Composite Amplifiers Part Two: The Composite Amplifier Choice

The second part of our series explaining the nature, characteristics, benefits, and caveats of composite amplifiers.

This is the second part in our series explaining the nature, characteristics, benefits, and caveats of composite amplifiers. The four parts of the series include:

The Composite Amplifier

A cascaded amplifier system is a multiple-stage amplifier system in which the output of a given amplifier is used to drive the input of the next successive stage. Each stage of a cascaded amplifier system typically has negative feedback. This is sometimes described as local negative feedback. Now we shall examine a composite amplifier.

A composite amplifier combines the best aspects of two (or more) different amplifiers into one high-performance amplifier. Upon inspection, individual amplifier stages may use local negative feedback to establish their voltage gains, but there will also be an overall negative feedback loop wrapped around the entire system.

The defining characteristic of a composite amplifier is its topology. Specifically, there is an overall negative feedback loop that connects the cascaded amplifier system output back to the amplifier system input. The amplifier system usually includes at least two amplifier stages. This arrangement permits the inclusion of amplifiers that offer specific system performance advantages that exceed the capabilities provided by a single amplifier choice. Therefore, the term “composite” is invoked.

A composite amplifier consists of two or more operational amplifiers cascaded together with a negative-feedback loop around the entire network. Composite amplifiers (also described as nested-feedback amplifiers) widen the bandwidth at high gain, boost slew rates, lower distortion, reduce output DC offset, and lessen noise.

Typical Op Amp Selection Categories

There are always tradeoffs. Exchanges. “I want to lose weight, but I also want muscle definition. Oh, wait—muscles are denser. I will probably not lose as much weight as I want.” Electronics works the same way. Consider the categories that follow:

Precision is associated with the input DC offset voltage (VOS). It should be small. The drift with temperature should be low. A large DC voltage gain is desirable along with small input bias currents.

Speed refers to having a large, small-signal bandwidth and a large slew rate. (Slew rate defines how quickly the op amp output voltage can change in response to an abrupt input signal change). Generally, operating currents run larger.

Power requirements are critical in battery-operated applications. The quiescent current (IQ) should be small. At low operating voltages, the ability to provide a rail-to-rail output voltage becomes important to provide a large output dynamic range.

Large-Output-Current Buffers typically provide unity gain, high speed, and wide bandwidth with the capability to deliver output currents of tens to hundreds of milliamperes. They are usually touted for having the ability to drive capacitive loads, such as those associated with loads connected using shielded and coaxial cables.

An AD711 op amp is to be used to provide a voltage gain of -3 (9.54 dB). The AD711 is described as a “a precision, low-cost, high-speed BiFET op amp.” It also has a small value of input offset voltage. Multisim was used to determine the input offset voltage as shown in Figure 9(a). The non-inverting amplifier was constructed as indicated in Figure 9(b). The input offset is about 0.5 mV. The inverting amplifier has a noise gain of 4, so the output voltage offset of approximately 2 mV makes sense. Again, by convention the noise (DC offset or AC noise) equivalent is applied to the non-inverting input terminal. The noise gain is the gain provided by the non-inverting amplifier.

Figure 9(a). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 9(a). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.
Figure 9(b). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 9(b). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

A 0.2-V peak signal source is added as illustrated in Figure 10(a). The oscilloscope display shows the peak output voltage is about 0.6 V as expected. The Bode plotter (Figure 10(b)) indicates a bandwidth of roughly 1.54 MHz.

Figure 10. Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 10. Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

So far, we see a low output DC offset, excellent gain precision, and a respectable bandwidth. Here is the rest of the problem: The amplifier must deliver the same results if it is driving a 10 Ω load. A 10 Ω load requires the amplifier to source and sink 60 mA. (When an output is sourcing current, current flows out of the output terminal. By the transistor convention, these currents are called negative on data sheets. Currents that flow into the output are described as sinking and are labeled as positive values on data sheets.) This situation is provided in Figure 11(a). Notice that clipping occurs. The output voltage is limited to about ±0.4 V. One solution is to incorporate a BUF602 within the feedback loop as shown in Figure 11(b). This is a true composite amplifier. The BUF602 is described as a “closed-loop high-speed buffer.” It has an extremely wide bandwidth of 1000 MHz, can provide 60 mA of output current continuously, and can provide peak output currents of ±350 mA. The buffer has unity voltage gain and a wide bandwidth. “Wide bandwidth” means its pole is many megahertz away from the 1.54 MHz pole of the AD711. So, oscillation or frequency response peaking should not be a problem (this will be explained in detail later).

Figure 11(a). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 11(a). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

In this composite amplifier, the AD711 provides the DC precision and wide bandwidth. The BUF602 provides the output current drive. The overall feedback is wrapped around the cascaded devices. Since the BUF602 is a unity-gain, non-inverting amplifier with a very wide bandwidth, it does not disturb the operation of the AD711 appreciably.

Figure 11(b). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 11(b). Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Observe in Figure 11(b) the clipping is removed, and the output is again ±0.6 V. This means the composite amplifier is providing 60 mA peak output current. The bandwidth of the composite amplifier was obtained by using Multisim as shown in Figure 12.

Figure 12. Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 12. Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

The bandwidth decreased slightly (from 1.54 MHz to 1.26 MHz) but remains respectable. If the gain of the buffer remains unity, the system bandwidth will remain constant. However, in this instance, the gain of the buffer decreases slightly. This causes the gain required of the AD711 to increase. If the gain-bandwidth product of the AD711 remains constant, its bandwidth must decrease.

More on Composite Amplifiers

Composite amplifiers can offer many advantages in addition to providing a large output drive current. Wider bandwidth, lower output DC offset, and even lower noise can be achieved using a composite amplifier. In contrast, a cascaded amplifier will not offer lower output DC offset nor reduce output noise.

Figure 13 shows a composite amplifier that includes two separate op amps labeled AR1 and AR2.  Local negative feedback is provided by R4 and R3. They establish the closed-loop voltage gain of AR2 to be 1 + R4/R3. Within the composite amplifier we observe that AR1 has no local feedback.  Within the confines of the composite amplifier, AR1 appears to be open loop. However, it will develop an effective voltage gain to satisfy the overall voltage gain requirements. The overall negative feedback establishes the voltage gain of the composite amplifier to be 1 + R2/R1. A summary is provided by the points below:

  • The overall voltage gain of the composite amplifier is:

  • The voltage gain of the second stage is determined by its local negative feedback and is given by:

  • The first stage has no local negative feedback, but its effective voltage gain will adjust. It can be determined by the relationship:

Figure 13. Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Figure 13. Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists.

Review and Conclusions

When a large voltage gain is required, a cascaded amplifier can provide the requisite voltage gain while maintaining a respectable bandwidth. However, each amplifier stage contributes to the output DC offset and the AC output noise. Typically, each amplifier employs local negative feedback to establish its voltage gain and bandwidth. A composite amplifier usually incorporates stages that provide distinct features. For example, one stage might provide DC precision while another may provide output current drive. Local negative feedback as well as overall negative feedback are present in composite amplifiers. Typically, the input stage does not incorporate local negative feedback. Its effective voltage gain will adjust as required to accommodate the gain of the other amplifier stages and the overall voltage gain requirements.

In part three, we’ll discuss amplitude response peaking and the potential for oscillation. A more intuitive—as opposed to a mathematical—approach will be used.