The advantages of closed-loop audio architectures in the high-definition television (HDTV) space have been well proven, with the vast majority of analog-input Class-D amplifiers being closed-loop. As the market now transitions to digital-input amplifiers (I2S/PCM Serial I/Fs) and the ever-present pressures of cost, time-to-market and performance intensify, closed-loop architectures are becoming even more compelling. This article provides a high-level overview of closed-loop architectures and addresses three of their primary advantages in the HDTV space: higher damping factor; increased power supply noise immunity; and better electromagnetic compatibility (EMC) performance.
Closed-loop architectures overview
In the audio world, the debate over closed-loop versus open-loop architectures has raged on for many years. Depending on the end-application or customer preference, valid arguments for both architectures can be made. In the HDTV space, closed-loop amplifiers have proven to be the clear winner. But in high-end audio, the war still rages. The main advantages of closed-loop architectures include improved linearity, gain stabilization, increased bandwidth, and reduced output impedance. Some of the disadvantages include the potential for decreased stability, decreased gain, and added complexity.
View full size
Figure 1: Closed-loop block diagram
Conceptually, it helps to think of a closed-loop amplifier in terms of "pre-distortion" (Figure 1). The feedback network samples the amplifier's output, which consists of both the amplified signal and any non-linear distortions introduced to the signal by the amplifier or power supply. The output sample is then attenuated and inverted before being recombined with the incoming source signal. The resultant signal out of the summing node (Point A) is an attenuated input signal with inverted "pre-distortion" in the areas where non-linearities from the amplifier and power supply were previously added. The amplifier then gains up the signal, adding in non-linear distortions. Since the source signal was pre-distorted via the feedback network, you get a canceling effect of the pre-distortion + distortion, yielding a very linear signal.
This is the fundamental advantage of negative feedback, whereby you have a mechanism that dynamically adjusts for non-linearities in the system. In an open-loop architecture, this mechanism does not exist. Thus, the performance requirements placed on the amplifier's linearity and supply regulation are much higher, which typically translates into increased cost and/or performance tradeoffs.
Next: Damping factor
|
