Class A is non-switching. It produces a full power envelope whether you are using it or not,that's why they get so hot.
Class A/B runs in Class A during low power operation and then switches to Class B for more demanding power needs. Class B amps adjust the power envelope to the power required for the music. Less heat. This is a very simplified explanation, but should give you a general idea.
I don't know what a "power envelope" is, but the difference between class "A" and "B" is how the signal is split when it is amplified.
In a class A amplifier, the entire signal, split into parts, + and -, is run through two identical amplifying circuits that do NOT split the signal at all. The whole signal is put through a transistor group, but it is never split into smaller components.
In a class B amplifier, the signal is split in two, which are in phase. There are now four amplifying circuits, two for both + and -. It is then recombined after amplification, and put across the binding posts.
Both of these have their advantages and disadvantages. Class A is the purists method. The signal is never tampered with, so it will remain cleaner. Class B splits the signal and then recombines it, which can create crossover distortion. But in a class A amp, the transitors are made to be driven much harder, which generates more heat. The transistors in a class B design do not have to be run at such high gain levels, which in turn allows them to run cooler.
A class "A/B" amp is simply one that runs class "A" up to a certain power level, where the amp switches from A to B operation, allowing for lower heat levels. The catch is that most A/B amps set the power level of this switching very low, making the amp, in essence, a class B amp (unless one is using high-sensitivity speakers). Hope this helps. Cheers!
Pass labs, Krell and Gryphoon are are some of the Class A crowd,FYI. Just becuase its class A doesn't mean its superior though. A well done class B can have lower distortion than a poorly done class A-although different kinds of distortion, or at least proportions of. There's about 8 key sources of distortion in a class B circuit (I'm using Douglas Self as my reference), whereas class A is freed of many of these. So in essence Class A does hold the potential for the lowest possible distortion. The price to pay for class A? Very inefficient: 50% is the maximum theoretical efficiency for a Class A ampilfier, and in real world operating conditions its more like 20%. So a 100wpc stereo amp is going to have to dissipate/waste an additional 800 watts of power in the form of heat. And heatsinking and the bigger power transformer are expensive. Hence the attraction to going with class B=more efficient=cheaper to build and if done well can still have very low distortion levels. As the others have pointed out, "class" a/b is just an amplifier that switches between the two classes, and in so doing usually has more distortion than if it was well optimized and run in B mode only. "class" a/b isn't really a good idea, but there's still alot of'em out there.
Preamps on the other hand are usually, or at least far more commonly done in class A. Its financially practical and easier to do. It just with those big powerful solid-state amps does the parts cost bill start racking up real quick. I bet some of those big Krell's with all the heatsinking--there's probably $1,000 to $2,000 for the casing and heatsinking alone, before the guts/actual circuit are even put in it. I don't know what manufacturer's pay, especially w/ bulk purchasing, but Class A can get expensive, for what some would consider only marginal gains in performance. There are other classes, but most aren't used in home audio. Class D is about the only other one you'll see pop-up now and then. The rest are for radio frequency use and what not, and a few oddballs that are used in home audio--I think there is/was a class T somewhere. Class C,G,H, and S are the others off hand.
No, if I were a betting man, I wouldn't put money on any of these answers. Why do I get the feeling class of operation is being confused with single-ended vs. push-pull? Where is any mention of bias level? What is this stuff about "adjusting power envelopes" and "gain levels" and "switching"? No, no, no. I'd bet none of this is really correct (if I were a betting man), just more confusing. I can feel it in my gut! Well intended, yes, and touching on some valid points, but nowhere a fundamentally sound, reasonably comprehensive answer to the technical question asked. Like where a sine wave ends up relative to a zero-crossing line on a graph depending on its amplitude, and what that implies. Or how output devices operate. And what is non-linearity. But I am not qualified to explain it all either, so I'll wait for some EE to come along and set things straight. Or surely there must exist some good links for this stuff?
I have a reciever with a class "G" rating. What does a class "G" do differently from "A"-"A/B"-ect? I would appreciate any info.
Solid state devices (ie: output transistor) work by using the base current to control the collector current. Vacuum tube devices (ie: power tube) work by converting a voltage input into an output current.
With respect to SOLID STATE devices:
CLASS A bias means the device base current is set to allow collector current to flow through the device. This basically means there is current flow at all times, regardless of signal input.
CLASS B bias means the device base current is set to zero and therefore NO collector current flows through the device and the device is cut-off. When an input signal is present the base current will lift and allow collector current to flow and the device is now conducting.
With respect to VACUUM TUBE devices:
CLASS A bias means the grid voltage is set to allow anode current to flow at all times.
CLASS B bias means the the bias point has been set to a point where the tube has almost stopped conducting. This results in a larger input signal being required to drive the tube and an output current for only half of the waveform. To amplify each half of the waveform it is necessary to use 2 tube devices in a "push pull" arrangement, with one device taking care of one half of the waveform and the other device taking care of the the other half.
