Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.


Share this Page URL
Help

CHAPTER 10: Class-G Power Amplifiers > ADDING TWO-POLE COMPENSATION - Pg. 324

PART IV Power Amplifiers DERIVING A NEW KIND OF AMPLIFIER: CLASS-A C A conventional Class-B power amplifier can be almost instantly converted to push­pull Class-A simply by increasing the bias voltage to make the required quiescent current flow. This is the only real circuit change, though naturally major increases in heat-sinking and power-supply capability are required for practical use. Exactly the same principle applies to the Class-G amplifier. In the book Self On Audio (2006, p. 293) I suggested a new and much more flexible system for classifying amplifier types and here it comes in very handy. Describing Class-G operation as Class-B C immediately indicates that only a bias increase is required to transform it into Class-A C, and a new type of amplifier is born. This amplifier configuration combines the superb linearity of classic Class-A up to the transition level, with only minor distortion artifacts occurring at higher levels, as demonstrated for Class-B C above. Using Class-A means that the simple V be -multiplier bias generator can be replaced with precise nega- tive-feedback control of the quiescent current. There is no reason why an amplifier could not be con- figured as a Class-G Trimodal, i.e. manually switchable between Classes A and B. That would indeed be an interesting machine. In Figure 10.19 is shown the THD plot for such an A C amplifier working at 20 and 30 W into 8 . At 20 W the distortion is very low indeed, no higher than a pure Class-A amplifier. At 30 W the transition gain-steps appear, but the THD remains very well controlled, and no higher than a Blameless Class-B design. Note that as in Class-B, when the THD does start to rise it only does so at 6 dB/octave. The qui- escent current was set to 1.5 A. Figure 10.20 reveals the THD residual during A C operation. There are absolutely no crossover arti- facts, and the small disturbances that do occur happen at such a high signal level that I really do think