Op-Amps: Part I

What is an Op-amp?

These are the philosophical questions that test our souls.  (Get out your #2 pencils!)

Figure 1 – A Texas Instruments op-amp in 8 pin dual in-line package (DIP)

Op-Amp – short for operational amplifier.  I’m sure you’re familiar with what an amplifier is: you plug your small signal in at one end, and it makes it louder at the other.  The operational part, however, references their initial intended use of peforming mathematical operations.  So it’s a special amplifier that’s good at doing precise calculations, which is why it’s used in almost every modern electronic audio product, from mixers to digital converters.

A picture of a typical op-amp in an 8 pin DIP package is shown in Figure 1, and the schematic representation of a single op-amp (how you’d draw it so others know what you’re talking about) is shown in Figure 2.

Figure 2 – Standard op-amp symbol

The basic op-amp has 5 pins, so there are 3 unused pins in the 8 pin DIP (Dual In-line Package).  Two of these pins are power supply pins – where the amplifier gets its power from (Vs+ and Vs-). Two of these pins are input signal pins – where you put in your small signal (Vi+ and Vi-).  And the final pin is the output pin – where the amplified signal come out (Vo).  The opamp does the simple mathematical function of Vo = A * [(Vi+) - (Vi-)], or more plainly, it takes the difference between its input signals, gives this a gain of A, and presents it at the output.

A common single op-amp layout is shown in Figure 3.  Keep in mind, however, that layouts may differ and you should always check the data sheet of the model you are using to see how it has been laid out.

Figure 3 – A single op-amp layout on an 8 pin DIP

So if the amplifier is all set up and packed into its little chip already, why can’t we just put a signal on the input pins and get what we want out the other end?  This is because the gain (A) is usually at least 100,000.  To put this in perspective, your typical home stereo takes an input signal of 1V and puts out an output signal of 50V to drive your speakers.  That’s only a gain of 50, and its already pretty darn loud.  Most microphone amplifiers have a maximum gain of +60dB, which is only a gain of 1,000.

So what’s the use of such a ridiculously large gain?  The answer is negative feedback.  By adding some resistors to your op-amp, you can trade some of that gain for precision and stability.  How exactly this works will be covered in the next section, but it’s the basic principle that makes most analog sound processing possible.  It was first concieved in 1928 by Harold Black for telephone lines [1]. So stand proud, audio applications have driven the electronics industry from the beginning!   Just to name a couple – the first vacuum tube Lee DeForest invented was called the “Audion“, and Hewlett-Packard started Silicon Valley with their first product, the 200A, an audio oscillator.

HP 200A - Based on Bill Hewlett’s master’s thesis

Lee DeForest’s ‘Triode’ Audion

Before there was negative feedback you relied upon the gain of your vacuum tube, which could vary greatly depending upon temperature or service life. Even with today’s precision transistor fabrication facilities, the gain of a single transistor varies from part to part, as many analog synthesizer enthusiasts can attest to.  This gain can also vary with the frequency range of signals you put in, so the entire range can become peaky and distorted. But, with negative feedback, the gain is no longer set by the inherant gain of the device, but rather by the external resistors you use.  And we are really good at making accurate resistors.  So the internal gain of your opamp (A) can vary from 100,000 to 200,000, and it won’t matter much at all, because the external components are doing the work.

This reliance on external components not only gives accuracy, it also gives versatility and ease of use.  Basically, someone has done all the difficult bits of transistor design for you, and abstracted it away to simple math.  This is similar to how modern programming languages take the difficulty of setting bits inside of a computer and abstract it away into english.  You can just buy the op-amp and start building, which is why we are going to focus so heavily on them here at Open Music Labs – you really can do just about anything with them.  By combining them with resistors and other components, you can build filters, multipliers, logarithmic converters, fullwave rectifiers, mixers, and much more.  So if you’re interested in building any of these things, look around the site for more information, and if it’s not here, let us know and we’ll rectify the situation.

— Footnotes —

Most of the historical information herein is taken from Analog Devices’ “Op Amp Applications” seminar notes, edited by Walt Jung, the author of the ever useful “IC Op-Amp Cookbook”.  Both are worth reading, with the “IC Op-Amp Cookbook” being more accessible.  The seminar notes can be downloaded directly from Analog’s website: