Amplificación Clase D: ¿válida para música en Alta Fidelidad?

The problem was speed: Class-D amps sample the input audio waveform hundreds of thousands of times a second before amplifying it, and it wasn't until the 1990s that cheap and reliable MOSFETs became available that were fast enough to sample a waveform at such high frequencies, says Bruno Putzeys, the chief engineer at Hypex Electronics. Hypex, in Groningen, Netherlands, is a leading maker of audiophile class-D amplifier modules, which are incorporated into products sold by other firms...

Tripath, in San Jose, Calif., was the first to introduce a class-D amplifier chip, the TA1101, in 1996. It was used in Apple's celebrated Power Mac G4 Cube. Tripath's first big hit, the TA2020, in 1998, could funnel 20 watts per channel into 4-ohm speakers (see "25 Microchips That Shook the World"). It was used in ministereos and early flat-screen TVs. "We were shipping millions of chips every quarter to companies like Samsung, Panasonic, Toshiba, Sanyo, NEC, Onkyo, and Sony," says Adya Tripathi, who was the company's president and CEO.

Large stocks of Tripath chips are still available. They feed an e-Bay market for tiny, ultracheap (as low as $15) amps from China and also a thriving business among hobbyists and DIYers...

And for the technically inclined, amplifier modules and kits allow anyone with modest soldering skills to get class-D sound, highly precise and detailed, for a few hundred dollars. One of our favorite kits is the SDS-224 kit from Class D Audio, with its robust power supply...

To the Ear

So how about sound? The output filter's highish HF output impedance, when uncorrected (in amplifiers without post-filter-feedback, ie. all digital ones and many analog ones too) is responsible for the oft- quoted tube-like warmth and air. At low frequencies the filter impedance is low, resulting in a commanding, dynamic bass. Because switching amplifiers previously had a reputation for sounding harsh (due to people who hadn't heard them but presumed that switching couldn't mean anything else), about every modern entry in the field was heralded as the "first audiophile class D". Read a review about the Bel Canto Evo or the Sharp 1-bit (which is analog, btw) to see what I mean. Keep in mind that if a device sounds radically different from what you've held in high esteem previously, there's usually something fishy going on.

Unfortunately, while warm and airy is nice, it isn't all you need for real audiophile sound. I like to think audio components should sound neutral and transparent too. A frequency response that wanders 10dB off the line at 20kHz isn't conducive to neutrality. High THD isn't good for transparency, especially when it goes up with frequency. Ergo the Tact Millennium (which has flat and low THD but a nonflat frequency response) sounds transparent but not neutral, and your average transistor amp (which has a low output impedance but sharply increasing THD) sounds neutral but not transparent...

Different Forms of Class D Amplification

Though we may think that “digital” means all the circuits in a digital amplifier work in on/off pulses, in fact there are a number of different types, including digital amplifiers that have analog elements.

A digital amplifier will have either analog or digital inputs. Good digital amplifiers with analog inputs can use analog feedback networks to lower the amplifier’s distortion, in much the same way that a Class A/B analog amplifier uses a negative feedback network to lessen the distortion. However, a digital amplifier that accepts only a digital input must rely on the incoming digital signal to lower distortion.

Feedback Networks

Why feedback networks? The reason they are used is that all parts in an amplifier have “tolerances”, which means that any particular part has a range or value in which it operates. Anyone who has examined such basic parts as resistors may have noticed they are specified as being “5%” or “10%” resistors, which means the specified resistor value is accurate within a range of 5% or 10%, respectively. Consequently, because of these variations in parts, a feedback network “looks at” the outgoing signal from the amplifiee -- the one that goes to your loudspeakers --and compares it to the incoming audio signal at the amplifier’s inputs. Any deviation in value away from the incoming signal is a distortion, so the negative feedback network applies inverse correction to compensate.

There are even differences in the operation of digital amplifiers.

For example, the “ICE” digital amplifiers developed by the Ice Power division of Denmark’s Bang & Olufsen use a very complex negative feedback system due to parts tolerances. B&O holds patents on its “ICE” amplifier, which is basically a Class D switching design (Pulse Width Modulator) with variants that B&O claims reduces distortion to levels associated with Class A amps, while retaining the high efficiency of Class D switching designs.



“IR” (International Rectifier) is the system used by Axiom Audio in its A1400 digital amplifier. Axiom worked with International Rectifier to keep parts tolerances held to the very minimum amount, so that very little negative feedback would be used to correct for anomalies in the output. This approach also made the amplifier more robust in its operation without being subject to oscillations or instability.

Axiom and IR developed new silicon output devices that drive the MOSFETs in the output stage in such a way as to produce a perfect Pulse Width Modulated square wave at the output before the reconstruction filter.

Pros and Cons of ICE and IR Digital Amplifiers

One of the downsides to using a complex negative feedback network in a digital amplifier of the type used in ICE designs is a potential loss of efficiency (around 83%). Performance may also suffer because of a slower clock rate.

