Explaining amplifier topologies is not a simple task, but I feel it is important, as hungry marketing departments continue to pray on unsuspecting consumers with word games.
All amplifiers do the same thing. They take a signal, and increase the voltage swing. Let's say we have a good quality deck, with a good strong preamp, say 4 Volt. At absolute full volume, we will get a musical waveform where the peaks are about plus and minus 2 Volts (although some are much more). I have attached a picture of a sine wave being fed into my oscilloscope.
Now here is were things start to get complicated. An amplifier takes this signal, and increases the voltage swing. So, if we had a 100 watt amplifier, playing into a 4 ohm load, the voltage swing would be about 57 volts peak to peak. I know you want to know how the heck I get 100 watts from that voltage. Ok, follow along..
Take your measured peak to peak voltage, let's say in this case it is 57 volts. First divide it in half, so we get 28.5 volts. Now take the RMS value of that. (RMS means root mean square, and is a factor (Square root of 2) that finds the equivalent DC level of a sine wave. So we take 28.5 and divide by the root of 2 or 1.414, we get just over 20 volts (20.155586987270155586987270155587 to be exact! (so there!)).
Now, we use the power formula which says the power = (voltage x voltage) / resistance. So we multiply 20 x 20 and divide the answer by 4 (our load). 20 x 20 is 400, and 400 divided by 4 is 100. AHA! 100 watts from 57 volts! Cool eh?
Ok, so you are bored to tears, stick with me as we move along.. It will all make sense.
The first amplifier topology we will investigate is called Class A. This is one of the simplest systems, as it really only requires one output device per channel. Oh yeah, output devices. That's what topology is all about. Every amplifier has a set of voltage rails. These are high voltages (say 30 or so for a 100 watt amp) in the amp which feed the output devices. The output devices are essentially valves or switches that will allow the rail voltage to be passed to the load (speaker) by varying amounts. The amount is controlled by the input signal. If you put a small signal to the output device (Bipolar Transistor or Mosfet), it puts out a larger signal. I know it's weird, but that's how it works. Anyways, in a Class A amplifier, we have one voltage rail, and in our 100 watt example, this would have to be at about 60 volts. We set up a single output device with the one leg at ground, and the other connected to the positive (60 volt) rail. The input signal is connected to the last terminal of the device. The load is connected across the switching device and referenced to ground.
The above diagram is taken from one of my college textbooks. If you replace the 10V voltage with 60, and the resistor on the right that's labelled as 4.7kohm with our speaker, you have the essential schematic of an amplifier. Don't try building this though, the resistor values are designed for the 4.7k load, not 4 ohms.
Now, here is the slightly hard to understand part. There is a capacitor between the output of the circuit, and the load. This is called a DC blocking cap. The name is self explanatory. When no signal is being passed to the load (speaker), the voltage between the output and ground is 30 volts (approx). The cap blocks this voltage from getting to the load, and the load see's zero volts. Now, as we provide an input signal to the circuit, the output voltage moves up and down in sync with the input, and this alternating voltage is passed through the cap (with the 30v DC component removed) and the speaker moves back and forth. As you can see from the design, we can swing almost 30 volts in either direction from the resting voltage.
The advantage of this type of circuit is that the voltage transition up and down is very smooth compared to other topologies. The disadvantage is that when no signal is being passed, the circuit is turned half way on and draws a lot of current, which is dissipated as heat through the transistor.
Class B circuits
In a class B circuit, we have two rail voltages. For our 100 watt amp, lets say they are around +30 Volts, and -30 volts. As you can see in the following image, there are two output devices, with the load connected between them.
In this circuit, we only turn one transistor on at a time. One handles the positive half of the wave, and the other the negative half. If you look back at the picture of the oscilloscope, when the voltage is above the center line (0 volts, or Ground), the top transistor is turned on, and the output voltage goes upwards towards the +30 Volt rail voltage. As the waveform comes down to the zero level, the top transistor shuts off, and as the signal goes negative, the bottom transistor turns on. This cycle repeats as the signal moves up and down.
The advantage to this system is that when there is no input signal, the circuit is mostly turned off. There are some small voltages needed to keep the devices ready to go when the signal arrives, and so on. So this is quite efficient compared to Class A.
The disadvantage is that when you switch from one device to the other, you have the potential for noise, an over or under lap. Ie, it's not quite as smooth as the Class A. Most designers have made the adjustment of the Biasing voltages (the voltage required to turn the transistor on, and get it ready to track the input signal) very accurate, so very little crossover distortion occurs. Over 75% of the amps on the market use Class B.