Ask a Question

Rajni

Note from My Teacher - Amplifier Efficiency I have sensed a fair amount of confusion (based on e-mail from readers) concerning amplifier efficiency. In general, efficiency of a device is defined as the amount of useful output power divided by the amount of required input power. The maximum efficiency any circuit can have is 100%, for amplifiers this value can never be met. In general, audio amplifiers are not particularly efficient. Class A amplifiers are the least efficient. Class AB amplifiers are better, with a maximum theoretical efficiency of around 78%. Well designed Class D amplifiers can approach 95%, a very good figure! High efficiency means less power is wasted in the form of heat. This means smaller heat sinks, less weight, and more output for a given input. However, there is more to this discussion than the simple numbers just given. Why are amplifiers not 100% efficient? For an amplifier to be 100% efficient, all of the AC power taken from the wall outlet would have to end up as useful output (power available to drive a speaker). This can never be realized because some power will always be required by the input stage (although this is fairly small). In most designs, the bulk of the power loss occurs in the output stage of the amplifier. This article can only go into a very high level explanation…. Basically, the output devices (transistors) in an output stage of most amplifiers act like variable resistors, kind of like a water faucet that varies flow rate in accordance with the music (class D is different as we have discussed previously). Remember we talked about rail voltages of a power supply? Basically, this voltage is dropped across the output device and the load (speaker) depending on the level of the output signal at any given moment. At the instant when the volume is high, there is relatively little voltage drop across the output device (but a lot of current flowing through it). When the output is small, there is a large drop across the output device (but very little current flow). Ohm’s law dictates how things work here: power = voltage times current = current squared times resistance. Basically, any time there is a significant current flowing and a significant voltage drop across a device, heat will result. So, to make an efficient amplifier, one needs to find a way to make the output device not dissipate much power. Class A amplifiers are bad because there is almost always a lot of voltage drop AND current flowing through the output device. Class AB is better but it still has the same basic problem. Class D amps are very efficient because the output device spends very little time in a state where there is both a significant voltage drop on the device and a large current simultaneously flowing through it. Basically, Class D amplifiers operate their output devices as switches, where the other amplifier classes operate the output devices in the linear mode, and this will inevitably lead to power loss in the device. There is a point I want to make about amplifier efficiency that many people fail to properly understand. I’ll use the common class AB amp for this discussion. Amplifiers of class AB are said to have an efficiency of around 75%. The key here is to understand that this is the maximum possible efficiency! This efficiency (for a well designed unit) can only be met under certain operating conditions! In practice, with real music signals, the actual efficiency will be significantly less than the maximum theoretical value. In actuality, the efficiency of an amplifier is different at every moment in time (for a time varying signal like music). The efficiency of a class AB amp is best when the amplifier is putting out a signal just at the threshold of clipping! This is because at that instant the current through the output device is high, but the voltage drop across it is low. The efficiency of the amplifier at lower operating levels is less than the maximum efficiency. So, in truth it is difficult to say that an amplifier has a specific efficiency as that value will depend on what the nature of the output signal is. However, for a given signal, amplifiers of different classes and the same output rating clearly have different efficiencies. The clear standout for efficiency in audio amplifiers is held by the Class D design (because these amplifiers operate the output devices as switches, meaning very little power is ever dissipated in the output devices). Amplifier efficiency is generally assumed to be stated for resistive loads. Speakers are reactive loads, meaning they are basically a resistance with the added characteristic of inductance or capacitance (one or the other at any given time). A speaker will never be totally reactive (meaning it will never be totally inductive or total capacitive, it will always have a resistive component). Ideal inductors and capacitors dissipate no power (these devices store electrical energy), however they DO allow current to flow. For illustrative purposes, let’s say that we have a very strange speaker that is 8 ohms inductive (100% inductive). If we hook this up to an amplifier will current flow (assuming the volume is up)? Yes! However, the inductor cannot dissipate any power (and hence this speaker makes no sound). Ohm’s law dictates how things work here. Basically, we will have the same current flowing through an 8 ohm inductor as would be in an 8 ohm speaker, HOWEVER, the current through the inductor has a different phase than it would through a resistive load. Ohms law states that the current around any circuit loop is the same in each component and the voltage drops in the loop must sum to zero. So, we have voltage drop across the inductor and current flowing through it, however because the current and voltage are 90 degrees out of phase there IS NO POWER dissipated by the inductor. HOWEVER, there IS power being dissipated in the output device of the amplifier EVEN THOUGH WE HAVE NO POWER BEING DISSPATED IN THE INDUCTOR (the strange speaker in this example)! What I am trying to explain in this long winded section is that inductive (and capacitive) loads can place added demands on the output devices of amplifiers! As I mentioned, no real speaker will be 100% inductive (or capacitive), but under certain conditions (conditions that are not uncommon in actual use) speakers can have an inductive (or capacitive) component that can definitely add to the stress on the output device of an amplifier. Reactive loads will basically increase the power dissipated in the output device in an amplifier. The result is that heat sinks must be larger, devices need to have higher ratings than what otherwise might at first seem adequate. If you take nothing else away from this section, realize that (a) reactive loads (speakers) can reduce the efficiency of amplifiers and they can cause added stress to the output stage of amplifiers (they can make your amp overheat prematurely). Now consider this: most speakers are only about 5% efficient! So we have an amp that might be 50% efficient and then that power is going to a speaker that is only 5% efficient! Suffice to say that the vast majority of AC power that amp takes from the wall socket never gets converted to useful sound (almost all of it ends up as heat)!