while loudspeakers are usually tested with a pink noise signal. The above graph shows the difference. For the sine wave signal, an amplifier must reproduce peaks that are only 3 dB above the rms voltage level. For the noise test signal, an amplifier must reproduce peaks that are 6 dB above the rms voltage level. This means that, for a loudspeaker rated at 100 W, the peaks during its power test were at least 400 W. For an amplifier rated at 100 W, the peaks during its power test were only 200 W. In other words, the peak power capability that can be assumed from amplifier and loudspeaker power ratings differs by a factor of two! A 100 W amplifier on a 100 W loudspeaker could not reproduce the loudspeaker's power test without clipping the amplifier. At least a 200 W amplifier would be needed to reproduce the 400 W peaks for a 100 W average noise signal.
Given the foregoing difference in amplifier and loudspeaker power ratings, a rule of thumb can be applied for selecting an amplifier. When the full capability of the loudspeaker is needed to achieve the desired acoustic output levels, use an amplifier that is twice the loudspeaker's average power rating. This allows the amplifier to reproduce peaks 6 dB above the specified power rating. This is consistent with the amplifier size used for the loudspeaker power test. However, this does not guarantee immunity from damage or failure.
The third issue concerns preventing damage to or failure of a loudspeaker. This is rarely related to either the loudspeaker's power rating or the size amplifier actually used. Preventing damage or failure means operating an audio system so that a loudspeaker is not stressed beyond its limits. Operated improperly, damage or failure can easily occur with an amplifier that is sized well below a loudspeaker's power rating. Contrarily, when operated properly, damage or failure can be avoided even with an amplifier sized well in excess of a loudspeaker's power rating.
Proper audio system operation includes being aware of the types of audio signals being reproduced, controlling output levels accordingly, and operating all electronic equipment so that its maximum outputs are not exceeded. Examples of improper operation include:
Equalization boosts at frequencies beyond the operating frequency range of the loudspeaker.
Large equalization boosts within the operating frequency range of the loudspeaker.
Allowing electronic clipping anywhere in the system, including the mixing console, signal processing equipment, or the power amplifiers.
Applying heavy compression to audio signals.
“Pushing” loudspeakers to levels just below obvious distortion. Inexplicably, many operators use this point as their operating level, usually resulting in eventual loudspeaker failures.
Reproducing sustained tones, like synthesizer notes, near full amplifier output.
Each of the foregoing can easily result in damage to or failure of a loudspeaker, regardless of the loudspeaker's power rating or the size amplifier used. It is the responsibility of the audio system operator to ensure that all equipment in the system is operated within its capabilities. That is the only way to ensure that loudspeakers are not stressed to the point of damage or failure.
Most power tests done by professional manufacturers are really voltage tests. The quantity invariably measured for such tests is the rms or average voltage of the signal. Power is calculated then calculated with the formula:
Power = voltage squared/nominal impedance
This works fine for power amplifiers, where the load impedance is a resistor. However, for loudspeakers the result of this calculation is not power for the following reasons:
The nominal impedance is rarely equal to the real impedance, which actually varies considerably with frequency.
The loudspeaker is usually a reactive load, meaning it behaves electrically, depending on the frequency, as either an inductor or capacitor. Voltage and current are not in “sync” with reactive loads, so the actual power cannot be calculated without knowing how far they are out of sync. This is measured as the phase angle between the voltage and current. Thus, for reactive loads, the power formula is:
power = (voltage squared × cosine phase angle)/impedance
The truth is, no one measures the voltage to current phase angle for loudspeakers. As such, one cannot solve this equation.
Amplifier and loudspeaker power ratings are really surrogate numbers for what is actually measured: voltage. However, power ratings are useful as a matter of accepted convention and convenience when comparing either amplifiers or loudspeakers. Just keep in mind several things.
Loudspeaker power ratings do not represent real watts.
Loudspeaker power and amplifier power do not mean the same thing.
Differences in power ratings by a factor of up to about 1.5 are usually insignificant. It takes about 10 times the power to make sound twice as loud. This is one reason why concert systems have huge amounts of amplifier power. For example, 50,000 W is only four times louder than 50 W.
No foolproof criterion exists for selecting an amplifier that can prevent loudspeaker damage or failure.
Operator competence is the primary means to prevent loudspeaker damage or failure.
Technical Services Manager
Eastern Acoustic Works
All Designers, Technicians, Manufacturers, Distributors, Groupies, Hangers-On, & Entertainment Technology Geeks:
Got an idea you want to share with your peers? An important industry issue you want to address? Or something you just want to get off your chest? Entertainment Design is always looking for more contributors to its monthly On Lighting, On Audio, and On Projection columns. If you can write and want to share your views with ED readers, please send your ideas to David Johnson at firstname.lastname@example.org.