Power and Watt
The amplifier takes electrical power from the socket, shapes it according to the music signal from the preamplifier and thus drives the loudspeaker. Thats amplification. This amount of power is measured in Watts [W].
Unfortunately, the marketing departments of the electronics groups have hyped the performance data far too much, especially in the lower consumer segment. Nearly all consumer devices somehow reach several hundred watts. Thanks to the technical short-term peak load "PMPO", several thousand watts of power are possible even with the cheapest loudspeakers. Without these loudspeakers are inevitably particularly loud and/or sound good.
Please: Ignore this number. A good stereospeaker will take less then 1 Watt power for good listening volume. Big hornspeakers will even be painfull loud with this amount of Power. Lets have a honstest look....
The aim of the power amplifier is to amplify the signal to a level that is sufficient to satisfy the loudspeakers. And here lies the main problem. Boxes are complex loads. Their resistance changes significantly depending on the frequency (pitch) applied. Thanks to the resonance frequency of the bass driver, an 8 Ohm speaker can sometimes reach a 50 Ohm value. But that depends on the speaker. Of course, the amplifier must not be distracted by this and must be able to maintain the same voltage in all situations. Loudspeaker manufacturers assume that the power amplifier can sufficiently replace an "ideal voltage source", i.e. it does this without grumbling.
When low complex impedances and high volume levels coincide, it is not uncommon for the amplifier to suddenly have to deliver extremely high short-term power in an attempt to maintain this voltage. And this brings me to the most frequently asked question; How many watts does the amplifier need?
Answer: That depends on the speaker and the volume. Loudspeakers have a different efficiency (or sound pressure level). That means: with the same power applied, different loudspeakers are differently loud. A high efficiency is no guarantee for good or bad sound.
The usual compact loudspeaker has an efficiency of 80 - 88 dB at one Watt. "dB" is a logarithmic pseudo measure. In acoustics, it is usually equated with an aurally correct indication of volume - this explanation is not quite correct, but should suffice at this point. A volume of 88 dB means that you have to speak quite loudly to be able to converse in this background noise. This value is usually measured at a distance of one meter from the loudspeaker. Large floorstanding loudspeakers usually have 88 to 93 dB at one watt of efficiency, large horn loudspeakers sometimes manage the 110 dB at one watt. At this level you can miss a percussion drill at close range. We are still talking about one watt here.
If you supply a standard compact speaker with an efficiency of 88 dB with one watt and listen to music in a room that is not too large, this will easily suffice for normal levels. The problem is with dynamic peaks, which demand very high power from the amplifier even for short periods of time. If the amplifier can deliver about 10 watts of clean sine wave, many commercially available compact speakers can be driven to volumes even at musical impulse peaks where the rattling of doors is a cause for concern.
Those who feel attracted by the astonishingly high watt specifications of some products should consider the following: The ear works non-linearly, for a volume doubling (i.e. 10 dB more), 10 times the wattage is needed. The actual difference between an amplifier with 50 and one with 100 watts will therefore not be very big in practice.
And: watts are not equal to watts. It is especially popular to make a PMPO specification for cheap products. This is a "peak pulse power". I have written that especially with dynamic music, peak pulse powers of several hundred watts can occur, so it seems reasonable to specify the PMPO power. On closer examination, however, "reasonable" is not necessarily the word of choice - the technical term "nameplate cheese" seems more appropriate.
The PMPO power refers to very short peaks, here values of over 1000 watts can be reached even with very cheap power amplifiers. If the somewhat longer impulse peaks in classical music actually suddenly need 100 watts from such power amplifiers for a somewhat longer period of time, many cheap power amplifiers will, however, quit under heavy smoke formation. Some amplifier manufacturers in the higher price segments also advertise high peak power, but also specify the period of time over which such peaks can be maintained. This makes a real difference for very loud and dynamic music, at least for low efficiency loudspeakers.
The PMPO specifications of most devices are completely useless for evaluation. Even the cheapest amplifiers, which get down on their knees at moderate volumes, achieve PMPO power levels of well over 500 watts.
However, one finds the so-called "sine wave power". Or also the "continuous power" or "RMS power". These figures indicate which power levels can be achieved constantly. And here, conjuring up utopian wattages is a little more difficult - but unfortunately still possible. If an amplifier has 10 watts sine, it probably really has 10 watts. And with most commercially available loudspeakers it should be sufficient for generous interpretations of room volume.
The situation becomes more difficult with critical speakers in large rooms. Some electrostatics
The relationship between loudspeaker and power amplifier can be clearly described by the analogy with water in Figure 2:
Figure 2: Electrc and Watercurrents
Shown are two water glasses at different heights. These are connected by a thin compensating pipe through which water flows by gravity.
The analogy to the terms used in electrical engineering is as follows:
- The electrical potentials φ1 and φ2 are analogously the absolute height of the respective water glass.
- The electrical voltage U is the difference φ1 - φ2 of the two potentials. The greater the difference between the two potentials, the greater the voltage - and the higher the current flow through the compensation tube.
- The electrical resistance corresponds to the flow resistance of the pipe.
- The electric current corresponds to the amount of water that flows
- The converted power P is the product of voltage and current. If a water turbine were installed in the pipe, the electrical power could be used technically.
- The German electricity grid constantly refills the water in the upper glass with the water in the upper glass. If the electricity network is integrated into the view, the flow of electricity is inevitably closed!
Particularly critical for the power amplifier is a large compensating tube or a loudspeaker with low resistance, here especially much power is converted.
The damping factor is another variable that makes a good amplifier, the value is dimensionless (has no unit). The value can only be given in connection with a load, e.g. the usual 8 Ohm. Furthermore it is frequency dependent.
The damping factor of an amplifier is calculated by the ratio of the loudspeaker impedance to the output impedance of the amplifier.
The amplifier is an ideal voltage source and supplies different water levels or a voltage at the loudspeaker terminals. A real voltage source can be imagined as an ideal voltage source with a resistor connected in series. This is called internal resistance.
Since there are also crossovers and loudspeaker cables before the loudspeaker, their impedances must be added to the internal resistance of the amplifier. The somewhat odd statement that cables should have an inherent sound is often justified with this connection. This can really make a difference when bell wire is laid over long distances, but you are on the safe side with normal DIY products. The blind tests I know of have not been able to prove cable sound to this day.
Instead of the damping factor, the internal resistance of the amplifier can be indicated. Since a high damping factor looks better on the data sheet than a low internal resistance, this is usually omitted.
The effects of the damping factor on the mid/high frequency are tiny, because the responsible drivers are operated far beyond their resonance frequency and have a high internal damping.
The damping factor makes audible differences in the bass range, where this "damping" forces the sluggish, wildly deflecting diaphragm of the bass driver (which plays near its resonance frequency) to follow the signal path. In the event of overshoot, the voice coil located in the magnetic field induces a voltage - but the power amplifier also wants to control the voltage level on the speaker lines.
If the voice coil makes movements in the magnetic field of the driver magnet (see speaker sketch) that do not correspond to the signal, a current is induced in the voice coil. The power amplifier must "stall" this process, i.e. create a short circuit for this return current. The decay behaviour of the diaphragm can be improved enormously at this short circuit, i.e. a low internal resistance of the amplifier. But the transient response becomes rather slow, so it would be better to let the amplifier idle. The internal resistance should therefore be chosen sensibly.
This means: theoretically, a certain internal resistance of the amplifier would be ideal for each loudspeaker chassis. This is the one where the impulse behaviour of the speaker is best. With good active speakers, where each driver has its own output stage, this can theoretically be set up.
Since a hifi amplifier should be compatible with all passive speakers available on the market, this becomes more difficult. The common loudspeaker manufacturers usually assume that the internal resistance of the amplifier is very low and take this into account in the design. There's no need to be afraid of very low internal resistances, after all the resistances of the crossovers and cables are added to this. So when buying an amplifier a high damping factor is usually good.
With a very low damping factor (high internal resistance) the bass becomes rather sluggish and spongy. Written this sounds worse than it is. Even with good tube amplifiers the damping factor is usually a power of ten below that of cheaper transistor units. Hifi freaks affectionately call the result "unobtrusive warm sound".
With active loudspeakers the power amplifier works directly on the voice coil, with passive concepts it also has to deal with the impedances of the crossovers and strings. More details under "Speakers".
If someone tries to apply his school knowledge to the damping factor here: With a low damping factor (less than 1) the power amp is practically an "ideal current source" . So it must not be operated in idle mode (important for a few tube amplifiers). Short-circuit is possible here.
An output stage with a high damping factor (greater than 1) is an "ideal voltage source" (this includes most tubes and actually all transistor amplifiers) . Ideal voltage sources must not be short-circuited under any circumstances. Idling is unproblematic.
The frequency range or the slew rate
Hearing takes frequencies from about 20 to 20000 Hz. Most amplifiers easily amplify sounds from about 10 to 50,000 Hz. Sometimes even much more. You do not need to worry about this value.
Occasionally the call for "fast" amplifiers appears. What is meant is that the amplifier amplifies the signal from the CD player cleanly and does not lag behind. In layman's terms: a sluggish amplifier would amplify signals in the same way that a digger arm would try to mimic the wing beat of a hummingbird. This comparison may be figurative, but it is as limp as a three-legged dog with crutches.
Already at about 12000 Hz the human ear is hardly able to distinguish between a square wave and a sinusoidal signal. Amplifiers that work up to over 100,000 Hz to reproduce the signal as accurately as possible and to follow the signal as "fast" as possible are of no use to human ears. The ability "to wave the excavator arm around" is described by the term "slew rate".
The damage caused to amplifiers by "slowness" is described by Transient InterModulation (TIM) distortion. With current circuit technology, this value loses interest.
The signal coming out of the amplifier should only be the signal from the CD player. Nothing should be added. It is now desirable to have a unit of measurement for these unwanted three inputs.
A sinusoidal signal of frequency f0 receives new frequencies, 2f0, 3f0, 4f0 ..., also known as K2, K3, K4 .... The fundamental oscillation is called the first harmonic. The new obwer waves as second harmonic, third harmonic, etc. These frequencies are a multiple of the actual fundamental wave.
Why take the trouble to filter out the multiples of the fundamental wave from the "three inputs"? In practice these values appear much more than the other "three inputs". Therefore they are examined separately.
The ratio of the RMS values of these "harmonics" to the total RMS value is called THD (Total Harmonic Distortion). Up to a certain point, distortion with even-numbered ratios is perceived as pleasant, i.e. the "K "s with even numbers.
For electric guitars, for example, tube-based guitar amplifiers deliberately focus on harmonic distortion.
But there is also a "malicious distortion". These are the distortions that have no relation to the input signal. So newly added frequencies that lie between the harmonics. These are perceived by the ear as very unpleasant.
Unfortunately, a normal transistor amplifier tends to produce distortion with odd Ks, (i.e. 3.5.7...), the distortion factor should remain below 0.1 percent (if the amplifier is not too heavily loaded). Since the hearing threshold is above this value even under favourable frame conditions, it can be overlooked with disinterest when comparing devices. Only tube amplifiers deliver a higher, but much more harmonic distortion. Mainly K2. Some music lovers swear by these "dirty effects", although they can also be produced with digital filters. But the latter is of course frowned upon by real hifi fans.
Since transistor amplifiers only produce an audible amount of distortion when hopelessly overloaded, you can rely on transistor devices without fear. Furthermore: not everyone prefers the "warm" clinking tube sound.