I have often read that Class A amps capture the soul of music better than other designs and audio reviewers wax eloquently about the sonic attributes of Class A amps.
Yet speaker manufacturers at audio shows more often display their product with Class AB amps and audio reviewers certainly have sung the praises of Class AB amps.
So is there any truth in Class A amps capturing the soul of music better than Class AB amps or is this just personal preference?
Michael G.
I don't know about 'the soul of the music' but to my ears some Class A amps sound more (I'm struggling for the right word here) natural and realistic. You pay a price though for this, not just upon initial purchase but with high electricity bills. For me the best compromise are amps running Class A up to a certain limit, say 30 to 50 watts such as Luxman's fabulous M-600A, and then switching to Class A/B if you feel like making loud noises above that limit. Speaker efficiency enters into this equation too but that's another story. One warning should be given. Some Class A amps, even low powered ones such as Musical Fidelity once assaulted the market with, run so hot and are so under equipped with heat sinking they self destruct before you've managed to fry a few eggs on top of them.
This topic is WAY over my head technically, but I recently had some interesting conversations with the folks at Mark Levinson about the new technology they are using in their new No. 53 mono power amplifier, which they claim has the benefits of "switching" amlifiers (e.g. class A/B) without the drawbacks. Here's what they say about the drawbacks of switching amplifiers on their website:
"Switching power amplifiers are so named because they switch the output devices on and off in very rapid succession, mimicking the input signal. One set of output devices drives the positive half of the waveform, and a separate set drives the negative half. The result is less power being wasted as heat, because the workload is essentially cut in half. Unfortunately, it also creates some significant design challenges in terms of how to manage the switching noise – the noise created from the output devices constantly turning on and off – as well as a phenomenon called “dead bands.” Traditionally, these have been the two contributing factors to switching power amplifiers’ reputation for inferior sound quality. "
I've heard the Levinson No. 53s driving a pair of Revel Ultima Studio 2s at a recent dealer event and they did sound fantastic. However, I've never heard a Class A amp so I can't make a comparison.
You can read a little more on this on the Mark Levinson website if you're interested:
http://www.marklevinson.com/ProductDetails.aspx?prdid=1
Cheers,
Cam
Let's not confuse anyone; "switching/Class-D" amplifiers are NOT the same as traditional Class-AB amplifiers. Let's see if we can sort this out without getting too damned technical.
In a Class-A amplifier, all active signal-processing devices are in conduction at all times, thereby keeping the active devices operating in their linear region of operation, resulting in the least inherent distortion levels. Class-A amplifiers must operate with their output devices biased to constantly deliver the maximum output current at all times; if the current isn't delivered into the loudspeaker load, the power is simply dissipated within the output stage; that's why Class-A amplifiers run so damned hot!
In a Class-AB amplifier, the output stage splits the workload between two active devices (or banks of active devices); one active device handling the processing of the positive portion of the audio waveform while the other active device handles the processing of the negative portion of the audio waveform. The "tricky" part of a Class-AB output stage is during the hand-off between the positive active devices and the negative active devices; a lot of the "magic" in high-end Class-AB amplifiers is in how this sensitive region of operation is managed. Most high-end Class-AB amplifiers are configured to deliver the first few watts of output power in Class-A mode, thereby providing that Class-A signal-transfer characteristic for all but the highest signal levels without having to maintain the standing bias current of a Class-A output stage. In most Class-AB amplifiers, only the output stage operates in Class-AB; the input/gain stages and driver stages are operated in Class-A mode of operation (it's only a few watts of power-dissipation to keep all of the other stages in Class-A mode).
OK, now "switching/Class-D" amplifiers work in a manner that is very orthogonal to Class-A/Class-AB amplifiers. BTW, "Class-D" does not mean "digital"; "D" was simply next letter available (after "A", "B", "C") when classes of active device operation were being assigned. In a switching amplifier, the output devices are not operated as linear devices and only have two states; the output devices are either "on" or they're "off". To output a 0-volt DC signal, the switching amplifier's output devices must symmetrically switch between "on" (the full positive power-supply voltage) and "off" (the full negative power-supply voltage); the resulting output averages O-volts DC. The longer the output stage is "on" during the switching cycle, the more positive the output voltage at the amplifier output and the longer the output stage is "off" during the switching cycle, the more negative the output voltage at the amplifier output. Using the variations in how long the output devices are "on" to recreate the input signal is a modulation technique know as pulse-width-modulation (http://en.wikipedia.org/wiki/Pulse-width_modulation). The switching frequency of a "switching/Class-D" amplifier is an important operational parameter since the resulting switching-noise is superimposed on top of the desired audio output signal and the switching-noise has to be separated/filtered-out of the amplifier output signal. The higher the switching frequency, the easier it is to remove the switching-noise from the amplifier output signal since lower-order passive filters can be employed without wrecking the audio signal. The Mark Levinson "switching" amplifier uses a 1MHz switching frequency, which is about twice the conventional switching amplifier's switching speed and is almost certainly responsible for part of the ML's enhanced sonic characteristics.