Reference

Designing Audio Power Amplifiers, Second Edition, by Bob Cordell, Routledge/Focal Press, 2019. 776pp. $160.00, hard cover; $97.95, paperback.

I first met Bob Cordell at clinics he gave at the last audio show Stereophile organized, Home Entertainment 2007, in Manhattan. At those clinics, Bob shared his views on why amplifier measurements are not always good at predicting differences in sound quality. So when I was scheduled to host a seminar—"Amplifiers: Do Measurements Matter?"—at this year's now-canceled AXPONA, Cordell was on the short list of designers I felt would offer valuable insight.

Bob describes himself as "both an engineer and a DIY hobbyist, and I love to mix the two." He has been deeply involved in audio since designing vacuum-tube amplifiers in his teens, branching out in the years since into designs using bipolar transistors and MOSFETs, as well as designing audio test equipment. Bob has worked at Bell Laboratories and other telecommunications companies, and also consults in the audio industry. Probably only Nelson Pass and John Curl know as much as Bob Cordell does about amplifier circuits.

The first edition of Designing Audio Power Amplifiers was published by McGraw-Hill in 2011. For the long-awaited second edition, Cordell has rewritten many of the original chapters and added five new ones, on such subjects as low-noise design, professional power amplifiers, and switch-mode power supplies (footnote 1).

This massive book—36 chapters, each with a comprehensive list of references—is divided into six parts. In the first part, Bob covers the basics of power amplifier design, including discussions of noise, damping factor, negative feedback, and slew rate. Each of these preliminary sections is succinct, providing a clear picture of what an amplifier is supposed to do and how it can do it. The specifics of circuit design and operational parameters are introduced by taking the reader through the design of an actual amplifier.

Modeling a typical 50W solid-state amplifier circuit follows in the third chapter, with examples of how the performance of each stage could be improved, illustrated using SPICE calculations of total harmonic distortion. (A complete discussion of SPICE simulations of circuits and components is offered in Part Four.) The next chapter is new to this edition: a case study of actually building a practical, 125W amplifier, starting from first principles and continuing to prototyping, construction, and testing. It ends with ideas for "upgrades," such as using a DC servo to eliminate the need for series coupling capacitors.

Those ideas are examined in detail in the second part of the book, "Advanced Power Amplifier Design Techniques," which also examines the essential subjects of feedback compensation and dealing with crossover distortion, as well as how error correction can be used to give very low THD across the audioband.

Part Three, "Real-World Design Considerations," examines how an amplifier copes with life in—yes, the real world, with discussions on such things as power supply design, grounding, output-stage thermal design and stability, short-circuit protection, control of clipping, and "Civilized Amplifier Behavior." In addition to SPICE simulation, Part Four includes chapters on measurements and instrumentation.

Part Five, "Topics in Amplifier Design," begins by examining the pros and cons of negative feedback, a controversial subject, as you will appreciate if you look at Stereophile's measurements of amplifiers that use high, low, or no feedback (footnote 2). It is clear from the outset where Bob stands. "Negative feedback has gotten a mostly undeserved bad rap. Much of this because poorly designed solid-state amplifiers of the 1970s happened to use large amounts of negative feedback. . . . Some designers carelessly believed that negative feedback could be used to linearize a design that was inherently not very linear to begin with; . . . the negative feedback allowed designers to make bad choices or cut corners." Amen to that—and to this: "[T]he open-loop bandwidth of a feedback amplifier should extend to the highest audio frequencies."

Bob continues his analysis of feedback by examining the various ways it can cause problems. One of those is something I have written about: A speaker cable can act as an antenna injecting RF noise into the amplifier via the feedback network (footnote 3). Bob is skeptical about this potential problem but does admit that this is a good reason to use an input stage with good signal-handling capability up to high frequencies.

The book concludes with four chapters on class-D amplifiers, including how such designs are best measured. "The stumbling block to adoption of class-D amplifiers in the past has been sound quality," Bob writes, summing up this section. "This has changed dramatically in recent years but still has a ways to go for high-end audio."

I know some engineers who will take exception to that statement!

Designing Audio Power Amplifiers is not a book to be devoured in one sitting. On the contrary: Each chapter requires days of thought. It's dense and expensive—but it's also essential reading for anyone fascinated by the superficially simple idea of how to make a small electrical signal powerful enough to drive a loudspeaker without degrading that signal in the process.

---

Footnote 1: More information on Designing Audio Power Amplifiers can be found here. A one-minute video interview with the author can be found here.

Footnote 2: See Martin Colloms's primer on this subject.

Footnote 3: See my discussions of this mechanism here and here.