CLASS AB bias is somewhere in between the above 2 classes with usually a small amount of bias applied
The subject needs more than a quick simple explanation to understand. I hope the above reads OK and I have not further confused anyone..??
Sorry you guys didn't like my "power envelope" description.
Amplifier Design Principles
This section will detail design philosophies of amplifiers and the various classes of amplifiers applied in audio. For example:
Class A - transistor amp conducts for the entire cycle of input signal, conduction angle 360 deg. Runs hot, as the transistors in the power amp are on all the time, but has high sound quality.
Class B - positive and negative halves of the signal dealt with by different parts of the circuit, the output devices switching continually. Runs cooler, but the sound is not as pure.
Class AB - biasing the transistor amp at a non-zero DC current much smaller than the peek current of the signal source. Second transistor conducts during negative half cycle of waveform and the currents from the 2 transistors are combined at the load. A compromise between sound quality of Class A and efficiency of Class B. Most amp designs employ this method.
This is a direct copy from the Audioholics web site.
CLASS G is when the power supply voltage is changed from a low level to a high level when large output level swings are required. So for most output levels the lower power supply value is used and will only switch to the higher power supply for large signal peaks. Hitachi originated this class of configuration.
Perhaps this will help resolve the explanations. In class A, the output devices conduct more or less as the input varies, but never cut off. In other classes, output stages can be cut off (not conducting) during art of the input signal variation while the other half are conducting. That is why one could say that "different parts" of the signal are being handled by "different parts" of the circuitry. This is not a precise technical explanation, but intended to resolve the verbage. The real deal performance wise is whether the output stage is operating on the most linear portion of its characteristic curve. In class A, they are. :)
It is indeed Hitachi I have. Is this a good thing? Apparently it didn't do so well, or was it not cost effective? Please forgive my ignorance, I've had this thing for twenty years + or - and I've never had any info about it.
I'll think about this over a USDA Class A sirloin steak.
Now that we all have a good idea (?) of the class A, B, A/B, was there something else I missed? Oh yeah, private first class :) The bottom line is, just go out and listen to the amp/preamp. If you like the way it sounds then buy it.
Robert Harley's book "The Complete Guide To High-End Audio", Second Edition, proves very useful on topics such as this. Amazon.com has it if you can't find it locally.
Class "A" is pretty accuratley described above. You only need 1 transistor in theory but would use more for more power. The bias level (operating voltage on the) transistors keeps them "on" all the time regardless of input signal. This is very innefficient and generates a lot of heat. It achieves a max. efficiency of 25%. This means for 100 watts output, you need at least 400 watts out of your wall. Gryphon amps and the Pass Aleph use this method. Many manufacturers, such as the newer Krell amps, say that they have "class A" but they are fibbing, what they do is have a "sliding bias". So at low input levels you have a low bias which improves efficiency, and then when the volume is turned up, you up the bias. This would be good if you had a buffer and could therefore predict the input. Your amp is not class A unless it is very, very hot with no input.
Class "B" is academic and not used for audio amps, this class will only turn "on" a transistor if there is either a + or - input to it. However, near the zero point it will be "off" because of diode voltage drop charactertistic of the transistor. You need a + and a - transistor for class B (only 1 transistor needed for class A) One transistor handles the + phase of the signal, and if you need - amplification, which you do, the other handles that. But in the crossover region, (.7 v or less closer to zero for a BJT xistor) you will get nothing, called crossover distortion. Humans are very sensitive to crossover distortion and it sounds very bad, painfully bad, so class B amps are NEVER EVER used. As an aside, I know a study that was done where class B was used as a video amp to see if it would cause visual pain but none resulted.
Class AB- the most common amplifier out there. This is a class B amplifier with the transistors biased to at least the crossover point, say .7 volts. This means there are at least two transistors, 1 for + and 1 for - that are biased "on" for very low voltages. They are efficient and run in class "A" for less than .7 volt inputs, but class B for all others, which is ok, since the crossover distortion is not around anymore. The problem with AB is that you need to closely match your + and - transistors or they will work differently on the same signal causing mucho trouble.
Why bias it at .7 when you can go higher? A class AB amp might be "highly class A" which means that instead of biasing it for the crossover distortion (.7 volts say) it is biased for like 4 volts, in both the + and - transistors. This means that its operating in class A for that part and "B" beyond, so its less efficient class A for low level inputs but more efficient when you crank it up. The newer Pass amps use this.
At least some of the Pass amps are classA/B. The X 150 is class A up to 15 watts/channel, then slides over to "B" on up.
About the Pass amps, I believe that's what I said. The amps that slide the bias up around the crossover are "class AB" but claim "high class A bias" within the class "AB" range. Its very confusing of couse. For example, Pass is introducing pure class A amps also in the X series that do not do any class AB.