Engineering terms used in the marketing of guitar amplifiers are a kind of jargon that is often confusing and misleading to guitarists so this article aims to help readers understand the technical principles of single ended and Class A. I also wanted to go further than just explaining the theory so I've included a section providing some practical insight into the way a single ended Class A valve circuit distorts and what this sounds like. This is a big subject so I'll probably end up re writing and adding to parts of it until I get the full story across.
Terms like “Class A” and Class “AB” are used in every guitar amp specification. Guitarists encounter these when considering and comparing amplifiers, although these are engineering descriptions used to differentiate particular configurations of power amplifier output and nothing to do with sound or music.
Put simply, the terms Class A and Class AB define the way the valves in the power output stage are biased, and no more than that. In a Class A circuit each valve is biased at a mid point. This means there is always current flowing through the anode even with no input signal, and when amplifying a signal, the average current flowing in the output load is more or less constant
Class AB amplifiers use at least one pair of valves each with a small amount of bias and a push-pull configuration. In a push pull configuration the signal is split in two, the positive going and negative going parts are amplified separately and then re-combined in the output load. There is a Class B (and a Class C and D) by the way, but in the guitar amp world these configurations are not used.
Although Class A tended to be typical in the early days of valve amplifiers it is now only used in a small proportion of valve amps. Class AB designs have been used in the majority of guitar amplifiers since the 1960's.
Historically power amps used in audio equipment were “single ended” Class A designs, when the overriding requirement for designers to meet was low cost, and simplicity in manufacturing or maintenance. In those days high levels of volume or high audio precision .were not typically demanded in the consumer market.
As the industry developed, the demand for louder amps became a main driver in design, and this could be achieved far more effectively with new class AB designs, So during the 1960's the class AB became the standard.
So, the choice of Class A or Class AB is fundamental when designing an amp and the decision to use one or the other is made for very good reasons and to meet a clear objective.
Class A fell out of favour because it has several down sides; it is inefficient because 50% of the power generated by the valves is dissipated as heat rather than used in reproducing the signal. This also means that to design a Class A amplifier with the equivalent output power to a Class AB requires more valves, larger power supplies and therefore makes the Class A amplifier inherently larger, heavier and more expensive than the Class AB.
So, why then does anybody still manufacture or use Class A amplifiers when on paper there appears to be no obvious benefit ?. Well perhaps the only reason would have to be that they sound different to Class AB amps and recognising that discrimination in sound and tone is a very subjective and personal thing it appears that some people prefer the sound of Class A amplifiers although many others have never had the opportunity to consider the difference and some don't really care.
Marketers put additional spin on this terminology, for example they may proudly refer to an amp as “Pure Class A”, or promote Class A as being the pinnacle of exclusivity and bestowed on it a mythical kudos which indicates the ultimate in tone or playability for the most discerning of guitarist.. They may also refer to the “Class A Sound”, and that description can also be seen applied to some patches available on multi amp digital modelling products which don't necessarily rely at all on Class A circuitry to reproduce sound.
Before attempting to explain why Class A amps might sound different, I will introduce another term often associated with Class A , and that is “Single Ended”.
What is “Single Ended” ?
The term “Single Ended”, refers to the topology of a particular output configuration of a power amp, whether it is a transistor or valve circuit. This term is seen infrequently in guitar amp specifications. The original Fender Champ is a well known example of a 1950's single ended class A design. These days there are very few guitar amp products using single ended designs, but amps such as THD's Univalve and Bivalve have a large following,
In a single ended valve circuit, the whole signal flows through a single circuit comprising of the anode(s) of the output valve(s) and the primary of the output transformer. See fig 3.
Figure 3 – simplified single ended power amplifier.
This is contrasted with a “push pull” output which splits a signal into two and drives the secondary of the output transformer differentially. The push pull output was developed as a way to minimise distortion in the output such that distortion components introduced in each half of the output stage , being roughly equal and opposite , would be cancelled out as the signal is recombined in the output load ( transformer primary in the case of a valve amp or loudspeaker in a transistor amp). The other push pull benefit is to significantly reduce the magnetic flux produced in the primary by the primary dc currents and enabling smaller transformer construction.
A typical single ended output has one output valve, but could have several valves wired in parallel. The valve will be cathode biased, meaning that a fixed value resistor in series with the cathode enables the valve to automatically settle at an appropriate bias voltage, - often referred to as self biasing. The biasing is Class A (it can't be anything else) which means that the bias point is set at a midway point allowing the signal voltage at the anode to swing anywhere between the lower cut off point and the higher saturation point.
“Why would single ended Class A sound different ?”
At lower signal levels the amplification will be linear which means no distortion to the signal. When the signal level at the anode is approaching the extremes of that range the characteristic of the valve becomes progressively non-linear and therefore signal distortion increases. The distortion is typically asymmetric and this results in by-products which are mainly even multiples of the fundamental frequencies in the original signal i.e . even harmonic distortion . The predominant harmonic is the second harmonic and this gives a pleasant and musical distortion for guitar because the distortion components are an octave above the original signal.
Compare this with a push pull circuit where even harmonic distortion products are cancelled out in the transformer primary and never reach the loudspeaker. When the amp is driven hard into non linearity, the distortion to the signal is more symmetrical and distortion products are mainly odd harmonics. The third harmonic is predominant and can give the distortion edge and bite but the upper order harmonics can sound positively nasty.
Another type of distortion present in the AB design and not the class A is “crossover distortion” which is caused each time one valve in the pair starts to conduct and take over from the other. The products are actually high order odd harmonics , this distortion sounds bad and is more obvious at low signal levels than high levels. The crossover distortion needs to be minimised by good bias adjustment in the output valves.
Generally speaking the distortions produced by valve power amps may be pleasant or un-pleasant and while the designer can adapt the circuit to balance this in the most musically pleasing way, designers still have to ensure that their circuits will be safe and that the stress on components will not reduce their life to an unacceptable level, so everything becomes a trade off and results in different models of amp with similar output configurations sounding and behaving differently when over driven..
Here are some links to recordings of the different distortions described above.
Class A single ended overdriven sinewaves, scope trace and frequency analysis
Class AB linear sinewave, scope trace and frequency analysis
Class AB overdriven sinewaves, scope trace and frequency analysis
