Transistor Amplifiers , Dennis Feucht, Innovatia, 2016, ISBN: 9781682736265; $60 US.
Transistor amplifier literature has existed for well over a half century. Why another book on an old subject? What is lacking in the analog circuits literature are detailed examples that give newer engineers gaining confidence in design some exposure to the thought sequences of an experienced engineer. The last half of this book presents the detailed design of about a dozen amplifiers of 2 to 6 transistors (one being a bandgap reference) by “walking through” their design considerations at an engineering level of detail as an actual design activity. Several amplifiers are designed in this book, step-by-step, with explanation. A few are shown below.
One amplifies temperature: a bandgap reference, Paul Brokaw style:
While going through design details, the actual circuits – most based on a monolithic NPN-PNP array, the CA3096 – are built, measured (with measurement considerations included such as probe loading and transmission-line sources), and compared to the theoretical calculations, then assessed. The following book excerpt (page 316) is of cascode amplifier 3, shown above. Power-supply rejection is measured and calculated, and the results compared.
PSR is analyzed by considering each supply as an input port and finding the gain from it to the output. No new concepts are required to perform this kind of quasistatic analysis. Measurements of the prototype of amplifier 3 resulted in the following PSR data:
The amplifier amplifies variations in the +12 V supply by a factor of –1.86. It is even more sensitive to –12 V. Variations in either are amplified instead of rejected, and this is a weakness of the circuit.
By inspecting the circuit structure, V – has a direct (×1) effect through RL2 on νO . For Q1, as V – decreases slightly, it is attenuated by the RE divider and then amplified. The attenuation is
The gain magnitude of Q1 as a CB stage is the same as the CE including the above attenuation. Thus the overall gain of V – through Q1 to νO is about –20/6 = –3.33. The sign opposes the positive change through RL2 and the combined gain is thus –3.33 + 1 = –2.33, not too different from the measured value.
Considerations that involve calculation from design formulas derived in the first half of the book include
- static design (biasing) for maximum dynamic range and bandwidth
- apportionment of gain between stages
- stage interaction and isolation to achieve circuit modularization. A new method of handling pole separation is developed.
- temperature drift, thermal design and on-chip thermal feedback
- beta sensitivity
- linearity (bit-accuracy approximation)
- power-supply rejection
- intrinsic noise
- large-signal (slew-rate) limitations
- multiple dynamic aspects including dual-path dynamics and
- impedance gyrations in the high-frequency region of transistor operation, shown graphically below for the example of a capacitively-loaded emitter with base R and C. The base resistance is referred to the emitter as a parallel LR which can resonate with the emitter capacitance in the high-frequency region (between fβ and fΤ ) of the transistor. (The same theory applies to FETs.)
Most of the time, algebra and a calculator produce reasonable approximations of design parameters, including bandwidth, but the succession of designs shows where calculator analysis (mainly of bandwidth) begins to fail and a simulator is needed. Design covers what is to be done before turning on a circuit simulator.
The first half of the book presents the fundaments of analog circuits in a refined and clarifying way, covering both passive and active circuits principles. This book also is unusual in that it attempts to address a very broad range of readership, from experienced engineers to those aspiring to be engineers, including astute young pre-university students who are good at math and science. Consequently, the challenge is taken to develop circuits concepts with a minimum of advanced mathematics as such. The mathematical basis for Laplace transforms, differential equations, and most of differential and integral calculus (no less advanced calculus) is not actually needed to think in the complex-frequency domain, which mainly involves complex-number algebra. I have included just enough calculus to bridge the gap for the pre-calculus student yet hopefully not annoy engineers too much. While chapter one starts out slow and easy, the book ends at a level suitable for an advanced analog circuits class or engineers in industry who want to find out what they might have missed.
The last chapter expands the scope to mixed-signal circuits and develops both quasistatic and dynamic transfer functions for sampling circuits, including the important cases of DACs and ADCs. There is some wider discussion about design and how engineering is different from science, a 3D view of circuits (structure, behavior, and function), seeing the circuits in the math of “low entropy” equations, the challenge of making the circuit perform beyond that of the components, about why structural or physical models are superior (most of the time) to black-box or behavioral models, the two-part reduction theorem (β and μ transforms), Middlebrook’s EET and ZEET analysis methods, and some recommended books with commentary.
This book is intended to be complementary to my Analog Circuit Design book-set while also enabling newer designers to access it. Even the simpler circuits in this book involve problems that experienced electronics engineers can ponder. I hope that for them, a few useful insights will be added to their already considerable knowledge of transistor amplifiers. Writing this book has done that for me!
Book details can be accessed at Innovatia where there is also a link to the printer purchase webpage.