In an IR type of digital design, which uses very little negative feedback or none at all, the clock rate is higher and efficiency increases. Moreover the high efficiency is combined with high power delivery and higher overall resolution. At full output, Axiom’s A1400 digital amp runs at about 95% efficiency (by comparison, class A/B analog amplifiers run between 50% and 60% efficiency; the remainder is wasted in heat).

the t.amp TA2400 MK-X power amplifier - 2x 1200W/4 Ohms, 2x 650W/8 Ohms thomann RMS, 19", input sensitivity 0,775V, slew rate 60V/µs, damping factor: 500:1/1kHz/8Ohm, short circuit protection, limiter, THD <0.05%/1kHz, 2U. In: jack+XLR, out: Speakon & binding post. Depth: 44cm. Weight: 19,7kg. Very good quality/value ratio!

maty escribió:
¿Qué amplificadores clase T escuchas? Apenas hay info sobre ellos, lo que me intriga. El siguiente me ha sonado siempre muy bien las veces que lo he escuchado en la sala de Andreu, claro que a volúmenes muy altos y en campo lejano. Cuando escuché un par de Behringer B2031P en campo cercano fue con el AV ¿Denon/Yamaha? A ver si lo escucho con ellas y con mis cajas Denon modificadas.

slippery escribió:Madre mía que bacalá tienes encima Maty.... T-Amp es la marca de Thomann, son productos chinos OEM marcados con ese nombre. Y el 2.400 es de transistores de toda la vida.

En los the t.amp ¿(re)diseñados? por Thomann y fabricados en China, los TSA son clase D. Y los TA ¿clase H?

Los Tripath con una clase T especial basada en una modificación de la clase D, de diseño registrado, como sucede con la ICE de B&O y la clase D.

Otra fuente de lío es que la empresa Tripath registró el nombre "Class-T" para los amplificacores Tripath, los famosos T2020, T2021...

Class-G amplifiers (which use "rail switching" to decrease power consumption and increase efficiency) are more efficient than class-AB amplifiers. These amplifiers provide several power rails at different voltages and switch between them as the signal output approaches each level. Thus, the amplifier increases efficiency by reducing the wasted power at the output transistors. Class-G amplifiers are more efficient than class AB but less efficient when compared to class D, without the negative EMI effects of class D.

Class-H amplifiers take the idea of class G one step further creating an infinitely variable supply rail. This is done by modulating the supply rails so that the rails are only a few volts larger than the output signal at any given time. The output stage operates at its maximum efficiency all the time. Switched-mode power supplies can be used to create the tracking rails. Significant efficiency gains can be achieved but with the drawback of more complicated supply design and reduced THD performance. In common designs, a voltage drop of about 10V is maintained over the output transistors in Class H circuits. The picture above shows positive supply voltage of the output stage and the voltage at the speaker output. The boost of the supply voltage is shown for a real music signal.

In the class-D amplifier the input signal is converted to a sequence of higher voltage output pulses. The averaged-over-time power values of these pulses are directly proportional to the instantaneous amplitude of the input signal. The frequency of the output pulses is typically ten or more times the highest frequency in the input signal to be amplified. The output pulses contain inaccurate spectral components (that is, the pulse frequency and its harmonics) which must be removed by a low-pass passive filter. The resulting filtered signal is then an amplified replica of the input.

These amplifiers use pulse width modulation, pulse density modulation (sometimes referred to as pulse frequency modulation) or a more advanced form of modulation such as Delta-sigma modulation (for example, in the Analog Devices AD1990 class-D audio power amplifier). Output stages such as those used in pulse generators are examples of class-D amplifiers. The term class D is usually applied to devices intended to reproduce signals with a bandwidth well below the switching frequency.

Class-D amplifiers can be controlled by either analog or digital circuits. The digital control introduces additional distortion called quantization error caused by its conversion of the input signal to a digital value.

• clase G y H son alimentados conmutando en varios (G) o infinitos niveles (H) mientras que

• clase D y T la señal de entrada (información) es conmutada en muchos niveles.

maty escribió:El problema es que menté el the t.amp y después profundicé, con lo que algunos pueden liarse.

¿Es correcto lo siguiente?

En los the t.amp ¿(re)diseñados? por Thomann y fabricados en China, los TSA son clase D. Y los TA ¿clase H?

No, no es correcto. Como te he dicho son etapas OEM marcadas por Thomann con su marca.

Los Tripath con una clase T especial basada en una modificación de la clase D, de diseño registrado, como sucede con la ICE de B&O y la clase D.

Otra fuente de lío es que la empresa Tripath registró el nombre "Class-T" para los amplificacores Tripath, los famosos T2020, T2021...

En su día, los tripath fueron puestos de moda por Thomman/b] gracias al [b]The SI T-AMP (model 5066).

No, esos T-Amp no tienen nada que ver Thomann, nos dió a casi todos por comprarlo hace 6 ó 7 años, eran muy baratos.
-> http://www.tnt-audio.com/ampli/t-amp_e.html

-> http://www.tnt-audio.com/ampli/tripath_amps_e.html



Como para no liarse...

luisbustilloojeda escribió:



luis