Errrr.... it's a bit more complicated than that..... and the Mesa article is sort of geared towards tube amps, where the terms are often used a little differently.
The "real" amplifier classes are Class A, Class A/B, Class B, and Class C.... and they refer to which output devices are conducting how much of the time.
"Class D" refers to designs where the output devices are either all the way on or all the way off.... but there are a lot of different types of circuits that fit that description.
"Class G" and "Class H" refer to how the power supply is connected to the amplifier.....
So you could have a "Class A/B Class H" amplifier, like our XPA Gen3 amplifier modules.....
You could also have a "Class B Class H" amplifier - although I've never seen one..... and a "Class D Class H" circuit wouldn't make much sense (although it's theoretically possible).
"SE" is
NOT an amplifier class; SE means "single ended" - which refers to a type of circuit.... but the term is usually used differently for solid state and tube circuitry.
It solid state circuitry it almost always refers to a circuit that runs off of a single-ended power supply.
In tube circuitry, most of which runs off a single-ended power supply, the term SE usually refers to the configuration of the output stage.
"Complementary" is something else altogether.
It refers to using a "matched but opposite" pair of transistors at similar places in the circuit.
For example, with output transistors, it might mean using an NPN transistor between the output and the + rail, and a PNP transistor between the output and the - rail.
In the beginning of solid state, the number and type of transistors available was limited, so the same transistor was often used for both.
And, when using the same transistor in different circuit positions, the circuitry around each must be different, so they were
NOT "complementary".
Certain circuits were designed to be as symmetrical as possible given this limitation; they were referred to as "quasi-complementary".
(You can generally achieve lower distortion will a fully-complementary circuit..)
None of our Class A/B amplifiers is single ended (they run off + and - supply rails, which makes them "split-rail", which is the "opposite" of single ended).
They are also "complementary symmetry" - which refers to their using matching but opposite NPN and PNP transistors at various points.
Our XPA-1 Gen2 is technically a "high-bias Class A/B design".
This means that,
WHEN YOU CONSIDER OPERATION INSIDE ITS CLASS-A POWER RANGE, it is Class A.
But,
WHEN YOU CONSIDER OPERATION ABOVE ITS CLASS-A POWER RANGE, it is Class A/B.
(Note that
ALL Class A/B amps have a Class A operating range; on the XPA-1 Gen2, when you flip the Class A switch, it extends the range over which the amp is operating in Class A to cover most listening conditions.)
Incidentally....
A fully differential amplifier consists of two identical amplifier channels operated as a bridge.
In this configuration, it
IS possible to use two amplifiers that operate single-ended rather than from a split rail.
(The DC portion of the output signals cancels out - so you don't have to use an output capacitor or transformer to block it).
HOWEVER, all of our differential amps are bridged, and all of them operate from a split-rail supply (so they aren't single-ended).
In regards to
mgbpuff's comment: "Most people regard single ended amps to be the epitome of Class A with its high 2nd order harmonics and all other harmonics due to its non symmetrical nature".
He has a sort of point.... but it's not that simple.
Most people who like
SINGLE ENDED TRIODE amplifiers like them because of the high levels of second harmonic distortion caused by their single-ended design.
In this (tube) context, the "single ended" refers to the configuration of the output devices and not of the power supply.
A push-pull tube device can be designed to operate in almost any class..... however a single-ended tube device can
ONLY be operated in Class A.
(So pretty much all SET amplifiers, and all tube preamps, which operate single-ended, are Class A.)
However, in push-pull tube designs, and in solid state designs, both of which can operate in any class, Class A is preferred because it produces lower distortion (and less unpleasant types of distortion).
Class A implies continuous conduction of the power device due to high bias whether it is a single ended design or a push-pull (Class A/B) two power device (complementary) design. In the case of Class A/B (push-pull), both complementary devices are biased high so that conduction never turns off in either. The XPA-1 series is a bridged type amp design, essentially there is a whole amplifier dedicated to the positive differential path and another for the negative differential path (it's essentially a hugh push-pull design). It can be biased for class A (full bias), B (no bias with accompanying cross over distortion), or A/B (only enough bias to avoid crossover distortion).
Basically there's Class A which means power devices are in continuous conduction, and then there is circuit architecture, which can be single ended, push-pull, or two deferentially bridged amps. The symmetry of the last two architectures causes a different harmonic distortion residue (no or little even even harmonics).
Most people regard single ended amps to be the epitome of Class A with its high 2nd order harmonics and all other harmonics due to its non symmetrical nature.
I'm shooting from the hip here, as I am not and have never been an electronics designer. So, Keith, as he usually does can set me straight if I am off base.
So what's the difference between push pull (complementary/two?) vs single ended design. Does that mean there is only one device?
