A review of some popular types of op amps, plus a close look at a specific application
BY TAMARA SCHMITZ
Senior Principal Applications Engineer and Applications Manager for Light Sensors
Intersil
www.intersil.com
Are op amps all the same? Why are some better than others at certain tasks? How are some op amps better than others? Below, we review some of the most popular types of op amps, then take a close-in look at a specific application. The application we’ll use is a guitar and microphone amplifier, mixer, filter and speaker driver. Could the same op amp work in all these functions? The answer is this: not likely and definitely not one that I’ve been able to find. If an op amp had all the characteristics needed to fulfill all of those roles, it would be too expensive to use. Let’s find out why.
Choosing an op amp can be a daunting task. Within a very short period of time, the op amp adventurer will run across terms like voltage feedback and current feedback, precision and high speed, instrumentation and buffer, low power and low noise, and specialty parts like current sense amps and power amps, just to name a few. A comprehensive discussion of all of these types would make for a book and probably a sleep aid.
Fig. 1. Block diagram of a simple two2-input audio amplifier.
The first concept to understand is that op amps are typically designed to fill a specific need. Handheld devices run on batteries and need to use components that are low power. Communication devices need to use radio frequencies to have reasonable size antennas, so their components are largely high speed. Military devices need to be extremely reliable, so they have components tested to extreme temperatures and in a range of test setups.
How does this apply to op amps? It will help to work with an example. Look at the system in Figure 1. This is the block diagram of a two-input guitar and vocal amplifier. Conveniently, each of the five blocks in this system can be built with an op amp.
Before we step through the blocks, let’s acknowledge the subset of operational amplifiers we will be considering. Since this is an audio application, frequencies above 20 kHz are not needed since the human ear can’t hear those frequencies. It follows that we won’t need any “high speed” amplifiers since those run at frequencies far above 20 kHz — like 1 MHz, 100 MHz, or even a few gigahertz (1,000,000,000 Hz).
A second assumption we are going to make is that all of the amplifiers in our system are voltage feedback. This is the default type of op amp. (Personal note: I didn’t learn about current feedback amplifiers until I spent time in industry.) Current feedback amplifiers use a different type of feedback signal than standard voltage feedback amplifiers.
Predictably, the feedback is a current instead of a voltage. This special feedback allows the amplifier to operate very quickly. In this application, it doesn’t make sense to choose the added performance and expense of a current feedback amplifier.
By the way, if an amplifier datasheet doesn’t specify whether it is voltage feedback or current feedback, check out the impedance of the inverting input. If the inverting input is high impedance, then it is voltage feedback. Conversely, if the input impedance is low, it is current feedback. If the type of feedback is not specified, it’s a pretty safe guess that it is a voltage feedback amplifier.
So far, we know that the op amps in our system will be low (audio) frequency and have the standard voltage feedback topology. Now we need to look at the specifics of the system.
Non-inverting feedback’s benefits
The microphone amp and guitar amp are very similar. Though they are shown in Fig. 1 as if they have a single input, they will be standard two-input and one-output op amps. While they can be configured with inverting feedback or non-inverting feedback, I strongly suggest non-inverting feedback.
Non-inverting feedback circuits separate the circuit feedback from the input. In doing so, they also ensure that the input device, a microphone or guitar in this case, see a high input impedance. This is a convenient place to put gain (amplification) as well, since you probably want to be able to control the loudness of each input separately.
Remember that hearing responds to logarithmic changes, not linear ones. Therefore, if a potentiometer (variable resistor) is used to vary the gain of the input circuit, make sure to use a logarithmic potentiometer, not a linear one.
Since the signals are small at the input, we would like the noise to be low as well or it might swamp out our signals. In radio circuits or in very sensitive applications, it is crucial to use a low-noise amplifier at the input stage. In our simple audio amplifier, it is a little less critical. In other systems, the input amplifier may be an instrumentation amp. This is an extremely balanced two-input amplifier with high input impedance, like the ISL28617.
The next stage in our flow is the “summer.” The summing amplifier is another common place to put some variable gain. Now, the output from the microphone input amplifier and the guitar input amplifier are attached together at the inverting input of the amplifier to act as a summer. A feedback resistor from the output to that same inverting input can be varied to provide system gain.
The requirements on this amplifier are not very stringent. It doesn’t need to be low noise, since the signals are most likely 10 or 100 times greater than they are at the input. They don’t need to have high bandwidth since this is an audio system. Also, they don’t need to drive large output currents since they are driving the input of another op amp. The requirements on the filter stage amplifier are similar to those on the summer.
Another mid-system amplifier possibility is an amplifier that might drive a converter. This particular system doesn’t convert the analog systems to digital. However, if it did, then the amplifier before the converter must have sufficient resolution to match the number of digital bits to be generated.
The last amplifier in the audio amplifier drives a speaker. This action takes a much larger amount of current than all of the other stages. Also, since the signal has been amplified by the time it reaches this point in the system, it is at its largest. That means the output will be moving the fastest, since the guitar and the (hopefully) melodic voice are composed of sine waves. That means the best choice is an amplifier with sufficient slew rate, which is defined as the volts per microsecond that the output can change. This is the most glaring mistake that shows when designers try to use the same amplifier for any and all op amp applications. Basic amplifiers like the OP-07 or the 741 may barely succeed in the beginning and middle stages of the audio amp, but they fall flat at this output stage. They are simply unable to drive the output fast enough — a limitation in slew rate.
A side amplifier that can be used is a current sense amplifier. Current sense amps are used within power supply lines to monitor the current consumed in a circuit or system. They can be placed near the supply or near a ground connection, if they are designed like the ISL28006.
All kinds of different amplifiers noted above are available now and are getting better all the time. There really is no substitute for carefully reading the datasheet and making sure you choose a part that fits a system’s needs. With practice and with time, the trade-offs and optimizations become second nature. Don’t forget the insights that can be gained by running a few simulations, too. Using a program like Intersil’s iSim simulator can help you choose the right component, build the right filter, and ensure system design success. ■
Advertisement
Learn more about Intersil
